Risk Factors and Different Therapeutic Patterns of Late-Onset Hemorrhagic Cystitis after Haploidentical Allogeneic Stem Cell Transplantation

Introduction

Late-onset hemorrhagic cystitis (LOHC) is a common complication after allogeneic hematopoietic stem cell transplantation (allo-HSCT) with an incidence ranging from 6.5% to 70% and leads to prolonged hospitalization and even death (Silva Lde P et al, Haematologica, 2010; LAM et al, Transpl Infect Dis, 2017). The pathogenesis of LOHC remains obscure, but in our previous study, viral infections and acute graft-versus-host disease (aGVHD) have frequently been shown to be associated with its development (Han et al., Am. J. Hematol, 2014). Therapeutic strategies for LOHC are still not standardized due to the complicated clinical background and given the paradoxical phenomenon that immunosuppressive agents, such as steroids, are necessary for aGVHD but cause an immunosuppressive state. We aimed to investigate the incidence, risk factors and outcomes of LOHC after allogeneic stem cell transplantation. Additionally, for the first time, we propose four therapeutic patterns and their impact on prognosis of LOHC to aid in clinical decision making.

Methods

This retrospective, nested, case-control study reviewed data from 2158 patients who received allo-HSCT from January 2014 to December 2018. In total, 364 (16.87%) patients were diagnosed with LOHC. Individual matching was performed by randomly selecting 3 controls from the same cohort for each identified LOHC patient according to the time of allo-HSCT. Standard basic measures included hyperhydration, forced diuresis, insertion of a vesical catheter for intermittent or continuous bladder irrigation and evacuation of clots. Patients with grade III-IV LOHC were divided into four groups according to immunosuppressant use (steroid, MMF and Basiliximab) 1 week before and after the onset of LOHC: an intensified-intensified (I-I) group, intensified-not intensified (I-N) group, not intensified-intensified (N-I) group and not intensified-not intensified (N-N) group. "Intensified" was defined as an increase in any immunosuppressant, and "not intensified" was defined as maintenance or tapering of all immunosuppressants. Data concerning the baseline characteristics, incidence, treatment patterns and outcomes were recorded.

Results

In total, 364 patients developed LOHC at a median of 29 days (range, 23-37.75 days), with a cumulative incidence of 16.87%. AML (HR =0.493 95% CI, 0.244-0.997; p=0.049), AA (HR =0.444, 95% CI,0.18-1.096; p=0.078), HLA match(HR =0.149, 95% CI, 0.036-0.616; p=0.009), days to platelet engraftment(HR =1.023, 95% CI, 0.998-1.049;p=0.069)and CMV viremia (HR =2.365, 95% CI, 1.179,4.733;p=0.015)were associated with LOHC. Only HLA match (HR =0.111, 95% CI, 0.014-0.878; p=0.037)remained significant in the multivariate analysis. 6.5%, 12.9%, 48.39% and 32.26% patients with III-IV LOHC were divided into I-I, I-N, N-I and N-N groups, respectively. There were no significant differences in overall survival (p=0.05) and the incidence of CMV viremia (p=0.187) among the four groups. The incidence of relapse (p=0.007) and the incidence of aGVHD (p=0.01) was highest in the I-I group, followed by the N-I, I-N and N-N groups. 50% of the patients in N-I group showed the improvement of LOHC, which is significantly highest among the four groups(p=0.000). However, no statistical significance was found regarding the CMV viremia turning shade or the improvement of aGVHD. Non-relapse mortality (NRM) was observed in 25% of patients without resolution of LOHC. NRM was 0% among patients without intensified immunosuppression but was 25%, 0% and 50% among those with intensified immunosuppression before LOHC, after LOHC and before and after LOHC, respectively.

Conclusion

The independent risk factors for LOHC after allo-HSCT were HLA mismatch. Avoiding intensified immunosuppression that damages endothelium or reactivates the related virus may improve transplant outcome. The N-N pattern could reduce the incidence of CMV viremia or aGVHD in the LOHC patients, and the N-I pattern might be more promising as a therapeutic strategy for LOHC.

Disclosures

No relevant conflicts of interest to declare.