A Large Cohort Comparison of 2005 National Institutes of Health Chronic Graft-Versus-Host Disease Staging Criteria to 2014 National Institutes of Health Chronic Graft-Versus-Host Disease Staging Criteria

Introduction

Chronic Graft-versus-Host Disease (cGvHD) is a complication of allogeneic stem cell transplantation. 2005 NIH Staging Criteria (SC) of cGvHD emerged as result of an international effort to provide uniform standards for clinicians and researchers. In 2014, NIH cGvHD SC were revised based on gathered scientific evidence. This study assessed the impact of 2014 revision of NIH cGvHD SC on Organ and Global Score distributions, and associations to functional and quality of life measures (QOL).

Methods

Patients (pts) with cGvHD participating in a natural history study of cGvHD (NCT00092235) between October 2004 and November 2016 were evaluated and followed for survival. NIH cGvHD organ and global severity scores (0-3) were prospectively assigned using the 2005 NIH cGvHD SC. Supervised by a physician experienced in cGvHD, two selected team members rescored pts retrospectively with the 2014 SC based on the original 2005 NIH cGvHD scoring forms and data obtained from assessments by clinician subspecialists. A total of 284 pts were evaluated. Distributions of organ-specific and global severity scores measured by the 2005 and 2014 NIH cGvHD SC were compared. 2014 NIH cGvHD SC were tested for associations with functional and QOL measures: Lee symptom scale (LSS), Short Form 36 (SF36), 2-minute walk test (2MW), grip strength (GS), joint range of motion (ROM), Human Activity Profile (HAP), mean Subspecialist Evaluation Score (mSSE), and Functional Assessment of Cancer Therapy-Bone Marrow Transplant (FACT-BMT), using the model established by Baird et al. to validate the original 2005 NIH SC in a subset of this cohort (BBMT 2013; 19:632). Survival analysis was performed to determine potential predictors of overall survival.

Results

Pts had a median age of 46 yrs (range 19-71); 37% were receiving moderate and 40% high levels of systemic immunosuppression. Pts scored by 2014 NIH SC tended to have slightly lower global severity scores in comparison to 2005 NIH SC (75% vs 72% pts with severe cGvHD). Lung and liver scores tended to be lower, 2014 NIH cGvHD organ-specific and global severity scores are shown in Table 1. 2014 NIH skin score was associated with reduced GS and ROM (both p<0.0001) and higher LSS (p=0.0001) and mSSE (p<0.0001). 2014 NIH lung score showed significant associations with lower values for 2MW, SF36 physical component summary score (PCS), and HAP maximum activity score (MAS), as well as higher mSSE (all p<0.0001) and a trend towards association with reduced GS (p=0.002). 2014 NIH global severity score (24% moderate, 72% severe in this population) was positively associated with mSSE (p<0.0001) and showed trends toward association with higher LSS and lower HAP adjusted activity score (AAS).

Five-year overall survival in this cohort was 70.9% (95% CI: 64.8-76.1%) with a median potential follow-up of 79 months. The final Cox survival model included 2014 NIH Lung score (3 vs 0-2, HR=2.67, 95% CI: 1.47-4.84; p=0.0012), time from cGvHD diagnosis to consent (4+ yrs vs <4 yrs, HR=0.53, 95% CI: 0.31-0.94, p=0.028), and Karnofsky performance status (KPS, 80-100% vs <80%, HR=0.48, 95% CI: 0.30-0.77; p=0.0020). In the analysis of the 2005 NIH cGvHD SC, Baird and colleagues reported a Cox survival model including 2005 NIH lung score, time from cGvHD diagnosis to consent, and absolute eosinophil count, but not KPS.

The associations between NIH Global score and the NIH lung score and patient-reported outcome measures are similar to those in the Baird et al. analysis of 2005 NIH cGvHD SC. In contrast to Baird et al. statistically significant associations are now seen between global severity, skin score and lung scores and performance-based measures including GS and 2MW. This suggests that scoring refinements introduced in 2014 have strengthened the validity of the NIH SC.

Conclusions

In this cross-sectional cohort severely affected by cGvHD, the 2014 NIH cGvHD SC organ-specific and global scores were associated with functional and QOL measures, and likely reflect cGvHD burden. Refinements in describing lung and liver cGvHD resulted in lower scores for these organs using the 2014 NIH cGvHD SC compared to the 2005 NIH cGvHD SC. These findings illustrate that 2014 NIH cGvHD SC are valid measures for cGvHD-related burden and warrant study within both clinical care and prospective trials.

Table

View largeDownload slide

Table

View largeDownload slide

Close modal