The response is what counts

Science Spotlight

The response is what counts

From the Othus Group, Public Health Sciences Division

Jan. 21, 2019

Acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) are generally characterized by an excess of immature or dysfunctional blood cells. Therapeutic options for these diseases aim to restore normal blood cell counts and include some that are referred to as intensive (curative-intent), such as the cytotoxic “7 + 3” chemotherapy regimen, or less intensive (not curative-intent) options, such as the hypomethylating drug azacytidine. While complete remission (CR) is the goal of both approaches, less complete responses still occur. However, whether survival benefits differ by treatment type or treatment response has not been fully assessed. In a recent publication in the journal Leukemia, Dr. Megan Othus in the Public Health Sciences Division and colleagues reported results from a study of survival benefits by type of treatment in AML and MDS patients.

To conduct their analyses, the authors utilized data from four separate SWOG Cancer Research Network trials that varied in patient demographics and/or treatment regimen. Three of these studies included AML patients and one included MDS patients. The selected studies allowed the authors to specifically evaluate whether treatment regimen (7 + 3 versus azacytidine) influenced overall survival in patients who achieved a remission response. To also investigate whether survival outcomes vary by remission response, the authors separately analyzed patients who achieved CR, CR with incomplete platelet or neutrophil count recovery (CRi; only in the AML studies), or hematologic improvement (HI; only in the MDS study).

Graphical representation of overall survival of AML and MDS patients categorized by complete remission (CR) or no CR in four SWOG trials.

Overall survival of AML and MDS patients categorized by complete remission (CR) or no CR in four SWOG trials.

Image provided by Dr. Megan Othus

Among the four studies, CR rates ranged from 21% to 70%. The achievement rates of CRi ranged from 4% to 14%, and 16% of patients in the MDS study achieved HI. In comparing patients that achieved CR with those who did not, the authors found that overall survival was significantly better for the CR groups in all four studies (see Figure). The absolute survival benefit of CR varied among the studies. However, follow-up statistical tests indicated that the relative benefit between CR and relative overall survival (as measured by the hazard ratio) did not significantly differ among the different studies. Dr. Othus summarized the findings, “we found that the absolute benefit of a complete remission (comparing measures such as 1-year or 3-year survival) was larger among patients who received curative-intent “intensive” 7+3 therapy compared to patients who received “less-intensive” azacitidine-based therapy. But the overall prognosis for patients receiving the different therapies was quite different, and the relative benefit (comparing measures of ratios of survival) of complete remission was similar between the therapies.”

To assess whether a best response of CRi in AML is associated with survival benefits, the authors conducted separate analyses to compare overall survival among patients with CR, CRi, and no response. Overall survival trended towards being shorter in CRi compared to CR patients, although the difference was not statistically significant. However, overall survival was not significantly better in CRi patients compared to those with no response. These results suggest that the greatest survival advantage comes with the achievement of CR as compared to lesser responses. Similar analyses were conducted to assess benefits of HI as compared to CR in MDS patients. When compared to CR patients the results were mixed. One analysis demonstrated worse survival with HI, but a separate, different type of analysis demonstrated no difference between the groups. When compared to patients without a treatment response, HI patients were found to have significantly better overall survival.

Overall, this work provides new insight on whether treatments of varying intensity differentially affect survival outcomes of AML and MDS patients who achieved complete or partial remission. This is important because categories of treatment response may combine CR, CRi, and HI in determining the overall response to a treatment. The authors suggest that follow-up studies should assess whether these various treatment responses (CR, CRi, and HI) are differentially associated with improvements in quality of life. Such work would help further clarify the extent to which CR is a distinct response from CRi and HI.


Fred Hutch/UW Cancer Consortium members Megan Othus, Frederick Appelbaum, and Eli Estey contributed to this research.

This research was supported by the National Institutes of Health.

Othus M, Sekeres MA, Nand S, Garcia-Manero G, Appelbaum FR, Erba HP, Estey E. 2018. Relative survival following response to 7 + 3 versus azacytidine is similar in acute myeloid leukemia and high-risk myelodysplastic syndromes: an analysis of four SWOG studies. Leukemia. doi: 10.1038/s41375-018-0275-x