Can bone marrow cellularity help in predicting prognosis in myelodysplastic syndromes?

Abstract

Objectives: To ascertain the relevance of bone marrow cellularity (BMC) to the interpretation of blast percentage (blast%) in MDS prognostication.

Methods: We compared survival prediction based on blast% adjusted to different levels of cellularity, compared to the survival based on the original IPSS-R blast% grouping.

Results: We analyzed 355 consecutive MDS patients. Cellularity, in and of itself or its interaction with blast%, was not associated with overall survival (OS). In a small subset of patients with a hypercellular marrow (15%; n = 26), dismal prognosis was observed at lower levels of blast%. For these cases OS was similar to higher IPSS-R blast groups. For example, within the Intermediate group (blast% 5%-10%), those with a hypercellular marrow and >6% blasts had an OS of 10 m similar to 16 m in the High (blast% 10%-19%) blast group. These changes did not translate into a significant improvement in overall prognostic power of a cellularity-adjusted IPSS-R (C index 0.71 vs. 0.70).

Conclusion: Adjusting blast% to cellularity did not improve prognostication. However, within IPSS-R-defined blast groups, a small subset of patients with relatively higher blast% and hypercellularity may have a worse prognosis than expected.

Keywords: MDS; bone marrow cellularity; bone marrow failure; myelodysplastic syndromes; prediction modeling in cancer.