Intermediate-dose posttransplantation cyclophosphamide for myeloablative HLA-haploidentical bone marrow transplantation

Abstract

High-dose posttransplantation cyclophosphamide (HD-PTCy), given at 50 mg/kg/day on days +3/+4, is a standard-of-care graft-versus-host disease (GVHD) prophylaxis for allogeneic hematopoietic cell transplantation (HCT). Our murine MHC-haploidentical HCT studies suggested intermediate-dose PTCy produces superior GVHD control compared with HD-PTCy and PTCy is maximally effective on day +4. We conducted a single-institutional prospective phase 1/2 trial to reduce PTCy dosing to 25 mg/kg/day on days +3/+4 or on day +4 only for myeloablative HLA-haploidentical bone marrow HCT using PTCy, sirolimus, and mycophenolate mofetil. Among 35 patients, 89% were ethnic/racial minorities, 46% had high/very-high-risk disease, and median comorbidity score was 3. The phase 1 dose-limiting-toxicity, grade III-IV acute GVHD, was not observed after either reduced-PTCy dose level. PTCy 25 mg/kg/day on days +3/+4 (intermediate-dose (ID)-PTCy; n = 23), the phase 2 dose, resulted in no grade II-IV acute GVHD; 2-year cumulative incidences of chronic GVHD requiring systemic immunosuppression, nonrelapse mortality, and relapse were 13%, 17%, and 22%, and 2-year overall survival, disease-free survival, and GVHD-free/relapse-free survival were 61%, 61%, and 52%. In exploratory analysis compared with HD-PTCy (n = 5), ID-PTCy resulted in significantly faster engraftment and T-cell reconstitution, fewer transfusions, less mucositis, and reduced severity of BK-virus-associated cystitis/urethritis; area-under-the-curve exposure of 4-hydroxycyclophosphamide (4HCY), a key cyclophosphamide metabolite, correlated with these outcomes but not with chronic GVHD occurrence. Ideal-body-weight-based PTCy dosing best approximated 4HCY exposure. ID-PTCy is effective and has apparent clinical benefits compared with HD-PTCy. Before broader implementation, further studies are needed to confirm these findings and define optimal PTCy dosing across various donor/graft types. This trial was registered at www.clinicaltrials.gov as #NCT03983850.

Graphical abstract

Figure 1.

Treatment schema and aGVHD. (A) Treatment schema. (B) Maximal grade of aGVHD among evaluable patients (n = 5 each for DL1 and DL3; n = 20 for DL2). Patients without aGVHD but with death, graft failure, or relapse before day +60 were not included, as per the protocol these patients were considered inevaluable for the primary endpoint.

Figure 2.

Temperature curves and peri-engraftment symptoms. (A) All temperature readings over the course of the first 21 days for all phase 1 patients are shown. Fevers were more common for patients in DL3 (PTCy 25 mg/kg on day +4 only) and were noted to be higher and more prolonged beyond the time of engraftment. (B) Fevers, rash, and transaminitis were common during the peri-engraftment period for all 3 dose levels, but engraftment syndrome criteria were met only by 3 patients: 1 at DL2, and 2 at DL3 (all 3 with fever, rash, and transaminitis); none of these patients required treatment for engraftment syndrome and the symptoms quickly improved without intervention. ^aIncludes 1 patient with prolonged engraftment time and fevers early in course of neutrophil recovery (>4 days from day of engraftment); ^b1 patient with symptoms that were attributable to sinusoidal obstructive syndrome. This patient was the first DL3 patient and had severe and rapid increase in transaminases along with concomitant fever on day +10 and received 1 dose of tocilizumab before it was determined that he was having sinusoidal obstruction syndrome during this peri-engraftment period as the cause of the transaminase elevation. This was the only patient with immunosuppressive treatment before day +21 beyond the intended PTCy, sirolimus, and mycophenolate mofetil.

Figure 3.

GVHD outcomes and survival probabilities for ID-PTCy–treated patients. (A-B) Cumulative incidences of (A) all aGVHD and grade II-IV aGVHD and (B) all cGVHD and cGVHD requiring systemic therapy. (C) Prophylactic and therapeutic systemic immunosuppression use for GVHD in the first posttransplant year. Patients with graft failure were excluded. Patients were counted as being on immunosuppression if alive and still taking immunosuppression on the last day of each month. The number of patients in each group at each time point are shown. (D-E) Cumulative incidences of (D) relapse and (E) nonrelapse mortality. (F-I) Kaplan-Meier estimates of (F) disease-free survival, (G) overall survival, (H) overall survival stratified by the DRI, and (I) GVHD-free, relapse-free survival (GRFS). CI, confidence interval.

Figure 4.

Faster engraftment, less transfusions, and less severe mucositis with ID-PTCy than with HD-PTCy. (A-E) Results by treatment group are shown. (A-B) Day of neutrophil and platelet engraftment. (C-D) Numbers of red blood cell (RBC) and platelet transfusions in the first 30 days after HCT. Of note, transfusion thresholds for patients were 7 g/dL for hemoglobin and 10 × 10³/μL for platelets unless clinical circumstances required higher thresholds (eg, fever or bleeding). (E) Duration of mucositis (including oral, pharyngeal, and rectal) as graded by common terminology criteria for adverse events version 5.0. Patients who did not engraft were not included for panels A-E. (F) Donor myeloid (CD33⁺) and T-cell (CD3⁺) chimerism results for all patients. One patient in DL2 with early malignancy relapse had insufficient cells at day +28 for CD33⁺ bead-sorted chimerism but flow cytometrically sorted chimerism of CD14⁺ cells at day +22 had showed 99% donor chimerism, and that result is included for the day +28 timepoint. Otherwise, patients with peripheral blood counts insufficient to perform chimerism analyses were categorized as not reported (NR). ∗P<.05; ∗∗P<.01.

Figure 5.

Faster T-cell reconstitution and less severe BK virus–associated cystitis/urethritis with ID-PTCy than with HD-PTCy. (A) CD4⁺ and (B) CD8⁺ T-cell recovery over 2 years after HCT for each dose level. Both CD4⁺ and CD8⁺ T-cell counts were significantly higher for patients in DL2 and DL3 than those treated with HD-PTCy at day +14 and also at other early post-transplant time periods, before counts numerically normalized by day +42. T-cell counts were evaluated before PTCy administration on days +3 and +4, before starting sirolimus and mycophenolate mofetil on day +5, on day +7, weekly for 100 days, and then at days +180, +270, +365, +545 (1.5 years), and +730 (2 years) after transplant. Up to day +21, samples were collected on the day labeled. Windows for subsequent samples are as follows: days +28 to +99 ± 3 days; days +180 and +270 ± 14 days; and all following time points ± 30 days. Patients who did not engraft and those with persistent mixed T-cell chimerism were not included. (C) Duration of symptomatic BK virus–associated cystitis/urethritis and hemorrhagic cystitis. BK cystitis symptoms were defined by the presence of dysuria, urinary urgency, urinary frequency, bladder spasms, or macroscopic hematuria occurring within the first 180 days after HCT in patients with BK viruria. (D) BK virus levels in urine and blood on serial monitoring in the first 100 days after transplant. Mean values with standard error of the means are shown. Statistical comparisons shown are of the AUC of weekly testing over the first 100 days. For panels C-D, patients included are known to be at risk for BK cystitis, defined by BK virus detection in the urine or blood at any point on serial monitoring; 2 patients in DL2 without symptoms but also without any detectable BK virus in urine or blood at any timepoint were excluded in case they were unexposed and not at risk for reactivation. Patients who did not engraft, those with follow-up of <100 days because of relapse or NRM (unrelated to BK virus), or with BK virus–associated cystitis/urethritis symptoms at the time of initiation of conditioning for HCT also were excluded. ∗P < .05; ∗∗P < .01.

Figure 6.

Pharmacokinetics of 4HCY correlate with IBW PTCy dosing and the apparent clinical benefits of reduced-dose PTCy. (A-B) The area-under-the-curve (AUC) of cyclophosphamide (CY) and its active metabolite 4HCY were measured for patients in all DLs although 2 DL2 patients had missing samples precluding AUC calculations. ∗P < .05; ∗∗P < .01. (C-D) For each patient, the relationship between PTCy weight-based dose and 4HCY AUC were compared using either the (C) ABW or (D) IBW for the PTCy dose calculation. Spearman correlation coefficients (ρ) and associated P values are shown. Colors reflect the race/ethnicity of individual patients. Green = White/Hispanic; red = White/Non-Hispanic; yellow = African-American; orange = Asian; light blue = American Indian/Alaskan Native; dark blue = multiracial. (E-H) Measured 4HCY exposures significantly correlated with clinical outcomes that appeared superior for ID-PTCy compared with HD-PTCy, including (E) time to neutrophil engraftment, (F) duration of mucositis (oral, pharyngeal, and/or rectal), and (G) absolute number of CD4⁺ T cells in the blood on day +14. Patients with graft failure, mixed T-cell chimerism, or those receiving tocilizumab (n = 1) were excluded from panels E-G. (H) No difference was seen in 4HCY exposure among evaluable ID-PTCy–treated patients with regard to the development of cGVHD.