Introduction

Allogeneic hematopoietic stem cell transplants (HSCTs) are a curative option for high-risk hematologic malignancies. The numbers of alternative donor transplants [unrelated donor transplants (UDTs) and haploidentical transplants (HITs)] have increased in recent years because of improved graft-versus-host disease (GVHD) prophylaxis and comparable outcomes with matched-sibling HSCTs^1,2. However, these transplants are limited by an increased incidence of GVHD, delayed immune reconstitution, and prolonged immunosuppression, leading to secondary infections^3–5. Viral reactivations pose a major threat in the early post-transplant phase as they may lead to organ involvement, increase the risk of developing GVHD, and cause poor graft function⁶. Data on the incidence and outcomes of viral reactivations in alternate donor transplants are limited, particularly in South Asian countries. We analyzed the incidence, risk factors, and outcomes of viral reactivations in alternative donor transplants in our HSCT unit to determine the exact burden of viral infections, their impact on survival in this subset of patients, and the utility of viral monitoring in these patients.

Materials and Methods

This was a retrospective study of viral reactivation profiles of all alternative donor HSCT recipients in a tertiary cancer care center in India between January 1, 2009 and February 29, 2020.

Transplant procedure

The conditioning regimen used was either cyclophosphamide (60 mg/kg × 2 days) along with myeloablative total body irradiation (> 8 Gy) or fludarabine-based reduced-intensity chemotherapy. Reduced intensity regimens used were fludarabine (150 mg/m²) in combination with either treosulfan (30-42 g/m²) or melphalan (140 mg/m²) or TBI (4 Gy to 8 Gy) or busulfan (9.6-12.8 mg/kg IV). All patients received T-cell replete grafts. Rabbit anti-thymocyte globulin (ATG) (2.5 mg/kg × 1 or 2 days) or alemtuzumab (Campath) was used in UDTs. Calcineurin inhibitors (CNIs), along with methotrexate, were used as prophylaxis for GVHD in UDTs. Post-transplant cyclophosphamide (50 mg/kg on days +3 and +4) along with CNIs and mycophenolate mofetil was used for HITs. Antifungal prophylaxis with posaconazole or voriconazole was administered to all patients. Acute GVHD was treated with methyl prednisolone (2 mg/kg) or an equivalent, and etanercept was the preferred second-line agent in steroid-refractory cases.

Virus monitoring

Cytomegalovirus (CMV) was monitored twice weekly from the start of conditioning until the patient was on immunosuppressants. After 2011, adenovirus (ADV), BK virus, and Epstein-Barr virus (EBV) were monitored weekly after engraftment until day 100. Virus monitoring was conducted using reverse transcription quantitative PCR (RT-qPCR) in peripheral blood for CMV and EBV and in peripheral blood and urine for ADV and BK virus. Stool was tested for CMV (quantitative) and ADV (qualitative) when clinically indicated. Until 2018, CMV reactivation was defined as the occurrence of two consecutive quantitative PCR readings > 500 copies/mL. Due to the observed high incidence of CMV infection in the haploidentical transplant (HIT) cohort, this threshold was revised post-2018 to a more sensitive criterion―two consecutive CMV PCR readings > 150 copies/mL. Resolution of CMV reactivation was defined as two consecutive undetectable CMV PCR results following the initiation of pre-emptive antiviral therapy. If CMV DNAemia reappeared above the defined threshold after achieving undetectable levels, it was classified as a second episode of CMV reactivation in the same patient. Colonoscopy with biopsy was performed when clinically indicated and technically feasible. A diagnosis of CMV colitis was established by histopathological evidence of CMV inclusion bodies or positive CMV PCR on colonic mucosal biopsies.

Parvovirus and human herpes virus-6 (HHV-6) were monitored when clinically indicated. ADV, EBV, BK virus, or parvovirus reactivation was defined as > 150 copies/mL in blood/urine/stool. Monitoring was continued until 3 weeks after discontinuation of immunosuppressants and restarted at the onset of acute or chronic GVHD or the use of systemic steroids. Hepatitis B virus (HBV) and hepatitis C virus (HCV) serology were tested pre-transplantation in all patients. Hepatitis serology and viral load were tested post-transplantation in case the patient developed transaminitis.

Viral prophylaxis and treatment

All patients received oral acyclovir (400 mg three times a day) prophylaxis against herpes and varicella virus from the start of conditioning until 1-year post-transplantation or for an extended duration in the case of herpes reactivation. Intravenous (i.v.) ganciclovir (5 mg/kg/dose twice a day, reduced to 2.5 mg/kg/dose if creatinine clearance < 60 mL/min) was used as a first-line pre-emptive agent for CMV reactivation. In case of grade 3 cytopenia precluding the use of ganciclovir or ganciclovir failure, artesunate or leflunomide was used. Foscarnet and cidofovir were used rarely because of unavailability. Asymptomatic adenoviremia and asymptomatic adenoviral shedding in urine were monitored, and treatment was initiated when adenoviral disease was probable or demonstrated. Cidofovir (5 mg/kg i.v.) was administered (whenever it could be procured) weekly for a maximum of 4 weeks, followed by every 2 weeks until resolution of adenoviral disease. Hemorrhagic cystitis due to BK virus reactivation was managed with supportive treatment including hyperhydration, analgesia, tapering of immunosuppression, and intravesical granulocyte colony-stimulating factor. Post-transplant lymphoproliferative disorder (PTLD) resulting from EBV reactivation was managed with weekly rituximab (375 mg/m²). Patients who had serologic evidence of past HBV infection [defined as positivity for anti-Hepatits B core (HBc) immunoglobin (Ig)M or IgG, or positivity for anti-Hepatitis B surface (HBs) antibodies in the absence of a history of vaccination] were started on prophylaxis with oral entecavir 0.5 mg or lamivudine 100 mg once daily pre-transplantation. The choice of entecavir versus lamivudine was based on HBV DNA (with lamivudine being administered to HBV DNA-negative patients and entecavir in case of HBV DNA-positivity with any titer). Entecavir or lamivudine for HBV reactivation was initiated in case of transaminitis with new-onset seropositivity (as defined previously⁷,) post-transplantation.

Viral DNA detection and quantitation

CMV DNA was isolated and identified using a CMV PCR assay used at our center, as described previously⁸. ADV, EBV, parvovirus, HHV-6, and BK virus DNA was extracted using Roche High Pure Viral RNA kit (Roche Diagnostics, Mannheim, Germany) and quantified using RT-qPCR in a CFX 96 thermal cycler (Bio-Rad Laboratories, Hercules, CA, USA) or a Qiagen Rotor-Gene Q5 real-time PCR system (QIAGEN GmbH, Hilden, Germany), as per the manufacturers’ instructions. The limit of detection was 10 to 5 × 10⁹ copies/mL for ADV, 50 to 5 × 10¹⁰ copies/mL for BK virus and parvovirus, and 25 to 2 × 10⁹ IU/μL for EBV. HHV-6 reactivation was diagnosed using qualitative PCR. Anti-HBs and Anti-HBc IgM/total were tested using a chemiluminescent microparticle immunoassay and enzyme-linked fluorescence assay, respectively, whereas HBV DNA was quantified using RT-qPCR, with a detection limit of 50 to 6.38 × 10⁸ copies/mL.

Statistical analysis

All data were entered in a Microsoft excel spreadsheet and statistically analyzed using SPSS version 21.0. The primary endpoint was the incidence of viral reactivation. The secondary endpoint was overall survival. Survival analysis was conducted using the Kaplan-Meier method, and survival rates were compared using the log-rank test. Clinical, transplant, and viral characteristics are represented as median with range, and proportions are expressed as percentages with 95% confidence intervals (CIs) where apt. The odds ratio (OR) was used to assess the strength of the association between various risk factors and viral reactivations. Institutional Ethics committee (IEC III, ACTREC, Tata Memorial Centre- Project # 901144) approved this study. Waiver of consent was approved as this was a retrospective analysis which involved only data collection and reporting of outcomes of standard of care treatment.

Results

In total, 703 transplants were conducted between January 1, 2009 and 29 February 29, 2020. Of these, 334 were allogeneic transplants, including 87 alternative donor transplants, encompassing 26 UDTs and 61 HITs. Among the 26 UDTs, 21 were fully human leukocyte antigen-matched transplants, three were mismatched (9/10) unrelated transplants, and two were 5/6 cord transplants. Peripheral blood stem cells were the predominant source (n=83; 96%), whereas bone marrow was the source in two patients and cord blood in another two patients. Fludarabine-based reduced-intensity conditioning was provided to 93% of patients (81/87), and cyclophosphamide and myeloablative total body irradiation conditioning to six patients. The baseline characteristics of the patients, their diseases, and transplant characteristics are provided in Table 1.

CMV reactivation

There were 92 episodes of CMV reactivation in 70 of the 87 transplants in total (incidence, 80.46%; 95% CI, 70%-88%). We identified 28 episodes in 21 UDTs, all of which were treated pre-emptively. Ganciclovir was used as the first-line agent for pre-emptive treatment in 20 (71%) episodes, artesunate in six episodes, and foscarnet and cidofovir in one episode each. Two patients with UDTs developed CMV colitis. Among HITs, we found 64 episodes of CMV reactivation in 49 patients. Ganciclovir was used as the first line in 39 episodes (61%), artesunate in four episodes (6%), foscarnet in two (3%) episodes, and leflunomide in 19 episodes (30%). CMV colitis was observed in four HIT patients, of whom three developed colitis within 100 days. Patients who had CMV reactivation were four times more likely to have acute GVHD during the reactivation period (OR, 4.32; 95% CI, 1.361-13.67; p=0.013). The details of CMV reactivation and risk factors according to multivariate analysis are shown in Table 2 and 3, respectively.

ADV reactivation

Twenty-one of eighty patients who were tested weekly for viral reactivations developed ADV reactivation (incidence, 26.25%; 95% CI, 17%-37%), including 12 of 25 UDT (48%) and 9 of 55 (16%) HIT patients. Four patients (5% of those tested) developed ADV disease and were treated with cidofovir: two had ADV colitis, one had ADV cystitis, and one had ADV colitis along with central nervous system involvement. Two of these four patients died with ADV disease. Patients who had ADV reactivation were thrice as likely to have received ATG compared to those who did not have reactivation (OR, 3.54; 95% CI, 1.24-10.06; p=0.03). Moreover, these patients were eight times more likely to have acute GVHD during the ADV reactivation episode (OR, 7.90; 95% CI, 2.49-25; p=0.0004). The characteristics of adenoviremia and adenoviruria are presented in Table 2, and the risk factors are shown in Table 3.

EBV reactivation

EBV reactivation occurred in 25 of 68 patients tested (incidence, 36.7%; 95% CI, 25%-49%), including 10 (50%) of 20 UDT and 15 (31%) of 48 HIT patients. The characteristics and risk factors are presented in Tables 2 and 3, respectively. Two patients developed PTLD requiring therapy with weekly rituximab, one of whom died due to uncontrolled PTLD. The remainder of the patients were asymptomatic and did not require any specific intervention, and EBV copies spontaneously declined after tapering or withdrawal of immunosuppressants.

BK virus reactivation

BK virus reactivation was observed in 59 of 71 patients tested (incidence, 83.1%; 95% CI, 72%-90%), including 19 of 26 (73%) UDT and 40 of 45 (88%) HIT patients. BK virus reactivation characteristics in blood and urine are presented in Table 2. Grade I/II hemorrhagic cystitis was observed in 7 recipients and grade III/IV in 8 recipients, as per the Droller grading system⁹. The remainder were asymptomatic with no evidence of microscopic or gross hematuria. All patients were treated with intravenous hydration. None received cidofovir for BK virus reactivation. Foley catheterization and bladder irrigation with cold saline were performed in three patients, and intravesical granulocyte colony-stimulating factor was administered to one patient. Presence of acute GVHD was associated with higher rate of hemorrhagic cystitis (OR, 5.29; 95% CI, 1.37-20.37; p=0.015) (Table 3).

Parvovirus reactivation

Two of the twelve patients tested developed parvovirus reactivation. One was a 9-year-old child with symptomatic anemia and a drop in hemoglobin from 8.6 g% to 4.7 g% over 15 days. The highest parvovirus load in peripheral blood was 3.0 × 10⁹ copies/mL on day +217 after transplantation. Hemoglobin improved to 7.4 g% within 2 weeks of initiation of i.v. immunoglobulin therapy. The second patient had asymptomatic parvo-viremia (3.9 × 10³ copies/mL), which resolved without any intervention.

HHV-6 reactivation

One of the fifteen patients tested had HHV-6 reactivation as indicated by qPCR on day +40. The patient died on day 57 after transplantation owing to a fungal brain abscess without clearance of HHV-6.

HBV reactivation

Thirty-one patients (35%) were started on lamivudine or entecavir prophylaxis because of positive HBV serology pre-transplantation. In total, 18 patients (20.68%; 95% CI, 12.75%-30.71%) underwent UDT or HIT and developed post-transplant HBV reactivation (new-onset positive serology with transaminitis), of whom eight developed reactivation within 100 days (Table 4). Of these 18 patients, only two tested HBV DNA-positive in peripheral blood. There were no cases of co-existing liver GVHD/chronic GVHD. Entecavir was started in eight and lamivudine in nine of these patients for post-transplant HBV reactivation and continued for 2 years post-transplantation. HCV reactivation was not observed in this cohort.

The number of patients with Parvovirus reactivations, HHV-6 reactivations and hepatitis B DNA positivity were too low to show statistically robust correlation with risk factors and hence these were not analyzed.

Outcomes

We identified a total of 196 viral reactivations (excluding herpes simplex and herpes zoster) in 87 patients who received alternative donor SCTs. The median overall survival for the entire cohort was 52.14 months (95% CI, 43.8-60.4 months). Despite the high incidence of CMV reactivation (80%), CMV end-organ involvement was observed in only four patients, indicating that pre-emptive treatment was largely successful, and there was no direct mortality because of CMV disease. Two patients and one patient died because of ADV disease and EBV-related PTLD, respectively. The cumulative incidence of major viral infections is shown in Figure. Viral reactivations had no statistically significant effect on overall survival in this cohort.

Discussion

We systematically monitored UDT and HIT patients between 2009 and 2020 for viral reactivations. Data on viral reactivations post alternate donor transplants from South Asian countries is sparse. The incidence of CMV reactivation was in our very high, around 80%, in both HITs and UDTs. This is much higher compared to incidence in western countries which ranges from 35-50%^10–12. The CMV reactivation rates are much higher in other Asian countries- and range from 55-60%¹³. The high reactivation rate in Indian patients is likely because of a high rate of donor/recipient (D+/R+) CMV seropositivity (88%) in our ethnic group, as described in the literature^10,11. Several other studies have reported a high risk of reactivation of 30%-71% in UDTs and HITs^12,14,15 because of delayed immune reconstitution and ex vivo T-cell depletion. In a study by George et al., fludarabine-based reduced-intensity conditioning was associated with a high risk of CMV reactivation^16,17. Our cohort predominantly received fludarabine-based reduced-intensity conditioning (n=81, 93.10%). The majority of reactivations occurred in the immediate post-engraftment phase, and acute GVHD occurred in 50 patients (57.47%), 90% of whom developed CMV reactivation. Steroid-refractory acute GVHD was observed in 15 of 50 (30%) patients, of whom 12 (80%) developed CMV reactivation. This is because steroid and immunosuppressive therapies further increase the risk and is concordant with previous studies in which acute GVHD, human leukocyte antigen disparity^6,18, showed a higher risk of CMV reactivation. Lymphopenia has been associated higher risk of CMV end organ disease and death in both allogeneic and autologous transplant setting^19,20. Despite the high rate of reactivation, CMV end-organ involvement was observed in only 8.6% of patients who experienced reactivation. There was no direct mortality attributed to CMV pneumonitis/hepatitis or enterocolitis. Active qPCR-based CMV surveillance twice a week and prompt initiation of pre-emptive anti-CMV therapy reduced the incidence of life-threatening CMV disease. However, secondary bacterial and fungal infections were a major cause of mortality in this cohort.

As primary ADV infection occurs in childhood, the risk of reactivation is generally increased in pediatric allogenic transplants, with an incidence of 2.5%-21%^21,22. The current study cohort comprised a predominantly young-adult population, with a median age of 25 years, and the cumulative incidence of ADV infection was 24.13% (in the entire cohort of 87 alternative donor transplants). Concomitant CMV reactivation was observed in 15 of 21 patients (71.4%). These findings are consistent with findings in several other studies in which alternative donor SCT, acute GVHD, prolonged immunosuppression, and an absolute lymphocyte count < 200/mm³ were significant risk factors for ADV reactivation^23,24.

The cumulative incidence of EBV reactivation was 28.73% in our entire cohort. Twelve of twenty-five (48%) patients who received ATG developed EBV-DNAemia. Reduced-intensity conditioning, in vivo/ex vivo T-cell depletion, cord and UDT, and acute grade II to IV GVHD are well-established risk factors for EBV-DNAemia and PTLD^25–28. Although the incidence of PTLD reportedly increases following UDT or HIT^29,30, we encountered PTLD in only two recipients. This may be because most HIT patients (61/87) received post-transplant cyclophosphamide rather than ATG in our cohort³⁰.

BK virus reactivation was the second most common viral reactivation after CMV reactivation, and the cumulative incidence was 59%, likely because of the high proportion of HITs³². The majority of patients with BK virus reactivation were asymptomatic (61%). According to the literature, routine BK virus monitoring as a screening tool for hemorrhagic cystitis has yielded variable results; therefore, its clinical utility has been questioned^33,34. Accordingly, the European conference on infections in leukemia (ECIL)-6 guidelines do not recommend routine BK virus monitoring in allo-HSCT recipients³⁵.

Studies have reported a high HHV-6 incidence of up to 30% in the pediatric population and of 7%-30% in adult transplant recipients, particularly in the early post-transplant phase, because of insufficient virus-specific T-cell immune responses^36,37. In our cohort, the incidence of reactivation was low: 1.1% for HHV-6 and 2.2% for parvovirus. This is because routine surveillance was not conducted for these viruses; monitoring was conducted only when clinically indicated. The incidence of HBV reactivation in our cohort was 20%, which is in line with findings by reported by Hammond et al.³⁸. Antiviral therapy for isolated anti-HBs antibody positivity was initiated in recipients without a history of vaccination and with deranged liver enzymes, although data to suggest treatment initiation are sparse in this setting⁷. The early initiation of antiviral therapy probably limited long-term complications due to HBV reactivation. To the best of our knowledge, the present dataset is one of the largest datasets on viral reactivations in an alternate donor transplant setting in the Indian population. Based on our study, the regular monitoring of ADV, BK virus, and EBV may have limited utility due to the low incidence of symptomatic disease requiring therapeutic intervention. Cytomegalovirus and hepatitis B are the important viruses of concern. This is a notable finding, and resources need to be focused on the detection these viruses, particularly in countries associated with financial constraints, such as India and other South Asian countries, rather than on the regular monitoring of all infectious viruses. This study was limited by its retrospective nature, heterogeneity of the study population in terms of the diagnosis of underlying hematological malignancies, and the long study period, which may have resulted in changes in the standards of monitoring and treatment of viral reactivations during the study period. We also acknowledge that the age group of the cohort varied from 5 years to 53 years, and the reactivation rates will also be affected by the age group of transplanted patients. Nevertheless, our data provide useful insights into incidence and burden of viral reactivation, viral reactivation patterns in alternative donor transplants in the south Asian context.

In conclusion, the incidence of CMV reactivation in alternative donor HSCTs is high (80%) in the Indian population, and acute GVHD is a significant risk factor for CMV, ADV, or EBV reactivation. ATG use is a significant risk factor for ADV reactivation. The rates of ADV, BK virus, and EBV reactivation requiring intervention are low. Better prophylactic strategies are required to prevent CMV reactivation in alternative donor transplants.

Author Contributions

AG and VM: data collection, analysis, and writing of the draft; SP, SM, AC, NJ, AB, LR, and BB: patient care and monitoring; VB, SB, RT, and SKumar: laboratory work and viral PCRs; SKanan: statistical inputs; GB: reviewing and editing the manuscript; NK: concept and overall study mentorship; all authors: input on manuscript and editing. AK and VM contributed equally to this work.

Conflicts of Interest

The author declares no conflict of interest. Disclosure form provided by the author are available on the website. NK is one of the Editors of Blood Cell Therapy. He was not involved in the editorial evaluation or decision to accept this article for publication.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author on reasonable request.