ENGG390 – Thayer School of Engineering

Final Report
Prototype Development Plan for ReviveFlow Corporation

 

Project by: Aaron Gjerde

aaron.gjerde@dartmouth.edu

612-669-4315

September 12, 2008

 

Faculty Advisor: J. Andrew Daubenspeck

 

Sponsor: ReviveFlow Corporation c/o Jim Bleck


 

 

Table of Contents

 TOC \o "1-2" Executive Summary                                                                    PAGEREF _Toc208854904 \h 2

Methodology                                                                                PAGEREF _Toc208854905 \h 2

Work Plan                                                                                                               PAGEREF _Toc208854906 \h 2

Change Summary                                                                                                    PAGEREF _Toc208854907 \h 2

Timeline                                                                                                                  PAGEREF _Toc208854908 \h 2

Work Accomplished                                                                     PAGEREF _Toc208854909 \h 2

System Model                                                                                                         PAGEREF _Toc208854910 \h 2

Failure Modes and Effects Analysis                                                                         PAGEREF _Toc208854911 \h 2

Specifications                                                                                                         PAGEREF _Toc208854912 \h 2

Component Selection                                                                                              PAGEREF _Toc208854913 \h 2

Request for Quotation – RFQ                                                                                   PAGEREF _Toc208854914 \h 2

Research Summary                                                                                                 PAGEREF _Toc208854915 \h 2

Clinical Background                                                                    PAGEREF _Toc208854916 \h 2

Anatomy                                                                                                                  PAGEREF _Toc208854917 \h 2

Physiology                                                                                                              PAGEREF _Toc208854918 \h 2

Appendix A                                                                                  PAGEREF _Toc208854919 \h 2

Stroke Information                                                                                                  PAGEREF _Toc208854920 \h 2

Appendix B                                                                                  PAGEREF _Toc208854921 \h 2

Request for Quotation                                                                                             PAGEREF _Toc208854922 \h 2

Appendix C                                                                                  PAGEREF _Toc208854923 \h 2

Mathematical Model                                                                                                PAGEREF _Toc208854924 \h 2

Appendix D                                                                                  PAGEREF _Toc208854925 \h 2

Component Chart                                                                                                    PAGEREF _Toc208854926 \h 2

Appendix E                                                                                   PAGEREF _Toc208854927 \h 2

Animal Model Research Summary                                                                           PAGEREF _Toc208854928 \h 2

Appendix F                                                                                   PAGEREF _Toc208854929 \h 2

Patent Summaries                                                                                                   PAGEREF _Toc208854930 \h 2

Appendix G                                                                                  PAGEREF _Toc208854931 \h 2

Company Summaries                                                                                              PAGEREF _Toc208854932 \h 2

Appendix H                                                                                  PAGEREF _Toc208854933 \h 2

Failure Modes and Effects Analysis                                                                         PAGEREF _Toc208854934 \h 2

Appendix I                                                                                    PAGEREF _Toc208854935 \h 2

Cerebral Arteries                                                                                                     PAGEREF _Toc208854936 \h 2

Appendix J                                                                                   PAGEREF _Toc208854937 \h 2

Cerebral Venous Sinuses                                                                                        PAGEREF _Toc208854938 \h 2

Appendix K                                                                                  PAGEREF _Toc208854939 \h 2

Neurophysiological Parameter Values                                                                     PAGEREF _Toc208854940 \h 2

Appendix L                                                                                   PAGEREF _Toc208854941 \h 2

Updated Task List                                                                                                    PAGEREF _Toc208854942 \h 2

Appendix M                                                                                  PAGEREF _Toc208854943 \h 2

Works Cited                                                                                                             PAGEREF _Toc208854944 \h 2

 


 

 

List of Tables and Figures

 TOC \c "Table" Table 1: Specification Matrix................................................................................................. PAGEREF _Toc208854843 \h 2

Table 2: Component Chart..................................................................................................... PAGEREF _Toc208854844 \h 2

Table 3: Animal Model Matrix................................................................................................ PAGEREF _Toc208854845 \h 2

Table 4: CoAxia Funding Summary....................................................................................... PAGEREF _Toc208854846 \h 2

Table 5: CoAxia Issued Patents............................................................................................. PAGEREF _Toc208854847 \h 2

Table 6: CoAxia Published Patent Applications..................................................................... PAGEREF _Toc208854848 \h 2

Table 7: Severity, Occurrence, Detection definitions.............................................................. PAGEREF _Toc208854849 \h 2

Table 8: FMEA....................................................................................................................... PAGEREF _Toc208854850 \h 2

 

 TOC \h \z \c "Figure" Figure 1: Stroke core and penumbra. PAGEREF _Toc208854851 \h 2

Figure 2: Types of stroke. PAGEREF _Toc208854852 \h 2

Figure 3: Stroke incidence. PAGEREF _Toc208854853 \h 2

Figure 4: NIH research budget allocation. PAGEREF _Toc208854854 \h 2

Figure 5: tPA odds ratio. PAGEREF _Toc208854855 \h 2

Figure 6: ReviveFlow Timeline. PAGEREF _Toc208854856 \h 2

Figure 7: Timeline. PAGEREF _Toc208854857 \h 2

Figure 8: Venous compliance and pressure. PAGEREF _Toc208854858 \h 1

Figure 9: Cranial anatomy. PAGEREF _Toc208854859 \h 2

Figure 10: Corrosion cast of cerebral vasculatureError! Bookmark not defined. PAGEREF _Toc208854860 \h 2

Figure 11: Embolus vs. ThrombusError! Bookmark not defined. PAGEREF _Toc208854861 \h 2

Figure 12: Vessel cross-sectional area. PAGEREF _Toc208854862 \h 1

Figure 13: Vessel as a Starling resistor PAGEREF _Toc208854863 \h 2

Figure 14: Notation for reversed flow.. PAGEREF _Toc208854864 \h 2

Figure 15: Artery vs Vein compliance. PAGEREF _Toc208854865 \h 1

Figure 16: CSF pathways. PAGEREF _Toc208854866 \h 1

Figure 17: Stroke as 3rd leading cause of death. PAGEREF _Toc208854867 \h 2

Figure 18: Elkins et al: historical and projected US stroke deaths. PAGEREF _Toc208854868 \h 2

Figure 19: ReviveFlow concept drawing. PAGEREF _Toc208854869 \h 2

Figure 20: Catheter detail PAGEREF _Toc208854870 \h 2

Figure 21: Canine vascular anatomy. PAGEREF _Toc208854871 \h 2

Figure 22: Stella model PAGEREF _Toc208854872 \h 2

Figure 23: Ursino electrical (A) and mechanical (B) analogs. PAGEREF _Toc208854873 \h 2

Figure 24: Lakin et al (2003) whole body model PAGEREF _Toc208854874 \h 2

Figure 25: Stevens model PAGEREF _Toc208854875 \h 2

Figure 26: Anatomic Variation of Arterial Supply in Dogs. PAGEREF _Toc208854876 \h 2

Figure 27: Reversed flow to body part "P". PAGEREF _Toc208854877 \h 2

Figure 28: ReviveFlow diagram with switching circuit PAGEREF _Toc208854878 \h 2

Figure 29: Design Mentor invention illustration. PAGEREF _Toc208854879 \h 2

Figure 30: Design Mentor assembly diagram.. PAGEREF _Toc208854880 \h 2

Figure 31: Neuroperfusion catheter PAGEREF _Toc208854881 \h 2

Figure 32: Neuroperfusion catheter intracranial location. PAGEREF _Toc208854882 \h 2

Figure 33: Alsius coolgard patent illustration. PAGEREF _Toc208854883 \h 2

Figure 34: Catheters placed near aortic arch. PAGEREF _Toc208854884 \h 2

Figure 35: Occluding catheter reverses blood flow.. PAGEREF _Toc208854885 \h 2

Figure 36: Frazee et al (1989) experimental setup. PAGEREF _Toc208854886 \h 1

Figure 37: Alsius Timeline. PAGEREF _Toc208854887 \h 2

Figure 38: Retroperfusion Systems Inc. overview.. PAGEREF _Toc208854888 \h 2

Figure 39: CoAxia's NeuroFlo Catheter PAGEREF _Toc208854889 \h 2

Figure 40: CoAxia Clinical Study Sites. PAGEREF _Toc208854890 \h 2

Figure 41: Arteries to the Brain. PAGEREF _Toc208854891 \h 2

Figure 42: Circle of Willis. PAGEREF _Toc208854892 \h 2

Figure 43: Cerebral arteries. PAGEREF _Toc208854893 \h 2

Figure 44a(above), b(below): Cerebral Venous Sinuses. PAGEREF _Toc208854894 \h 2

Figure 45: Frequency distribution of CBF values in 21 humans. PAGEREF _Toc208854895 \h 2

Figure 46: Mean pressure in vasculature. PAGEREF _Toc208854896 \h 2

Figure 47: Intracranial Compliance Curve. PAGEREF _Toc208854897 \h 2

Figure 48: ICP frequency distribution. PAGEREF _Toc208854898 \h 2

Figure 49: Cranial Inflows and Outflows. PAGEREF _Toc208854899 \h 2

Figure 50: Change in ICV during one cardiac cycle. PAGEREF _Toc208854900 \h 2

Figure 51: Important pressure relationships with CBF. PAGEREF _Toc208854901 \h 2

Figure 52: Feedback responses over time. PAGEREF _Toc208854902 \h 2

 


 

 


 

 

Executive Summary

Someone experiences a stroke every 45 seconds in the United States, on average[1]. The project got off to a late start, but ran mostly ahead of schedule.  All committed deliverables are included in the attached appendices of this report.  Changes to scope representing minor risks to the timeline have been encountered, but careful consideration for scope additions and planned risk mitigation strategies helped prevent serious issues.  Mostly this involved working with the sponsor to agree to acceptable contingencies if certain information was not available, such as the animal model, and by working with the sponsor to identify limits to scope, such as choosing a simple mathematical model.  Extensive literature and textbook reviews coupled with assistance from Thayer School Professors Hoopes and Daubenspeck as well as interviews with neurosurgeons, intensivists, and anesthesiologists from Dartmouth-Hitchcock Medical Center helped assure that clinical issues were accounted for.

The work completed to date builds on the previous accomplishments and provides critical deliverables for the sponsor.  The sponsor has met with an interested venture capital firm to fund the prototype and first experiment, which this specification defines. Therefore the RFQ was submitted to the vendor in draft form ahead of schedule. The timeline of this project aligned well with the sponsor as their next meeting with the investors is on September 16.

 

 

Stroke Overview

Figure  SEQ Figure \* ARABIC 1: Stroke core and penumbra[2]

Strokes are caused by a blockage of blood flow or ruptured blood vessels in the brain, which prevents oxygenated arterial blood from reaching the capillaries where oxygen can perfuse the brain.  The specific location of the blockage or rupture is called the “core” and the oxygen-deprived area surrounding the core is the called the “penumbra,” or the ischemic region as shown in  REF _Ref208854745 \h Figure 1.  The resulting damaged area from a stroke is called the area of infarction, which is generally considered irreversible.  Infarction size depends on the severity, duration, and location of the core.  Although the penumbra has reduced oxygen and nutrient supply, it is considered tissue that can be saved.  Thus, the goal of most stroke therapies is to perfuse the penumbra as quickly as possible to reduce the potential infarction.  Removing a clot results in “reperfusion,” which can cause secondary brain damage if it occurs too abruptly.

Figure  SEQ Figure \* ARABIC 2: Types of stroke[3]

Strokes are categorized into two types: ischemic and hemorrhagic.  Ischemic strokes are the result of blockage in a cerebral artery, resulting an ischemic, or oxygen starved, zone.  Ischemic strokes are by far the most common, occurring in 87 percent of US cases.  Hemorrhagic strokes occur when a blood vessel in the brain ruptures and leaks either into the intracerebral space within the brain or the subarachnoid space surrounding the brain, as shown in  REF _Ref208854759 \h Figure 2.  Intracerebral episodes account for approximately 9 percent of US stroke cases and are often related to hypertension (high blood pressure). Subarachnoid hemorrhagic strokes account for the remaining 4 percent of US stroke cases. NOTEREF _Ref82498879 \f \h 1

Figure  SEQ Figure \* ARABIC 3: Stroke incidence NOTEREF _Ref208224416 \f \h 80

Approximately 700,000 strokes occur in the United States each year, resulting in 280,000 deaths, making stroke the third leading cause of death in the United States.  Worldwide, stroke claims 4.4 million deaths each year, which is approximately 9% of all deaths.  Stroke is a leading cause of long-term disability in the United States[4], and was estimated to cost $57 billion directly and indirectly in 2005.[5]  Although the National Vital Statistics Report shows a 6.8% drop in deaths due to stroke since 1958[6], Elkins et al project that deaths due to ischemic stroke will nearly double between 2002 and 2032[7] with a total cost projected by Brown et al to exceed $2 trillion from 2005-2050.[8]

To address this significant problem, the U.S. National Institute of Neurological Disorders and Stroke (NINDS) budgeted to spend $1.6 billion dollars on stroke-related research in 2008[9].  Although significant, the American Heart Association believes this is inadequate as it represents only 1% of the total research dollars spent by the National Institutes of Health (NIH). NOTEREF _Ref208286966 \f \h 10

Figure  SEQ Figure \* ARABIC 4: NIH research budget allocation[10]

Although estimates are unavailable, tremendous resources have been invested over the past fifty years in stroke prevention, detection, and treatment.  As an indicator of the effort, a search for the word “stroke” in the journal article titles on HighWire generates 207,773 results and over 3 million on Google Scholar.  Despite the tremendous amount of past research and current resources allocated to ongoing efforts, effective solutions remain elusive.

Problem Statement

The tragic loss of life and disability due to stroke and the current and projected enormous costs to the United States healthcare system combine to create a high priority for developing effective stroke treatments.  Despite decades of exhaustive research, a solution that is widely available does not currently exist.  Tissue Plasminogen Activator (tPA) by Genentech is the only FDA-approved treatment for stroke currently available.  Although proven effective, tPA must be administered intravenously within three hours of stroke onset (see odds ratio in  REF _Ref208284682 \h Figure 5) and is only effective for ischemic stroke, which limits its use to less than 5% of stroke patients.  Additionally, tPA is associated with numerous significant adverse negative outcomes.[11]  Several endovascular treatments are under development and various stages of clinical trial, but none are widely available commercially. NOTEREF _Ref208809256 \f \h 69, NOTEREF _Ref208809590 \f \h 70

Figure  SEQ Figure \* ARABIC 5: tPA odds ratio[12]

Solution

ReviveFlow hopes to improve upon the significant shortcomings of the few existing treatments by effectively treating multiple forms of stroke with a larger time window for therapy deployment.  ReviveFlow Corporation is a startup company whose innovative stroke therapy proposes to reverse cerebral blood flow to perfuse the brain with oxygenated blood during a stroke.  The ReviveFlow solution involves several pumps, multiple lumen balloon catheters, and a control system with a switch to reverse the flow of blood in the brain.  This reversed flow is proposed to perfuse the ischemic capillary bed via the venous system and may also dislodge the embolus.  ReviveFlow prefers to use existing components and technology wherever possible.  REF _Ref82272540 \h Appendix B shows concept drawings of the experimental setup optimized for canines.  The goal of the first experiment will be to prove that a blood flow reversal system can be deployed in an otherwise healthy animal without causing harm to the animal.  If the safety and feasibility test is successful, a follow up test of efficacy will be performed on large animals with a stroke model.

Sponsor Plan

In the timeline below from the ReviveFlow business plan, the first stage titled research is where the scope of this project fits.  The alpha prototype will be built by an outside vendor based upon the research, prototype plan, and RFQ produced by this project. 

Process Timeline

Figure  SEQ Figure \* ARABIC 6: ReviveFlow Timeline


 

 

Methodology

The overall project methodology defined in the proposal remains unchanged with four phases and associated milestones: Initiate, Define, Implement, and Finalize.  Each phase of the project is laid out with major tasks listed in  REF _Ref208854945 \h Appendix L.  Below is a description of the tasks completed during each project phase and a summary of changes to plan.

Work Plan

Phase 1: Initiate

The goal of the first phase was to select and propose a project.  The first step was to identify prospective projects.  Once a short list of three viable projects was identified, a decision was made to select the most promising project.  A pre-proposal was submitted describing the project.  The first project selected fell through after the pre-proposal due to major changes in direction within the sponsor.  A second project was quickly identified, although the work was initiated later than had been hoped. 

The second step occurred immediately, which was to perform preliminary research to understand the problem and develop a work plan.  This included conference calls and in person meetings with the sponsor in Boston.  Several patents were reviewed, some provided by the sponsor and reviewed in great detail, and others were noted as probably relevant and in need of further investigation.  The sponsor also provided a business plan and some technical drawings which were also studied in order to get up to speed on the project.

The work plan was presented in the proposal and included a description of the work to be accomplished, broken down into four phases, each with associated milestones and deliverables.  A high level description of the major tasks that would lead to the successful generation of the proposed deliverables was also listed in the proposal.  A faculty advisor was selected and confirmed, based on his knowledge of physiological systems modeling, as that task was identified as the most challenging and risky.

Measures of success were subsequently described in an email to Professor Graves as providing the promised deliverables on time in with content that satisfied sponsor expectations.

The proposal document was the deliverable defined for phase 1 and was provided on time to the sponsor, faculty advisor, and program director.

Phase 2: Define

Phase two defined the problem in more detail and generated the first set of deliverables, two of which were ahead of schedule.  Research was performed regarding the clinical background of stroke, physiological impact of proposed solution, fluid mechanics of blood, and effect of reversed pressure gradient on oxygen diffusion in capillaries.  Much more research work was completed than had been anticipated, so the research task from the implement stage was moved up to this stage, which was a change from the proposal.  The objective was to learn about cerebral anatomy and neurophysiology such that the relevant parameters could be identified both for the mathematical model, but ultimately for a system specification.  Existing theoretical models in literature that described intracranial hemodynamics were investigated and several possible models identified as candidates for implementation. 

A failure modes and effects analysis (FMEA) was completed based on the initial notes and documentation provided by the ReviveFlow team.  This helped identify items to be included in the specification, which was also drafted and distributed to the ReviveFlow Medical Advisory Board.  Due to fortunate scheduling, the task of reviewing the specification draft was moved up to this section and the feedback was incorporated into the specification for the first progress report.  This cost one extra business day due to the need to travel to Boston to review the specification, but it pushed the project ahead of schedule.

Phase 3: Implement

The work effort during this stage was primarily additional research to help understand and implement a model.  Since the major review of the specification was complete, only minor changes were needed before the specification was submitted to the vendor for review.  The vendor responded with questions, which were primarily conceptual and didn’t result in any changes to the specification. 

Additional research system model enabled implementation of an existing lumped parameter model by Ursino using Stella software with the assistance of Professor Daubenspeck.  This enabled the identification of potential physiological issues that could be introduced by reversed cerebral flow, which were recorded as metrics to be monitored during the experiment in the specification chart.

Possible components for the prototype device were identified and are listed in  REF _Ref82274805 \h Appendix D.  The remaining patents were reviewed and summaries have been provided to the sponsor, which are included in  REF _Ref82274393 \h Appendix F.

The deliverables defined in the proposal (revised specification, component selection, Progress Report 2) were completed and submitted to Thayer School and ReviveFlow according to schedule.  In addition, a basic model of intracranial fluid flows was also implemented.

Phase 4: Finalize

The major task for the final report involved analyzing and summarizing previous work.  Industry research summaries were completed and are presented in  REF _Ref82273713 \h Appendix G.  Information was gathered to help the sponsor make a decision about which animal model to use and is available for reference in  REF _Ref82507039 \h Appendix E.  The Request for Quotation was submitted in draft form to InterPlex Medical and an updated draft is included in  REF _Ref82272540 \h Appendix B.  Feedback from Interplex was obtained and is summarized in the Work Accomplished section of this report.  Although the animal model remains a question and many specifications do not have precise metrics or tests defined, the specification draft was needed immediately by the sponsor and included enough information to enable the vendor to provide a budgetary estimate to ReviveFlow for inclusion in their investor presentation.

All proposed deliverables are complete and were submitted according to plan, except phase one, which was delivered one business day after plan, although it included significant work that was delivered well ahead of schedule.  Two additional tasks were added to the original plan.  The first was to gather information about animal models to help ReviveFlow decide which model is best for their purposes.  The second was to further pursue a mathematical model that could actually model reverse cerebral blood flow.  I met Professor Scott Stevens in his office at Champlain College and we modified his model using Mathematica until we were able to fully switch arterial and venous flows, which is summarized in the Work Accomplished section of this report. 

Change Summary

Changes to scope and timeline have occurred and each was well-justified and received approval from the sponsor or Thayer School program director, wherever appropriate.  The first change to timeline occurred with the opportunity to move up the sponsor medical board review of the specification.  The impact of this change was that the first progress report was submitted one business day later than planned.  This change was approved in advance via email by Professor Graves and ultimately pushed the project ahead schedule.

Changes since the first progress report were additions to scope but did not impact the timeline.  The sponsor found a venture capital firm that expressed interest in funding the experiment for which this project is laying the groundwork.  This created additional tasks such as identifying relevant factors for selecting an animal model.  Thus, I accepted an increase in scope to help the sponsor answer this question by reviewing journal articles, contacting researchers at Dartmouth and other institutions, and summarizing the results of that information for the sponsor.

The original project scope included the implementation of a simple model, which was accomplished with a model by Ursino.  However, I wanted to see if an existing whole body mathematical model developed at the University of Vermont could predict what would happen in a situation of retrograde cerebral blood flow.  These last two tasks were agreed to because they were of great value to the sponsor, they were personally interesting to me, and the project was ahead of schedule.

Timeline

            The progress completed for each phase is depicted in the chart below, which was updated for this progress report.  The red line indicates current date.  Blue bars indicate work completed and grey bars represent work planned. 

Figure  SEQ Figure \* ARABIC 7: Timeline


 

 

Work Accomplished

The goal of this project is to develop a plan that will enable a vendor to build a proof of concept implementation, or alpha prototype, of the ReviveFlow system for a large animal test.  In particular, this includes creating a basic system model, a failure modes and effects analysis ,a comprehensive system specification, a list of potential system components, and a Request for Quotation (RFQ) for potential vendors.

System Model

After identifying salient parameters by learning the relevant anatomy and physiology (summarized in the Clinical Background section of this paper), various existing system models were evaluated for their relevance and simplicity.  Existing models were chosen as sufficient time was not available to develop and implement a model from scratch nor would it have been as valuable to the sponsor to create a simple model from scratch when well characterized and tested models already exist.

The mathematical model by Ursino shown as an example in Appendix B of the original proposal and referenced in the first progress report[13] was evaluated as too complicated to implement for the purposes of this project.  Another paper Ursino co-authored[14] indicated that perhaps another model was available for educational purposes, so an email was sent to Mauro Ursino in attempt to obtain an implemented copy of the model.  Professor Ursino replied with a generous offer to send the model to me, however it was written using software that was not available for my use. 

Additional research identified many attempts to mathematically describe the complex non-linear relationships of cerebral hemodynamics, which are much too numerous to list here.  For example, another paper by Ursino described the different physiological factors and their relationships[15], which enabled derivation of the equations which described his other models[16],[17] and taught me the physiological background behind the equations.  Typical of most studies was the utilization of lumped parameters, such as in the work by Olufsen et al (2001)[18].  However Lakin et al (2003) NOTEREF _Ref80621638 \f \h 57 point out that the assumptions of lumped models limit their applicability and therefore developed a whole-body model, shown as example in  REF _Ref82397595 \h Appendix C.  It was for this reason that I contacted Lakin, the result of which is described later in this section.

Ursino Model Implementation

For this project, a simpler model also by Ursino NOTEREF _Ref206918866 \f \h 38 was implemented for this report and can also be seen in  REF _Ref82397628 \h Appendix C.  This model has been well characterized and has good fit between actual values from animal tests versus the predicted values from the model.  However, the assumptions underlying the model make it such that it will not predict physiological behavior accurately in a reversed flow situation.  This is primarily due to the fact that no venous compliance is included in the model, which is expected to be of great importance in the situation where the pressure gradient is reversed and the venous side receives higher pressure.  Cirovic et al (2003) provided some evidence for this concern when they described the importance and effects of venous compliance on intracranial pressure-volume relationships[19]

Text Box: Figure 8: Venous compliance and pressure
 

 

The model also assumes intracranial pressure is equal to the venous pressure, which may not be strictly true in the case of reversed flow.  That this model and others linked ICP and venous pressure (Pv) is an observation worth further investigation.  It indicates that intracranial pressure is closely tied to venous pressure and is likely to increase with the elevated venous pressure likely needed to create a retrograde pressure gradient.  Most likely is that while ICP and Pv will track each other proportionally, they will not be the same value, due to the venous compliance that will act as the proportionality constant between the two pressures as depicted in  REF _Ref208829927 \h Figure 8. 

 

Further, the one-way valves modeled for cerebral spinal fluid generation and absorption represent the choroid plexus (generation) and arachnoid villi (reabsorption).  These physical structures are also driven by pressure gradient and may thus cease generating CSF or continue generating CSF, but stop reabsorbing via the arachnoid villi.  In either case, Dr. Lollis thought this would not have significant negative impact acutely, but the model suggests that these factors could combine to create a dangerous chronic situation as cerebral edema could lead to herniation of the brain through the foramen magnum (see  REF _Ref206923758 \h Figure 9), which is the same risk that occurs during hydrocephalus.

Figure  SEQ Figure \* ARABIC 9: Cranial anatomy

While the model implemented in this report is not capable of predicting the specific circumstances of the ReviveFlow system, it has provided valuable insight into possible issues to monitor and has identified other resources that may help the team eliminate unnecessary or inconclusive experiments.  The experiment should be based on a thorough review of current literature, should be based on current scientific knowledge, and should have predictive value.  A result from a predictive theoretical model will add tremendous value to the planned experiments as actual measured results can be compared with predicted and discussed.  Further, a model can improve the company’s public image by eliminating unnecessary sacrifice of animals and can save costs by reducing the amount of investigation that needs to occur.  A model can also speed up the experimental process and reduce the overall cost while providing an additional layer of validation.  In addition to the clear business benefits of improved accuracy, reduced risk, reduced cost, and shortened time, if the model suggests the concept of reversed cerebral flow is feasible, it will also provide theoretical support for the efficacy of the proposed system, which can also help convince investors to support the effort.

The idea of modeling blood flows through the circle of willis discussed in the first progress report was abandoned in favor of the more complete models of flow outlined herein.  The broader model enables a discussion regarding possible outcomes, which provide better value for ReviveFlow than a simple fluid mechanics model of a partially occluded circle of willis.

Thus a mathematical model was identified, implemented, and has yielded valuable information for this project, which satisfied the requirements for this project at the time of the second progress report.  However, at that time of the second progress report, a theoretical prediction for the particular case of reversed flow had not been performed. 

Stevens Model Implementation

As mentioned earlier, I contacted Professor William Lakin at the University of Vermont in Burlington to request a meeting to investigate adapting his model.  The objective was to see if it could theoretically predict the physiological outcome of reverse cerebral blood flow.  He responded promptly that he had retired on September 1 and was not available.  However, Professor Lakin introduced me to Professor Scott Stevens, his research collaborator and professor of mathematics at Champlain College in Burlington.  I was able to meet with Professor Stevens in his office and unfortunately, the whole body model was written in an old version of software that crashed whenever it ran.  Professor Stevens had more recently adapted the complex whole-body model for a specific application of modeling the effects of microgravity on ICP in astronauts, which was built using Mathematica and was much simpler than the whole-body model with only seven intracranial compartments.[20] 

Several coding changes were required to get the model to work in reverse.  Mostly these involved eliminating the calculations relating to angle of incline and microgravity effects.  To further simplify the model we also limited the results to steady-state.

Results

Due to the time required to modify the model structure and definitions, we were only able to run one simulation with Professor Stevens’ model.  Thus we were only able to switch arterial and venous completely such that Pa = Pv1 = 92 mmHg and Pv = Pa1 = 5.4 mmHg.  These values were used because they were the values Professor Stevens had used in his model based on other literature.  The results were consistent with the discussion in the Clinical Background section of this paper; ICP rose to a clinically dangerous level of 73.8 mmHg, Pv1 was maintained at 89.6 mmHg, and the resulting Pa1 was 15.4 mmHg.  Capillary pressure (Pc) was a clinically unacceptable 62.8 mmHg (versus ~20 mmHg normal) and the pressure in the brain tissue was 73.7 mmHg (versus 8-15 mmHg). 

Discussion

If this was a patient, he/she would probably be unconscious, suffering from brain herniation through the papillaformen such that their eyes would be bulging out, and also brain herniation through formen magnum and would be paralyzed. 

The best I could think of for comparing the model’s predicted values were with Frazee’s studies.  It is important to note the significant difference that Frazee was only increasing the venous pressure and was not manipulating arterial pressure.  This fact notwithstanding, his results match the model rather well.

Frazee’s earliest study investigated cerebral retroperfusion pressures (Pv1) as high as 176 mmHg and he writes that “intracerebral pressures [ICP] of >70mmHg were regularly accompanied by a deterioration in the EEG.” NOTEREF _Ref80621266 \f \h 54  Frazee also noted an average retrograde arterial pressure (Pa1) of 95 mmHg, which was 9 mmHg higher than baseline mean arterial pressure (MAP or Pa).  This increase in arterial pressure by 9 mmHg is similar to what Steven’s model predicts, as the model showed an increase in Pa1 of 10 mmHg (15.4 – 5.4).   Frazee did not measure capillary pressure.  In this initial study he concluded that retroperfusion pressures (Pv1) should be less than 70 mmHg, which seems to indicate that perhaps Frazee had observed that ICP and Pv1 are in fact closely associated.  In later studies he concluded that 20 mmHg was the most efficacious retroperfusion value, which is likely in baboons to be slightly higher than capillary pressure, so that he would perfuse the capillary bed but not have much flow into the arterioles.  Since his catheter balloons were pulsed, the blood would exit the capillaries via normal venous outflow pathways.

Another important limitation of the model is that it does not include the effects of collateral venous outflow, although Andeweg states that the importance of this in humans has been overstated in the past. NOTEREF _Ref208817434 \f \h 34

Recommendation

With additional time modifying the model programming code, Professor Steven and I believe his model is capable of predicting initial values for the retrograde cerebral pressure gradient (Pa1 and Pv1).

 

Failure Modes and Effects Analysis

A draft failure analysis document by ReviveFlow was expanded and quantified using a failure mode and effects analysis (FMEA), which is presented in  REF _Ref82275461 \h Appendix H.  The FMEA is prefaced with Severity, Occurrence, and Detection metrics that were defined specifically for the ReviveFlow application.  Failure modes are grouped by functional groups as defined by the ReviveFlow team and metrics were given both engineering and clinical consideration.  The resulting risk priority number (RPN) is the result of multiplying the three scores from Severity, Occurrence, and Detection and provides a useful priority ranking in which to address the potential failure modes.

The FMEA is intended to establish a baseline, which meets current FDA and other international regulatory guidelines (such as CE Mark in Europe), for future device design exercises.  It is intended to be a living document with increasing utility over time as more is learned about the product from modeling, design tests, system tests, and actual experiments.  When each failure mode is addressed or better elucidated, the effects and actions can be updated as well as the scores, which can help to prioritize the next most critical failure mode to address.

The FMEA was provided to the sponsor in soft copy for this purpose at the time of the first progress report and was accepted as complete. 

Specifications

As reported in progress report 1, version one of the specification was developed and presented to the Medical Board of Advisors for ReviveFlow ahead of schedule.  Feedback from the medical board of advisors was incorporated into the specification and the second revision was presented in progress report one ahead of schedule.  Revision three of the specification thus only needed minor refinements and was distributed to Interplex Medical, the vendor selected by ReviveFlow.  The specification as submitted to the vendor, which is version controlled and includes a change log, is shown in  REF _Ref82272540 \h Appendix B.  Also included were concept drawings by Jim Bleck and Aaron Gjerde for the ReviveFlow system (see  REF _Ref82272540 \h Appendix B).  Interplex Medical responded with an initial estimate for budgeting purposes of $100,000 for an acute model that does not require sterilization and $250,000 for a sterile, chronic model.

Inputs for the specification were the original failure analysis from ReviveFlow, research of current literature, and particularly, the system model and physiology background.

One particular issue is that the specifications need to be tailored for the specific animal model chosen.  A major research effort was initiated since the first project report to determine which model to use for the experiment.  Initial surgical proof of concept experiments using anastomosis by Dr. Sameh Mesallum, the inventor of the ReviveFlow system, were performed on dogs.  However a porcine model was also suggested for consideration by InterPlex Medical.  When I initial reported Dr. Frazee’s work with retrograde transvenous neuroperfusion in baboons NOTEREF _Ref80621266 \f \h 54, NOTEREF _Ref80625930 \f \h 55, NOTEREF _Ref206926374 \f \h 56 to the team, another viable stroke animal model was brought to light.  An excellent article by Dr. Judy Huang[21] at Columbia University generated much discussion in favor of baboons because of its excellent instructive detail and establishment of a well-characterized baseline condition.  Experience with stroke models in rats at Harvard added rats to the list of possible animal models.  Each model has important costs and benefits that need to be weighed, such as possible physical challenges, mechanical limitations, anatomical peculiarities, and ethical treatment concerns.  Although not exhaustive, an overview of major decision factors are summarized in text below and in matrix form in  REF _Ref82507039 \h Appendix E.

I met with P. Jack Hoopes of Dartmouth Medical School to obtain additional information about this issue.  Professor Hoopes suggested that while rats were perhaps best characterized and had highly homogeneous genetic composition, it may be physically challenging to place the catheters and a larger animal such as a dog might mitigate these challenges.  However, cerebral vasculature in dogs varies from humans, so the experiment would be focused on proof of principle and extrapolating results to relevance in humans will be limited.  Professor Hoopes also suggested the possibility that the experiment could be performed at Dartmouth.  The ReviveFlow team was grateful for the offer and pending the outcome of the upcoming investor meeting, additional investigation work regarding study hypothesis, protocol, and location will be performed.

I also discussed the animal model questions with Dr. Scott Lollis, chief resident of the section of neurosurgery at Dartmouth-Hitchcock Medical Center.  Dr. Lollis said from previous experience with mice that a silk suture was used to occlude an artery one branch beyond the internal carotid, suggesting that mice in particular, but probably also rats, would have cerebral vasculature that is too small to maneuver a triple lumen balloon catheter.  In a journal article on this topic, the suture diameter was only 0.15mm (5-0 silk suture)[22], which seems reasonable when compared to MCA diameter values found in other literature listed in  REF _Ref82508396 \h Table 3. 

A literature search was preformed to collect the data summarized in  REF _Ref82508396 \h Table 3 which gives average values for some of the relevant arteries for animal stroke models.  Other relevant factors such as ethical concerns, arterial and venous anatomy, and cost are also summarized in a forced rank decision matrix to help ReviveFlow prioritize their decision.  The next step will be to optimize the specification for the chosen animal model.  Because an animal model decision has not been made as of the due date for this paper, a range of reasonable values found in literature for various small and large animals is provided in the specification, as was proposed in progress report two.

Baboons, and other non-human primates such as monkeys, have similar blood supply to other primates, including humans, but are costly, increase timelines, and require special facilities.  One difference is that humans tend to have two anterior cerebral arteries whereas non-human primates tend only to have one ACA.  Another difference described by Lake et al in their excellent cerebral anatomy comparison between primates is that non-human primates have “a ‘common inferior cerebellar artery’ which bifurcates to form the anterior inferior cerebellar and posterior cerebellar arteries.  In humans, these arteries arise separately from the basilar and vertebral arteries, respectively.”[23]  Further, baboons, vervet monkeys, and bushbabies have essentially the same venous sinus drainage as humans but they also have significant additional sinuses not found in man that gives them a far more extensive and complex drainage network than man.  These are described in excellent detail in the study mentioned above by Lake et al NOTEREF _Ref82508018 \f \h 23.

Dogs, like primates, have a complete circle of willis, although also like non-human primates, dog arterial supply has relevant differences from humans.  These differences are described in excellent detail in a journal article by Brenowitz and Yonas (1990)[24] and focus on a much larger role for the ACA in dogs than the MCA in humans, and a higher level of anastomses between perforating arteries.  Dogs, like non-human primates, also have a very extensive collateral circulation[25], which combined with ACA versus MCA dominance, have made experimental canine stroke models difficult to create.  Brenowitz and Yonas also characterized the anatomic variation they found near the circle of willis, which is included herein in  REF _Ref82510174 \h Figure 26.

 

Component Selection

A list of possible components discovered during research was compiled and is provided for reference in  REF _Ref82274805 \h Appendix D.  The sponsor was pleased that a strong prospect for a flow-switching valve by Bio-chem Fluidics that utilizes that same design sketched in concept drawings.  A company that specializes in dialysis pump parts and rebuilding services was also identified as a prospect for the blood pumps.  Many options for clinical or lab monitoring equipment were listed so that the experiment results could be recorded, evaluated, and communicated.  Of particular interest is a company called VasSol that makes magnetic resonance angiography software that assembles quantifiable, interactive 3D models to facilitate vessel identification and flow calculations.  The vendor selected to build the prototype, Interplex Medical, has extensive experience with catheters and thus no additional work was required of this project by the sponsor regarding catheters.  The component chart was complete at the time of the second progress report and no new additions have been made.

Request for Quotation

Request for quotation (RFQ) scope remains unchanged with vendor scope, budget, timeline, and specification for fabrication of the alpha prototype to use in the large animal test.  A draft document was created and distributed to the vendor, Interplex Medical.  The timeline and budget remain under discussion with the ReviveFlow team and potential investors and are included in the attached RFQ as a placeholder.  The initial plan was to review this draft with the ReviveFlow team and to distribute to potential vendors for the final report.  However, the meeting with the venture capital firm on September 16 increased the urgency of the specification and a less formal draft RFQ was required immediately and the numbers need only to be rough estimates as opposed to precisely defined for the purposes of the investor presentation.  This enabled efforts to be focused on gathering data for the animal model decision and to spend a little more time trying to find and modify another existing theoretical model for reverse cerebral blood flow.

Research Summary

Patent Research

Relevant patents were reviewed from ReviveFlow (US 2008/0177245 A1), Design Mentor (US 7,238,165 B2), Neuroperfusion (US 5,908,407), Alsius (US 6,386,202 B1), and CoAxia (US 6,736,790 B2 and US 2007/0198049 A1)  Thomas Kardos.  A brief description of each patent is provided in  REF _Ref82274393 \h Appendix F.  Although additional related patents from these companies and others exist, a thorough patent search was performed by the intellectual property counsel retained by ReviveFlow and further detailed patent searching was deemed duplicative and unnecessary for this project.

The ReviveFlow patent provided important background information about the project and helped establish understanding of what the extent of system functionality could be.  Many different flow switching scenarios to various body parts are discussed in the patent, with a preferred embodiment suggesting reversed flow of blood to the brain.

This work provided meaningful background information for the project.  The Frazee patents I discovered were meaningful to the sponsor as they had previously not been referenced in the ReviveFlow patent and the company’s intellectual property counsel was notified accordingly.  This aspect of the project was complete for the second progress report.

Industry Research

The patent research revealed several companies have investigated, or appear to be investigating, technologies and techniques similar to ReviveFlow.  Additional research was performed regarding these companies in order to identify specifications that were important to them and to learn from their progress and mistakes.  The goal for this project was to provide the most effective design specification and RFQ possible, however this effort can also serve as a foundation for ReviveFlow to analyze whether or not these companies may also be possible competitors or collaborators.  The companies of primary interest were Neuroperfusion/Alsius and CoAxia.

Neuroperfusion conducted extensive testing in the 1980s and 1990s of what they called “retrograde transvenous neuroperfusion” (RTN), which involved inserting a catheter into the femoral artery and pumping oxygenated blood into the venous sinuses via a pulsatile occluding balloon catheter.  Early experiments in baboons and a small human trial showed that it was possible to perfuse the ischemic area from the venous side and that infarction was reduced.  However, the procedure was abandoned due to system complexity, lack of meaningful improvement in the deployment time window, and the lure of a potentially more profitable application for intravascular catheter technology.  The new application was intravascular cooling, which resulted in the transformation of the company into Alsius, which launched its CoolGard product in 2004.  Other than a patent for a simplified system filed by the Principal Investigator of the clinical study, no additional work appears to have been performed in this arena by any of the original team members.  See  REF _Ref82273713 \h Appendix G for a detailed summary and timeline.

CoAxia is a more recent company and is currently testing a catheter-based retrograde flow system called NeuroFlo that has similarities to both ReviveFlow and Neuroperfusion.  Like Neuroperfusion, NeuroFlo draws blood from the femoral artery and supplies the blood via the venous system to the cranium.  Like ReviveFlow, the NeuroFlo system does not get inserted deep into intracranial vessels and is instead placed in the descending aorta.  The primary difference between the ReviveFlow proposed system and NeuroFlo is that ReviveFlow intends to provide a return on the arterial side for non-oxygenated blood that has passed the capillaries.  CoAxia is well funded and is led by a team of industry veterans.  The current activities involve extensive national and international clinical trials to determine if the system is effective when deployed in a time window as long as twenty-four hours after stroke onset.  A more detailed summary is included in  REF _Ref82274188 \h Appendix G.


 

 

 

 

Clinical Background

In order to solve the problem of stroke, it is important to understand the clinical basics of cerebral anatomy and neurophysiology.  Given the enormous complexity of these topics, only an abbreviated background is given herein with the goal of explaining the relevant human systems with sufficient depth as to isolate variables and define parameters suitable for lumping, such that the reader will be able to understand the mathematical model presented in the next section.

Anatomy

Cerebral vasculature, whether in humans or animals, is highly detailed and complex and varies within and between species.  Fluid flows and tissue behavior are often non-linear and measurement scales range from nanometers to centimeters.  Upon visual inspection of the corrosion cast in  REF _Ref208855163 \h Figure 10, it is possible to surmise that many parallel pathways exist.  The question becomes, what parameters are relevant in the case of stroke, and particularly in the case of reversed cerebral blood flow?

Figure  SEQ Figure \* ARABIC 10: Corrosion cast of cerebral vasculature[26]

To answer this question, the revelant cerebral anatomy is presented followed by a summary of salient neurophysiology.  Emphasis is placed on characterizing the parameters and their relationships using mathematical equations, which help establish understanding of the system interactions and are the basis for the mathematical models described in this paper.

Arterial Supply

Blood flows from the heart into the aorta and up toward the head via the common carotid arteries as shown in  REF _Ref82358314 \h Figure 41.  The common carotid bifurcates into the external carotid artery, which goes toward the face extracranially, and the internal carotid, which supplies the deeper brain tissues intracranially.  The other main supply pathway is via the vertebral arteries, which converge to become the basilar artery.  Together, the internal carotid arteries (ICA) and basilar artery comprise the main supply of oxygenated blood to the brain.

These supply arteries converge again at the base of the brain in what is called the circle of willis, where approximately 80% of blood flow to the brain is distributed.  The circle of willis has three primary outflow pathways via the anterior cerebral arteries (ACA), the middle cerebral arteries (MCA), and the posterior cerebral arteries (PCA) (see  REF _Ref82498217 \h Figure 42).  Anastomoses are parallel pathways for blood and two of the five intracranial arterial anastomoses are found in the circle of willis.  These are the anterior communicating artery that links the bilateral anterior cerebral arteries and the bilateral posterior communicating arteries that link the posterior cerebral artery to the internal carotid artery on each side of the brain.  The third relevant anastomosis to the ReviveFlow system is the set of three communicating arteries between the internal and external carotid arteries on each side of the brain.  The remaining two cerebral arterial anastomoses relate to the subclavian artery, which supply the vertebral arteries and upper extremities.  It is important to note that many small perforating arteries branch off of all the main arteries at any point and vary anatomically from person to person.

Arteries are thick-walled, rigid under positive pressure, and serve the high-pressure side of the vascular system. Arteries will vasoconstrict or vasodilate in order to maintain a constant wall shear stress of about 15 dynes cm-2,[27] which is proportional to the velocity gradient at the wall and the viscosity of blood, as given by  REF _Ref82399481 \h Equation 1[28].  According to Glogov et al (1988), this value does not seem to vary either between or within animal species.

Equation  SEQ Equation \* ARABIC 1: wall shear stress

The blood pressure inside the artery creates forces that distend the artery, which generates hoop stress in the circumferential direction.  Arterial hoop stress is also maintained at a relatively constant value of 105 Pa NOTEREF _Ref82400107 \f \h 27 and can be estimated using the Law of LaPlace where P is transmural pressure, R is vessel radius, and t represents wall thickness. NOTEREF _Ref82498036 \f \h 28 As with wall shear stress, this value does not vary significantly between or within animal species. NOTEREF _Ref82400107 \f \h 27

Equation  SEQ Equation \* ARABIC 2: Law of LaPlace

Figure  SEQ Figure \* ARABIC 11: Embolus vs. Thrombus[29]

Despite their strength and ability to withstand high positive pressures, arteries collapse easily in the absence of internal pressure to hold them open. Arterioles have the greatest amount of strength and account for most cerebral vasoconstriction and vasodilation regulation.[30] 

Most ischemic strokes are caused by blood clots, which can form when blood shears or stagnates.  At this point the clot is stationary and referred to as a thrombus.  Once the clot has dislodged and is traveling through the vascular system, it is referred to as an embolus.  These emboli are a common cause of ischemic strokes.  As a clot travels to the brain via the internal carotid arteries, it is intuitive to see that the most likely path to follow is into the middle cerebral artery (see  REF _Ref82498217 \h Figure 42), which is indeed the case, as approximately 80% of all ischemic strokes occur in the MCA, due to the MCA dominance (most blood flow travels via MCA) in humans.  The inner diameter of an artery shrinks as the clot travels into the arteriole and approaches the capillaries.  Depending on the size of the clot, the artery inner diameter, and the path the embolus takes, it is likely to become stuck prior to reaching the capillary bed at which point it causes an ischemic stroke, also referred to as a “brain attack.”

Capillary Exchange

In normal (non-stroke) conditions, after the blood leaves the arteries, it flows through the capillaries where small perforations allow oxygen, nutrients, and fluids to diffuse into the surrounding tissue while waste is picked up by the blood.  By convention, oxygenated blood in arteries is depicted in red and veins with non-oxygenated blood are depicted in blue.

Capillary diameters range from 4um to 80um and vary between species.  Red blood cell diameters tend to be slightly smaller than the capillaries through which they pass, bending as they travel.  The rate of blood flow into the capillary is actively managed by the vasoconstriction or vasodilation of the arteriole just prior to entering the capillary.[31]  Capillary length is typically 0.3 to 1 mm and thus blood only spends 1 to 3 seconds in the capillary. NOTEREF _Ref82500321 \f \h 33

Text Box: Figure 12: Vessel cross-sectional area
 

 

Although the aorta is the largest blood vessel in the system and capillaries are very small, the collective capillary cross-sectional surface area is 1,000 times greater than the aorta.   REF _Ref208824262 \h Figure 12 shows the relative areas of each section of the vascular system.

 

Venous Return

Veins are thin-walled and much more compliant than arteries, which gives veins their tremendous capacitive ability.  Indeed, veins hold the largest percentage of blood in the vascular system.  Despite this important role, the venous side of the vascular system is often an afterthought and largely ignored.  This is evidenced by many easily found drawings and a plethora of studies elucidating the arterial system with only a rare mention of veins.  As anecdotal evidence, the Atlas of Human Anatomy by Frank Netter[32]has eleven slides depicting the arterial system and only four slides of the cerebral venous system. Guyton and Hall’s Textbook of Physiology acknowledges that this has been the case but is changing:

“For years, the veins were considered to be nothing more than passageways for flow of blood to the heart, but it has become apparent that [veins] perform other special functions that are necessary for operation of the circulation.”[33]

Despite the lack of attention in the past, an excellent article on the anatomy of collateral venous outflow was written by Andeweg (1996) NOTEREF _Ref208817434 \f \h 34.  It is highly recommended for the ReviveFlow team.  Andeweg explains that in the latter half of this century, angiographic descriptions of venous anatomy led to “an unjustified belief that practically everywhere cerebral veins have sufficient anastomoses had evolved.”  He goes on to counter that belief by stating that “so-called intracerebral venous anastomses through the centrum semiovale towards the convexity are nonexistent or negligible.”[34]

In the situation of reversed flow, the venous side and its behavior become much more important.  Following is a summary of venous anatomy to help establish a basis for the physiology and system model discussions in this paper. 

Upon exiting the capillaries, blood enters the venules which gradually increase in size to become veins which are collected in the venous sinuses of the brain.  The main sinuses are depicted in  REF _Ref80621144 \h Figure 44 and are described below.  The superior sagittal sinus (SSS) collects blood along the midline of the cranium near the outer surface and empties into the transverse sinus at the posterior base of the skull.  Parallel to the superior sagittal sinus, but deeper inside the brain, is the inferior sagittal sinus which connects to the straight sinus.  The straight sinus also terminates in the transverse sinus near the SSS.  The SSS is connected by numerous redundant bridging veins to the basal vein of Rosenthal, which terminates in the transverse sinus on the posterior side and the cavernous sinuses on the anterior side.  Like nearly all other sinuses, the cavernous sinuses also connect to the transverse sinus via the petrosal sinuses.  Finally, after all sinuses have emptied into the transverse sinus, blood flows bilaterally through the sigmoid sinuses into the internal jugular veins in the neck and to the vena cava where the blood returns to the heart and lungs.

Some veins, in particular the lower extremities, have one-way valves that prevent backflow.  Much like the venous system itself, cerebral venous valves have received very little research attention.  However one study identified “vertebral venous valves at the junction of the vertebral vein and the brachiocephalic vein, which move synchronously with the internal jugular venous valves.”[35]  Although the internal jugular valves may affect catheter insertion, both of these valves are below where the balloon occlusions are planned, such they should not pose an issue to a situation of reversed cerebral blood flow.

Physiology

Many physiological parameter values are interdependent, and feedback loops are often non-linear, making intuition difficult to develop.  The cranium is generally considered isochoric (Kellie-Monroe doctrine) and rigid with only a few access apertures.  The primary constituents in the cranial cavity are the brain (80% of volume; 1400ml), blood (10%; 150ml), and cerebrospinal fluid (10%; 150ml).[36]  The primary parameters and regulation mechanisms that are impacted by a reversed cerebral blood flow are likely to the cerebral blood pressures and resulting cerebral blood flow (CBF), cerebrospinal fluid (CSF) pressure and flow, the nonlinear relationship between pressure and volume in the craniospinal cavity, the starling resistor mechanism for the cerebral veins, and the central nervous system ischemic response.

Cerebral Blood Flow (CBF)

A fluid, such as blood, will flow through a vessel if the pressure at the source end is higher than the pressure at the destination.  The velocity is determined by the magnitude of the difference between the two end pressures and the resistance of the vessel.  The direction of flow is determined by which end of the vessel has the higher pressure, as the fluid will flow from high to low pressure.  This pressure difference (ΔP) in antegrade (normal) flow is calculated as arterial pressure (Pa) minus venous pressure (Pv), where Pa> Pv and blood flows from arterial to venous systems. Pa varies depending upon the cardiac output cycle, and in most cases for the sake of this project, the mean arterial pressure (MAP) will be used for Pa.  Typical values for each vascular segment are provided for reference in  REF _Ref82426831 \h Figure 46, note that arterioles contribute the largest drop in pressure in the vascular system.

Equation  SEQ Equation \* ARABIC 3: Pressure difference

Figure  SEQ Figure \* ARABIC 13: Vessel as a Starling resistor

Discuss Starling resistor pressure gradient direction. 

The relationship between flow (Q) and pressure difference (ΔP) is analogous to Ohm’s law for electrical circuits (ΔV= I × R), where I=Q and V=P.

Equation  SEQ Equation \* ARABIC 4: Flow

Reversed pressure and flow will be notated using a superscript prime, such that the value for reversed arterial pressure will be denoted as Pa1, reversed venous pressure will be Pv1, and the resulting pressure difference will be ΔP1.  Reversed flow will be denoted as Q1.

Figure  SEQ Figure \* ARABIC 14: Notation for reversed flow

Blood flow (Q) is also inversely proportional to vessel resistance, which is the primary mechanism of flow regulation, particularly administered by the arterioles that constrict or dilate to increase or decrease resistance respectively.  A frequency distribution of human CBF values can be referenced in  REF _Ref82500183 \h Figure 45.  With regard to stroke, Rohl et al note that “studies indicate that the tissue viability threshold of irreversible damage (i.e., the infarction core) is a CBF less than 12 mL/100 g/minute, whereas the CBF of the penumbra is between 12 and 18 mL/100 g/minute.”[37]

Resistance is a function of the viscosity (η) of the fluid as well as the length and radius of the vessel through which the fluid flows.  It is important to note that radius has a very powerful effect on resistance (R) as resistance is inversely proportional the fourth power of radius (r) shown in the Poiseuille equation. 

Equation  SEQ Equation \* ARABIC 5: Poiseuille equation

This is of particular relevance to ReviveFlow because when a catheter is inserted in the internal carotid artery, for example, the double lumen will divide that blood flow path in half and due to catheter tube wall thickness, will reduce the effective radius by more than 50%.  The radius of the ICA is given in  REF _Ref82508396 \h Table 3 as 5cm, π and η are constant, and assuming the catheter is fully inserted in the artery such that length does not change, we can estimate the change in resistance. 

 QUOTE                 Normal internal carotid resistance

 QUOTE                   Resistance inside catheter

Percent change in resistance= QUOTE  

Thus we can see that a reduction in radius of the pathway for blood by 60% has a powerful increase of over 3800% on the resistance to flow.

Velocity (v) of flow is a function of radius as well, and is customarily described by the rate of volumetric displacement [ml/sec] (Q) divided by cross-sectional area (A) of the vessel where A=πr2.

Equation  SEQ Equation \* ARABIC 6: Velocity of Blood Flow

GuytonHallFig15-1.jpgVessel compliance is the measure of vessel capacitance and describes what volume (V) can be held at a given pressure (P).  Compliance is the slope of the volume-pressure graph for a particular vessel.  Thick-walled, mostly rigid arteries are less compliant than thin-walled veins.  A systemic vein is about 24 times more compliant than a corresponding artery because the vein is approximately eight times more distensible and holds about three times more volume (8 × 3 = 24). NOTEREF _Ref82500321 \f \h 33

Equation  SEQ Equation \* ARABIC 7: Vessel Compliance

Text Box: Figure 15: Artery vs Vein compliance 3308D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600380032003500300030003300320031000000 
 

 

 

 REF _Ref208820578 \h Figure 15 graphically compares venous and arterial compliance.  Upon inspection, it is easy to see how dramatically different they are by comparing the slope of each line, which is compliance (also referred to as capacitance).

Normal blood flow is laminar, especially in the microcirculation, although certain conditions, such as branch points, stenoses, and thrombi, can cause turbulence.  The presence of turbulence in blood flow can be predicted by calculating the Reynold’s number, which is dependent upon the density of blood (ρ), vessel diameter (d), blood flow velocity (v), and the viscosity of blood (η).  Due to the wide range of vessel geometries in humans, the Reynold’s number for blood can range from 1 in capillaries to 4000 in the aorta, although a typical systemic value is around 2000 whereby values less than 2000 indicate increasing likelihood of laminar flow and values over 2000 indicate increasing likelihood of turbulence. NOTEREF _Ref82414066 \f \h 30

Equation  SEQ Equation \* ARABIC 8: Reynold’s number definition

Turbulence is undesirable physiologically, because blood clots can be generated as a result of pooling and stagnation and plaques can also be dislodged, such as in the case of blood turbulently flowing past a calcified stenotic aortic valve.

Algebraic manipulation of  REF _Ref208824854 \h Equation 8 can put the calculation for Reynold’s number in terms on radius and length, as shown in  REF _Ref208825129 \h Equation 9.

Equation  SEQ Equation \* ARABIC 9: Reynold's number

Again, the effect of radius is very powerful; small increases in radius are magnified as large increases in the Reynold’s number.  Strictly considering the Reynold’s number and turbulence, this is good news for the ReviveFlow proposed system because the catheters will be smaller in diameter than the arteries in which they are inserted.  The trade-off is with resistance, as was demonstrated earlier. 

This is a very useful equation for ReviveFlow, as it can be used during catheter design to predict if turbulence might be a problem.  Building on the previous example, when placing a double-lumen catheter in the internal carotid in humans, we can calculate NR during normal flow and inside the catheter. 

 

Cerebrospinal Fluid (CSF)

Text Box: Figure 16: CSF pathways 3308D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600380032003500300030003300320031000000 
 

 

GuytonHallFig64-4.jpgCSF is created when nutrient-rich fluid diffuses from capillaries to the choroid plexus deep inside the brain into the third, fourth, and lateral ventricles. CSF circulates in the subarachnoid space inside the cranium to cushion the brain in the event of an impact and to distribute nutrients.  The subarachnoid space also extends through the foramen magnum and surrounds the spinal cord.  The diagram in  REF _Ref208827531 \h Figure 16 provides an overview of these flows.  CSF is generated at a rate of about 500 ml/day, NOTEREF _Ref82414066 \f \h 30 which is about three to four times the total volume of CSF at any time. NOTEREF _Ref82500321 \f \h 33 

 

CSF reabsorption takes place in the arachnoid villi of the cerebral dural sinuses, such as the superior sagittal sinus and others mentioned in the venous anatomy section of this paper. The generation and reabsorption process is driven by pressure gradient.  Capillary pressures of about 20 mmHg transport fluids to the subarachnoid space.  This space has a lower pressure than capillary pressure, and is usually equivalent to intracranial pressure (~15 mmHg).  CSF is easily reabsorbed via the arachnoid villi into the venous sinuses which have a much lower pressure of about 6 mmHg.  This is a unidirectional process as the arachnoid villi operate as one-way valves.

CSF circulation obstruction, CSF overproduction, or insufficient reabsorption can result in hydrocephalus, which is increased CSF in the ventricles.36  The effects of increased volume of CSF will be discussed in the next section about intracranial pressure and the cranial compliance curve.

Intracranial Pressure (ICP)

Intracranial pressure (ICP) is an important neurophysiological parameter that is directly related to mean arterial pressure (MAP) and cerebral perfusion pressure (CPP), which reflects the ability of the brain tissue to be perfused.  The three pressures are related by  REF _Ref82501096 \h Equation 10. 

Equation  SEQ Equation \* ARABIC 10

More specifically, ICP is the pressure inside the cranium and is typically in equilibrium with the pressure inside the brain, venous sinus pressure, and subarachnoid (CSF) pressure.

Recall that the skull is rigid, has a roughly fixed volume, and that it’s three constituents are blood, CSF and brain.  Both blood and CSF are roughly incompressible fluids, leaving the brain as the only compressible constituent.  Thus, if the volume of one of these constituents increases, in order to maintain ICP, others need to decrease in volume to maintain equilibrium.  This is known as the Kellie-Monroe doctrine and is commonly depicted by the so-called cranial compliance curve, which can be reviewed in  REF _Ref82422977 \h Figure 47.  It shows that a very small change in volume results is a large change in ICP.  Because the curve is customarily plotted as volume versus pressure, it is technically an elastance curve, which is the inverse of compliance.

For example, by following the compliance curve in  REF _Ref82422977 \h Figure 47, it can be seen that a change of only 20 ml (about 4 teaspoons) in volume results in an ICP increase of 50 mmHg above normal, which might be 15 mmHg, for a total ICP of 65 mmHg.  A normal value for MAP is 80mmHg, which in this case, would mean CPP would be 80-65=15 mmHg.  CPP normal values fall in the range of 50-150 mmHg NOTEREF _Ref82424311 \f \h 13, which means the patient in this example has grossly inadequate cerebral oxygen supply and their brain is quickly dying.

Beyond the effect on CPP, increased ICP due to increased intracranial volume (ICV) means that the three intracranial constituents will try to find a way out of the cranial vault, which occurs in the small apertures, or foramen, in the skull.  Increased ICP will collapse venous sinuses and the brain itself will herniate through the openings, typically behind the eyes and at the foramen magnum where the spinal cord exits the skull, causing a variety of undesirable clinical conditions.  REF _Ref82424645 \h Figure 50 shows that in one cardiac cycle, a normal change in intracranial volume is about 5 ml, and the compliance curve in  REF _Ref82422977 \h Figure 47 shows us that a change in volume of 5 ml is easily managed.

Autoregulation is the term given to physiological feedback loops.  An important autoregulation mechanism that governs cerebral blood flow is the partial pressure of carbon dioxide (PaCO2).   Summarize blood gas partial pressures, CPP and CBF relationships.Chart already added to appendix.

Most of the sensors for the cardiovascular autoregulation mechanisms depicted in  REF _Ref82504431 \h Figure 52 are located inside the peripheral vascular system, outside the brain.  However, one important autoregulation mechanism that is located in the brain and could effect the ReviveFlow system is the CNS ischemic response.  As can be seen by the CNS ischemic response gain in  REF _Ref82504431 \h Figure 52, the CNS ischemic response is a very strong response.  It doesn’t activate until a very low threshold is reached, which is why it is sometimes referred to as the “last stand” or “last ditch effort” of the brain to save itself. 

Although the exact trigger mechanism is not known for certain, one theory is that slowly flowing blood flows is too slow to remove carbon dioxide away from the vasomotor cortex of the lower brain stem.  This stimulates a powerful sympathetic response in the brain’s medulla that tells the heart to pump as hard as possible to increase systemic arterial pressure to its maximum. Guyton and Hall describe the CNS ischemic response as “one of the most powerful of all the activators of the sympathetic vasoconstrictor system.” NOTEREF _Ref82500321 \f \h 33  They go on to describe that the CNS ischemic response begins to be activated when cerebral arterial pressure drops below 60 mmHg and maximizes its response when Pa is around 15 to 20 mmHg.  Because the trigger for this mechanism is not fully understood and previous efforts by Frazee et al and CoAxia have not reversed flow or dropped pressures on the arterial side of the brain, it is possible that the fully reversed flow will trigger this response with Pa values that will likely be less than 20 mmHg.  Pharmakinetic treatments may be able to mitigate this response and should be further researched.


 

 

 

Appendix A

Stroke Information

Figure  SEQ Figure \* ARABIC 17: Stroke as 3rd leading cause of death

 

Figure  SEQ Figure \* ARABIC 18: Elkins et al: historical and projected US stroke deaths7


 

 

Appendix B

Request for Quotation

Scope

The scope of this RFQ is to develop a proof-of-principle alpha prototype for a system that will reverse the flow of blood to the brain.  The objective is to test the safety and feasibility of the system in an experiment performed in a large animal model.

The initial specification has been developed for the ReviveFlow team and is provided herein as a starting point for the system.  Upon commencement of the proposed project, the specification and design history record should be maintained by the vendor for the duration of the project and returned to ReviveFlow as deliverables when complete.

Bench testing should be conducted to validate the device performs according to specifications prior to animal experimentation.  The results of the bench test should be provided to ReviveFlow as a deliverable which will enable a go/no-go decision prior to commencing the animal experiment.

The vendor will be expected to provide a team member who is experienced with animal experiments, catheters, and pumps to provide prototype system support during the experiment, which may be performed at one of the following locations: Boston, Massachusetts; Hanover, New Hampshire; or Cairo, Egypt.  Other locations may be defined and vendor recommendations are welcome.

The prototype should be optimized for a specific animal model, as noted in the specification.  The current leading choices for a suitable animal model are porcine or canine.  Feline and rat are also under consideration.  Non-human primates offer certain advantages, but are not deemed feasible due to costs and care concerns at this stage of development.  Vendor recommendations for animal model are welcome.

Timeline

Please provide a prototype development timeline with relevant milestones.  The timeline should not exceed six months and should begin no later than December 1, 2008.  Please include any required lead-time for vendor engagement prior to start of work.

Budget

Initial budget estimates of $100,000 for an acute model and $250,000 for a chronic model have been discussed and a detailed estimate and resource plan should be submitted to ReviveFlow.


 

 

 

 

Concept Drawings

 

ReviveFlow concept drawing

 Catheter detail 

Canine vascular anatomy


 

 

Specification

Table  SEQ Table \* ARABIC 1: Specification Matrix

Source

Specification

Justification

Quantification

Test

Catheter

Pump

Device

Procedure

Other Eqpt

Alpha Prototype

Notes

JB

Optimized for animal model

To obtain meaningful test results

Animal model probably dog, need investigate porcine and rat for feasibility
Frazee, Huang: Baboon

n/a

X

X

X

X

X

Reqd

 

R_F_IIaiv1

Pressure sensors distal and proximal to balloon

Monitor pressure in vasculature on both sides of balloon occlusion to prevent vasculate damage

Pressure outputs:
Venous: 0-10mmHG
Arterial: 0-200mmHg

+/- 1mmHG sensitivity

X

 

X

 

 

Reqd

 

Sam

Maintain sterility

Patient safety, infection prevention, ensure experimental data not impacted by infection factors

Design spec: disposible, not autoclavable.

 

X

X

X

 

 

Reqd

 

R_P

Control of catheter placement

Need to direct oxygenated blood to ischemic side of occlusion to perfuse

Location: bilateral jugular bulbs & internal carotids.  Likely direct jugular approach for experiment.  Clinical implementation likely femoral approach.
Frazee 1989: bilateral sigmoid sinuses
Frazee 1997: superior sagittal sinus (SSS) & transverse sinus (TS)

Visual confirmation using fluoroscopic guidance

 

 

X

X

X

Reqd

Direct jugular & carotid vs femoral access.
Above venous valve & basal valve?

Sam

Flow to be switched gradually

To prevent sudden pressure or flow changes in brain

90 seconds - should be modifiable and to be validated during animal test

Count time while operating switch.

 

X

X

X

 

Reqd

Manual for alpha prototype

R_F_1biii

Power source to be used globally

USA development team and Egyptian animal test team

50-60Hz, 100-240V, outlet receptacle differences

Verify meets voltage and current standards

 

 

X

 

X

Reqd

 

R_F_Iai4

Minimal mixing of blood when switching flows

No continuous mixing; keep venous and arterial blood separated

Minimum possible; <<1%

?

 

 

X

 

 

Reqd

 

R_F_1dbiv

System can be run continuously

Sufficient time to reverse ischemia, remove or break up clot or enable new collateral flow, and weening process

1 week (7 days) max

Run machine for seven days continuously

 

 

X

 

 

Reqd

 

R_F_IVbii2

Monitor patient blood count

To watch for possible blood shearing or hemolyzation when blood is run through device.  Vendor to perform bench test of alpha prototype to validate system performance

Before baseline blood compared to after pumping through device for X hours.

CBC (Complete Blood Count): # RBC, # WBC, hemoglobin, hematocrit, mean corpuscular volume (MCV)

 

 

X

 

 

Reqd

Not during procedure.  Not specific device functionality; bench test to validate specs met.

R_F_IVciii3, R_F_VIIbiv

Minimize pooling and prevent of stagnation of blood

To prevent coagulation of blood and thrombosis.

Design spec

 

 

 

X

 

 

Reqd

Possible utilization of existing technology that has been shown to not pool or stagnate.

R_F_VcIII1a

Minimize flow resistance in catheter circuit

To prevent shearing while maintaining appropriate pressure/volume balance

Catheter length, ID and Critical Reynolds #
Frazee 1989: length = 100cm

Calculate reynolds number and compare to lit. values

 

 

X

 

 

Reqd

 

JB

Number of devices

Need meaningful experimental data

Estimate 10-50 catheter sets for animal experiements

Verify quantity received

 

 

X

 

 

Reqd

RFQ

R_F_VIaiii1

Place catheters using image guidance

Patient safety: to prevent access site complications, such as bleeding from ateriotomoy/venotomy

Must be visible in vivo under fluoroscopic guidance

Place device in tissue and observe under fluoroscopy

X

 

X

X

X

Reqd

 

R_F_Iaii4, R_F_Iiaiv3,4,5,6

Automatic control and monitoring of balloon inflation pressure with manual override.

Enable blood to flow in normal pathways. Control with override to prevent vessel damage due to overinflation

Define balloon inflation pressure range, sensitivity, safety factor.
Frazee 1989: 8sec infl, 2 sec defl. cycled

Monitor balloon pressure

X

 

X

X

 

Reqd

Manual control only for alpha

Sam

Blood flows must not be turbulent

To prevent emboli and/or thrombi

Below critical reynolds number systemwide.

Calculate reynolds number for each system

X

X

X

 

 

Reqd

 

R_F_Vid

Catheter introducer with multiple ports

Enable use of multiple intra-arterial catheters simultaneously, such as dialysis or autolysis or other.

4 ports

Insert 4 catheters simultaneously

X

 

X

 

 

Reqd

Use off the shelf sheath or introducer

AMG

Catheter design

desire off the shelf

OD, lumen qty, lumen size, flexibility, compressibility, visible under which imaging modalities, able to coat with drugs, sterile, durability

Vendor to propose

X

 

X

 

 

Reqd

 

R_P

Control of pressure

To manage perfusion via changing pressure/flow parameters

Deliver vasoactive medications

Verify ability to deliver vasoactive drugs

 

X

X

X

 

Reqd

Secondary control to pressure via pressure/volume curve

R_P, R_F_VIIciii

Control of flow rate (pulsed or continuous)

To manage perfusion via changing pressure/flow parameters and detection of system failure (no flow in hub, across catheters)

Maintain 5L/min
Min = 1.5L/min (-70%)
Max = 7L/min (+30%)

Transcranial Doppler Ultrasound at Middle Cerebral Artery (TCD @ MCA)

 

X

X

 

X

Reqd

Primary factor to control for Pressure/Volume relationship. Manual control.

AMG

Pulsatile pump

to mimic physiological heart beat

Use existing solutions; characterize like heart motion

 

 

X

X

 

 

Reqd

 Venous compliance may favor continuous pump

R_F_Iai1

Control of stroke volume and frequency (if pulsed flow)

To manage perfusion via changing pressure/flow parameters

Maintain 5L/min
Normal: 70cc x 70bpm
Possible: 35cc x 140bpm or
10cc x 500bpm or
20cc x 250bpm

Verify with tissue vitality monitor.

 

X

X

 

X

Reqd

via pressure/volume curve.  Manual control.

R_F_Iai2, R_F_IIIaiv4

Control of flow direction

To manage perfusion direction. To dislodge clots & pull into filters (filters in future scope)

Detect direction of flow (antegrade vs retrograde) real time

Verify with tissue vitality monitor.

 

X

X

 

X

Reqd

 

R_P

Control of each inflow pathway

To manage perfusion direction.

Pressure, flow rate & direction, frequency & stroke volume (if pulsed)

Verify with tissue vitality monitor.

 

X

X

 

 

Reqd

 

R_P

Control of each outflow pathway

To manage perfusion direction.

Pressure, flow rate & direction, frequency & stroke volume (if pulsed)

Verify with tissue vitality monitor.

 

X

X

 

 

Reqd

 

AMG

Pump design

To create and control pressure gradient. desire off the shelf

Four pumps proposed

Verify quantity of pumps deployed

 

X

X

 

 

Reqd

 

R_F_1biii

Automated (software controlled) weening mechanism/protocol with manual override

to promote collateralization by slowly decreasing blood flow to ischemic/infarcted area(s) without causing greater tissue infarction.

a. through flow/pressure measurements, hub can find the optimal flow/pressure relationship for maximizing blood flow past occlusion.
b. Flow/pressure should be automatically or manually adjusted to decrease amount of blood flow and promote collateral circulation formation
c. Hub software should have weaning protocols involved with switching arterial/venous direction so as to promote collateral formation in patient’s intrinsic artery to venous circulatory direction

 

 

 

X

X

 

NR

 

R_F_VcIII1b

Coordinate pumping function with intrinsic heart pumping

To minimize increased afterload on heart due to increased blood flow resistance

Continuous pumping, intermittent (synchronized with electrocardiogram or pressure changes in arterial circulation) ?

 

 

 

X

 

X

NR

NR for alpha prototype

R_F_VIIIa

MRI compatibility

For use during MRI imaging procedures

Non-ferrous materials.

Verify materials are non-ferrous.  Apply magnet and observe no attraction force.

 

 

X

 

X

NR

 

R_F_IIIaiv1, R_F_IIIaiv4

Filter blood, reversibly

To remove clots

Filter metrics: particle size, selectivity?, volume, flow rate, other?  Where located in system?  Reversible trapping?

 

 

 

X

 

 

NR

Research Gore technology from Parodi procedure

R_F_Iai3

Able to cycle flow direction

To maximize ischemic tissue perfusion and enable weaning procedure.

Detect flow direction and change frequency

TCD

 

 

X

 

 

NR

 

R_F_1bIII

Backup power source

Ensure patient safety and machine reliability during power outages

 

 

 

 

X

 

 

NR

 

R_F_1biii

System must be reliable

Patient safety and system effectiveness

best effort for alpha prototypes.  MTBF for production later on.

Software and system testing

 

 

X

 

 

NR

 

R_F_IIIaiv2

Intravascular embolization protection devices proximal and distal to balloons

Catch dangerous clots if they exist and become dislodged by catheter

 

 

 

 

X

 

 

NR

 

R_F_VIIaii1

Air filters within hub at site just before blood exits to intravascular circulation

Catch air within circuit and prevent intravascular air embolization

 

 

 

 

X

 

 

NR

 

R_F_VIIciii4

Must be able to stop flow in catheters

Patient safety in the event of hub failure to prevent blood loss from patient into device

Clamps on compressible catheters?  Valves?

 

 

 

X

 

 

NR

 

R_P

Control of chemical/
pharmacological composition

 

 

 

 

 

X

 

 

NR

 

R_P

Catheters directly coupled or via flow control apparatus

 

 

 

 

 

X

 

 

NR

 

JB

Oxygenate blood

Need means to control oxygen level in arterial blood to be delivered to brain

AG: arterial blood already oxygenated?

 

 

 

X

 

X

N2H

Nice to have; NR

AMG

Balloon dimensions

Must fit thru introducer and vessels, and inflate to controllably occlude specified cerebral veins/arteries

Frazee 1989: 10 mm wide x 15 mm long
Research Gore calculations for size

Vendor to propose

X

 

X

 

 

Reqd

 

Sam

Able to trace path of blood flow

Experimental endpoint: show where blood went when flow is reversed

Record tracings in pump and at catheter tips.

Radioactive marker in blood?  Sequential CT-scans?

 

 

?

X

X

Reqd

i.e. Inject at distal tip of IJ and record at pump.

R_F_IIaiv2

Flow rate and direction sensors distal and proximal to balloon

Monitor flow in vasculature on both sides of balloon occlusion to prevent vasculate damage

Direction and rate outputs

Maybe TCD or sensor located in catheter

 

 

?

 

?

Reqd

 

R_F_VIaiii2

Intraoperative arteriotomy closure while removing arterial sheaths

Patient safety: to prevent access site complications, such as bleeding from ateriotomy/venotomy

Perclose device to be used

 

 

 

 

X

X

Reqd

Perclose device to be used.

AMG

Anesthetic

Patient comfort: sedate patient, prevent movement, pain control

Define drugs, concentrations, volumes, boluses, etc.
Frazee 1989: 10 mg/kg im ketamine, 0.4 mg im atropine, 5-10 mg/kg thiopental sodium, 0.5-1.5% halothane inhalation, 0.1 mg/kg q 2h pancuronium bromide, 2000 u q 2h heparin

 

 

 

 

X

X

Reqd

RFQ

AMG

Method for inducing stroke in animal model

Need to create ischemic condition to test reviveflow system

Probably use aneurysm clips if dog or porcine. 
Frazee 1989: Surgical snare occlusion & catheter embolization with 2x4mm gel-foam sponge
Frazee 1997: Aneurysm clips

 

 

 

 

X

X

Reqd

RFQ

AMG

Location for induced stroke in animal model

Need to create ischemic condition to test reviveflow system

Most common stroke is MCAO.
Frazee 1989: Occlude MCA (middle cerebral artery)
Frazee 1997: Occlude right ICA (internal carotid artery) & A-1 segment of both ACA (anterior cerebral arteries)

 

 

 

 

X

X

Reqd

RFQ

AMG

Ability to observe ischemia severity

To verify occlusion is properly placed and procedure/device effectiveness

Frazee 1989: EEG, 28 channels, 50% chg in amplitude of any freq band vs baseline considered ischemic.
Frazee 1997: SSEP (somatosensory evoked potential) monitoring, see study for definition of ischemic

CT and/or EEG (not MRI)

 

 

 

X

X

Reqd

Clinical endpoint

R_F_1aii5

Able to measure flow/pressure in cerebral circulation

Patient management to ensure perfusion and safety.  And collect device data to compare with standard clinical patient management.

MAP via BP cuff or A-line, ICP via ventriculostomy or ICP via CVP vs MAP estimate, CBFV via TCD at MCA

Measure and compare

 

 

 

X

 

Reqd

 

R_F_Vibi, R_F_VIIbiv

Deliver systemic anti-coagulation drug(s)

Patient safety: to prevent thrombosis/embolization caused by large bore catheter

Per standard operating room procedures; not in device scope for alpha prototype.

 

 

 

 

X

 

Reqd

 

JB

Access points for artery & vein

Best practice for arterial and venous access

Femoral artery and vein
Frazee 1989: supply from femoral artery and catheters introduced via bilateral femoral veins & direct jugular

Probably prefer femoral

 

 

 

X

 

Reqd

RFQ

R_F_Vcii2

Monitor patient pulse oximeter during procedure

To watch for heart failure or pulmonary edema

pulse ox output; clinically defined values

Pulse oximeter for specific animal model

 

 

 

X

X

Reqd

 

R_P, R_F_IVaiv1

Control blood temperature

Hypothermia to prevent fever, lower brain metabolism, and minimize infarction

Provide heating or cooling of blood within hub OR add ice to cool as needed.

 

 

 

 

X

X

NR

Standard OR cooling procedures and patient monitoring.

R_F_1aii6, R_F_1biv3

Automated identification of optimal patient parameters

Real-time, automated patient management to achieve optimal perfusion of ischemic tissue

Flow/pressure measurements integrated w/ software to find optimal conditions (stroke vol, freq, vaso meds)

 

 

 

 

X

X

NR

Manual monitor and drug delivery only for alpha

R_F_IVciii1, R_F_VIIbiv

Coat device linings with heparin

To prevent coagulation of blood and thrombosis within hub.

 

 

 

 

 

X

X

NR

Manual delivery of heparin for alpha prototype, coating NR.

R_F_VIc

Monitor perfusion of limb distal to catheter while catheter is in place

Prevent ischemia of limb distal to arteriotomy site due to percutaneous large bore catheter

Pulse volume recordings, remove catheter if sign of ischemia

 

 

 

 

X

X

NR

 

R_P, R_F_Iai5, R_F_IIIaiv3

Drug delivery access port for each outflow

Deliver vasodilators or vasopressors thru device for optimal ischemic tissue perfusion and to manage bradycardia/hypotension caused by carotid body manipulation.  Or anticoagulant delivery.

Per standard operating room procedures; not in device scope for alpha prototype.

 

 

 

 

X

 

NR

 

R_F_VIIciii

Automatic balloon deflation if sensors detect no flow in hub or across catheters, with manual override to deflate

Patient safety in response to hub failure; normal cerebral circulation is automatically restored

 

 

 

 

 

X

 

NR

Manual deflation control

R_F_VIIciii

Alarms to sound at pressure and flow thresholds

Patient safety: notify operator of danger condition

Define pressure, flow rate (direction?) thresholds

Manual for alpha

 

 

 

 

 

NR

No alarms for alpha

R_F_1biii

ICU staff able to monitor system

ICU staff likely users of system in highly monitored environment with many critical alarms and requirements

 

Ease of use?  Documented procedures needed to know?  Intuitive controls?

 

 

 

 

 

NR

Not in scope for alpha

R_F_Vbiii2,3

Antibiotic coatings on catheters and hub surfaces

To prevent infection

 

 

 

 

 

 

 

NR

Future possinbility; NR for alpha prototype

R_F_1bII

Able to identify if/when arterial occlusion is removed

Determine when to remove catheters.

(notes) Removed by medication, mechanical disruption, autolysis

Angiogram or maybe ultrasound at MCA

 

 

 

X

X

?

May need for clinical endpoint

R_F_1bi

Able to identify if/when collateral circulation develops due to arterial occlusion

Determine when to remove catheters.

Karen Dudich to research literature on cerebral collateral flow development time

Angiogram, 64-slice CT scans?  MRI/MRA

 

 

 

?

?

?

 

Change Log

Version

Date

Author

Change Summary

1

7/30/2008

Aaron Gjerde

Initial specification based on failure analysis document, conversations with AG & JB, and review of patent

2

8/2/2008

Aaron Gjerde

Revision based on feedback from 7/31/2008 meeting with SM, KD, JB, RR, AG

3

8/10/2008

Aaron Gjerde

Minor revisions and updates, mostly to fill in the "Test" column before sending to Interplex.

3

8/14/2008

Aaron Gjerde

Added pulsatile pump specification as required for prototype

3

8/19/2008

Aaron Gjerde

Animal models noted, clarified catheter location, approach. Added CBC test details.

3

9/9/2008

Aaron Gjerde

Added note to pulsatile pump, corrected spelling and typos.

 

 


 

 

Appendix C

Mathematical Model

Stella Model

 REF _Ref206917478 \h Figure 22 is an implementation of the model described by Ursino et al (1997) in the article titled “A simple mathematical model of the interaction between intracranial pressure and cerebral hemodynamics.”[38]

Figure  SEQ Figure \* ARABIC 22: Stella model

Equations

Below are the equations and variable definitions as implemented in Stella.  Units are indicated inside curly brackets {}.  Initial values are taken from Ursino et al (1997).

Ca(t) = Ca(t - dt) + (dCa\dt) * dt

INIT Ca = 0.15 {ml/mmHg}

 

INFLOWS:

dCa\dt = 1/Tau*(-Ca+sigmaGx)

Pic(t) = Pic(t - dt) + (dPic\dt) * dt

INIT Pic = 9.5 {mmHg}

 

INFLOWS:

dPic\dt = (ke*Pic)/(1+Ca*ke*Pic)*((Ca*dPa\dt+dCa\dt*(Pa-Pic)+(Pc-Pic)/Rf-(Pic-Pvs)/R0))

Can = 0.15 {ml/mmHg}

deltaCa = IF(x<0) THEN(deltaCa1) ELSE(deltaCa2)

deltaCa1 = 0.75 {ml/mmHg}

deltaCa2 = 0.075 {ml/mmHg}

dPa\dt = 0 {because not pulsatile pump}

G = 1.5 {ml*mmHg^-18100% CBF change ^-1}

ke = 0.11 {1/ml}

kr = 4.91*10^4 {mmHg^3 * sec / ml}

ksigma = deltaCa/4

Pa = 100 {mmHg}

Pc = (Pa*Rpv+Pic*Ra)/(Rpv+Ra) {mmHg}

Pvs = 6 {mmHg}

q = (Pa-Pc)/Ra {ml/sec}

qn = 12.5 {ml/sec}

R0 = 526.3 {mmHg*sec/ml}

Ra = kr*Ca^2/Va^2 {mmHg-sec/ml}

Rf = 2.38*10^3 {mmHg*sec/ml}

Rpv = 1.24 {mmHg*sec/ml}

sigmaGx = ((Can+deltaCa/2)+(Can-deltaCa/2)*EXP(G*x/ksigma))/(1+EXP(g*x/ksigma))

Tau = 20 {sec}

Va = Ca*(Pa-Pic) {ml}

x = (q-qn)/qn {normalized}

 
 

 

Ursino Simple Model

The following schematics are from Ursino et al (1997) and include both an electronic analog (A) and mechanical analog (B) of the implemented model.  The only factor not implemented was I_i, which is the rate of infusion of cerebral spinal fluid.  Its omission was justified because was used to compare the model to a clinical tool called the pressure volume index (PVI) and is not needed for this project.

Figure  SEQ Figure \* ARABIC 23: Ursino electrical (A) and mechanical (B) analogs

 

Lakin Whole Body Model

Figure  SEQ Figure \* ARABIC 24: Lakin et al (2003) whole body model NOTEREF _Ref80621638 \f \h 57

 

Stevens 7 Compartment Cranium Model

Stevens7.jpg

Figure  SEQ Figure \* ARABIC 25: Stevens model NOTEREF _Ref208853436 \f \h 20

 

Programming Code for this model consists of 12 “M-files” totaling over 700 lines of code and due to length is not included herein, but is available for inspection upon request.  I apologize for any inconvenience or lack of clarity this might cause.

 


 

 

Appendix D

Component Chart

Table  SEQ Table \* ARABIC 2: Component Chart

Image

Manufacturer

Product

Description

Weblink

 Notes

Moor

moorFLPI

full field blood perfusion imaging system

www.moor.co.uk

Excellent temporal resolution; up to 25 frames/sec

Moor

moorLDLS

Rapid Laser Doppler Blood Flow Imaging

www.moor.co.uk

50 x 64 pixels in 5 seconds (full hand every 5 seconds)

Moor

moorLDI2

laser Doppler blood perfusion imager

www.moor.co.uk

map tissue blood flow over areas from 5cm x 5cm up to 50cm x 50cm with 256 x 256 pixel resolution. Max resolution=100um

Moor

DRT4

Laser Doppler monitor

www.moor.co.uk

Standard clinical laser Doppler monitor, no imaging

Moor

moorLAB

Basic laser Doppler monitor

www.moor.co.uk

Basic lab laser Doppler monitor, no imaging

Transonic

PXL Flow Sensor

Ultrasonic flow probe for Blood or other fluids

www.transonic.com

ultrasonic transit time technology measures volume flow with external sensors designed for tubing

Transonic

Physiogear

Wireless flow telemetry

www.transonic.com

for conscious and untethered canine model

Transonic

Rodent model flow probes

Real time volume flow probes for rodents

www.transonic.com

Invasive, direct, real time measurement of flow.  Only cerebral application available is carotid artery.

Transonic

Laser Doppler Perfusion monitor

Tissue perfusion and flow monitoring

www.transonic.com

Real time assessment for acute and chronic applications.

Discovery Technology International

Laser Doppler perfusion monitor

Single or multi-channel laser Doppler tissue blood perfusion monitor

www.discovtech.com, www.oxford.optronics.com

Real time assessment at up to 4 sites. Can combine with temp and pO2 sensors.

Perimed

Periflux System 5000 with rat or mouse probe

Continuous blood flow and perfusion monitor

www.perimed.se

Used to verify MCAO placement in rats & mice via perfusion values

Perimed

Periscan PIM 3

Laser Doppler imager

www.perimed.se

Open skull applications in animals or humans

Biopac Systems Inc.

Cerebral Blood Flow monitoring system

Laser Doppler perfusion monitor

www.biopac.com

Used to measure subcutaneous micro-vascular blood flow

ADInstruments

PowerLab systems and multiple transducers

Data acquisition for many physiological parameters including blood flow, pO2 perfusion, others

www.adinstruments.com

Laser Doppler blood flow meter, transit time ultrasonic flow meter, others.

VasSol

NOVA

MRI/MRA-based noninvasive measurement of volumetric blood flow rate in cerebral vasculature

www.vassolinc.com

interactive 3D images fully rotatable, 360° view of vasculature enables precise identification of each vessel for volumetric blood flow calculation

GE/Philips/Siemens

64-slice CT

Volume rendered vasculature as 3D model

www.gehealthcare.comwww.philips.comwww.siemens.com

Limited temporal resolution, 1 model every several seconds, could be time elapse sequence

Medtronic Perfusion Systems

Bio-pump plus

Centrifugal blood pump

www.medtronic.com

Designed specifically for pumping blood; low shear, minimal stagnation, etc.

Instech

Harvard Pulsatile Blood Pumps

Simulated ventricular heart pumping action for rat and canine models

www.instechlabs.com

For isolated organ perfusion, whole body perfusion and blood transfer

World Heart

NovaCor LVAS

Component for Left Ventricle Assist System

www.worldheart.com

Component in a system, not available directly

Abiomed

AB5000 Ventricle

Blood Pump component in ventricle assist device LVAD

www.abiomed.com

Component in a system, not available directly

Ventracor

Ventrassist LVAD

Blood Pump component in ventricle assist device

www.ventracor.com

Component in a system, not available directly

Thoratec

Various pump components for ventricular assist devices

Blood Pump component in ventricle assist device

www.thoratec.com

Component in a system, not available directly

Berlin Heart

INCOR pump

Impeller driven archimedian screw-type blood pump for LVAD

www.berlinheart.de

Component in a system, not available directly

Dialysis Parts and Supplies

Various replacement pump assemblies, switches, and manifolds

Dialysis machine replacement parts including pumps, switches and manifolds

www.dpsitech.com

Compare specifications

Flo-Tec Inc.

Sells dialysis pumps and spare parts

Also provide pump rebuilding services

www.dialysispumps.com

Inquire directly

Manifolds_large

Boston Scientific

Vaxcel Plus Chronic Dialysis Manifold or MORSE Manifold

Large bore multiport manifold

www.bostonscientific.com

 

Qosina

Various stopcocks and manifolds

Search for stopcock yielded 79 results with varying specifications

www.qosina.com

Catheter parts, check valves, and other parts

Bio-chem Fluidics

Omnifit P/N 1114 Manual Valves and fluid switches

Stackable, various configurations

www.biochemfluidics.com

Looks very promising; matches Bleck switch designs

Bio-chem Fluidics

Gradient and flow selection valves

Solenoid controlled switches

www.biochemfluidics.com

Also make pumps, but not for blood specifically

Hemodynamics AG

Transcranial Doppler Simulator software

Simulation software for learning techniques and limitations of TCD

www.hemodynamics.com

Basic version is a free download

Karolinksa Institutet

3D interactive website

Cerebral vasculature

http://3d-brain.ki.se/atlas/vascular_supply.html

Excellent neurophysiology resource

 

 

 


 

 

Appendix E

Animal Model Research Summary

Table  SEQ Table \* ARABIC 3: Animal Model Matrix

Animal Model

Rat

Cat

Canine

Porcine

Baboon

Human

Vascular Average Diameters (mm)

Internal Carotid Artery (ICA)

0.3[39]

 

1.8[40]

2.5[41]

4.0±0.5[42]

5.0±0.5[43]

Middle Cerebral Artery (MCA) proximal

0.239

0.4±0.15[44],[45]

1.0[46]

 

 

2.9±0.3[47],[48]

Anterior Cerebral Artery (ACA)

0.239

 

0.6[49]

1.0[50]

 

 

Cerebral Capillaries

 

 

 

 

0.007±0.0013[51]

0.01

 

Figure  SEQ Figure \* ARABIC 26: Anatomic Variation of Arterial Supply in Dogs NOTEREF _Ref82510142 \f \h 24

Although

 


 

 

Appendix F

Patent Summaries

ReviveFlow

Patent No.:         WO 2008/089264 A1, US 2008/0177245 A1

Filing Date:          January 16, 2008

Inventor(s):        Sameh Mesallum

Assignee:            ReviveFlow Corporation, Chemsford, MA

Abstract:              Device and method for switching flow direction of fluid to a body part.

Figure  SEQ Figure \* ARABIC 27: Reversed flow to body part "P"

 

Figure  SEQ Figure \* ARABIC 28: ReviveFlow diagram with switching circuit

 


 

 

 

Design Mentor

Patent No.:         US 7,238,165 B2

Filing Date:          February 21, 2003

Inventor(s):        Douglas Vincent, Matthew Murphy

Assignee:            Design Mentor, Pelham, NH

Abstract:              Pump that provides physiological pulsatile flow.

Figure  SEQ Figure \* ARABIC 29: Design Mentor invention illustration

Figure  SEQ Figure \* ARABIC 30: Design Mentor assembly diagram


 

 

Neuroperfusion Inc.

Patent No.:         US 5,908,407[52]

Filing Date:          July 25, 1997

Inventor(s):        John G Frazee, David Cornett, Scott Evans

Assignee:            Neuroperfusion, Inc., Irvine, CA

Abstract:              Catheter with two inflatable balloons and segment between balloons that provides pressurized blood to veins .

Figure  SEQ Figure \* ARABIC 31: Neuroperfusion catheter           

 

Figure  SEQ Figure \* ARABIC 32: Neuroperfusion catheter intracranial location


 

 

 

Alsius Inc.

Patent No.:         US 6,386,202 B1[53]

Filing Date:          May 25, 1999

Inventor(s):        John G Frazee

Assignee:            UCLA, Oakland, CA

Abstract:              Continuation in part of previous patent assigned to Neuroperfusion.  Catheter with inflatable balloons and segment between balloons that provides pressurized and cooled blood to cerebral veins allowing partial antegrade and retrograde flow.  Emphasis on cooling.

Figure  SEQ Figure \* ARABIC 33: Alsius coolgard patent illustration

CoAxia Inc.

Patent No.:         US 6,736,790 B2

Filing Date:          July 11, 2001

Inventor(s):        Denise Barbut, Russell Patterson

Assignee:            CoAxia, Maple Grove, MN

Abstract:              Cooled and recirculated oxygenated cerebral blood from arterial to venous vasculature for the purpose of cardiac arrest treatment or patient support during aortic surgical procedures.  Occlusions placed near aortic arch instead of intracranial.

Figure  SEQ Figure \* ARABIC 34: Catheters placed near aortic arch

Patent No.:         US 2007/0198049 A1

Filing Date:          April 17, 2007

Inventor(s):        Denise Barbut, Russell Patterson

Assignee:            CoAxia, Maple Grove, MN

Abstract:              Catheter for reversing flow in a vessel to prevent distal embolization during various vascular surgery procedures.

Figure  SEQ Figure \* ARABIC 35: Occluding catheter reverses blood flow


 

 

Appendix G

Company Summaries

Neuroperfusion/Alsius

Neuroperfusion Inc. was inspired by research by John G. Frazee MD[54][55] at UCLA that began in 1986 NOTEREF _Ref80625705 \f \h 58 and continued through the 1990s, although efforts appear to have tapered off after 2000.  The technique they developed was referred to as Retrograde Transvenous Neuroperfusion (RTN) and has been used in experiments in over 100 baboons and eight FDA-approved human trials.[56]  The first published study by Frazee et al (1989) NOTEREF _Ref80621266 \f \h 54 described a technique using a specially designed catheter that was inserted into the femoral artery to draw oxygenated blood.  The catheter and pump system was made by Retroperfusion Systems Inc and was originally designed for coronary retroperfusion.  An original system overview can be seen in  REF _Ref208843315 \h Figure 38.  For Frazee experiment, the arterial blood was pumped through another catheter routed via the jugular vein into the cranium.  The distal ends of these catheters were located in the bilateral sigmoid sinuses such that the two balloons at the tips were located on each side of the superior sagittal sinus, as shown in  REF _Ref208842898 \h Figure 36.

Text Box: Figure 36: Frazee et al (1989) experimental setup
 

 

These balloons received pulsatile inflation cycles of 8 seconds inflated and 2 seconds deflated.  The inflated balloons created an almost complete occlusion while oxygenated blood from the femoral artery was pumped retrograde at about 20 mmHg into the venous sinuses, which normally have an average venous pressure of 6 mmHg[57].  It is important to note that nothing was done to reduce the normally high systemic pressure of arterial blood, which is approximately 80 mmHg NOTEREF _Ref80621638 \f \h 57.  When the balloons deflated, the blood flowed antegrade via normal pathways.  It is postulated by Frazee, but unverified, that a possibly large percentage of blood flowed through collateral drainage while the balloons were inflated. NOTEREF _Ref80621266 \f \h 54  Upon inspection of the cerebral venous sinuses in baboons and many alternate pathways in  REF _Ref80621144 \h Figure 44, his postulate appears reasonable.  This is further validated by a study by Lake et al NOTEREF _Ref82508018 \f \h 23 that describes baboons as having an extraordinarily extensive and complex collateral venous outflow.

 

The second study by Frazee et al (1997) NOTEREF _Ref80625930 \f \h 55 on Retrograde Transvenous Neuroperfusion (RTN) was intended to refine the technique and to identify the limitations of the procedure.  Different techniques were used for creating arterial occlusions and different locations for the balloons were examined.  The objective for using different occlusion techniques and locations was to create ischemia with different levels of severity.  Although middle cerebral arterial occlusion (MCAO) is the most common manifestation of stroke in humans, more severe models such as right internal carotid artery (ICA) and bilateral anterior cerebral arteries (ACA) were also occluded.  Catheter locations investigated were a single catheter in the superior sagittal sinus (SSS) and bilateral catheters in the transverse sinus.  The results showed that the more severe stroke models can kill the baboon very quickly and were revised after the first attempts to a less severe model.  The bilateral catheter location in the transverse sinus was shown to be superior to the single SSS location as well as earlier bilateral sigmoid sinuses. NOTEREF _Ref80621266 \f \h 54  It was discussed that this location enabled better perfusion of both superficial and deep brain structures because blood could flow through both the straight and superior sagittal sinuses.

This second study also confirmed the importance of avoiding high intrasinus pressure, which was defined in the first study NOTEREF _Ref80621266 \f \h 54 as greater than 20 mmHg.  The average value given by Lakin et al (2002) for intracranial capillaries was 20 mmHg, which would indicate that a retrograde pressure gradient would begin to exist between the intracranial veins and sinuses when the intrasinus pressure exceeds 20 mmHg.  See  REF _Ref208843676 \h Figure 24 for the average pressure for each body compartment Lakin et al (2003) identified.

The third of four studies produced by Frazee reported excellent results with a mean infarction volume in the untreated group of 8.8+/-3.1% versus 0.3+/-0.2% in the group of baboons treated with RTN[58].  The study also reported that the blinded, independent neurological score for the treated group was 99.2 and the combined mean score for the untreated group was 66.4 (p<0.015) NOTEREF _Ref206926374 \f \h 56.

Preliminary results of the study were presented in 1989 at multiple conferences[59],[60] and the final results were presented on Thursday, February 6, 1997 at the 22nd International Joint Conference on Stroke and Cerebral Circulation where it received significant attention from the media.  Within days, news articles were published in various publications including the Seattle Times[61], New York Times[62], Los Angeles Daily News[63], and the Daily Bruin[64] at UCLA.  The concept became popular enough that an amateur fiction writer featured a stroke patient undergoing neuroperfusion treatment in her manuscript.[65]

A timeline depicting the major research, regulatory, financial, and commercial events at Neuroperfusion/Alsius is included in  REF _Ref207083437 \h Figure 37.

John G Frazee MD at UCLA did extensive testing on over 100 baboons and 8 human subject in the 1990s using a system designed by a company called Neuroperfusion, Inc., which appears to be a spin-off of Retroperfusion Systems Inc., which went out of business in 1990[66] but is mentioned in the SEC S-1 documents for Alsius[67].  The Neuroperfusion system created a reverse cerebral pressure gradient similar to the goal of the ReviveFlow system, but it differed significantly in that it created a reverse gradient only by applying a sufficiently high pressure on the venous side to overcome the capillary pressure, thus reversing the gradient and resulting flow to the capillary bed only.  ReviveFlow seeks to not only create a higher positive pressure on the venous side, but to also create a lower pressure on the arterial side, thus fully reversing CBF.  It has been proposed that Pa1 could be a negative pressure, but the likely outcome of a negative pressure (suction) applied to arteries would be arterial collapse.  Thus, in order to reverse flow, a pressure lower than that of the venous side will be needed to reverse the gradient, but the arterial pressure will still need to be positive.  It is theorized by the ReviveFlow team that this lack of reduced pressure on the arterial side and a fully reversed cerebral pressure gradient is why Frazee’s experiments were ultimately not completed. 

Neuroperfusion Inc. no longer exists as a company and was merged with another company to become Alsius, which is a maker of intravascular cooling devices.  In a recent article in the Orange County Register, another clue is given as to why Neuroperfusion did not pursue further development of the RTN device:

“a member of the company’s scientific advisory board proposed a more profitable use of the catheters, Worthen said. Stroke specialist Dr. Camilo Gomez of the University of Alabama suggested that, with a modest design change, the catheters could create medically-induced hypothermia.”[68]

In addition to a potentially more profitable application, Dr. Camilo Gomez gave another clue about the possible limitation of RTN when he commented in a study, “Preliminary results have suggested that this technique may be of temporizing value, allowing time for the implementation of more definitive flow restoring strategies.”[69]  Thus it appeared that RTN may not have been effective without adjunct therapies.

Yet another reason for not pursuing RTN development further was stated by Frazee himself in a meta-analysis study he co-authored regarding interventional stroke therapies with Leary et al:[70]

“retrograde perfusion of ischemic brain tissue was beneficial in preventing and reversing serious injury in the baboon stroke model. NOTEREF _Ref80625705 \f \h 58  Additionally, it appeared safe and feasible but technically challenging in a small clinical trial in human patients in the late 1990s. NOTEREF _Ref206926374 \f \h 56

The “technical complexity” weakness of RTN is corroborated implicitly by Thomas Kardos, who also worked with Frazeeand was a former employee of Retroperfusion Systems Inc.  The second Frazee patent[71] references a study led by Thomas Kardos as primary investigator that was completed in 1993[72] and was funded, at least in part, by an NIH grant[73].  Although it has not been possible to obtain this study to know conclusively what its results were, Thomas Kardos filed for a patent in 2000 for a “Simplified Cerebral Retroperfusion Apparatus and Method” and states in his prior art discussion:

“Based on prior art, retroperfusion of cerebral ischemia has been successfully performed on animals and on at least one clinical trial.  Prior equipment included a multiplicity of catheters and was sub-optimal due to reperfusion injury and the complexity of placing multiple catheters.”[74]

Another patent by Lundquist and Kardos[75] is also cited by the Kardos/Davidner patent in the prior art discussion which lends addition insight into the Frazee system shortcomings:

“the methods and devices described by Frazee et al and/or Lundquist et al to date have failed to reduce the potential reperfusion injury which occurs when prolonged ischemic tissue is suddenly perfused with fully oxygenated blood, causing an oxygen blast and subsequent oxygen free radical formation resulting in additional tissue necrosis beyond that caused during the ischemic period.” NOTEREF _Ref81736348 \f \h 74

The Davidner/Kardos patent gives one more insight in the limits of the time delay when the prior art discussion states:

“Thus, in the past, cerebral retroperfusion has been impractical due to the multiplicity of catheters required and/or insufficient effectiveness due to the short time window under which the technique must be applied when reperfusion injured is not mitigated.  NOTEREF _Ref81736348 \f \h 74

This reason of insufficient time window was also corroborated again by Frazee in an abstract of an oral presentation he gave at the 2002 Congress of Neurological Surgeons where he stated that after nine human patients in the Phase 1 Clinical Trial, “Our results indicate that neuroperfusion can be performed within a seven hour window and can reverse patients’ clinical deficits and reduce or eliminate their recovery period.”[76]

Frazee’s RTN technique appears to have effectively perfused the ischemic penumbra and extended the treatment time window, but was too complicated to deploy, still suffered from a short time window, and appeared to be less profitable than other opportunities at the time.


 

 

Neuroperfusion/Alsius Timeline

 

 

Retroperfusion Systems Inc. overview[77]

 


 

 

CoAxia

CoAxia has developed a cerebral perfusion augmentation therapy as well and is a current player in the race for interventional stroke treatment.  The CoAxia technique has similarities to both the work performed by Frazee and that proposed by ReviveFlow in that it seeks to perfuse the ischemic penumbra.  The company has closed three rounds of funding (see  REF _Ref208220394 \h Table 4) with Canaan Partners, Prism Ventureworks, Baird Venture Partners, Affinity Capital, and Johnson and Johnson Development Corporation.[78]

Table  SEQ Table \* ARABIC 4: CoAxia Funding Summary[79]

Series

Amount (millions)

Date

A

$17.6

2004

B

$11

Sept 2006

C

$11.5

April 2008

 

CoAxia was founded in 1999 by Denise Barbut MD who was the Director of the Stroke Research Program at Cornell University Medical College.  According to searches performed on the United States Patent Office web database on September 3, 2008, the company has obtained an impressive list of thirty-nine issued patents (see  REF _Ref208221184 \h Table 5) and has an additional twenty-seven published applications pending (see  REF _Ref208844731 \h Table 6).  Several of these appear to be continuations in part, however a detailed search has already been performed by the ReviveFlow intellectual property counsel and was not repeated for this paper.

The NeuroFlo Catheter is the device developed by CoAxia (shown in  REF _Ref208844757 \h Figure 39) which is similar to the Frazee catheter with its double-balloon and multiple lumen design as well as that it forces arterial blood retrograde into the cerebral venous system.  However, the NeuroFlo catheter is not pulsed like Frazee’s and it is placed in the aortic arch instead of intracranially like both Frazee and Reviveflow.  CoAxia does not appear to provide for a full reversal of flow in the same way that ReviveFlow proposes.  The device is inserted via the femoral artery into the aorta.  The two balloons are manually inflated and about 70% of the blood at the descending aorta is diverted from the lower extremities to the cerebral collaterals.  After forty-five minutes, the balloons are deflated and the device is removed.  The clinical outcome is that existing collateral flow pathways in the brain are dormant before and during the stroke and the NeuroFlo device “activates” these pathways.

Figure  SEQ Figure \* ARABIC 39: CoAxia's NeuroFlo Catheter

 

The first clinical studies in humans commenced in April 2002.  There are currently two national and international FDA-approved studies conducted at the locations shown in  REF _Ref208223107 \h Figure 40.  In September 2007, Andrew Weiss told Medtech Insight that 80% of patients treated showed blood flow improvements in the penumbra in imaging studies.[80]  The first study is a 300-patient randomized, multicenter study called SENTIS that tests effectiveness for use up to ten hours after stroke onset.  Asuming positive results from this study, the pre-market application (PMA) is expected to be submitted to the FDA in 2009 or 2010.  The second study, called Flo24, is testing efficacy for use up to 24 hours after stroke onset.  Additional studies for patients that fail tPA and/or Merci Retriever therapies are also underway. NOTEREF _Ref208224416 \f \h 80

Figure  SEQ Figure \* ARABIC 40: CoAxia Clinical Study Sites

 

Table  SEQ Table \* ARABIC 5: CoAxia Issued Patents

 

PAT. NO.

Title

1

7,374,561

Devices and methods for preventing distal embolization during interventional procedures

2

7,340,298

Enhancement of cerebral blood flow by electrical nerve stimulation

3

7,166,097

Cerebral perfusion augmentation

4

7,150,736

Cerebral perfusion augmentation

5

6,942,686

Regulation of cerebral blood flow by temperature change-induced vasodilatation

6

6,896,663

Retrograde venous perfusion with isolation of cerebral circulation

7

6,887,227

Devices and methods for preventing distal embolization from the vertebrobasilar artery using flow reversal

8

6,878,140

Methods for flow augmentation in patients with occlusive cerebrovascular disease

9

6,866,647

Aortic shunt with spinal perfusion and cooling device

10

6,848,448

Devices and methods for cerebral perfusion augmentation

11

6,840,949

Devices and methods for preventing distal embolization using flow reversal in arteries having collateral blood flow

12

6,837,881

Devices and methods for preventing distal embolization using flow reversal by partial occlusion of the brachiocephalic artery

13

6,830,579

Devices and methods for preventing distal embolization using flow reversal and perfusion augmentation within the cerebral vasculature

14

6,817,985

Intravascular spinal perfusion and cooling for use during aortic surgery

15

6,796,992

Cerebral perfusion augmentation

16

6,767,345

Partial aortic occlusion devices and methods for renal and coronary perfusion augmentation

17

6,758,832

Medical device for intrathecal cerebral cooling and methods of use

18

6,743,196

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

19

6,712,806

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

20

6,635,046

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

21

6,626,886

Devices and methods for preventing distal embolization during interventional procedures

22

6,623,471

Devices and methods for preventing distal embolization from the internal carotid artery using flow reversal by partial occlusion of the external carotid artery

23

6,595,980

Devices and methods for preventing distal embolization using flow reversal by occlusion of the brachiocephalic artery

24

6,595,963

Aortic shunt for selective cerebral perfusion in stroke and cardiac arrest

25

6,592,557

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

26

6,565,552

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

27

6,558,356

Medical device for flow augmentation in patients with occlusive cerebrovascular disease and methods of use

28

6,555,057

Intravascular methods and apparatus for isolation and selective cooling of the cerebral vasculature during surgical procedures

29

6,533,800

Devices and methods for preventing distal embolization using flow reversal in arteries having collateral blood flow

30

6,530,894

Aortic shunt with spinal perfusion and cooling device

31

6,383,172

Retrograde venous perfusion with isolation of cerebral circulation

32

6,379,331

Medical device for selective intrathecal spinal cooling in aortic surgery and spinal trauma

33

6,355,010

Intravascular spinal perfusion and cooling for use during aortic surgery

34

6,312,444

Medical device for removing thromboembolic material from cerebral arteries and methods of use

35

6,231,551

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

36

6,217,552

Medical device for selective intrathecal spinal cooling in aortic surgery and spinal trauma

37

6,165,199

Medical device for removing thromboembolic material from cerebral arteries and methods of use

38

6,161,547

Medical device for flow augmentation in patients with occlusive cerebrovascular disease and methods of use

39

6,146,370

Devices and methods for preventing distal embolization from the internal carotid artery using flow reversal by partial occlusion of the external carotid artery

 

Table  SEQ Table \* ARABIC 6: CoAxia Published Patent Applications

 

PAT. NO.

Title

1

20070239135

Cerebral perfusion augmentation

2

20070198049

Devices and methods for preventing distal embolization using flow reversal in arteries having collateral blood flow

3

20070135793

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

4

20070118095

Cerebral perfusion augmentation

5

20060035948

Benzopyran derivatives substituted with secondary amines including tetrazole, method for the preparation thereof and pharmaceutical compositions containing them

6

20050159640

Cerebral perfusion augmentation

7

20050149112

Devices and methods for preventing distal embolization using flow reversal in arteries having collateral blood flow

8

20050090854

Devices and methods for preventing distal embolization using flow reversal and perfusion augmentation within the cerebral vasculature

9

20050085685

Cerebral perfusion augmentation

10

20050070838

Intravascular spinal perfusion and cooling for use during aortic surgery

11

20040243058

Medical device for selective intrathecal spinal cooling in aortic surgery and spinal trauma

12

20040127885

Devices and methods for preventing distal embolization during interventional procedures

13

20040006299

Aortic shunt for selective cerebral perfusion in stroke and cardiac arrest

14

20030195382

Cerebral perfusion augmentation

15

20030171769

Devices and methods for preventing distal embolization using flow reversal in arteries having collateral blood flow

16

20030144624

Aortic shunt with spinal perfusion and cooling device

17

20030097036

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

18

20030083617

Partial aortic occlusion devices and methods for renal and coronary perfusion augmentation

19

20030023200

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

20

20020165573

Devices and methods for preventing distal embolization using flow reversal and perfusion augmentation within the cerebral vasculature

21

20020128586

Retrograde venous perfusion with isolation of cerebral circulation

22

20020115982

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

23

20020091356

Medical device for intrathecal cerebral cooling and methods of use

24

20020052620

Medical device for removing thromboembolic material from cerebral arteries and methods of use

25

20010041860

Partial aortic occlusion devices and methods for cerebral perfusion augmentation

26

20010038807

Method and system for selective or isolated integrate cerebral perfusion and cooling

27

20010020159

Medical device for selective intrathecal spinal cooling in aortic surgery and spinal trauma

 


 

 

Appendix H

Failure Modes and Effects Analysis

Table  SEQ Table \* ARABIC 7: Severity, Occurrence, Detection definitions

Rank

Severity

Occurrence

Detection

5

Death

Every procedure; 95%+ confidence

Uncertain; defect not detectable

4

Disability

Occurs often 75% or more

Remote; defect unlikely to be detected

3

Minor Complication

Occurs 50% of the time

Moderate; 50% chance of detection

2

Productivity

Occurs occasionally, 25% or less

High; likely to be detected

1

Convenience

Rarely; <5% of the time

Certain to be detected

 

Table  SEQ Table \* ARABIC 8: FMEA

Function

Potential Failure mode

Potential Effect(s) of Failure

Severity

Potential Causes of Failure

Occurrence

Current Controls

Detection

RPN

Recommended Action(s)

Brain

Brain not perfused through venous system

Ischemic infarction eventually resulting in tissue death

5

Venous escape from hi pressure created by arterial occlusion (balloon) and/or collateral drainage

3

CLINICAL SOLUTIONS
1. Alter pressure or flow based on pressure/volume relationship
2. Revert to orthodromic flow
3. Cycle flow direction
4. Deliver vasoactive meds
DEVICE
Balloon design, location, inflation pressure.
Stroke volume, frequency, flow direction.
OTHER
Remove occlusion
1. by medication
2. by mechanical disruption
3. by autolysis

3

45

Device Specifications to include ability to:
1. Change stroke volume, pump frequency (flow rate)
2. Change flow direction
3. Change balloon inflation pressure
4. Measure cerebral flow rate
5. Calculate cerebral perfusion pressure
6. Software to process measurements and suggest optimal conditions for:
a. Stroke volume
b. Frequency
c. Vasoactive med delivery

Vasculature

Embolization of intravascular plaque

Causing pulmonary embolism if flow is toward heart, or ischemic stroke if flow is toward brain

5

Existing plaque is dislodged or clot formed due to blood stagnation, shearing or turbulence

2

Filters, anticoagulation medications administered either systemically or intra-apparatus

3

30

Follow current controls

Vessel complications

Thrombotic surface(s)

5

Due to rupture, tear, dissection, barotrauma, stricture, tortuosity, or loss of intimal lining

2

Pressure & flow sensors distal and proximal to balloon at tip of catheters and inside balloon

2

20

Investigate computer controlled balloon inflation with manual override

Unable to cannulate vessels

Unable to deliver intended device therapy

5

Due to pre-existing stenosis

1

Device to be placed by experienced neurointerventionalist under fluoroscopic guidance

2

10

If complete occlusion, procedure is either aborted or circuit created without cannulation/balloon inflation within stenosed vessel. Or if partial occlusion, attempt balloon inflation proximal to stenosis and adjust pressure/flow after inflation

Blood

Blood coagulation

Clot formation

5

Stagnation, contact with thrombogenic surfaces

2

Administer anticoagulants, design spec minimizes stagnation

2

20

Investigate coating device linings with heparin

Blood temperature changes

Hypothermia, coagulopathy, patient discomfort.  Hyperthermia, fever, increased metabolism, worsen infarction

5

Heat transfer from pump motors, heating due to shearing or turbulence, existing health problems

3

Standard OR procedures for patient cooling/heating, such as ice or warming blankets or other devices.

1

15

Investigate intravascular temperature control technologies

Blood shearing or hemolyzation

Blood loss

5

Exceeding critical Reynolds number or surgical intervention

3

Design specification based on current literature. 

1

15

Blood count test during procedure.  Procedure preparations for blood loss and replacement.

Alteration of Hemodynamics

Bradycardia or hypotension

5

Carotid body manipulation

2

Administer vasoactive medications

1

10

 

Access Site

Infection

Chronic wounds, inflammation, death.

5

Percutaneous vascular access, non-sterile eqpt or procedure

1

Systemic antibiotic administration

3

15

Investigate antibiotic coated device linings

Bleeding

Hematoma, retroperitoneal bleed, from arteriotomy venotomy

4

Unintended shearing or tearing of vessel during catheterization

1

Standard procedures for catheterization. image guidance.

2

8

Possible arteriotomy closure depending on size of access.  Investigate feasibility of multiple port introducer

Balloon inflation increase resistance to blood flow

Decreased cardiac output, heart failure, death

5

Increased afterload on the heart

1

Evaluate patient cardiac status prior to procedure for heart failure or low ejection fraction

1

5

Monitor pulse oximeter intraoperatively for heart failure or pulmonary edema.  Investigate coordinating pump cycle with heart.

Hub

Catheter rupture

Unintended mixed blood, thrombosis, clots, arterial blood does not make it to ischemic tissue, pressure changes, flow changes

5

Material failure due to defect, incorrect specification, or does not meet spec, or is used for unintended purpose outside spec.

1

Design specification based on current literature.

2

10

Thorough prototype and system testing.

Balloon rupture

Unintended mixed blood, thrombosis, clots, arterial blood does not make it to ischemic tissue, pressure changes, flow changes

5

Material failure due to defect, incorrect specification, or does not meet spec, or is used for unintended purpose outside spec.

1

Design specification based on current literature.

2

10

Thorough prototype and system testing.

Control system failure

System freeze, no inputs read, no outputs sent.

5

Code bugs

1

Control system testing

2

10

Ensure complete and accurate control system testing protocol.

Sensor malfunction or failure

Incorrect input data read by physician or control system

4

Material failure due to defect, incorrect specification, or does not meet spec, or is used for unintended purpose outside spec.

1

Design specification based on current literature.

2

8

Thorough prototype and system testing.

Pump failure

System stops pumping; potential risk to patient safety

5

Power outage, failed wiring, motor failure, linkage failure, others design dependent

1

Design specification based on current literature. Thorough prototype and system testing.

1

5

Auto-deflation of balloons, alarms, catheter shut-off/clamp to prevent blood loss from patient to device

Electrical power source outage

System stops pumping; potential risk to patient safety

5

Power outage, failed wiring,

1

Used in hospitals with back up power.  No internal controls.

1

5

Use in hospitals with back up power.  Create control system that is fail-safe to power interruption

 

 


 

 

Appendix I

Cerebral Arteries

Figure  SEQ Figure \* ARABIC 41: Arteries to the Brain[81]

 

 

 

Figure  SEQ Figure \* ARABIC 42: Circle of Willis[82]

 

Figure  SEQ Figure \* ARABIC 43: Cerebral arteries[83]

 


 

 

Appendix J

Cerebral Venous Sinuses

Lab1fig6

Figure  SEQ Figure \* ARABIC 44a(above), b(below): Cerebral Venous Sinuses[84]


 

 

Appendix K

Neurophysiological Parameter Values

Cerebral Blood Flow and Pressure

Figure  SEQ Figure \* ARABIC 45: Frequency distribution of CBF values in 21 humans[85]

Figure  SEQ Figure \* ARABIC 46: Mean pressure in vasculature NOTEREF _Ref82414066 \f \h 30

Intracranial Compliance Curve

Figure  SEQ Figure \* ARABIC 47: Intracranial Compliance Curve[86]

Intracranial Pressure (ICP)

Below is a comparison between ICP frequency distributions using direct measurement via ventriculostomy and indirect measurement via magnetic resonance images.

Figure  SEQ Figure \* ARABIC 48: ICP frequency distribution NOTEREF _Ref82422803 \f \h 85

Intracranial Volume (ICV)

 REF _Ref82425945 \h Figure 49 shows the total cranial inflows and outflows during one cardiac cycle.  Arterial inflow is indicated by filled circles, venous outflow is indicated by open circles, and the flow of CSF between cranium and spinal sac is indicated by diamonds.

Figure  SEQ Figure \* ARABIC 49: Cranial Inflows and Outflows NOTEREF _Ref82422803 \f \h 85

The above total flows result in a net total change in volume during one cardiac cycle seen below.

Figure  SEQ Figure \* ARABIC 50: Change in ICV during one cardiac cycle NOTEREF _Ref82422803 \f \h 85

CBF and Pressure Relationships

PaO2 represents the partial pressure of oxygen in arterial blood.  PaCO2 represents the partial pressure of carbon dioxide in arterial blood.  CPP and ICP are cerebral perfusion pressure and intracranial pressure as described in the physiology overview of this paper.

Figure  SEQ Figure \* ARABIC 51: Important pressure relationships with CBF[87]

Cardiovascular feedback responses over time

Figure  SEQ Figure \* ARABIC 52: Feedback responses over time33


 

 

Appendix L

Updated Task List

Completed tasks are indicated by a checkmark.  ü

Initiate

ü  Identify prospective projects

ü  Select Project

ü  Initial Research

      Conference calls, in person meetings with sponsor

      Literature and patent review

      Business Plan and associated background documents

ü  Project Planning

      Scope, Deliverables, Timeline, Resources

ü  Confirm Faculty Advisor

ü  Deliverable: Proposal

Define

ü  Establish baseline

ü  Create FMEA

ü  Draft initial specification

ü  Research academic/lab prior art (moved up from Implement Phase)

ü  Research theoretical system model

ü  Review specification with Medical Advisory Board (Moved up from Implement Phase)

ü  Deliverables: FMEA, Initial Specification, and Progress Report 1

Implement

ü  Research System Requirements

ü  Research academic/lab prior art

ü  Research theoretical system model

ü  Develop basic system model

ü  Component manufacturer/supplier research

ü  Patent search

ü  Revise and expand specification

ü  Review specification with Medical Advisory Board(Moved up to Define Phase)

ü  Review specification with potential vendors

ü  Research suitable animal model for experiment

ü  Deliverables: Revised specification, component selection, Progress Report 2

Finalize

ü  Create Request for Quotation

ü  Distribute Request for Quotation to potential vendors

ü  Recommend suitable animal model for experiment

ü  Deliverables: Final specification, RFQ, research summary, and Final Report

 

 


 

 

Appendix M

Works Cited


 


[1] Graham, Sharon M, Louise D McCullough, and Stephanie J Murphy. "Animal Models of Ischemic Stroke: Balancing Experimental Aims and Animal Care." Comparative Medicine 54.5 (Oct. 2004): 486-496.

[2] http://www.brainomix.com/images/img_stroke.jpg

[3] http://www.kenthospital.org/healthGate/images/si55551195.jpg

[4] http://www.strokecenter.org/patients/stats.htm

[5] United States Centers for Disease Control and Prevention. http://www.cdc.gov/Stroke/stroke_facts.htm

[6] Kung, Hsiang-Ching, et al. "Deaths: Final Data for 2005." National Vital Statistics Report 56.10 (Apr. 2008): 1-121. US Department of Health and Human Services, Center for Disease Control and Prevention.

[7] Elkins JS, Johnston, SC. Thirty-year projections for deaths from ischemic stroke in the United States. Stroke. 2003; 34: 2109-2113.

[8] Brown DL. Boden-Albala B. Langa KM. et. al. Projected costs of ischemic stroke in the United States. Neurology Oct 2006; 67: 1390-1395.

[9] United States. Cong. Senate Subcommittee on Labor-HHS-Education Appropriations. Mind, Brain, and Behavior FY2008 Budget Request. By Story C Landis. Ed. Richard J Turman. Washington DC: National Institute of Neurological Disorders and Stroke, n.d.

[10] Breaking Our Hearts: Still America’s No.1 Killer: NIH Funding for Heart and Stroke Research. Washington DC: American Heart Association, 2008.

[11] Demaerschalk, Bart M, MD. "Thrombolytic Therapy for Acute Ischemic Stroke: The Likelihood of Being Helped Versus Harmed." Stroke 38.8 (Aug. 2007): 2215-2216. Editorial

[12] http://www.brainomix.com/images/img_treatments.jpg

[13] Ursino, M, et al. "Cerebral hemodynamics during arterial and CO2 pressure changes: in vivo prediction by a mathematical model." American Journal of Physiology-Heart and Circulatory Physiology 279 (2000): H2439-H2455.

[14] Giannessi, Massimo, and Mauro Ursino. "A comprehensive Cerebrovascular Simulation Model for Teaching and Research." Unpublished Abstract. Department of Electronics, Computer Science and Systems, University of Bologna, Italy. email contact information included.

[15] Ursino, Mauro. "Mechanisms of Cerebral Blood Flow Regulation." Critical Reviews in Biomedical Engineering 18.4 (1991): 255-288.

[16] Ursino, M, et al. "Cerebral autoregulation and gas exchange studied using a human cardiopulmonary model." American Journal of Physiology: Heart Circulation Physiology 286.2 (Feb. 2004): H584-H601.

[17] Ursino, M, and M Giulioni. "Quantitative assessment of cerebral autoregulation from transcranial Doppler pulsatility: a computer simulation study." Medical Engineering & Physics 25 (2003): 655-666.

[18] Olufsen, Mette S, Ali Nadim, and Lewis A Lipsitz. "Dynamics of cerebral blood flow regulation explained using a lumped parameter model." American Journal of Physiology: Regulatory Integrative Comparative Physiology 282  (2002): R611-R622.

[19] Cirovic, S, C Walsh, and W D Fraser. "Mathematical study of the role of non-linear venous compliance in the cranial volume-pressure test." Medical & Biological Engineering & Computing 41 (2003): 579-588.

[20] Stevens, Scott A. "Mean pressures and flows in the human intracranial system, determined by mathematical simulations of a steady-state infusion test." Neurological Research 22 (2000): 809-814.

[21] Huang, Judy, et al. "A Modified Transorbital Baboon Model of Reperfused Stroke." Stroke 31 (2000): 3054-3063.

[22] Belayev, Ludmila, et al. "Middle cerebral artery occlusion in the mouse by intraluminal suture coated with poly-L-lysine: neurological and histological validation." Brain Research 833 (1999): 181-190.

[23] Lake, A R, et al. "Angiology of the the Brain of the Baboon Papio ursinus, the Vervet Monkey Cercopithecus pygerithrus, and the Bushbaby Galago senegalensis." The American Journal of Anatomy 187 (1990): 277-287. Excellent anatomical comparison of primate arterial and venous vasculature.

[24] Brenowitz, Gene, and Howard Yonas. "Selective Occlusion of Blood Supply to the Anterior Perforated Substance of the Dog: A Highly Reproducible Stroke Model." Surgical Neurology 33 (1990): 247-252. Good description of canine arterial supply and quantification of anatomic variation.

[25] Lawner, Pablo M, et al. "Hemodynamic and Clinicopathologic Verification of a Stroke Model in the Dog." Stroke 12.3 (1981): 313-316.

[26] http://911stroke.info/brainVesselsCorosionCast.jpg

[27] Glagov S, Zarins C, Giddens DP, Ku DN. 1988. Hemodynamics and atherosclerosis: insights and perspectives gained from studies of human arteries. Arch. Pathol. Lab. Med. 112:1018–31

[28] Ku, David. "Blood Flow in Arteries." Annual Reviews in Fluid Mechanics 29 (1997): 399-434.

[29] http://www.mdconsult.com/das/patient/body/0/0/10041/18120.jpg

[30] Constanzo, Linda S. Physiology. 1998. 3rd ed. Philadelphia: Saunders Elseiver, 2006.

[31] Fung, Y C, and B W Zweifach. "Microcirculation: mechanics of blood flow in capillaries." Annual Reviews in Fluid Mechanics 3 (1971): 189-210.

[32] Netter, Frank. Atlas of Human Anatomy. 3rd ed. East Hanover: Novartis, 2002.

[33] Guyton, Arthur C, MD, and John E Hall, PhD. Textbook of Medical Physiology. Philadelphia: Elsevier Saunders, 2006: pg 199, 287-288, 761-763.

[34] Andeweg, J. "The anatomy of collateral venous flow from the brain and its value in aetiological interpretation of intracranial pathology." Neuroradiology 38 (1996): 621-628. Excellent description of collateral venous anatomy and its relevance to retrograde flow.

[35] Chou, Chi-Hsiang, A-Ching Chao, and Han-Hwa Hu. "Ultrasonographic Evaluation of Vertebral Venous Valves." American Journal of Neuroradiology 23 (Sept. 2002): 1418–1420.

[36] Hill, Lisa, and Carl Gwinnutt. "Cerebral Blood Flow and Intracranial Pressure." Unpublished notes, date unknown. Cerebral Physiology II. Hope, United Kingdom. Internal review document, perhaps for residents, at the Royal Oldham Hospital in the UK.

[37] Roehl, Lisbeth, et al. "Time Evolution of Cerebral Perfusion and Apparent Diffusion Coefficient Measured by Magnetic Resonance Imaging in a Porcine Stroke Model." Journal of Magnetic Resonance Imaging 15 (2002): 123-129. Human CBF thresholds during stroke summary.

[38] Ursino, Mauro, and Carlo Alberto Lodi. "A simple mathematical model of the interaction between intracranial pressure and cerebral hemodynamics." Journal of Applied Physiology 82 (1997): 1256-1269.

[39] Koizumi, J, et al. "Experimental studies of ischemic brain edema, I: a new experimental model of cerebral embolism in rats in wh