In Vitro Hemocompatibility Testing: Continuing

Development of an Ovine Blood-Loop Assay

Yan Chen, PhDSr. Director, Biocompatibility

American Preclinical Services/NAMSA

SOT-MDCPSS Webinar, March 30, 2021

2CONFIDENTIAL2

DISCLAIMER

The material presented is for general information sharing only and is solely the opinion of the presenter.

3CONFIDENTIAL

THROMBOGENICITY TESTTHROMBOGENICITY TEST ISO 10993-4 thrombogenicity testing is widely used for meeting regulatory requirements for approval of blood-

contacting medical devices.• Thrombus deposition• Thromboembolism

In vivo thrombogenicity study• NAVI (Non-anticoagulated venous implant)• AVI (Anticoagulated venous implant)• NAAI (Non-Anticoagulated arterial implant)• AAI (Anticoagulated arterial implant)

A: TestB: LMCD (Legally Marketed Comparison Device)1: Femoral2: Jugular

ISO 10993-4 Selection of tests for interactions with blood (2017)

4CONFIDENTIAL

TRADITIONAL NAVI TEST FOR THROMBOGENICITY

ISO 10993-4 Selection of tests for interactions with blood (2017)

Beagle or Mongrel Dogs (N=2-3)

Non-heparinized Animals

Evaluation in the Jugular/Femoral vein

4 Hour Dwell Time

e f

5CONFIDENTIAL

CHALLENGES OF THE IN VIVO MODEL

Not internationally accepted by regulatory agencies

Controversy and caveats with the methodology• The implant location/anatomy• The implant technique • The extent of device-vessel wall contact (Tissue

Damage)• Time/incubation period • The explant technique• Hydrophilic surface• Statistical power• Inconsistent data for LMCD (legally marketed

comparison device)

1Wolf, M. F., and Anderson, J. M., 2012, “Practical Approach to Blood Compatibility Assessments: General Considerations and Standards,” Biocompatibility and Performance of Medical Devices, J.-P. Boutrand, ed., Woodhead Publishing, Oxford, UK.

Is the NAVI model clinically relevant?

N AVI L M C D

0

2 0

4 0

6 0

8 0

1 0 0

L M C D in N A V I

Th

rom

bu

s A

sses

smen

t (%

Su

rfac

e A

rea)

m e a n 3 2 ± 2 1 %n = 7 6

6CONFIDENTIAL6

IN VITRO MODEL – BLOOD LOOPLoop Configuration Temperature Control

Positive Control: Latex Tubing (Manually Abraded)

Negative Control: PU Catheter (Hydromer Coated)

8CONFIDENTIAL8

QUANTITATIVE THROMBUS EVALUATION

Insertion Site

= test device (Total Length = 10 cm)

= Thrombus formation

Thrombus on segment: for 50% coverage

for 100% coverage

Cross section View - Example for Calculating Coverage

𝟑𝟑.𝟏𝟏𝟏𝟏𝟏𝟏 𝒙𝒙 𝟓𝟓𝟓𝟓𝟓 + 𝟏𝟏𝟏𝟏𝟏𝟏 𝒙𝒙 𝟏𝟏𝟓𝟓𝟓𝟓𝟓𝟏𝟏𝟓𝟓𝟏𝟏𝟏𝟏

= 26% coverage

3.1cm 1.0 cm

Negative Control

Positive Control

9CONFIDENTIAL

STUDY ASSESSMENT

Thrombus formation • % surface area coverage

Non-adherent thrombus • Weights of non-adherent thrombus

Blood characterization• Complete blood counts (CBC)• Activated clotting time (ACT) • Platelet counts

10CONFIDENTIAL

IN VIVO – IN VITRO COMPARISON

NAVI (in vivo) Blood Loop (in vitro)

Implant positionand technique

• Animal anatomy variation• Experience based implant technique• Hard to check device deployment

• Uniform tubing diameter• Controlled deployment procedure

Device-vessel wall contact

• Tissue Factor • Minimal tubing wall contacting• No Tissue Factor components

Statistical power • N=2-3 • N= 3 – 9• Three donors

Controls • LMCD • LMCD• Positive Control• Negative Control

Risk for thromboembolism

• Downstream Organs (Heart and Lung)

• Unable to distinguish between test and LMCD

• Non-adherent thrombus quantification

• Separate loop to distinguish between test and LMCD

11CONFIDENTIAL

IN VITRO BLOOD LOOP - PUBLICATIONS

Journal of Medical Devices: Special Issue on Cardiovascular Device Development

12CONFIDENTIAL

3R’s AWARD FOR IN VITRO ALTERNATIVE ASSAY

13CONFIDENTIAL13

PERFORMANCE OF LMCD IN THE LOOP

L o o p L M C D P o s .C o n tr o ls N e g . C o n tr o ls

0

2 0

4 0

6 0

8 0

1 0 0

L M C D in L o o p w ith P o s it iv e a n d N e g a t iv e C o n tro ls

Th

rom

bu

s A

sses

smen

t (%

Su

rfac

e A

rea)

m e a n 1 1 ± 1 5 %n = 8 4

m e a n 2 .0 ± 4 .7 %n = 1 1 4

m e a n 8 2 ± 1 1 %n = 8 4

14CONFIDENTIAL14

RESULTS COMPARISON IN VIVO – IN VITRO

N AVI L M C D L o o p L M C D

0

2 0

4 0

6 0

8 0

1 0 0

L M C D in N A V I v s . L o o p

Th

rom

bu

s A

sses

smen

t (%

Su

rfac

e A

rea)

m e a n 3 2 ± 2 1 %n = 7 6 m e a n 1 1 ± 1 5 %

n = 8 4

15CONFIDENTIAL15

SCORE - PERCENTAGE SURFACE AREA

Table C.2 NAVI/AVI Scoring Scheme B (0-4)

N AVI L M C D L o o p L M C D

0

1

2

3

4

5

L M C D in N A V I v s . L o o p(T h ro m b o g e n ic ity S c o re )

Th

rom

bu

s A

sse

ss

me

nt

(Sc

ore

)

m e d ia n = 2

m e d ia n = 1

N AVI L M C D L o o p L M C D

0

2 0

4 0

6 0

8 0

1 0 0

L M C D in N A V I v s . L o o p(% S u r fa c e A re a )

Th

rom

bu

s A

sses

smen

t (%

Su

rfac

e A

rea)

m e a n 3 2 ± 2 1 %n = 7 6 m e a n 1 1 ± 1 5 %

n = 8 4

16CONFIDENTIAL16

CONCLUSIONS

The high frequency of thrombus formation leading to scores of ≥3 (≥ 51% surface areaoverage) for LMCDs highlights the deficiency in the performance of NAVI assay.

In comparison, the in vitro blood loop assay results where assay performance ofLMCDs was much more in line with their clinical performance and regulatory history.

In addition, the blood-loop assays also carries the enhanced support by concurrent useof well-behaved positive and negative controls to address the variability of the in vitroassay.

Overall, these results strongly supported the inability of the NAVI model to predictclinical risk. Alternative assays need to be developed and the regulatory acceptance isdesired.

17CONFIDENTIAL17

ACKNOWLEDGEMENT

Mark Smith, PhDChief Scientific Officer, Emeritus

APS

APS staff contributed to the blood loop projectAmber DargisSarah Howard

Heather AckersonTammy Fossum

18CONFIDENTIAL18

QUESTIONS

Thank you!

Contact: ychen@apsemail.com