Prognostic validity of AML composite model in predicting mortality

A newly proposed composite model, AML-CM, predicted early and late mortality in a Canadian patient population with acute myeloid leukaemia (AML) [1]. These findings emphasise the significance of assessing comorbidities and leukaemia-specific features prior to induction therapy for AML and tailoring targeted interventions to mitigate their risks.

Intensive chemotherapy is the standard of care for fit patients with newly diagnosed AML. However, assessing the impact of comorbidities on treatment outcomes when counselling patients remains a challenge. To this end, Dr Shannon Murphy (Dalhousie University, Canada) sought to validate the prognostic validity of AML-CM, which incorporates both comorbidities and leukaemia-specific features to predict overall mortality following administration of intensive induction chemotherapy.

The researchers retrospectively collected data on patients with newly diagnosed AML who received first induction chemotherapy in their treatment centres. Of 194 identified patients, 56% were male, median age at induction was 54 years (range 18-75 years), of which 38% were ≥60 years. Molecular/cytogenetic risk per ELN classification was as follows: 23% favourable risk, 37% intermediate risk, 34% high risk, and 6% unknown. The most frequently used regimen was the standard 3+7 regimen (daunorubicin and cytarabine). AML-CM scores were calculated as per previously published guidelines [2]. Logistic regression models were performed to analyse 8-week and 1-year mortality and competing risk regression to model overall survival after adjusting for known variables.

AML-CM (comparison AML-CM 10+ vs AML-CM 1-4) was predictive of 8-week mortality (OR 12; P=0.0003) and 1-year mortality (OR 11.76; P=0.0021). Overall survival was inversely proportional to increasing AML-CM scores (P<0.001); 1-year overall survival for patients with different AML-CM scores was as follows (see Figure):

• AML-CM 1-4: 50%;
• AML-CM 5-6: 30%;
• AML-CM 7-9: 18%; and
• AML-CM ≥10: 2%.

Figure: Overall survival by AML-CM score range [1]



Dr Murphy stated that AML-CM could be utilised to adjust for the impact of comorbidities by guiding patient selection in clinical trials investigating intensive AML therapies.

1. Murphy S, et al. The Prognostic Validity of the Acute Myeloid Leukemia Composite Model in Predicting Risks of One-Year Mortality Among Patients in Atlantic Canada: A Multi-Centre Experience. 62^nd ASH Annual Meeting, 5-8 December 2020. Abstract 999.
2. Sorror ML, et al. JAMA Oncol. 2017;3:1675-1682.

Posted on 18 February 202122 February 2021