Hemorrhagic Cystitis in Patients Undergoing Allogeneic Hematopoietic Cell Transplant with Post Transplant Cyclophosphamide As GvHD Prophylaxis

The use of post-transplant cyclophosphamide (PTCy) as graft-versus-host-disease (GvHD) prophylaxis after hematopoietic cell transplantation (HCT) has increased significantly over the past decade due to expansion of donor pool with haploidentical and mismatched unrelated donors. A recently completed phase 2 study, PROGRESS-2 (NCT02345850), has highlighted the efficacy of PTCy in the matched donor setting. Hemorrhagic cystitis (HC) is the most notable toxicity associated with high dose Cy. However, data specific to the incidence and severity of HC in the post HCT setting is sparse, with no consensus on best practices to prevent HC in patients receiving PTCy. Current strategies to prevent HC in PTCy setting have been adapted from data in pre-HCT Cy (conditioning setting), such as hyperhydration with forced diuresis, continuous bladder irrigation, mesna as an intermittent and continuous infusion, or a combination of these methods. The aim of our study was to describe the incidence and severity of HC in patients undergoing HCT with PTCy as GvHD prophylaxis, identifying potential risk factors and impact of HC on HCT outcomes.

We retrospectively reviewed 194 consecutive patients who underwent their first HCT with PTCy from 2014 to 2018 at our center. More than half of the patients (53%) received myeloablative conditioning regimen with majority receiving peripheral blood stem cells (81%) from haploidentical donors (96%). GvHD prophylaxis was unified with PTCy (50 mg/kg on Days +3 and +4), in addition to MMF (1 gm 3x daily starting on Day +5) and tacrolimus (1 mg as a continuous infusion daily starting on Day +5). Standard HC prophylaxis was hyperhydration with forced diuresis and mesna at 320% the daily dose of PTCy. Incidence of HC was based on physician documentation or presence of blood in urinalysis up to Day +100. To determine severity of HC, CTCAE 5.0 grading system was used.

Median age of patients was 45 years (range: 2-73), with 60% of patients being male. KPS was ≥80% in 83% of patient and 40% had HCT-CI of ≥2. The most common diagnoses included: AML (41%), ALL (24%) and MDS/MPN (19%). There were 55 patients who received ≥3 lines of therapy, 116 patients received <3 lines of therapy pre-HCT, and 23 patients did not receive chemotherapy before HCT. DRI was high/very high in 38% of patients. Incidence, median onset, median days until resolution, severity, and the types of intervention for treatment of HC are displayed in Table 1. At a median onset of 12 days (range: 3-71) after HCT, a total of 61 patients (31.4%) developed Grade 1-2 (88.5%) and Grade 3 (11.5%); none of the patients developed Grade 4 HC. Viral HC (presence of either BK [n=26] or adenovirus [n=0] in urine) and non-viral HC were developed at a median of 34 (range: 6-71) and 7 days (range: 3-59, p<0.001), respectively. Most patients improved with supportive care over a median of 9 days (range: 1-69). Seven patients (11.5%) required continuous bladder irrigation, and six received anti-viral therapy. In multivariate analysis, age above 60 years (HR 4.16, 95% CI 1.52-11.42, p = 0.006) and myeloablative conditioning (HR 2.44, 95% CI 0.99-6.05, p = 0.054 [trend]) were associated with higher risk for HC. Disease type and risk, performance status and prior chemotherapy history were not associated with higher risk of HC in our cohort. By multivariate analysis, the incidence of HC or severe HC by Day +100 post-HCT was not associated with overall survival (OS), disease-free survival (DFS), non-relapse mortality (NRM), or grade 3-4 acute GvHD.

In conclusion, hyperhydration with forced diuresis added to aggressive mesna dosing is an effective strategy in preventing severe HC in HCT patients receiving PTCy as GvHD prophylaxis. Incidence of Grade 3 or 4 HC was low and transient and did not impact HCT outcomes. Viral HC had a significantly later onset than non-viral HC, suggesting a different pathophysiology. Older age and myeloablative conditioning were independent factors for higher incidence of HC in our cohort.