Co-Morbidity Score at the Time of Transplant Determines Prognosis in Older Patients Who Develop Acute Graft-Vs.-Host Disease (GVHD).

Older age patients are increasingly being considered for allogeneic hematopoietic cell transplants (HCT). Concerns related to early treatment related mortality (TRM), namely from acute GVHD, often limit referrals. Outcomes for older age patients who develop acute GVHD remain unknown.

We performed a retrospective analysis of 83 adult patients enrolled onto two consecutive trials at MD Anderson to determine predictors for TRM at 6 months following systemic treatment of acute GVHD. These trials enrolled patients with newly diagnosed acute GVHD between 2000-2008 and included a single-center randomized trial of infliximab + methylprednisolone (MP) vs. MP alone and a multicenter trial of MP + one of 4 agents (mycophenolate, etanercept, denileukin diftitox or pentostatin).

Median age was 49 yrs (range, 20-70 yrs) and 63% of patients (pts) were male. 52% had acute leukemia and 31% were in remission at the time of transplant. Myeloablative condition and/ or peripheral blood HCT were performed in 52% and 54% of pts, respectively. Matched sibling, matched unrelated and mismatched transplants were 49%, 40% and 11%. GVHD prophylaxis was tacrolimus/ methotrexate in 87% of pts. Grades I/II, III/IV and visceral-organ acute GVHD represented 68%, 32% and 60% of pts occurring at a median time of onset of 25 days post-HCT. Day 28 acute GVHD response (defined as complete or partial response) was 63%. The median co-morbidity score at the time of HCT was 3 (range 0-11). Co-morbidity scores >3 were more common in pts >50 yrs (47%) than in those '50 yrs (24%), p=0.03. The proportion of responders on day 28 was comparable in pts >50 yrs (60%) and those '50 yrs (64%), p=0.7. On univariate analysis, significant predictors of higher TRM rate at 6 months following initiation of systemic therapy for acute GVHD were lack of response on day 28 post therapy (Hazard Ratio (HR) 3.6, p= 0.004), age > 50 yrs (HR 2.9, p=0.03) and co-morbidity score >3 (HR 3.1, p=0.01). There was no significant impact on the rate of TRM for donor type, cell source, intensity of conditioning regimen, donor/recipient sex mismatch, disease status at the time of transplant, GVHD grade at the time of initiation of systemic therapy or protocol assigned therapy. To adjust for potential interaction and confounding effects, multivariate analysis was performed by classifying pts into mutually exclusive groups according to day 28 response status, age, and co-morbidity score (see table). The cumulative incidence of TRM was lowest in pts who were '50 yrs of age and who responded to first line therapy with co-morbidity scores not impacting outcomes. TRM was comparably low (15%) in the absence of co-morbidity scores >3 in pts >50 yrs who responded to first line therapy. In contrast, in the presence of co-morbidity scores >3, TRM rate was high in pts >50 yrs regardless of whether pts responded (40%) or did not respond (100%) to first line therapy.

The ability to withstand acute GVHD and/ or its therapy in pts older than 50 yrs depends on co-morbidity scores at the time of transplant.

No relevant conflicts of interest to declare.