Bone Marrow Cellularity during Induction Is Highly Predictive of Complete Remission in De Novo AML but Not in Secondary/Therapy-Related AML
Gang Zheng, Amy E DeZern, Keith W Pratz, Judith E Karp, Christopher D Gocke. Johns Hopkins University School of Medicine, Baltimore, MD
Background: The achievement of complete remission (CR) after induction therapy leads to improved survival AML patients. Data are contradictory on the significance of morphologic parameters in predicting CR, and most studies have examined patients early in induction therapy. We studied a uniformly treated cohort of AML patients at day 14 after start of induction chemotherapy to identify morphologic predictors of CR.
Design: 44 newly diagnosed, previously untreated AML patients with high risk features were studied. All patients received an induction regimen of flavopiridol, cytosine arabinoside and mitoxantrone (FLAM) at a single institution. Bone marrow trephine biopsies and aspirates at approximately day 14 of the first cycle were reviewed and clinical records were abstracted. Marrow cellularity was determined by a point counting method. The mean of the two independent pathologists' counts was calculated. Other factors such as fibrosis, blast count and lymphocyte count were also included in the correlation study.
Results: For de novo AML, low bone marrow cellularity is an excellent predictor for CR.
With a cut point at 9% cellularity, there is a specificity of 100% and a sensitivity of 79% in predicting CR (p<0.001, Fisher's exact). However, for therapy-related AML or AML derived from myelodysplastic syndrome or myeloproliferative neoplasm, day 14 bone marrow cellularity is a poor predictor of CR: although cellularity of less than 9% also has a specificity of 100% in predicting CR, its sensitivity is only 25% (p>0.2, Fisher's exact). In our study, other factors such as bone marrow fibrosis and lymphocyte count did not correlate with clinical outcome.
Conclusions: In a cohort of aggressively treated, high risk AML patients, morphologic assessment of the day 14 bone marrow is helpful in predicting the chance of CR when the AML is de novo. Marrow cellularity is not informative in secondary/therapy-related AML. This supports the concept of a biologic difference between primary and secondary AML. Cellularity assessment may allow for targeting of de novo AML patients who are unlikely to achieve remission and in need of more aggressive therapy.
Tuesday, March 20, 2012 9:30 AM
Poster Session III # 225, Tuesday Morning