Calculated NIH Response Correlates with Changes in Patient-Reported Symptoms but Not with Quality of Life: Results From the Chronic Gvhd Consortium

Abstract

Abstract 1988Background: In 2005, the National Institutes of Health (NIH) Consensus Conference recommended measurement tools to capture clinician or patient-assessed chronic GVHD manifestations. The NIH Conference also proposed a set of provisional definitions to calculate response using these measures. The correlation of the calculated NIH responses with patient-reported quality of life (QOL) or symptoms has not been examined. Ideally, improvement according to NIH response criteria would be associated with better QOL and fewer symptoms compared to patients who have stable disease or progress. Patients and methods: We studied 274 patients with chronic GVHD requiring systemic treatment who were enrolled in a multi-center, prospective, longitudinal, observational cohort. Treatment of chronic GVHD was not uniform for this study. 256 patients had follow-up visits at 6 months after enrollment. The 6 month time point was chosen to be consistent with many earlier phase II studies. Responses were classified according to the calculated NIH response algorithm (BBMT 2006;12:491) as complete, partial, stable or progression, separately in skin, eye, mouth, lung, liver and gastrointestinal (GI) tract as well as overall by comparing organ manifestations reported at the follow-up visit to the enrollment visit. Complete or partial response was considered “response”, and stable disease or progression as “no response”. QOL measures included the SF-36, FACT-BMT and Human Activities Profile (HAP) score. Symptom measures included the Lee symptom scale and 10-point global rating of symptoms. Linear regression models were used to estimate the change in measures associated with response vs. no response, adjusting for case type (incident vs. prevalent), donor/patient gender and NIH global severity at enrollment. Results: Patients with a median age of 52 years were enrolled at a median of 13 months from transplant. Of 274 patients, 247 (90%) had mobilized blood cell transplant, 132 (48%) had HLA-matched related donors, 138 (50%) had unrelated donors, and 153 (56%) had myeloablative conditioning. At 6 months, overall response rates were 10 (4%) CR, 72 (28%) PR, 14(6%) SD, and 157(62%) PD, for a CR+PR rate of 32%. Organ-specific CR+PR response rates were 31% for skin, 16% for eye, 26% for mouth, 23% for GI, and 31% for liver. Compared with no response, overall response was associated with improved overall symptom burden measures, but not with changes in QOL measures. For example, the Lee overall score was estimated to be 4.6 points lower in responders than in non-responders (Table 1). Organ-specific analyses showed that organ response was associated with improved symptom burden in skin, eye, mouth and GI except for mouth dryness (Table 2). Conclusion: The calculated organ-specific and overall NIH responses by the provisional algorithm correlated with corresponding organ-specific and overall changes in patient-reported symptom burden, but calculated overall response did not correlate with changes in patient-reported QOL. Our results suggest that the calculated NIH response is a surrogate measure for patient symptom burden in clinical trials.Table 1Estimated change in QOL and symptom measures associated with overall responseOutcomeClinically meaningful change1R vs. NR Estimated change (95% CI)PSF-36 PCS4.80.8 (−1.9–3.5).57SF-36 MCS5.20.1 (−2.7–2.9).93FACT-BMT102.7 (−1.8–7.1).24HAP MAS6.20.7 (−3.2–4.6).73HAP AAS8.41.6 (−2.6–5.8).45Lee symptom overall score6.2−4.6 (−7.3–−1.9).001Patient-rated overall symptoms2−0.9 (−1.6–−0.3).003R, response; NR, no response; PCS, physical component score; MCS, mental component score; MAS, maximum activity score; AAS, adjusted activity score.1Half-standard deviation of baseline score or 2-point change on a 10-point scale.Table 2Estimated change in symptom measures for each organ associated with organ responseOrganClinically meaningful changeR vs. NR Estimated change (95% CI)PLee symptom scaleSkin10.4−7.3 (−14–−0.8).03Eye14.8−16 (−27–−4.7).005Mouth13.8−16 (−24–−7.6)<.001GI (nutrition)6.2−7.1 (−13–−1.1).02Patient-rated symptomsSkin itching2−1.0 (−1.8–−0.1).03Eye problem2−1.3 (−2.4–−0.2).03Mouth dryness2−0.1 (−1.1–0.8).81Mouth pain2−1.1 (−1.9–−0.2).02Mouth sensitivity2−1.0 (−1.9–−0.1).03R, response; NR, no response. Disclosures:No relevant conflicts of interest to declare.