The making of indigenous vascular prosthesis

Ever since the historic use of Vinyon-N for arterial replacement by Voorhees et al1 in 1952, tubular grafts made from textile fabric have been firmly established in modern vascular surgery. However, technological data regarding their physical, chemical and toxicological properties which lead to biological healing after implantation of these grafts, were not available for manufacturing these in India. Hence, it was considered worthwhile to initiate planned development of arterial prosthetic grafts indigenously so that these could be made available readily at an acceptable cost to patients. This prosthesis was intended to serve as a permanently implanted device for treatment of patients with vascular pathologies like aneurysm or arterial occlusive disease, to replace or bypass the diseased blood vessel, so as to preserve life and/or restore vital blood supply to the affected organ(s).

As per the Joint council of the Society of Vascular Surgery and International Society of Cardio-vascular Surgery (SVS/ISCVS)2, the ideal characteristics of vascular prosthesis are indicated in Table I. The objective of this programme was the development of straight woven vascular prosthetic grafts of polyester in the size range of 10 to 25 mm internal diameter initially which should be comparable in safety and efficacy to the widely used commercially available vascular prostheses.

Table I Ideal characteristics of vascular prosthesis (as per Joint council of SVS/ISCVS)2

Material & Methods

Development & in vitro testing: Designed and fabricated at South India Textile Research Association (SITRA), Coimbatore, the main testing and clinical work of woven grafts was carried out at Sree Chitra Tirunal Institute for Medical Sciences and Technology (SCTIMST), Thiruvananthapuram, Kerala, India. The prosthesis was essentially made of polyester polyethylene terephthalate (PET). The design and developmental process included yarn preparation, weaving of the tubular fabric, crimping and sterilization. The fabric was a seamless woven tube of 75 denier texturised polyester yarn which was suitably crimped (Fig. 1) to provide large diameter vascular graft with a porosity of 200±50 ml/min/cm² of normal saline at 120 mmHg. The physical characteristics of the yarn, in terms of the yarn count, yarn twist and tensile strength to weave the grafts were determined by SITRA. The chemical composition of the material was analysed by infrared spectroscopy using spectrophotometer (Shimadzu, Japan), and differential thermal analysis at the laboratory for technical evaluation of biomaterials at the biomedical Division of SCTIMST. Toxicity/biocompatibility studies such as acute systemic toxicity, intracutaneous irritation, in vitro haemolysis and implantation in muscle were studied as per international standards3 for the vascular graft material. Acute systemic toxicity was studied in mice using physiological saline and cotton seed oil extracts of the material. Intracutaneous irritation was done by injecting the extracts of material and control intradermally into rabbits under ketamine anaesthesia. Grading of erythema and oedema of material in experimental and control rabbits were recorded at 24, 48 and 72 h. In vitro haemolysis was carried out with material and extract of the material using centrifugation (REMI R-8C DX, India) with rabbit blood. The muscle implantation was carried out in nine albino rabbits. The implanted animals were sacrificed at the end of one, four and 12 wk, the tissue with the implanted materials were collected, sectioned using RM2255 microtome (Leica, biosystems, India) and subjected to histopathological analysis (Axio Imager Z1 microscope, Carl Zeiss).

Close

Fig. 1

The test vascular polyester polyethylene terephthalate (PET) graft with internal diameter ranging from 8-30 mm.

Export to PPT

Pre-clinical animal testing: During 1982-1983, as part of the pilot study4, 30 pigs had received the test graft. Further, in vivo experiments in 1992-1994, reported herein, involved 15 pigs as per Ethics Committee requirements in accordance with draft document of Association for the Advancement of Medical Instrumentation5. Protocol was drawn to evaluate 11 mm test graft and 12 mm USCI DeBakey graft (USA) as concurrent controls (Fig. 2).

Close

Fig. 2

Protocol of pre-clinical testing in Yorkshire pigs using test and control grafts and the explantation data. CPB, cardio pulmonary bypass.

Export to PPT

Four to six months old large white Yorkshire pigs (Source: Government pig farm, Parassala, Thiruvananthapuram) weighing 40-45 kg, were used for each experiment after due conditioning as per animal testing guidelines6. Following overnight fasting and premedication, general anaesthesia was employed using thiopentone sodium (10 mg/kg) and muscle relaxants and maintained with endotracheal intubation and inhalation anaesthetic agents. Open cut down was employed to obtain central venous access through saphenous vein and arterial access through femoral artery. Continuous monitoring of heart rate, blood pressure, electrocardiogram (ECG) and arterial blood gases (ABG) at regular intervals formed the standard protocol. Left postero-lateral thoracotomy was performed and pleural cavity entered by excising 4^th or 5^th rib. Lung was gently retracted and mediastinal pleura over descending thoracic aorta incised and aorta dissected and looped. Dissection was continued as far as feasible towards suprahiatal aorta whenever long grafts had to be implanted. Heparin (1 mg/kg) was administered three minutes prior to clamping in some cases except where blood retrieval strategy was employed in which case 3 mg/kg was required7. Cardiotomy reservoir with suction device was utilized to retrieve shed blood during most procedures. After completion of arterial grafting (Fig. 3A), protamine was administered and haemostasis achieved. Thoracotomy was closed without indwelling chest drainage tube.

Close

Fig. 3

Replacement of segment of descending thoracic aorta in pig with the test graft using standard protocol via left postero-lateral thoracotomy (A). Gross healing characterstics of test vascular graft at explant post-6 month in vivo in pig in 8, 4 and 12 cm lengths respectively (B, C, D). Neointima was found to be smooth in short replacements while it was less regular and even, nonetheless thrombus-free, in long grafts.

Export to PPT

All animals except one could be extubated at the end of procedure and walked back to cage 4-6 h after completion of the procedure. They were given water and feeds from next day morning. Streptopenicillin injection was given for five days for infection prophylaxis. The surviving animals were followed up for three and six months as per protocol.

Grafts were explanted to study the healing characteristics, blood/tissue interactions and patency rates. Explantation was performed at three months or six months following graft insertion under general anaesthesia; 1 mg/kg heparin was administered to avoid misinterpretation of pre- or postmortem thrombus.

Phase I clinical trial:

(i) Design - Prospective randomized study was designed to prove non-inferiority of the test prosthesis with concurrent controls using commercially available prosthetic grafts. Objective patency assessment was made by non-invasive haemodynamic indices, duplex ultrasound and catheter angiography/CT angiography.

(ii) Parameters evaluated were (a) ease of pre-clotting; (b) pliability/conformability; (c) ease of suturing/suture retention; and (d) short-term patency at one year, long-term patency at five years and beyond.

(iii) Institutional Ethics committee (IEC) gave permission to use large diameter prosthesis of length less than 10 cm for a total of 10 patients during phase I trial as a matter of abundant caution. This restricted its use only for repair of abdominal aortic aneurysms (AAA) and coarctation of thoracic aorta (CoA). Special and detailed informed consent was obtained in every patient included in the study.

(iv) Methods - Between September 1998 and November 1999, 10 patients were included in phase I clinical trial using the test graft at SCTIMST, six of whom underwent repair for AAA and four for CoA (Table II). AAA involved the infra-renal segment of aorta which was repaired through xipho-pubic midline laparotomy and transperitoneal approach to the aorta. Standard inclusion graft technique was performed using appropriate (14-20 mm internal diameter) test graft using 3/0 monofilament polypropylene suture for proximal anastomosis and 4/0 suture for distal. Inferior mesenteric artery was re-implantated into the prosthesis using Carrel's patch technique8 in selected cases upon indication. CoA repair was performed through left postero-lateral thoracotomy. Excision of coarct segment and interposition graft placement using 14-16 mm ID test graft was done in one patient while left subclavian artery (LSA) to descending thoracic aorta bypass (jump graft) using 12-14 mm ID graft was done in the other three. The latter involved end-to-side anastomosis to the dilated LSA proximally and post-coarct aorta distally using 4/0 polypropylene suture. In the control group, eight patients with AAA and two with CoA underwent elective implantation of the control Dacron graft.

Table II Clinical details and results of phase I (controlled) clinical trial

Phase II clinical trial: Twenty two patients who underwent indexed vascular reconstructions at four designated hospitals viz. Medwin Hospital, Hyderabad; G. Kuppuswamy Naidu Memorial Hospital, Coimbatore; Medical College Hospital, and SCTIMST, Thiruvananthapuram; formed the basis of this part of the study conducted from August 2005 to September 2008. Patients chosen were all symptomatic and investigated, operated upon and followed up using an already established and common protocol. Age ranged from 19 to 74 yr with mean age of 55 yr; four patients were female (>4:1 male: female ratio). Clinical indications for graft implantations included CoA in three, middle aortic syndrome in one, thoraco-abdominal aortic aneurysm (TAAA) in two, AAA in 10, iliac artery occlusion in five and iliac artery aneurysm in one patient. Pre-operative ankle brachial index (ABI) was noted in all patients. Internal diameter of prosthesis used ranged from 10-18 mm.

Parameters were carefully noted regarding handling qualities of prosthesis and procedure related complications or mortality. Postoperative status regarding distal pulses and ABI were documented (Table III). Patients were followed up for a maximum period of five years using clinical and imaging documentation of graft patency and integrity.

Table III Clinical details and results of phase II (multicentric) clinical trial

Long-term follow up till date: Patients recruited in the clinical trials that involved implantation of the test graft at the institute as well as the control group were followed up by clinical examination and duplex ultrasound on yearly basis. CT scan/magnetic resonance angiography (MRA) were performed yearly for first two years and thereafter once in 2-3 years of follow up.

Parameters evaluated were (i) Dilatation and elongation - No progressive dilatation > 15 per cent/that of control grafts (as measured in systole). (ii) Structural stability - should be non-biodegradable. (iii) Anastomotic characteristics - Freedom from pseudoaneurysm formation. (iv) Infection - <1 per cent in intracavity grafts, <2 per cent for grafts crossing inguinal ligament and <3 per cent for extremity grafts.

Results

In vitro testing: The results of the study did not show any significant irritation or systemic toxicity with physiological saline and cotton seed oil extracts of the material. The percentage of haemolysis induced by the material and extract was under acceptable range. Results of the histopathological evaluation suggested that the material did not produce any histopathological changes. Hence the toxicity study concluded that the material was non toxic, non irritant, non haemolytic and biocompatible.

Pre-clinical animal testing

(i) Immediate - All animals, except one (15^th), were extubated at the end of procedure and walked back to their cage 6-8 h thereafter. The 15^th animal which underwent interposition graft using cardio-pulmonary bypass to support circulation succumbed to air embolism at the end of an otherwise satisfactory procedure. Another animal died on 2^nd post-operative day following hyperpyrexia of 104°F and at autopsy graft was noted to be patent. The third animal developed delayed paraplegia 24 h after operation that involved 45 min of aortic cross-clamping, had to be sacrificed and at autopsy the prosthetic graft was found to be patent and void of any thrombus. Hence the remaining 12 animals were available for explantation to study the healing characteristics.

(ii) Explantation data - Explantation was performed at three and six months after graft implantation under general anaesthesia. Animals weighed 55 and 65 kg, respectively; 1 mg/kg heparin was administered to avoid misinterpretation of pre- and post-mortem thrombus.

Gross examination showed that the grafts were well incorporated in a 1-4 mm thick perigraft capsule. No pseudoaneurysm was noted in either of the groups. Cut surface showed that suture lines were tidy and clean with no evidence of thrombus formation. The body of the prosthesis was glistening white with well organized neo-intima. However, neoimtima inside long grafts appeared irregular and uneven but well organized in contradistinction to smooth glistening thin lining in short replacements (Fig. 3B-D)

Histologically, the inner lining was thin and measured 0.6-1.0 mm, and showed uniform regular neointima of fibro-collagenous tissue, a few islands of calcification along with cellular infiltration of lymphocytes and a few foreign body giant cells (Fig. 4).

Close

Fig. 4

Low power microphotographs using Hematoxylin & Eosin (H&E) staining showing lack of thrombosis at anastomotic site (black arrow) with tissue ingrowth between the graft (A), and endothelial-lined neointima (green arrow) on the inner aspect of the graft (B) 6 months after implantation in porcine model.

Export to PPT

Phase I clinical trial: First patient following repair of AAA was discharged from hospital on 10^th day. At discharge, her blood pressure was controlled with 50 mg Atenolol (beta-aderenergic blocker) and her ankle brachial index was 1.1. She was maintained on aspirin and iron supplementations as well. CT scan prior to discharge showed patent graft, intact suture lines, smooth luminal outline and patent inferior mesenteric artery. She was evaluated in person five years after surgery. In 2013, a telephonic interview confirmed her wellbeing with no complaints regarding abdomen.

All patients survived major surgical procedure and made satisfactory early recovery except one patient (4^th) who underwent surgical repair for complex post-subclavian coarctation of aorta with large heavily calcified post-coarct aneurysm, who developed paraplegia. Prolonged aortic cross-clamp time of >60 min along with significant bleeding requiring blood transfusions led to spinal cord dysfunction. Minor sequelae in others included prolonged paralytic ileus in two and respiratory infection in one (Table II). All patients were discharged from hospital on or before 10^th postoperative day except the patient who developed paraplegia (20 days). All patients underwent Duplex ultrasound and CT scan before discharge from hospital which showed satisfactory repair, patent graft and smooth regular luminal surface of prosthesis.

Complete follow up was available for all 10 patients at three, six months, one year and yearly thereafter. All patients were in good health and active except the 4^th patient who did not recover from paraplegia inspite of aggressive physiotherapy. ABI at follow up ranged from 0.85 to 1.1. CT/conventional aortogram was done at discharge and on yearly follow up (Fig. 5). The inner diameter of the graft as measured in CT/aortogram was 0.5-0.7 mm less than the internal diameter of the appropriate graft size. This feature matched the control group of DuPont Dacron grafts used in similar locations.

Close

Fig. 5

Contrast-enhanced abdominal CT scan axial section of a symptomatic patient in phase I clinical trial showing an infra-renal abdominal aortic aneurysm (A). Abdominal contrast enhanced CT scan of the same patient taken 8 years after repair showing intact and patent test vascular graft (yellow arrow) (B). Follow-up imaging of an adult patient with coarctation of aorta treated by interposition grafting with test vascular graft as part of phase I trial. Digital subtraction angiogram (DSA) at 1 year (C) and CT angiogram at 6 years (D) showing preserved and patent graft (red arrows).

Export to PPT

All patients who received the test implant survived the procedures and recovered well except the patient who developed paraplegia; however, digital substraction angiography (DSA) at one year and CT aortogram at five years showed intact repair with patent graft. Host-graft interaction was reflected at three sites namely, aorta to graft interphase, outside the graft and its inner luminal surface. No untoward sequelae of excessive intimal hyperplasia or pseudoaneurysm formation were noted in and around the proximal and distal suture lines in these patients. The incorporation of the graft in the form of perigraft capsule appeared satisfactory. The luminal side of the grafts were clean and regular with no evidence of thrombus formation inside the test or control grafts.

Phase II clinical trial: Prosthesis was found to be surgeon friendly, with excellent suturing and suture retention qualities. All centres reported good graft handling qualities with pliability, lack of fraying, ease of pre-clotting and suturing and absence of excessive weeping following implantation similar to control Dacron grafts. Patients were electively ventilated for 4 to 24 h with mean of 8 h as dictated by their clinical condition. All 22 patients at 4 centres of study made satisfactory early recovery from appropriate procedures. Two patients (both TAAA) developed transient renal dysfunction (increase in serum creatinine >1 mg/dl higher than pre-operative level) eventually normalising at discharge. Follow up visits were planned at three and six months, one year and yearly thereafter. Apart from regular clinical assessment, duplex scan evaluation was performed in patients with graft implantation in abdomen. Check MR/CT angiogram was performed on follow up showed preserved patency with no evidence of graft-related complications (Fig. 6).

Close

Fig. 6

Follow up CT/MR angiograms of patients in the multicentric trials who underwent vascular reconstruction using test graft (white arrows), showing good results following open repair of abdominal aortic aneurysm at 5 years (A), aorto-femoral bypass for iliac occlusion at 4 years (B), and open repair of thoraco-abdominal aortic aneurysm at 6 years (C).

Export to PPT

One patient succumbed to stroke 18 months after aorto-femoral graft procedure; and another to road traffic accident leading to irreversible brain damage three years after AAA repair. Post-operative ABI at 1-4 years were 0.9 or more in 15, 0.8 in 4, and 0.7 in 3 (Table III).

Long-term follow up till date: Follow up was complete in eight patients at >10 yr after implantation of test vascular prosthesis in phase I trial and >5 yr in 14 patients in phase II trial. Three patients who were around 70 yr of age in 1998-1999 died of old age and cardiac events beyond five years after the operation. Present evaluation has confirmed excellent clinical status of the survivors and satisfactory graft function with no incidence of pseudoaneurysm, aneurysm, infection, thrombosis, dilatation or any other untoward sequelae with reference to implanted vascular prosthesis (Fig. 7)9, except one 32 yr old patient with thoraco-abdominal aortic aneurysm due to Takayasu's disease was noted to have pseudoaneurysm at distal aortic anastomosis five years following surgery requiring endovascular aortic stent grafting. Two patients were lost to follow up beyond one year.

Close

Fig. 7

CT angiogram, volume-rendered image at 14 years following jump graft bypass (white arrow) from left subclavian artery to descending thoracic aorta for coarctation of aorta (green arrow) using 14 mm test graft.

(Source: Ref. 9, Reproduced with permission).

Export to PPT

Discussion

This study reports indigenously fabricated polyester graft developed and extensively tested in vitro, than tested in vivo using large animal, prior to clinical study by implantation in patients. Each batch of grafts was put through physical (yarn count, twist and tensile strength), chemical (infra-red spectroscopy and thermal analysis) and toxicity (intracutaneous injection, haemolysin and intramuscular implantation in mice) tests with satisfactory results in accordance with standards laid down by Association for the Advancement of Medical Instrumentation (AAMI) and American National Standards Institute (ANSI)5.

Several animal models are described in literature for in vivo testing of prosthetic grafts. Classic animal experimental data reported by Adam Wesolowski delineated pigs as the most suitable model for in vivo experiments10. His original work on growing pigs weighing 20-39 kg formed the benchmark for graft research as implantation data of three months in growing pigs in terms of healing patterns and characteristics were noted to be equivalent to three years in man.

Pore size of prosthesis is essential for ingrowth of fibroblasts and capillary buds from peri-graft tissue to form and stabilize neointima in the graft body. However, larger pore size results in bleeding through its interstices. Hence the optimal pore size is a trade off that minimizes blood loss through interstices while permitting ingrowth of fibroblasts and capillaries to augment neointimal healing11. The woven test graft porosity was 200 ± 50 ml/cm²/min of normal saline at 120 mmHg. No bleeding problem was encountered through the prosthesis after blood pre-clotting when partial heparinisation (1 mg/kg) was employed. For the experiment performed under cardio-pulmonary bypass, plasma pre-clotted and autoclaved test graft appeared impervious to blood during implantation and thereafter.

Subsequent to general concerns of survival, toxicity, allergy and rejection, in vivo animal experiments primarily focused on study of explanted vascular grafts at varying intervals after implantation. Study of explanted graft, in turn, was further characterized by:

1. Structural integrity of the prosthesis and freedom from dilatation/pseudoaneurysm formation.

2. Patency in general, indicating the functional ability of graft in conducting blood for distal perfusion.

3. Perigraft capsule formation signifying the incorporation of foreign device into the host tissue.

4. Neointima formation, which constitutes the most critical determinant of optimal functioning of prosthetic graft in living body, the quality of which is inversely proportional to thrombogenicity. Major sources of endothelialisation are as follows; across suture lines by direct growth from either ends, by deposition of fibroblasts and wandering endothelial precursor cells (EPCs) in the bloodstream onto the platelet-fibrin layer on the inner side, or by transmural ingrowth of fibroblasts and capillary beds from perigraft tissue entering through the pores of woven/knitted fabric12.

Structural integrity was well preserved and the anastomotic suture lines appeared covered with a transparent fibrin layer separating the flowing blood in the prosthesis. All grafts were incorporated into the perigraft capsule and developed a compact layer of neointima, with a thickness from 0.1-1.0 mm depending upon the diameter of graft, size of parent vessel, distal run-off and, to a certain extent, the surgical technique. In all short segment grafts, neointima appeared very thin, pink, shiny and intimately adherent as expected as endothelialisation is easily possible from both ends of native aorta. However, in long grafts, neointima was thicker (1-1.2 mm), whitish, slightly irregular but still clean and free from thrombus.

In human studies, permanent implants like vascular prostheses must provide optimal safety for the recipient patient along with efficacy and performance for a long duration with uninterrupted functional integrity (Table IV)2. Reporting standards laid down by Ad-hoc committee of the Joint Council of Society of Vascular Surgery and International Society of Cardiovascular Surgery, North American Chapter, mandates a preliminary assessment at two years and final report at five years before multicentric trials can be resorted to2. Safety, efficacy and performance standards of the graft were found to be excellent in this study, even beyond six years and in some patients continuing beyond 14 years. Similarly, patients in multicentric trial were followed up for duration of at least five years. No device-related complications occurred in any of the patients in clinical trials.

Table IV Reporting standards generated by an Ad-hoc Committee of Joint Councils of Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter2

All patients had optimal graft performance status on follow up with continued patency and absence of thrombus formation, pseudoaneurysm formation, loss of tensile strength, progressive dilatation or infection. There was only one incidence of anastomotic pseudoaneurysm in the entire series which involved a 32 yr old patient who developed distal anastomotic pseudoaneurysm four years after an intact repair of thoraco-abdominal aortic aneurysm, who also had a previous history of carotid pseudoaneurysm following carotid bypass. The occurance of such anastomotic pseudoaneurysms is not uncommon in Takayasu's disease13 especially in the setting of long-term steroid usage. This follow up study over a protracted period of 6-14 years has proven safety to patients. The results compared favourably with published data on long-term biostability of polyester vascular prosthesis which reported 95 per cent 10 yr patency rates and structural failure rate of 0.2-3.0 per cent141516.

In conclusion, indigenously designed, developed and thereafter comprehensively performed clinical trials have provided robust data regarding safety and efficacy of indigenously developed vascular prosthesis during an extended study period of over a decade. In a total of 32 patients, with diverse clinical conditions and ages thus studied in the clinical trials encompassing both single-centre and multicentric phases, no complications or adverse events occurred. The technique of coating the prosthesis to render it impervious to blood at implantation17, thus obviating the need for preclotting prior to implantation, was introduced to the Chitra vascular graft as well. Laboratory testing and in vivo experiments in 30 pig models were completed and the coated graft will undergo clinical trial to prove its safety and efficacy before making it available on shelf.