Characteristics of Graft-Versus-Host Disease (GvHD) After Post-Transplantation Cyclophosphamide Versus Conventional GvHD Prophylaxis

Abstract

Post-transplantation cyclophosphamide (PTCy) has been shown to effectively control graft-versus-host disease (GvHD) in haploidentical (Haplo) transplantations. In this retrospective registry study, we compared GvHD organ distribution, severity, and outcomes in patients with GvHD occurring after Haplo transplantation with PTCy GvHD prophylaxis (Haplo/PTCy) versus HLA-matched unrelated donor transplantation with conventional prophylaxis (MUD/conventional). We evaluated 2 cohorts: patients with grade 2 to 4 acute GvHD (aGvHD) including 264 and 1163 recipients of Haplo and MUD transplants; and patients with any chronic GvHD (cGvHD) including 206 and 1018 recipients of Haplo and MUD transplants, respectively. In comparison with MUD/conventional transplantation ± antithymocyte globulin (ATG), grade 3-4 aGvHD (28% versus 39%, P = .001), stage 3-4 lower gastrointestinal (GI) tract aGvHD (14% versus 21%, P = .01), and chronic GI GvHD (21% versus 31%, P = .006) were less common after Haplo/PTCy transplantation. In patients with grade 2-4 aGvHD, cGvHD rate after Haplo/PTCY was also lower (hazard ratio [HR] = .4, P < .001) in comparison with MUD/conventional transplantation without ATG in the nonmyeloablative conditioning setting. Irrespective of the use of ATG, non-relapse mortality rate was lower (HR = .6, P = .01) after Haplo/PTCy transplantation, except for transplants that were from a female donor into a male recipient. In patients with cGvHD, irrespective of ATG use, Haplo/PTCy transplantation had lower non-relapse mortality rates (HR = .6, P = .04). Mortality rate was higher (HR = 1.6, P = .03) during, but not after (HR = .9, P = .6) the first 6 months after cGvHD diagnosis. Our results suggest that PTCy-based GvHD prophylaxis mitigates the development of GI GvHD and may translate into lower GvHD-related non-relapse mortality rate.

Keywords: Graft-versus-host disease; Non-relapse mortality; Post-transplantation cyclophosphamide; Prophylaxis.

Figure 1.. Patient population.

The study population consisted primarily of two cohorts: consecutive patients who developed 1) grade 2–4 acute GvHD and 2) those who developed de novo, progressive, or relapsing chronic GvHD after allogeneic stem cell transplantation from a haploidentical donor with post-transplant cyclophosphamide (PTCy) graft-versus-host disease prophylaxis or 8/8 HLA-matched unrelated donor with conventional GvHD prophylaxis performed between 2013–2017. Patients who developed grade 2–4 acute GvHD and chronic GvHD are included in both cohorts. SCT, stem cell transplant; PTCy, post-transplant cyclophosphamide; TAC, tacrolimus; MMF, mycophenolate mofetil; HLA, human leukocyte antigen; MTX, methotrexate; GvHD, graft-versus-host disease

Figure 2.. Outcomes in patients with grade 2–4 acute GvHD.

(A1) The cumulative incidence of non-relapse mortality by donor type in recipients of transplants that are not from a female donor to a male recipient by donor type, adjusted for grade 3–4 acute GvHD, recipient age ≥ 40 years, HCT-CI > 3, and seropositive recipient CMV status. (A2) The cumulative incidence of non-relapse mortality by donor type in recipients of transplants from a female donor to a male recipient by donor type, adjusted for grade 3–4 acute GVHD, recipient age ≥ 40 years, HCT-CI > 3, and seropositive recipient CMV status. (B1) The cumulative incidence of chronic GvHD by donor type in recipients of RIC/NMA conditioning, adjusted for grade 3–4 acute GvHD, high DRI, and KPS < 90. (B2) The cumulative incidence of chronic GvHD by donor type in recipients of myeloablative conditioning, adjusted for grade 3–4 acute GvHD, high DRI, and KPS < 90. (C) Actuarial OS by donor type, adjusted for grade 3–4 acute GvHD, high DRI, recipient age ≥ 40 years, HCT-CI > 3, and seropositive recipient CMV status. MUD, matched unrelated donor; RIC, reduced intensity conditioning; NMA, non-myeloablative; haplo, haploidentical

Figure 3.. Outcomes in patients with chronic GvHD.

(A) The cumulative incidence of NRM by donor type, adjusted for recipient age ≥ 40 years, HCT-CI > 3, and transplants from CMV-seronegative donors into a CMV-seropositive recipients. (B) Actuarial OS within the first 6 months after chronic GvHD diagnosis by donor type, adjusted for recipient age ≥ 60 years, high or very high DRI, HCT-CI > 3, and transplants from CMV-seronegative donors into a CMV-seropositive recipients. (C) Actuarial OS 6 months after chronic GvHD diagnosis by donor type, adjusted for recipient’s age ≥ 60 years, high or very high DRI, HCT-CI > 3, and transplants from a CMV-seronegative donor into a CMV-seropositive recipient.

Figure 3.. Outcomes in patients with chronic GvHD.

(A) The cumulative incidence of NRM by donor type, adjusted for recipient age ≥ 40 years, HCT-CI > 3, and transplants from CMV-seronegative donors into a CMV-seropositive recipients. (B) Actuarial OS within the first 6 months after chronic GvHD diagnosis by donor type, adjusted for recipient age ≥ 60 years, high or very high DRI, HCT-CI > 3, and transplants from CMV-seronegative donors into a CMV-seropositive recipients. (C) Actuarial OS 6 months after chronic GvHD diagnosis by donor type, adjusted for recipient’s age ≥ 60 years, high or very high DRI, HCT-CI > 3, and transplants from a CMV-seronegative donor into a CMV-seropositive recipient.