Abnormal Marrow Cellularity at Day 100 after Allogeneic Marrow Transplantation Impacts Overall Survival (OS) but Not Event-Free Survival (EFS)
Payal Sojitra, Pranav Gandhi, Ameet R Kini, Milind M Velankar, Scott Smith, Patrick J Stiff, Girish Venkataraman. Loyola University Medical Center, Maywood, IL
Background: Adequate marrow engraftment is critical for restitution of marrow trilineage hematopoietic elements and hence we hypothesized that abnormal age-adjusted marrow cellularity after allogeneic transplant may impact outcome. Therefore, the current study examined the relevance of marrow cellularity to outcome in day 100 marrows (d100) of patients in CR who underwent allogeneic bone marrow transplant (aBMT).
Design: From our aBMT database, we identified 80 patients (cord n=12; sibling n=22; MUD, n=46) transplanted (for AML, n=34; ALL, n=10, MDS, n=6; NHL, n=18, MPD, n=4; others, n=12) between 2008-2010, who had d100 marrow exam with concurrent biopsy available without evidence of disease and >98% donor chimerism. Clinicopathologic data including age/sex, and CD34 stem cell dose were extracted besides flow cytometry data. Marrow cellularity was visually estimated as a percentage and subcategorized as low (LC), normo (NC) or hypercellular (HC) based on expected cellularity for age. The outcome was modeled as EFS and overall survival (OS).
Results: The mean age of the cohort was 49 years (38 females, 42 males); mean follow up was 19 months (3.8-49.93 mo). During this period, there were 14 relapses with 29 all cause deaths. Thirty-seven (46.8%) cases showed NC marrow with a median OS of 20 mo while 38 (48.1%) were LC with a median OS 13.5 mo. In univariate outcome analyses, when comparing LC vs. NC, LC adversely impacted OS (log rank,p=0.03) but not EFS (p=0.09). Likewise, in the comparative analysis of LC+HC vs. NC groups, abnormal marrow cellularity retained its predictive ability in the EFS analysis. Two 3-parameter mutivariable Cox models were examined. In the first model including age, CD34 dose and cellularity (LC/NC only), only cellularity significantly impacted OS (see fig.), while in the second model including cellularity coded as NC vs other, only abnormal cellularity was retained as a significant adverse predictor of OS.
Conclusions: A simple measure of the visual estimate of age-adjusted cellularity status is effective at predicting outcome measures in disease free allotransplanted patients.
Monday, March 4, 2013 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 235, Monday Morning