Ferrara unfitness criteria provide good predictions for early mortality after intensive chemotherapy in acute myeloid leukemia (AML), according to study results published in the Journal of Clinical Oncology.

The Ferrara criteria are widely used to identify patients with AML who would be most suitable for intensive chemotherapy, however, the accuracy of these criteria are unknown.

Raffaele Palmieri, MD, of the Fred Hutchinson Cancer Research Center in Seattle, Washington, and colleagues evaluated the accuracy of the Ferrara criteria to predict early mortality and survival based on the fitness of adult patients with AML undergoing intensive chemotherapy. The predications from the Ferrara score were also compared with those of the previously developed treatment-related mortality (TRM) score.

Continue Reading

The investigators identified 703 adult patients (mean age, 60.5 years; range, 18.6-91.4; 56% men) with AML (81%) or other myeloid neoplasm presenting with at least 10% blasts in the blood and/or marrow (19%) who received curative-intent induction or reinduction chemotherapy with 7 + 3 cladribine, high-dose cytarabine, granulocyte colony-stimulating factor, and mitoxantrone (CLAG-M) or dose-reduced CLAG-M between at the Fred Hutchinson Cancer Research Center between January 2006 and January 2020.

Of the 703 patients, Ferrara and TRM criteria could be applied to 655 (93%) and 642 (91%) patients, respectively. Under the Ferrara criteria, 30% of the patients were classified as unfit for intensive chemotherapy (F-unfit), most frequently due to pulmonary function impairment (42%, 79/186). The overall survival of F-unfit patients (median, 4.8 months; 95% CI, 3.6-6.5) was significantly shorter than F-fit patients (median, 36.8 months; 95% CI, 27.4-73.0; P <.001).

The Ferrara score was more accurate compared with the TRM score at predicting day-28 and day-100 mortality (area under the receiver operating characteristic curve (AUC), 0.76 and 0.79 vs 0.66 and 0.62, respectively).

Adding covariates, including performance status and albumin, to the Ferrara assessment improved predictive accuracy (day-28 and day-100 mortality AUC, 0.84 and 0.85, respectively). Prediction of overall survival with a multivariate model, including the Ferrara and TRM scores as well as age, cytogenetic risk, relapses/refractory status, and albumin, was less accurate (c-statistic, 0.75).

The major limitations of the study were the retrospective design, missing data for some patients, and an inability to parse which criteria are most important for determining unfitness.

“Although we found substantially worse outcomes for F-unfit [vs] F-fit patients with intensive chemotherapy, well-controlled, ideally randomized studies will be required to determine whether outcomes in F-unfit patients are better with alternative, less intense therapies,” wrote the authors.

Disclosures: Some authors have declared affiliations with or received funding from the pharmaceutical industry. Please refer to the original study for a full list of disclosures.


Palmieri R, Othus M, Halpern AB, et al. Accuracy of SIE/SIES/GITMO consensus criteria for unfitness to predict early mortality after intensive chemotherapy in adults with AML or other high-grade myeloid neoplasm. J Clin Oncol. Published online October 8, 2020. doi:10.1200/JCO.20.01392