The Toll of Prophylaxis: Cytomegalovirus Infection in the Haploidentical Donor and Post-transplant Cyclophosphamide Era

Originally observed as a “peculiar interstitial pneumonia” seen in recipients of allogeneic hematopoietic stem cell transplantation (HSCT) and a major cause of death post-transplantation, human cytomegalovirus (CMV) infections have hindered the success of HSCT since the inception of the technique in the 1970s.¹  Despite advances in antiviral therapy, improved understanding of host immunity to CMV and the development of superior CMV diagnostic methods during the past 50 years, CMV infection continues to be one of the most common and life-threatening infectious complications post-HSCT and is strongly associated with overall higher nonrelapse mortality and lower overall survival (OS).^2,3  As the use of human leukocyte antigen (HLA) -haploidentical hematopoietic transplantation with post-transplant cyclophosphamide (HaploCy) as the backbone of graft-versus-host disease (GVHD) prophylaxis has expanded worldwide and now accounts for 16 percent of all HSCT,^4-6  several reports from individual transplant centers have suggested that HaploCy is associated with increased incidence of CMV infection as well as an earlier onset of CMV in the post-transplant course relative to patients receiving HLA-matched HSCT.^7-10  No large data set to date has examined the association between HaploCy and CMV infection. Moreover, it is unclear as to whether the donor source (HLA-haplo) or the PTCy-based GVHD prophylaxis is an increased risk factor for CMV infection.

In the current report, Dr. Scott Goldsmith and colleagues and the Center for International Blood and Marrow Transplant Research (CIBMTR) have tried to determine whether haploidentical donor source and/or PTCy-based prophylaxis increase the risk of CMV infection after HSCT by comparing outcomes to a control cohort of HLA-matched sibling donor transplant recipients treated with calcineurin-based GVHD prophylaxis. Data were reported to the CIBMTR between 2012 and 2017 for patients with acute myeloid leukemia, acute lymphoblastic leukemia, and myelodysplastic syndrome (MDS) who received HaploCy (n = 757), HLA-matched related sibling (Sib) with PTCy (SibCy; n = 403), and HLA-matched related siblings with calcineurin inhibitor–based GVHD prophylaxis (SibCNI, n = 1,605). CMV infection was defined as blood stream only (DNAemia) and/or tissue-invasive (end-organ) disease. The term “CMV reactivation” was not used in the study. Three major analyses were performed in this retrospective registry study.

First, the authors examined the incidence of CMV DNAemia and CMV disease at transplant day 180 in the three transplant cohorts (HaploCy, SibCy, and SibCNI). The incidence of CMV DNAemia was significantly higher by day 180 in patients receiving PTCy (HaploCy, 42% [99% CI, 37-46] and SibCy, 37% [99% CI, 31-43]) compared to patients receiving SibCNI (23% [99% CI, 20-26], p<.001). The median time to CMV reactivation was faster in recipients of PTCy (HaploCy, 38 days [range, 2-176 days], SibCy, 32 days [range, 5-136 days]) compared to SibCNI (42 days [range, 4-176 days], p<.001). The incidence of CMV organ disease was low for all patients; the cumulative incidence of CMV by day 100 was 2.8 percent (HaploCy), 3.4 percent (SibCy), and 1.4 percent (SibCNI). The incidence increased modestly by day 180, but remained nonsignificant between cohorts (p=.115).

In the second analysis — the “CMV serostatus analysis” — the three transplant cohorts (above) were divided further into three subpopulations based on donor/recipient (D/R) CMV serostatus and classified as follows: CMV-seropositive recipient (R⁺), CMV seropositive donor (D⁺/R^–), and CMV seronegative recipient and donor (D^–/R^–). The principal outcome of interest was CMV infection (DNAemia and/or organ disease). As expected, the incidence of CMV was highest for CMV seropositive recipients (R⁺); the day 100 cumulative incidence of CMV DNAemia among R⁺ patients was 51 percent (99% CI, 46-57) for HaploCy, 48 percent (99% CI, 41-56) for SibCy, and 29 percent for SibCNI. Donor CMV seropositivity (D⁺/R^–) was also associated with increased incidence of CMV DNAemia at day 100 (HaploCy, 21% [99% CI, 8-37]; SibCy, 12% [99% CI, 2-30]; and SibCNI, 11% [99% CI, 6-18]). The day 100 cumulative incidence of CMV DNAemia among D^–/R^– patients was low: 2.4 percent (99% CI, 0.2-7.1) for HaploCy, 2.6 percent (99% CI, 0-9.2) for SibCy, and 1.3 percent (99% CI, 0.2-3.4) for SibCNI. In multivariate analysis only CMV serostatus influenced CMV infection. The R⁺ patients (R⁺ HaploCy, R⁺ SibCy, and R⁺ SibCNI) had significantly higher risk of CMV infection as compared to the reference cohort of seronegative Sib HSCT with conventional GVHD prophylaxis (D^–/R^– SibCNI) with hazard ratios (HRs) of 50 (99% CI, 14-181), 48 (99% CI, 13-177), and 24 (99% CI, 7-82) among R⁺ HaploCy, R⁺ SibCy, and R⁺ SibCNI, respectively (p<.0001 for all comparisons). Significantly, among R⁺ subgroups, the use of PTCy for GVHD prophylaxis doubled the risk for CMV infections, and donor source (Haplo or matched sibling donor) did not impact the risk (R⁺ HaploCy vs. R⁺ SibCNI, HR 2.1 [99% CI, 1.4-3.1]; p<.0001; R⁺ SibCy vs. R⁺ SibCNI, HR 2.0 [99% CI, 1.3-3.1]; p=.0001).

The third analysis examined the effect of CMV DNAemia and CMV serostatus on transplant related outcomes that included OS, disease-free survival, cumulative incidences of relapse, non-relapse mortality (NRM), and GVHD by two years. Univariate analysis revealed that donor or recipient positive CMV serostatus was associated with inferior NRM at 100 days, one year, and two years. In multivariate analysis, CMV seropositive recipients (R⁺) in the HaploCy cohort had significantly higher risk of NRM and inferior OS compared to R⁺ SibCNI (HR, 1.4 [99% CI, 1.1-1.7], p=.0001). Additionally, the R⁺ HaploCy cohort had significantly inferior OS compared to D^–/R^– HaploCy cohort (HR, 1.6 [99% CI, 1.1-2.5]; p<.004).

In univariate analysis, there was no difference in the development of grade 2 to 4 acute GVHD among the three transplant cohorts or according to CMV serostatus. The incidence of chronic GVHD at six months was similar across the cohorts and CMV serostatus. The incidence of chronic GVHD at two years was lower in R⁺ HaploCy (HR, 0.68 [99% CI, 0.47-0.98]; p=.0064) and R⁺ SibCy (HR, 0.56 [99% CI, 0.39-0.81], p=.0001) compared to D^–/R^– SibCNI. However, only the R⁺ SibCy cohort had a significantly lower risk of chronic GVHD compared to R⁺ SibCNI. In patients without CMV DNAemia, HaploCy was associated with lower incidence of chronic GVHD (HR, 0.60 [99% CI, 0.42-0.85], p=.0002), and a non-significant trend was seen in patients with SibCy. Among patients with HaploCy, there was a higher incidence of chronic GVHD in those who developed CMV DNAemia as compared to those without CMV DNAemia (HR, 1.6 [99% CI, 1.0-2.3]; p=.006). The incidence of chronic GVHD for HaploCy patients who developed CMV was similar to the SibCNI cohort. Finally, CMV serostatus did not influence the incidence of relapse at two years in any of the transplant cohorts.