Outcomes of Hematopoietic Stem Cell Transplantation after Chronic Graft Versus Host Disease and Evaluation of Revised Failure Free Survival in a Retrospective Analysis from Two Referral Centers.

Chronic graft versus host disease (cGVHD) is one of the most fearsome complications after allogeneic hematopoietic stem cell transplantation (HSCT). Design of cGVHD-based clinical studies (both in retrospective and in prospective setting) remains problematic because of the lack of validated response criteria for treatment and of standardised indicators for evaluation of outcomes of patients with cGVHD. Indeed, classical endpoints (such as overall survival-OS- and event free survival-EFS-) are quite limited in their ability to capture cGVHD associated morbidity. Inamoto et al. (Blood 2014) proposed a composite endpoint (Failure free survival-FFS, defined by the absence of second-line treatment, non-relapse mortality (NRM), and recurrent malignancy during initial treatment of cGVHD), for the evaluation of initial treatment after cGVHD.

Here we reported outcomes of 296 transplanted patients consecutively diagnosed with cGVHD according to 2015 revised Consensus Conference National Institute of Health (NIH) criteria, and proposed a "revised" definition of FFS conceived as survival function including as events: 1) stopping of all systemic treatment for cGVHD 2) NRM and 3) recurrent malignancy after initial diagnosis of cGVHD.

Patients were transplanted at Federico II University in Naples (n=24) and Saint-Louis Hospital in Paris (n=272) between 2007 and 2014. Their characteristics are described in table 1.

Median age at transplant was 45.1 (range 5.4-68.3) years. Median duration of follow-up for survivors was: 57.7 (range 2.64-97.6) months.

Cumulative incidence (CI) of patients developing an acute GVHD (aGVHD) before cGVHD was 54.2% (95%CI48.2-59.7) for grade II-IV and 12.5% (95%CI9-16.6) for grade III-IV. CI of starting immunosuppressive therapy (IST) was 66.2% (95%CI 61-71.9) at 1 year and 84.1% (95%CI 80-88.4) at 2 years post HSCT. Median time of diagnosis of cGVHD (from transplant) was 6.7 (range 1.02-58.78) months. GVHD characteristics are shown in figure 1A, Band table 2.CI of stopping IST was 11.6% (95%CI 0.56-85.7) at 1 year and 26.3% (95%CI 20.9-32.1) at 3 years from diagnosis of cGVHD. Median time of stopping all systemic treatments was 29.47 months. Three-yearOS and EFS of the whole population was respectively 83% (95%CI-78.7-86.6) and 73% (95%CI 68.9-76.7). Three-year CI of relapse was 15.4% (95%CI 11.5-19.9). Factors influencing survival in univariate analysis were oral, gastrointestinal, and mucosal involvement of cGVHD (OS and EFS), NIH severity of cGVHD (only OS), subcategory of cGVHD (classical/overlap), steroid dose and number of agents at initial treatment of cGVHD (OS and EFS). COX regression model identified as risk factors of survival only prednisone equivalent dose>0.5 mg/kg/day (for OS) and over 2 agents for initial cGVHD treatment (for both OS and EFS), whereas the only factor impacting on CI of relapse was type of diagnosis. Neither stem cell source nor type of donor resulted associated with survival or CI or relapse in our analysis. FFS for patients requiring IST was 55% (95% CI 49.1-60.8) at 3 years from initial treatment. Factors influencing FFS in univariate analysis were: year of transplant, liver and mucosal involvement, and overlap subcategory at presentation of cGVHD. In multivariate model only year of transplant and mucosal involvement impacted on FFS.

Evaluation of transplanted patients developing cGVHD remains challenging especially in retrospective setting. Our analysis showed major outcomes post cGVHD diagnosis and pointed out that most of risk factors identified for survival are not modifiable. Moreover we proposed the application of a "revised" definition of FFS as alternative effective endpoint to retrospectively evaluate long-term outcomes of transplanted population developing cGVHD.

View largeDownload slide

View largeDownload slide

Close modal