Physician-Reported CR+PR at 6 Months Predicts Subsequent Survival in Patients with Chronic Gvhd

One of the greatest challenges to research progress in chronic graft-versus-host disease (GVHD) is lack of validated response criteria for complete response (CR), partial response (PR), stable disease (SD) and progressive disease (PD).  We compared four methods of assessing chronic GVHD response (calculated response proposed in the 2005 NIH chronic GVHD Consensus conference or clinician-reported CR+PR vs. SD+PD, and clinician- or patient-reported improvement, stability or worsening) for their ability to predict three subsequent clinical outcomes at 3 and 6 mos: failure free survival (FFS), defined as survival without addition of a new treatment, relapse or death; non-relapse mortality (NRM); and survival. Data were derived from 575 adult patients in a prospective observational study conducted by the Chronic GVHD Consortium at 9 centers. 342 (59%) were enrolled within 3 months (mos) of chronic GVHD diagnosis. The median time to enrollment was 11.9 mos after transplant (range 2.9 - 294), median FU after enrollment was 44 mos (range 0.9-76), and 149 (26%) have died. At enrollment, 302 (53%) had moderate severity cGVHD and 220 (38%) were classified as severe. Landmark analyses were performed for patients who were alive and had not failed at the three-month or six-month visit. We found that clinician-reported response at 3 mos was predictive of subsequent FFS (p<0.001), and the same response measure at 6 mos was predictive of subsequent OS (p = 0.007) and FFS (p=0.004). Patient-reported changes were associated with FFS after 3 mos but not with other outcomes. No response measures predicted NRM. The fact that FFS could be predicted by clinician- or patient-reported response measures might reflect that clinicians and patients control the main determinant of FFS, i.e., whether new systemic immunosuppression is added. Incorporation of the calculated CR + PR category did not significantly change the ability of the clinician-reported CR + PR to predict FFS or survival after 6 months.  We conclude that clinician-reported CR+PR is highly associated with OS after 6 months, which is independent of calculated CR+PR. Calculated response algorithms might be improved by understanding and incorporating clinician judgment in assessing response.

Table.

	FFS after 3 mos	NRM after 3 mos	Survival after 3 mos	FFS after 6 mos	NRM after 6 mos	Survival after 6 mos
NIH-calculated CR+PR vs. SD+PD	P=0.51	P=0.55	P=0.51	HR 0.73, 95% CI 0.53-1.02, p=0.06	P=0.24	HR 0.64, 95% CI 0.39-1.02, P=0.06
Clinician-reported CR+PR vs. SD+PD	HR 0.34, 95% CI 0.22-0.52, P<0.001	P=0.84	P=0.47	HR 0.61, 95% CI 0.44-0.85, P=0.004	P=0.06	HR 0.55, 95% CI 0.36-0.85, P=0.007
Clinician-reported Improvement vs. Stable vs. Worsening	P<0.001	P=0.57	P=0.82	P=0.28	P=0.62	P=0.60
Patient-reported Improvement vs. Stable vs. Worsening	P<0.001	P=0.62	P=0.36	P=0.08	P=0.33	P=0.44