Impact of pandemic COVID-19 lockdown on penetrating keratoplasty patients

Abstract

Purpose:

To study the impact of the COVID-19 lockdown on the regular follow-up of keratoplasty patients.

Methods:

This retrospective interventional case series included 30 patients who had immunological corneal endothelial rejection out of 190 patients who came for post-PKP follow-up between September 15, 2019, and September 30, 2020. The demographics, primary diagnosis, surgical technique, time of presentation, recovery of graft, associated ocular problems, and visual acuity at 1 month were analyzed. Forward stepwise (likelihood ratio) binary logistic regression was used to find significant variables.

Results:

The study population had 19 males (63.33%) and 11 females (36.67%). The mean age of the study group was 42.83 ± 18.89 (8–80) years. Of 30 patients, 19 (63.3%) presented before and 11 (36.7%) after the COVID-19 lockdown. Overall, 23 (77%) showed a reversal of graft rejection. Logistic regression showed that preoperative indications, large-sized grafts, and deep corneal vascularization were significant risk factors for non-resolution of graft rejection. It was noted that patients who presented to the hospital late had poor recovery (P = 0.002). The delay in the presentation was a significant risk factor for non-resolution of graft rejection (P < 0.01). Z-test for proportions revealed that the difference in the non-resolution of rejection on immediate or delayed treatment in patients presenting during lockdown (P = 0.002) was significant.

Conclusion:

This article is to highlight the impact of the COVID-19 lockdown on graft rejection recovery of PKP patients due to delays in follow-up. Early treatment helps in the recovery of graft transparency and the reversal of immunological graft rejection. Also, primary diagnosis, deep vascularization, and large-sized grafts were significant risk factors for non-resolution of graft rejection.

Penetrating keratoplasty is one of the most common forms of human tissue transplantation and has a high success rate.[1] Due to the recent progress in corneal grafting, the indications have gradually included high-risk indications as well.[2] Corneal opacity is reported to be one of the most common causes of blindness worldwide.[3,4] In India, the number of patients with unilateral corneal blindness was estimated to increase to >10 million by 2020.[3,5]

The term allograft rejection refers to a specific immunologic response of the host to the donor corneal tissue and is a serious complication of corneal transplant surgery.[3] It should be distinguished from other non-immune mediated graft failures, such as primary donor failure. Diagnosis of rejection should be made only if the signs appear after 2 weeks of postoperative clear graft.[6]

The high success rate in low-risk PKPs without immunosuppression is due to the immune privilege of the graft and anterior chamber.[7] Therefore, the risk of graft rejection following corneal dystrophies, keratoconus, and pseudophakic corneal edema is less.[8] Patients with deep corneal vascularization (more than two quadrants), herpes simplex keratitis, dry eyes, chemical injuries, and previously failed grafts are at higher risk of graft rejection.[4]

The incidence of allograft rejection after PKP as compared to anterior lamellar keratoplasty (ALK) is higher and causes serious endothelial damage and cell loss.[9] The recent times of COVID-19 have affected patients who had undergone penetrating keratoplasty and developed graft rejection. The delay in follow-up during the COVID-19 lockdown was unavoidable due to travel restrictions and the halt of public transport. Also, patients during travel restrictions may ignore the initial symptoms of graft rejection, which can be avoided with appropriate awareness and vigilance. Poor prognostic factors play a significant role in deciding the surgical outcome of corneal grafting. Early diagnosis and prompt management within time can save rejection from going into failure.[10]

This study investigated the various prognostic factors for corneal graft recovery post immunological graft rejection presenting during the COVID-19 lockdown in a tertiary referral eye care center in South India.

Methods

This was a retrospective study of 30 patients (eyes) who presented with signs of corneal graft rejection between September 21, 2019, and September 21, 2020, following PKP and were treated subsequently with topical, subconjunctival, and systemic steroids. Corneal graft rejection was defined as the appearance of one or more of the following: an endothelial rejection line, new keratic precipitates, anterior chamber cells, and/or corneal edema in the donor graft not previously noted. Medical records of enrolled patients were then studied, and relevant demographic, clinical, and surgical data at the time of presentation were analyzed for association with the outcome. The study was conducted in accordance with the tenets of the Declaration of Helsinki, and approval was taken from the ethics committee at our institution.

Patients were treated with either only topical, or with subconjunctival and oral corticosteroids. Prednisolone acetate 1% eye drops (Pred Forte, Allergan India Pvt. Ltd.) 1 hour were started immediately on diagnosing graft rejection. The dosage was tapered as per the response weekly. Subconjunctival injection of dexamethasone and oral prednisolone 1 mg/kg body weight (Wysolone, Pfizer Limited, India) tapered as the per clinical response was given in some patients. Patients with systemic disease, gastric ulcers, or poorly controlled diabetes were treated under a physician’s guidance.

The main outcome measure was the reversibility of corneal graft rejection. Rejection was considered to be reversed when clinical signs disappeared after the initiation of treatment. Prognostic factors studied were age, sex of the patient, primary diagnosis, deep vascularization, presence of anterior or posterior synechiae, type of surgery done, lens status, number of rejection episodes earlier, graft size, type of suturing done, presence or absence of sutures, presentation of the patient if immediate or delayed, the type of treatment initiated (topical only, topical with a subconjunctival injection, or both with oral steroids) and the distance of patient’s location from the hospital. The recorded visual acuity at the presentation and at the 1-month follow-up was also analyzed.

Statistical analysis

A forward stepwise (likelihood ratio) binary logistic regression was applied. Statistically significant differences were determined using logistic regression between the two groups. Statistical product and service solutions (SPSS) 16.0 software was used to apply statistical tools, and a P value of < 0.05 was considered statistically significant.

Results

The study population had 19 males (63.33%) and 11 females (36.67%). The age distribution of both male and female patients has been shown to be above, and the mean age of the subjects in the study group was 42.83 ± 18.89 years (8–80 years) [Table 1].

Table 1.

Patient demographics and prognostic factors for resolution of graft rejection

Age/Sex	Frequency	Percentage	Mean
Age			42.83±18.89 years
Males	19	63.33%
Females	11	36.67%
	Frequency	Percentage	Outcome_1 month*
Diagnosis before PKP
Keratoconus	3	10.0%	3, 0
Bullous keratopathy	3	10.0%	3, 0
Infective keratitis	7	23.3%	4, 3
Regrafts	4	13.3%	2, 2
Degeneration/dystrophy	6	20.0%	5, 1
Vascularized leukoma	7	23.3%	6, 1
Total	30	100%	23 (76.67%), 7 (23.33%)

*(n=resolved, n=non-resolved)

Corneal opacity due to presumed infective etiology or trauma (n = 7, 23%) and infective keratitis (n = 7, 23%) were the two most frequent primary diagnoses in our study group, which presented with corneal graft rejection post-PKP [Table 1]. Most eyes (n = 18, 60%) were operated on with PKP only and six patients each (20%) were operated on with IOL same sitting or some other additional procedure at a later date [Table 2].

Table 2.

Details of Surgery performed

	Frequency	Percentage
Type of surgery
Only PKP	18	60%
PKP + IOL same sitting	6	20%
PKP + Additional procedure later date	6	20%
Lens status
Phakic	16	53.3%
Pseudophakic	14	46.7%
Graft size
≤ 7.5 mm	8	26.7%
8.0 mm	13	43.3%
≥ 8.5 mm	9	30.0%
Suturing
Interrupted	18	60%
Continuous	12	40%
Sutures on/off
On	21	70%
Off	9	30%

The majority of patients had endothelial rejection on presentation (n = 25, 83.3%) and five eyes (16.7%) had mixed endothelial with stromal rejection [Table 2]. Thirteen grafts (43.3%) were 8.0 mm and nine (30%) were >8.5 mm. Eight grafts (26.7%) were <7.5 mm in size. Most eyes did not have any deep vascularization (n = 20, 66.7%) [Table 2]. Five eyes (16.7%) had deep vascularization in one quadrant, three (10%) in two quadrants, and one eye each (3.3%) in three and four quadrants.

The majority of patients who presented with corneal graft rejection were operated on with only penetrating keratoplasty (n = 18, 60%). Six patients (20%) were operated on with intraocular lens implantation (IOL) in the same sitting. Six patients (20%) underwent PKP with an additional procedure IOL implantation/repair of graft dehiscence subsequently on a later date. The majority of eyes (n = 21, 70%) did not have any sutures in place on presentation with corneal graft rejection. Nine eyes (30%) had sutures in place at the time of rejection [Table 3].

Table 3.

Details of graft rejection and treatment

Rejection Details	Frequency	Percentage(%)
Type of rejection
Endothelial	25	83.3%
Endothelial + stromal/sub-epithelial	5	16.7%
Time of rejection
Before COVID-19 lockdown	19	63.3%
During COVID-19 lockdown	11	36.7%
Previous rejection
0	12	40.0%
1	15	50.0%
≥2	3	10.0%
Presentation early/delayed
Early	21.0	70.0%
Delayed	9.0	30.0%
Treatment given
Topical	14	46.7%
Topical + subconjunctival	7	23.3%
Topical + subconjunctival+Oral	9	30.0%

Nineteen patients presented with rejection before COVID-19 lockdown (63.3%) and 11 (36.7%) patients during COVID-19 lockdown [Table 4].

Table 4.

Time of presentation and outcome of treatment

Time of presentation	Resolved	Non-resolved
Before COVID-19
Immediate	16	1
Delayed	0	2
During COVID-19 lockdown	4	0
Immediate
Delayed	3	4

Logistic regression analysis showed that preoperative indications (P < 0.01), large-sized grafts (P < 0.01), deep corneal vascularization (P < 0.05), and delayed reporting (P < 0.01) for treatment after the onset of graft rejection were significant risk factors for non-resolution of the rejection process [Table 5]. Factors including age, sex, distance from home town to the hospital, type of surgery, and additional surgical procedures were not found to be significant (P > 0.05) risk factors.

Table 5.

Special features and level of significance of selected independent variables by maximum likelihood method for the logistic regression equation

Variable	B	S.E	Wald	df	Sig.
Pri_diagnosis	52.613	5762.309	0.000	1	0.0010
Vas.Deep	-30.435	2825.914	0.000	1	0.0190
Graft size	150.747	15027.533	0.000	1	0.0001
Immed_Delay_Rx	-236.538	20937.442	0.000	1	0.0001
Constant	-91.483	24049.159	0.000	1	0.9970

The Wald test [Table 5] clearly shows that there are FOUR significant variables that can influence the recovery post-graft rejection in PKP eyes. The following logistic regression equation was obtained for the same.

Model score (Y) = -91.483 + 52.613 × primary diagnosis -30.435 × deep vascularization + 150.747 × graft Size-236.538 × immediate/delayed presentation

The variables chosen by the forward stepwise (likelihood) method also showed significant changes in − 2 the log-likelihood method as shown in Table 6 and further support the adequacy of the model.

Table 6.

Change in−2 Log-likelihood

Variable	Model Log likelihood	Change in -2 Log likelihood	df	Sig. of Change
Pri_Diagnosis	-7.124	11.475	1	0.001
Vas.Deep	-4.118	5.463	1	0.019
Graft_Size	-9.470	16.168	1	0.000
Immed_Delay_Rx	-15.071	27.368	1	0.000

There was a significant association between the time of presentation of patients (before and during COVID-19 lockdown) and immediate or delayed presentation/treatment (Pearson’s Chi-square test: P = 0.002) but did not have a significant correlation with the outcome at 1 month (P = 0.199) [Table 7]. Similarly, it was observed that there was a significant improvement in visual acuity in the presentation of graft rejection before and during the COVID-19 lockdown (Wilcoxon signed Ranks test: P = 0.001 and 0.028, respectively) as shown in Table 8.

Table 7.

Correlation between the time of presentation (before/during COVID-19 lockdown) and immediate/delayed treatment (Row 1)/outcome at 1 month (Row 2)

Pearson Chi-square test	Value	df	Asymptotic significance (two-sided)
1) Time_Presentation *Immed_Delay_Rx	9.357	1	0.002a
2) Time_Presentation* Outcome_1 month	1.648	1	0.199b

^aSignificant ^bNot significant

Table 8.

Correlation between the time of presentation (before/during COVID-19 lockdown) and change in visual acuity at 1 month

Wilcoxon signed Ranks test	VA_Post-VA_Pre
a. Time_rejection=Before Lockdown
Z	-3.535
Asymp. Sig. (two-tailed)	0.001*
b. Time_rejection=During Lockdown
Z	-2.201
Asymp. Sig. (two-tailed)	0.028*

*P=Significant

Z-test for proportions revealed that the difference in non-resolution of rejection on immediate or delayed treatment in patients presenting during the COVID-19 lockdown (P = 0.002) was significant but before COVID-19 lockdown was non-significant (P = 0.303).

The logistic regression model for prognostic risk factor identification is an effective method for predicting corneal graft (PKP) recovery post-immunological rejection. It has excellent sensitivity and good specificity as shown by the area under the ROC curve [Fig. 1]. The area under the curve was reported to be 99.7% and suggested that the current model is good enough to predict risk factors.

Figure 1.

Receiver-operating characteristic (ROC) for the forward stepwise (likelihood ratio) binary logistic regression model generated for significant predictors for resolution from graft rejection

Discussion

Graft rejection was treated with topical and systemic steroids. Of 30 patients, 23 (77%) showed reversal of graft rejection. The rate of reversibility has been reported to be around 51–63.8%.[11,12] In our study, logistic regression analysis showed that indications for penetrating keratoplasty, deep corneal vascularization, large-sized grafts, and a delay in the treatment of rejection were significant risk factors for non-resolution of the rejection process [Table 4]. Risk factors for non-resolution after graft rejection reported in the literature include patient age, donor age, the primary indication of penetrating keratoplasty, and previous episode graft rejection or failure.[11,13] Our study supports the results reported in these studies. Delay in treatment can be explained by the lack of transport facilities and movement restrictions during the COVID-19 lockdown. Earlier institution of treatment was reported to be associated with a better outcome.[12]

In our study, age and gender were not found to be significant risk factors. According to most studies, the young age of the recipient is associated with a higher rate of graft failure from rejection,[14,15,16] the sex of the recipient or donor and cause of donor death do not influence graft failure due to rejection.[14] Due to the small study population, gender predisposition for allograft rejection cannot be commented upon.

Corneal opacity due to the presumed infective etiology or trauma (n = 7, 23%) and infective keratitis (n = 7, 23%) were the two most frequent primary diagnoses in our study group, which presented with corneal graft rejection post-PKP. As also seen in our study, endothelial rejection is the most common type of corneal allograft rejection seen after PKP occurring in around 50% of cases.[2,17,18]

Indications for the primary surgery are significant predictors of reversal of allograft rejection. In our study, keratoconus, bullous keratopathy, and degenerations/dystrophies were associated with the reversal of the rejection process [Table 1]. In a multivariate analysis, a primary diagnosis other than keratoconus was found to be the strongest predictor for definitive graft rejection. Ten-year graft survival has been shown to be the best in grafts done for keratoconus (94%–100%),[19,20,21] whereas pseudophakic corneal edema and regrafts had less favorable 5-year survival rates (67.0% and 64.0%, respectively).[22] Infectious keratitis-related leukoma has been seen to have a relatively lower survival rate (around 49% 10-year survival).[19] Out of these, herpetic disease as an infectious etiology has a higher rate of failure because of the increased risk of neovascularization and also recurrence of the disease.[19,23]

Regrafts because of previous sensitization and deep vascularization are at higher risk of rejection. In our study, corneal opacities, regrafts, and active keratitis combined were responsible for 18 (60%) graft rejections [Table 1]. All these factors were responsible for 86% of non-reversal cases. The severity of deep corneal vascularization has been identified as a risk factor for graft rejection. Patients with two or more quadrants of deep corneal vascularization have been categorized as high risk for corneal transplant surgery. Studies suggest that host corneas can be divided into low, medium, and high risk on the basis of the number of quadrants of vascularization (avascular, 1–2 quadrants, and 3+ quadrants respectively).[24] According to the ACGR report, neovascularization in all four quadrants is associated with reduced graft survival (50%) at 4 years, whereas it was 73% in the absence of neovascularization.[25] The CCTS showed that neovascularization in all four quadrants was associated with double the risk of graft rejection.[14]

In our study, deep corneal vascularization was found to be a risk factor for non-reversal of graft rejection.

Large-sized grafts are at risk for immune reaction due to proximity to the limbal vessels and the antigenic load.[26] Regrafts having a larger diameter than previously failed grafts are also prone to graft rejection because of the same reason. Graft rejection in regrafts may be non-immunological[27] or immunological (due to neovascularization).[28] The host in the case of a regraft is sensitized. It has been shown that graft rejection in a regraft relates more to the number of blood vessels in the cornea than to the number of previous grafts.[28]

In our study, the type of surgery and additional surgical procedures were not found to be significant risk factors. The effect of lens status remains less clear. In the ACGR, the absence of the lens (aphakia) before or after full-thickness transplantation was associated with reduced graft survival,[25] whereas lens status was not found to be a significant risk factor for failure in the cornea donor study.[29]

Most patients presented early with signs of rejection (n = 21, 70%). Nine patients (30%) presented late, that is, more than 7 days after experiencing initial symptoms. The majority of patients were from within 25 km distance from the hospital (n = 15, 50%), six (20%) from 25 to 100 km distance, four (13.3%), and five (16.7%) from more than 200 km distance from the hospital. In our study, delayed reporting (P < 0.01) for treatment after the onset of graft rejection was a significant risk factor for non-resolution of the rejection process.

Patients diagnosed with corneal graft rejection were treated with either only topical (T) (n = 14, 46.7%), with T+ subconjunctival (Sc) (n = 7, 23.3%) and T+ Sc+ oral corticosteroids in 9 cases (30%) [Table 3]. Corticosteroids represent the principal medication in the management of corneal transplantation.[30] Prednisolone and dexamethasone are the most commonly used forms and these were the steroids used in this study. Loteprednol etabonate ophthalmic suspension was used by 9 to 11% of respondents at the 2004 Cornea Society meeting survey in low-risk grafts but rarely for high-risk grafts.[31] Furthermore, difluprednate, a novel strong synthetic steroid emulsion, has become widely accepted in the treatment of postoperative inflammation as well as in high-risk eyes.[32] High-risk corneal allografts need to be more intensively treated and for a longer duration, and the efficacy of prolonged use of topical steroids for the prevention of rejection after PKP has been shown.[33]

Early treatment was reported to be associated with a better outcome,[29] and longer rejection-treatment time has been shown to be associated with an increased risk of corneal decompensation post-rejection.[12] Our results support this finding. Figure 2 shows a patient presenting with acute graft rejection [Fig. 2a] and a patient under treatment with resolving rejection [Fig. 2b].

Figure 2.

(a) acute graft rejection, (b) resolving rejection

There are a number of studies on risk factors for corneal allograft rejection. The literature on the prognostic factors for the reversal of graft rejection is limited. The retrospective nature, small numbers, and limited follow-up after the initiation of treatment of graft rejection are limitations of the study.

Lockdown during the COVID-19 pandemic was imposed on March 25, 2020, throughout India, and this led to a significant delay in the follow-up and also the late presentation of graft rejection after penetrating keratoplasty. Also, remote locations and rural backgrounds can lead to failure in regular follow-ups and delayed presentation of corneal graft rejection.[34] The use of teleconsultation and telephonic contact to spread awareness among transplant recipients, especially during the COVID-19 pandemic has recently been mentioned.[35] This may help recognize the initial symptoms of rejection and may benefit patients through early treatment and increased chances of a reversal of the rejection process. Apprehension in the minds of patients to travel even after the lifting of travel restrictions in COVID-19 was also responsible for the delay in presentation. Reassuring patients of proper COVID-19 precautions in the hospital during their visit will also help in curbing the concern.

Conclusion

Our study demonstrates that the COVID-19 lockdown caused a significant delay in the presentation of patients, which had a significant impact on the non-resolution of immunological corneal graft rejection. Early treatment helps in the recovery of graft transparency and reversal of rejection, saving grafts from subsequent graft failure. Given the ever-increasing demand for corneal transplantation in the developing world, additional research in immune-mediated/non-immunological rejection mechanisms and their management is needed. Also, revised follow-up protocols, especially during pandemics such as COVID-19 should be a priority to improve risk prediction and outcomes in PKP.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.