EXPERIMENTAL AND CLINICAL TRANSPLANTATION

The impact of coronavirus disease (COVID-19) in solid-organ transplant (SOT) has so far been well described since the start of the pandemic.¹ Mortality rates among SOT patients with COVID-19 are reportedly high around the world.² However, comparisons of COVID-19 outcomes between SOT and non-SOT have yielded conflicting reports, as there is a divergence in conclusions among various published reports. Some comparisons have been made, mostly in nations from American and European regions. However, no such comparisons have been made in South East Asian regions. So far, severe acute respiratory syndrome coronavirus (SARS-CoV-2) has shown a wide diversity in clinical presentations and mortalities among different areas of the world³ and in different populations.^4,5 Hence, we conducted a single-center retrospective study in which we compared all admitted kidney transplant recipients (KTRs) with a non-KTR group during the first wave of the COVID-19 pandemic in India. This study should shed further light in understanding any associations between severity or mortality in COVID-19 and organ transplant. This will help in further stratification and triage of the most vulnerable groups in case of potential future waves.

Materials and Methods

Study design and settings
We conducted a retrospective cohort study of the data from the COVID-19 unit of the Institute of Kidney Diseases and Research Centre, Institute of Transplantation Sciences (Gujarat, Ahmedabad, India). This institute is located in the Western region of India and has the largest public sector transplant program. Here, more than 6000 renal transplants, including >1000 from deceased donations, have been performed. To meet the health care crisis during the pandemic, the institute was converted to a dedicated COVID-19 unit, halting all transplant activities. Outcomes of KTRs from this center have been previously published.⁶

Patient identification
All patients diagnosed or admitted as probable or confirmed COVID-19 in the dedicated COVID-19 ward were included in the study. The index case was on May 12, 2020, and the last case was January 11, 2021. For final analysis, 1627 registered cases were included, as 150 entries were removed for being duplicates, which was mainly because patients on dialysis had history of repeated registration for receiving dialysis sessions in the COVID-19 isolation ward. The patients who were given home isolation (n = 158) were also excluded from the study. Patients in the non-KTR group were followed until discharge from the hospital or death. Patients in the non-KTR group who were lost to follow-up (n = 17), left against medical advice (n = 10), and transferred to another hospital (n = 36) were also included in the outcome measurements. These patients were censored during analysis. The non-KTR group (n = 1448) included all patients other than KTRs who were registered in the hospital during the pandemic. Among this group, 804 had no comorbidities, 473 had hypertension, 283 had diabetes, 83 had cardiovascular diseases, 27 had a cerebrovascular accident, 46 had chronic lung disease, and 30 had chronic liver disease.

Outcome measurement
The primary outcome was to compare the in-hospital mortality rates of COVID-19 in KTRs with rates in a non-KTR group. The secondary outcome was to compare the severity of COVID-19 between the groups. Other outcome measures were differences in the predictive risk factors for mortality in the study.

Statistical analyses
We used SPSS version 25 for statistical analyses. A double-sided P < .05 was considered a priori for statistically significant difference in our study. All categorical values are expressed as numbers and percentage. All continuous data were checked for normality distribution through Q-Q plot, Shapiro Wilkins, and Kolmogorov-Smirnov test and thereafter expressed as mean (standard deviation) or median (interquartile range) as appropriate. Our univariate analysis was done with the chi-square test, the Fisher test, the t test, or the Mann-Whitney test depending on the type of data and the sample size. A propensity matched cohort was computed to eliminate the Berkson bias occurring from differences in the baseline characteristics of the 2 groups. The covariates entered in the match run included all baseline characteristics from Table 1. For calculating the propensity scores of the individual variables, a logistic regression was done with the nearest neighbor algorithm. Matching was done at a 1:3 ratio with caliper cutoff of 0.2. The method for matching was kept as random so as to remove any bias from the nearest neighbor method. No replacements were taken. Standard differences were computed to check for any residual imbalances in the matched cohort. There was no covariate with imbalances in the matched cohort. Time-to-event analyses for death and severe COVID-19 were described with Kaplan-Meier plots. A Cox proportional ratio with backward stepwise algorithm was fitted to find individual risk factors for mortality in the whole cohort. In the model, the variables with small sample sizes were removed. The strength of association was recorded as hazard ratio (HR) accompanying upper and lower bound of confidence interval (CI).

Table 1 shows differences in baseline features between the study groups and results after propensity matching. Overall, as seen from age distribution, the non-KTR group was clearly older than the KTR group (P < .01). There were more patients in the non-KTR group with baseline Eastern Cooperative Oncology Group (ECOG) score of 3 than in the KTR group (363 [25.1%] vs 14 [7.8%] patients; P < .01). Comorbidities in the non-KTR group also included 3 patients with solid-organ malignancy, 9 patients with hypothyroidism, 3 patients with orthopedic problems, 2 patients with gout, and 1 patient with chronic pancreatitis. There were smaller rates of patients with hypertension (473 [32.7%] vs 118 [65.9%]; P < .01) and cardiac diseases (83 [5.7%] vs 32 [17.9]; P < .01) in the non-KTR group versus the KTR group. After propensity matching, baseline characteristics of the KTR (n = 98) and non-KTR group (n = 167) were similar.

Table 2 shows the Cox proportional hazard model for predictors of mortality in our cohort. Being in the 61- to 70-year-old age group (HR = 1.5; 95% CI, 1.0-2.2; P = .04), the 71- to 80-year-old age group (HR = 1.64; 95% CI, 1.0-2.5; P = 0.02), and the >80-year-old age group (HR = 1.91; 95% CI, 1.1-3.1; P = .01) were associated with statistically significant incidence of mortality. There was no association with sex. Among comorbidities, KTR status (HR = 0.43; 95% CI, 0.26-0.71; P = .001) was not associated with mortality; rather, it showed a negative association with mortality. This can be explained by the younger age of KTRs: our analysis showed that age <60 years was not associated with mortality and the median age of KTRs was 44 years. The most important risk factor for mortality was age, even surpassing comorbid conditions. The admission ECOG score of 2 (HR = 36; 95% CI, 10.8-124; P < .01) or ?3 (HR = 58.6; 95% CI, 17.2-199; P < .01) was significantly associated with mortality. Of 436 deaths, 416 (28.7%) were in the non-KTR group and 20 (11.1%) were in the KTR group. Figure 1, A and B, shows the Kaplan Meier plot for mortality and severity of COVID-19 between the matched KTR and non-KTR groups. There was no difference in mortality between the 2 groups (log-rank, Mantel Cox test; P = .07).

Table 3 shows the management protocol and treatment regimen used in the study patients. The immunosuppression protocol in the KTR group included stopping (20%) or tailoring of antimetabolites (80%) and stopping or tailoring of calcineurin inhibitors (19%). Tacrolimus levels were normal in most patients (94%). The number of graft losses was 12 within a median follow-up of 3 months. Patients with baseline serum creatinine above 1.5 mg/dL constituted the graft losses (n = 1) in our report. When we compared mortality and severity in the KTR group (n = 179) versus patients on dialysis, severity and mortality were significantly higher in the dialysis group (Figure 1, C and D).

Figure 2 illustrates prognosis in terms of mortality with different induction agents (thymoglobulin and interleukin 2 blockers) and maintenance immuno-suppression (calcineurin inhibitor, antimetabolite, mechanistic target of rapamycin inhibitors). There were no statistical differences with regard to any of the regimens. History of recent antirejection therapy was also not associated with morality. However, baseline low-ejection fraction measured by echocardiography was associated with mortality.

COVID-19 outcomes in SOT patients have been well documented in previous studies.^7-9 The outcomes of COVID-19 in all types of organ transplant recipients, including liver,¹⁰ kidney, heart,¹¹ and lung,¹² have shown concerning results. Reports have shown higher chances of morbidity and mortality in SOT recipients. As shown in a US cohort,¹³ there were increased odds for mortality, invasive ventilation, and acute kidney injury in SOT compared with non-SOT patients. In a preliminary study from Spain,¹⁴ outcomes of COVID-19 in SOT recipients were more unfavorable compared with outcomes shown in general patients, although the difference was not statistically significant. In a Swedish report,¹⁵ in which lung transplant recipients were compared with general patients, higher mortality was shown in the transplant group. In a recent meta-analysis¹⁶ that included 265?839 participants,1485 of which were SOT recipients, there were increased odds for need of intensive care admission and mortality.

In contrast, high-level data have also demonstrated similar outcomes in SOT versus general patients. The initial such report from Spain¹⁷ showed similar outcomes in SOT recipients compared with the general population. A recently published US report¹⁸ showed a similar length of hospital stay and mortality and severity rates in SOT recipients compared with non-SOT patients. However, they also reported that declines in severity were more prominent for SOT recipients. In another large study from the USA, in which COVID-19 outcomes in 2307 SOT recipients were compared with 231?047 general patients after propensity matched scoring, there was no difference in 30-day or 60-day mortality rates. However, SOT recipients were shown to be more prone to developing acute kidney injury. In an Italian study,¹⁹ there was no difference in 30-day mortality between a SOT and a general population group.

The rationale behind our research was the lack of any data from this region of the world. The primary goal of conducting this study was to compare outcomes of a KTR cohort versus a non-KTR cohort in our center, which is located in India. Our center was a referral center for COVID-19 cases during the COVID-19 surge in the city and was converted to a COVID-19 center, thus ceasing all transplant activities. The comprehensive clinical, laboratory, and treatment profiles of KTRs have been published previously in our other reports.20-22 In our study, we found that there was no difference in mortality in a matched KTR versus non-KTR group. The overall mortality rate in the KTR group was 11.1% compared with 28.7% in the non-KTR group. In previous reports from other parts of the nation,23,24 the mortality rate among hospitalized general patients was less than that shown in our report. A possible explanation may be because the hospitalized mortality was in a referral center, which was working to cope with the overflow of cases in nearby COVID-19 centers. Another important observation from our report is the high odds ratio of mortality when ECOG score was high. Thus, a late referral or initiation of treatment could have significantly decreased the mortality in patients with COVID-19. Hence, the mortality cannot be represented to general population.

In our secondary analysis, we found that baseline graft dysfunction was associated with graft loss, and also baseline low ejection fraction was associated with mortality. There was no difference in prognosis of COVID-19 when we examined immunosuppres-sion used by patients in the study. In comparison, patients on wait lists or on dialysis had higher severity and mortality with COVID-19. These observations in KTRs are similar to those shown in earlier reports.^20,22

Limitations
A limitation of this study includes the relatively small sample size of our KTR group. Because of logistics involved and the overburdened health care infrastructure amid the COVID-19 pandemic, there may also be information bias for patients in the non-KTR group regarding preexisting comorbidities and ongoing medicines. Selection bias of more sick patients may have also occurred. Our KTR cohort was relatively younger compared with other reports, which partly explains the decreased in-hospital mortality in the KTR group.

Conclusions
Older age was the most important risk factor for in-hospital deaths in those with COVID-19. Kidney transplant status may not be independently associated with mortality. Further research, in different timelines of the pandemic and in different geographic areas, will strengthen our findings.

Volume: 19 Issue: 12 December 2021

FULL TEXT