|Year : 1974 | Volume
| Issue : 2 | Page : 11-16
Pathogenesis of corneal graft failure
AL Aurora1, RC Khandur1, G Singh2
1 Department of Pathology, Maulana Azad Medical College and Associated Hospitals, New Delhi, India
2 Department of Ophthalmology, Maulana Azad Medical College and Associated Hospitals, New Delhi, India
A L Aurora
Department of Pathology, Jawaharlal Institute of Postgraduate Medical Education and Research, Pondicherry-6
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Aurora A L, Khandur R C, Singh G. Pathogenesis of corneal graft failure. Indian J Ophthalmol 1974;22:11-6
Recent advances in corneal surgery and transplantation biology have considerably improved the prognosis in keratoplasty and have given new hope to cases of corneal blindness. However, very few systematic histopathologic studies have been carried out to elucidate the pathogensis of corneal graft failure. Except for a recent, chiefly clinical study from Ahmedabad by Dhanda and Kelevar  , no histopathological studies are available in the english litrature from South-East Asia despite the preponderance of corneal blindness in this part of the world.
The present study was undertaken to highlight the important pathological factors responsible for graft failure as seen in Northern India.
| Material and Methods|| |
The material was obtained from 100 cases operated for single or multiple corneal grafts between July, 1970 to December, 1972 at Irwin hospital, New Delhi, India. The time interval between the first and last graft varied from 1 day to 2 years. The patients ranged in age from 5 weeks to 70 years. There were 29 females and 71 males, the youngest patient of 5 weeks was a girl.
Of the 100 cases, 42 had single keratoplasty either penetrating (P. K.) or lamellar (L. K.). These included 9 cases of spontaneous ulcers*, 6 cases of traumatic ulcer, 6 cases of herpetic ulcers, I case of exanthematous ulcer and 20 cases of opacities. 7 of the opacity cases had simple macular or leucomatous opacity while 13 had leucoma adherans.
Fifty-eight cases had more than one graft, the last being a P.K. Eight cases had two grafts, L.K. followed by P. K.; 32 had two grafts, both P.K; and 18 cases had more than two penetrating grafts. Among these 18 multiple graft cases, two were given five grafts, one case had four grafts while all other cases had three grafts each. Thus 81 grafts (P. K.)were available for study from these fifty-eight cases.
The corneal buttons received were examined grossly with unaided eye and under the dissecting microscope. A representative piece was taken, processed and embedded in paraffin. Tae sections were cut at 5(l, routinely stained with Hematoxylin and Eosin (H & E) and by Mc Manus periodic acid Schiff method (PAS). Special stains wherever indicated were employed according to the standard procedures. 
| Observations|| |
The details of the follow up and fate of the grafts of the 42 cases who had single keratoplasty are given in [Table - 1]. It will be observed that despite the limited duration of follow up, more grafts in ulcer cases (50 per cent) became opaque as compared to the opacity cases (25 per cent). In case of deep or perforating ulcers where penetrating grafts had been given, the rate of failure was relatively higher than in cases with superficial ulcers given lamellar grafts. Of the 6 cases of herpetic ulcer, 2 had P.K. and 4 L.K. One penetrating and three lamellar grafts remained clear in the follow up period of 3 to 42 months, the other two grafts became opaque. In the single case of exanthematous ulcer, the graft became opaque during the follow up period of 3 months.
On the basis of histological studies, the factors responsible for graft failure in 81 grafts (P.K.) were arbitrarily categorized into the following eight groups.
I. Poor wound healing
In 35 of the 50 cases given two P.K. or multiple grafts, junction between the donor and recipient cornea could be histologically demonstrated. Poor wound healing was seen in 12 grafts. In three grafts epithelial downgrowth along the line of junction accompanied the stromal overgrowth from the junctional zone to form a postoraft membrane. In one case, an epithelial cyst had formed on the back of the cornea [Figure - 1]. In the fourth case, iris had prolapsed through the junctional zone. In another case, descemetocele occupied the junction. In two cases, the wound had ruptured. In one case Ps. pyocyanea along with Alkaligenes faecalis were isolated. In another four cases, stitch abscess had formed in the junctional zone. Staphylococcus pyogenes was isolated in one case. In the another case poor wound apposition was characterized by the presence of young granulation tissue, even though the graft was there for 12 days. In this case Staphylococcus pyogenes could be isolated from the conjunctival sac.
II. Infection with ulceration of the donor cornea
Donor cornea got infected in 25 grafts. Six were mycotic ulcers*. These ulcers were superficial in two cases, deep in another two and had perforated in two cases. Culture studies carried out in one case of superficial ulcer revealed Anetratum species besides Ps. pyocyanea. In one case of perforated ulcer, in an opium addict, the sections revealed branching septate hyphae in the most superficial parts of the substantia propria with break in the Bowman's membrane. Culture studies revealed Staphylococcus pyogenes and Penicillium species.
Of the other 19 ulcers, ten had perforation with prolapse of iris in three. There were four deep ulcers involving more than one-third thickness of the substantia propria. Five were superficial ulcers being confined to less than one third the thickness of the substantia propria. In 5 cases of perforated ulcers where culture studies had been carried out, Ps. pyocyanea was isolated in one case. Staphylococcus pyogenes was isolated in one case who had developed a corneal fistula. Coagulase negative Staphylococcus was isolated in two cases, and Ps. pyocyanea along with Staphylococcus coagulase negative in the last case. Culture studies were available in two cases of deep ulcers. These showed micrococci in one case and Ps. pyocyanea with Klebsiella in the second case. Culture studies carried out in one case of superficial ulcer showed Alkaligenes faecalis.
III. Infection without ulceration
Four grafts were characterised by considerable edema of the graft, and presence of many polymorphs in the substantia propria without ulceration. The culture studies from the conjunctival sac of three cases showed Staphylococcus pyogenes in one case and Proteus with Klebsiella species in the second case. In the third case, the culture was sterile.
IV. Edema of the graft
Edema of the graft without inflammatory cells was the chief feature in twelve cases. In one case, au earlier L. K. was followed by P. K. In this case, the posterior half of the substantia propria showed many edema cysts [Figure - 2].
V. Defective continuity of Descemet's membrane
In four grafts, Descemet's membranes of the donor and recipient corneas were curled up anteriorly on either side of the junction. In three of these cases, further complications had occurred. In one case, a small zone of stromal overgrowth also intervened between the two ends of Descemet's membrane [Figure - 3].
In the second case iris was adherent to the back of the junction. In the third case Descemet's membrane was missing over small central area as well as at the periphery. Anterior peripheral synechia was seen in the region devoid of Descemet's membrane.
VI. Leucoma adherens and anterior peripheral synechia
Leucoma adherens was responsible for graft failure in four grafts. Some degree of stromal overgrowth and vascularization were invariably present in these cases. Anterior peripheral synechia were detected in nine grafts.
Vascularization alone could be considered a cause of graft failure in seven cases. Variable degree of vascularization was also associated with graft ulcers, anterior synechia and leucoma adherens.
VIII. Diminution in the number of keratocytes of the graft
A marked decrease in the number of keratocytes in the deeper parts of P.K. was a prominent feature in four grafts. An interesting finding in a case given L.K. followed by P.K., was the marked diminution of keratocytes in the region of the earlier lamellar button [Figure - 4].
| Comments|| |
The histologic evidence of epithelial downgrowth, prolapse of iris or descemetocele along the line of junction clearly indicated poor apposition of donor and host corneas due to faulty surgical tecnique and accounted for 6 of 12 cases with poor wound healing. Winter  considered surgical fault as the primary cause in 15 of 36 cases in his series. Kurz and D'Amico  reported such complication in only 2 of their 41 cases.
Post-operative infection is an important cause of graft failure. Infection was obviously responsible for poor wound healing in at least five of our cases. Further in the culture studies carried out in 16 of the 32 grafts suspected of infection, Staphylococcus pyogenes and Ps. pyocyanea proved to be the most important offenders. In six cases fungal hyphae could be demonstrated within the donor cornea. In two of these cases hyphae were present only in the superficial lamellae of the substantia propria indicative of a secondary mycotic invasion in a pre-existing bacterial ulceration. In fact, in these cases, Ps. pyocyanea and Staphylococcus pyogenes were isolated and considered responsible for the ulceration. The high incidence of infection in our series is due to the fact that in nearly 50 per cent of our cases had initially a rapidly progressive corneal ulcer. Corneal graft was considered the only hope in these cases and risk taken to save the eyes.
Edema of the graft in the absence of inflammatory cells was an interesting finding in twelve cases. The formation of edema cysts in the posterior part of the cornea in one case of L.K. suggests the possibility that the L.K. disturbed the hydrodynamics of the cornea. The fluid seeping into the cornea from the anterior chamber, apparently could not find its normal course anteriorly resulting in waterlogging and edema cysts. Winter  did find marked edema in 20 per cent of his cases, but obviously not marked enough to form edema cysts. Rycroft  also mentioned edema as an important complication in his masterly Doyne's memorial lecture. However, there is no mention of the formation of edema cysts.
The defect in the continuity of the Descemet's membrane in four cases in the present series lays stress on the importance of handling the donor cornea with utmost respect and also emphasizes the proper apposition of the wound if the graft has to succeed. These defects favour the formation of post graft membrane.
The post-graft membrane to a variable extent was present in 40 percent of the grafts in the present series. These included cases with poor wound healing, defect in Descemet's membrane and anterior synechia. The anterior graft membrane highlighted by Rycroft  and Winter  was detected in 50 percent of the grafts in the present series.
The marked diminution of keratocytes in the donor cornea in five cases points to a subtle imbalance in the fluid micro-environment of the keratocytes and thus possibly of the nutrition between the host and the donor tissues. This finding has been mentioned earlier by Kurz and D'Amico  in 19 of their 41 cases. Winter  , Rycroft  , Hales  and Dhanda and Kalevar  did not record such a change.
The immunological aspects of corneal graft rejection have been recently discussed at length by French  . The immune reaction in the present series was indicated by the presence of plasma cells, lymphocytes and histiocytes in the junctional zone. The presence of edema with many lymphocytes and vascularization of the substania propria of the donor cornea was a prominent feature in only one case. Variable degree of edema and lymphocytic infiltrates seen in nine cases with anterior synechia could be a consequence of accompanying iritis.
It is obvious that majority of the factors responsible for graft failure are preventable and need greater precision in wound apposition and proper handling of the donor cornea.
| Summary|| |
One hundred cases of corneal blindness, in whom one or more keratoplasties has been carried out, were investigated for etiopathogenesis of graft failure. The factors of importance included infection, failure of proper wound apposition, defect in the continuity of Descemet's membrane and disturbed hydrodynamic's as the chief causes of graft failure.
| References|| |
Dhanda, R.P. and Kalevar, V., 1972, Inter. national Ophthalmology Clinics, 12,
French, M.E., 1972, In Casey, T.A., Corneal grafting.
63, Butterworths, London.
Hales, R.H. and Spencer, W.H., 1963, Arch. Ophrhal70,
Kurz, G.H., and D'Amico, R.A., 1968, Am. J. Ophthal, 66,
Luna, L.G., 1968, Manual of histologic staining methods of the Armed Forces Institute of Pathology,
3rd ed, Blakiston Division, McGraw-Hill Book Company, New York.
Rycroft, B., 1965, Trans. Ophthal Soc. U.K. 85,
Rycroft, P.V., 1965, Trans. Ophthal Soc, U.K. 85,
Winter, F.C., In Rycroft, P.V. 1969, CorneoPlastic Surgery
557. Pergamon Press, Oxford,
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4]
[Table - 1]