Introduction

Allogeneic stem cell transplantation (alloSCT) is a potentially curative treatment of a wide variety of malignant and nonmalignant disorders of hematopoiesis. Traditionally, the best outcomes of alloSCT have been reported when the donor is HLA matched.1 As a matter of fact, each patient has only a 25% chance of being HLA matched with a sibling and with the small-size family concept now seen in many countries, patients have only about one-third chance of having an HLA-matched donor, which demands the need for alternate donors. Although, according to several studies, the outcomes with matched unrelated donor (MUD) transplants rival those with matched sibling donor (MSD) transplant, cost associated with MUD transplants are huge, which is a major limiting factor for resource-restricted countries.2 With the emergence of haploSCT in the last decade, our wisdom on its shortcomings has expanded exponentially. An HLA-haploidentical donor is a related donor who shares exactly one HLA haplotype and differs by a variable number of HLA genes on the unshared haplotype, and the greatest edge is that it is almost always available for every patient. The Beijing Protocol was shown to be a reliable treatment strategy for haploidentical stem cell transplantation (haploSCT) patients.3 The most commonly used and known John Hopkins Protocol incorporating posttransplant cyclophosphamide (PTCy) had shown that high-dose Cy substantially alleviates alloreactivity to the extent that outcomes are constantly improving.4 However, until recently, various studies have proclaimed that haploSCT may not replace MSD transplants, but may be preferred to MUD transplants. Even so, future prospective comparisons with matched donors would be desirable.56 Nevertheless, it needs to be acknowledged that there is a discrepancy in the haploSCT outcomes when compared between developed countries and resource-limited countries. There is scarcity of data pertaining to haploSCT being performed in India. A study from South India had reported their experience of haploSCT in primary immune deficiency disorders in the pediatric population.7 More recently, in regard to hematological malignancies, a study on 149 patients (including 158 haploidentical transplants) stated that haploSCTs are feasible to be performed, but infectious complications are the major challenge.8 Also, not to forget that graft failure, nonrelapse mortality (NRM), GVHD, delayed immune reconstitution, and relapse remain significant concerns associated with haploSCTs.9 One of the additional challenges in haploidentical donor selection could be the presence of donor-specific antibodies (DSA) in recipients, which mandates cumbersome desensitization techniques.1011

We are, therefore, presenting our data of over 7 years on the challenges and outcomes of haploSCT performed with PTCy in patients with hematological disorders.

Patients and Methods

We performed a retrospective study in our bone marrow transplantation unit (BMTU) where all the patients with hematological disorders (benign and malignant) who underwent haploSCT with PTCy between 2015 and 2022 were included. All the patients were screened for anti-HLA antibodies, and only those donors who were nonreactive to the patient were taken as donors. PTCy was administered at a dose of 50 mg/kg/d on D + 3 and D + 4 of infusion of T-cell replete graft. All the patients received mesna in adequate doses along with forced alkaline diuresis. Mycophenolate mofetil (MMF) and calcineurin inhibitors (CNIs; cyclosporine or tacrolimus) were started on day +5. All patients received granulocyte colony-stimulating factor (G-CSF) from day +5 onward. Antifungal prophylaxis with posaconazole or voriconazole or echinocandins was administered as per the clinician's decision after considering previous history of fungal pneumonia and drug interactions with the conditioning regimen. Antiviral prophylaxis with acyclovir or valacyclovir was administered. Pneumocystis prophylaxis with co-trimoxazole was started once the stable engraftment was achieved. Chimerism was tested on days +30, +60, +100, 6 months, and 1-year posttransplant. In case of dropping chimerism, testing was done more frequently. Patients were followed up in the transplant clinic for monitoring of any symptoms and signs of GVHD, tapering of immunosuppression in case of no GVHD, and timely monitoring of disease status.

Data were collected from the medical records and the hospital information system. Data on the baseline demographics of the patients, donor characteristics, engraftment rates, peritransplant complications, graft versus host disease (GVHD), veno-occlusive disease (VOD), infections, relapse, NRM, and overall survival (OS) were captured and analyzed. The study was approved by the institutional review board and the ethical committee.

Statistics

All continuous variables with normal distribution were represented by mean. Non-normally distributed continuous variables were calculated as median. Qualitative variables were represented by percentage. Descriptive statistics was used for the calculation of median with range. Survival analysis was done with the Kaplan–Meier curve. Data analysis was carried out by SPSS software. A p value of <0.05 was considered statistically significant.

Results

Transplant Outcomes

Forty-eight (48.4%) patients had died at the last follow-up (Table 2). Deaths at less than 30 days, between 30 and 100 days, and beyond 100 days were 18 (37.5%), 11 (22.9%), and 19 (39.5%), respectively. Causes of death were sepsis (27 [56.2%]), relapse (10 [22.2%]), GVHD (4 [8.8%]), VOD (2 [4.4%]), fungal pneumonia (2 [4.4%]), primary graft rejection (2 [4.4%]), and HHV6 encephalitis (1 [2.2%]). Thirty-seven (37.3%) of 99 transplants (who had completed at least 100 days of follow-up) had NRM. NRM at less than 30 days, between 30 and100 days, and beyond 100 days was 18 (48.6%), 8 (21.6%), and 11 (29.7%), respectively. Eighteen (17.8%) patients relapsed (medullary = 16, CNS = 2) and the median days at relapse was 253 (60–835) days. Five (4.9%) patients received donor lymphocyte infusion (DLI) for relapse. Two (18.1%) patients had secondary graft rejection. After a median follow-up of 6 (0.5–89.5) months, 42 (42.4%) patients were alive in remission and 1 (1%) patient was alive in relapse. Eight (8%) patients (after relapse = 7, in CR = 1) were lost to follow-up. The median OS of the entire cohort is 18 ± 11.46 (0–40.77) months. One-year OS for the entire cohort is 56.7%, while the 2-year OS is 49.3%. The estimated 1-year progression-free survival (PFS) and event-free survival (EFS) for the cohort were 76.70 and 48.50%, respectively (Fig. 1). There was no statistically significant difference in PFS (p = 0.796) and OS (p = 0.988) in patients with pretransplant CR1 versus CR2 versus CR > 2 status.

Close

Fig. 1: (a) Overall survival (OS) of the entire cohort (1-year OS: 56.7%). (b) Progression-free survival of the entire cohort (1-year OS: 76.7%). (c) Event survival of the entire cohort (1-year OS: 48.5%).

Export to PPT

Discussion

Compelled by the lack of universal availability of donors for many patients, the option of haploSCT was explored. In the beginning, haploSCT was associated with inferior outcomes due to high probability of GVHD and graft rejection leading to compromised OS. With the introduction of PTCy strategy, which results in in vivo selective T-cell depletion, these problems have been taken care of to a large extent.13 In particular, haploSCT for hematological malignancies remains a challenge because of the more aggressive nature of the disease, accumulation of comorbidities due to prior therapies, and significant posttransplant complications. In addition, it is a difficult decision for clinicians in case of older individuals where most patients have at least a child available as donor, making haploSCT a readily available option for them. Additionally, there are challenges with haploSCT due to HLA disparity, which leads to delayed immune reconstitution.1415 The above-mentioned approach of PTCy has revolutionized the use of haploSCT in United States and many European countries.1617 Even developing countries are capable of doing haploSCT with this strategy as it overcomes the need for graft manipulation, which demands special expertise and cell separation devices. Various studies from developed countries have compared the outcome of haploSCT with MSD and MUD and had shown outcomes equivalent to MUD or even MSD.1819202122 There is a scarcity of literature from developing countries; hence, demonstrating similar outcomes is a challenge.

In India, a study by Jaiswal et al conducted on 40 haploidentical transplants involving both malignant and benign disorders over 3 years showed NRM at day + 100 of 15% and OS at 2 years of 55%. The major cause of NRM mentioned was infection with CR gram-negative bacilli accounting for 90% of the deaths.23 Another study from India, although done on primary immunodeficiency patients only (n =16), had shown 37.5% mortality with OS of 62.5% with a median follow-up of 23.3 months.7 In another study by George et al where 149 patients underwent 158 haploidentical transplants for both malignant (n = 85) and nonmalignant (n = 73) indications, the 1-year OS was 44% and 2-year OS was 39%.8 In comparison, a study by Bacigalupo et al on 459 consecutive patients with hematologic malignancies elicited a 4-year survival of 45% in the MSD group and 52% in the haploSCT group.24 In another study conducted in China, the OS and disease-free survival and relapse-free survival after MSD were significantly superior to those after haploSCT, although there was no significant difference in the transplant-related mortality (TRM) rate between the two.5

In the current retrospective study, 99 patients underwent haploidentical transplants for both malignant (95 [95.9%]) and nonmalignant (4 [4%]) indications with conditioning received as MA (34.6%), NMA (42.5%), and RIC (22.7%). Primary graft rejection was seen in nine (8.9%) patients, which is almost similar to the rate of 8% reported by George et al.8 In our cohort, acute and chronic GVHD were seen in 33 (32.6%) and 12 (11.8%) patients, with grade 3/5 acute GVHD seen in 11 (10.8%) and severe chronic GVHD in 3 (2.9%) cases. Our results are similar in the rates of development of acute GVHD (grade 3/4) in a study by Jaiswal et al,23 reported as 9.3%. Similarly, chronic GVHD was seen in 10% patients in their study, which coincides with our results. George et al had also reported acute GVHD in 32% patients.8 Comparably, Bacigalupo et al had showed an acute GVHD (grade 3/4) rate of 14%, with a trend toward less moderate to severe chronic GVHD in their haploSCT group.24 Relapses were seen in 18 (17.8%) patients in our cohort. In the above-mentioned study by Bacigalupo et al, cumulative incidence of relapse was 35% in the haploSCT group.24

In our experience, infections were a major problem with bacteremia seen in 42 (41.5%) cases (out of these, 83% were gram negative) and viral infections in 19 (18.8%) cases and the commonest virus was BK (8.9%). In another Indian study, gram-negative bacteremia has been reported as 41% and viral infections (CMV and BK) were seen in 68% patients.8 In our cohort, viral infections including BK and CMV were seen in 76% patients. The major cause of NRM was infection (56.2%). Similarly, George et al had realized that infectious complications were the major challenge with haploSCT.8 In contrast, Bacigalupo et al had reported infection-related mortality as low as 11%.24 In our study, the overall TRM was 37.3% and TRM at 100 days was 26.2%. The 1-year OS for the entire cohort was 56.7%, while the 2-year OS was 49.3%. One-year PFS and EFS were 76.7 and 48.5%, respectively. In the study by George et al, 1-year OS was 44.1% and 2-year OS was 39.2%.8 The Perugia group had reported 40% NRM with common causes being CMV and aspergillus.2526 The Durham group achieved 10.2% TRM at day 100 and 40% at 2 years of follow-up.27 Other significant posttransplant complications observed were CRS (88%; in contrast to another Indian study that reported CRS in 75% patients23), VOD (4.9%), TMA (16.8%), HC (6.9%), and posttransplant HLH (1.9%). Although neither our study nor any other Indian studies have made a comparison between haploidentical, MSD, and MUD transplants, there have been reports of comparison among the three groups in Western literature. A recent meta-analysis including 30 studies elicited that haploSCTs were associated with increased all-cause mortality and NRM compared with MSD, but similar in comparison to MUD.6

We acknowledge that there are shortcomings in our study, as the follow-up period was short. Long-term follow-up is required to reflect again on our results in the near future. In addition, we could not include data on posttransplant immune reconstitution due to unavailability of data.

Conclusion

HaploSCTs are an attractive option in alternate donor transplants. However, these are still associated with significant all-cause mortality and NRM with substantial GVHD rates. We are aiming to reduce our NRM rates but still facing the demons of resistant infections. Real-world experience from developing countries is still striving to achieve the optimistic potential highlighted in the Western literature.