Body Mass Index at 3 Weeks after Intensive Induction Regimen Is a Prognostic Factor for the Risk of Relapse and/or Failure in Patients with Acute Myeloid Leukemia

Introduction:

Cancer-related malnutrition occurred in 30 to 80 % of patients (pts) with cancer (Br J Cancer. 2010;102:966). Malnutrition affects well-being and it is associated with an impaired survival in pts with solid cancer (Eur J Cancer 2008;44:1238). By contrast, few studies focused on the nutritional status of adult patients with acute myeloid leukemia (AML) eligible to intensive therapy and its prognostic consequences.

Materials and methods:

Therefore, we systematically assessed nutritional status with the body mass index (BMI) before starting induction therapy and again 3 weeks later in AML patients who received intensive induction regimen from January 2008 to December 2011 at Amiens University Hospital. Weight loss, albumin and Glasgow prognostic score (GPS, Brit J Cancer, 2003; 89: 1028) were also monitored during hospitalization and enteral or parenteral nutrition was initiated in case of denutrition. The prognostic value associated with nutrition parameters was evaluated with the model of Fine and Gray (JASA, 94: 496 1999) in order to take the 2 competing risks of relapse/initial failure (R/F) or death without R/F into account.

Results:

A total of 106 AML pts (mean age 49.8 years [18-75 years], M/F=0.88) entered the study. Eleven pts presented with acute promyelocytic leukemia. Mean serum albumin concentration was 35.6 g/L (15.7-40.8). At treatment initiation, underweight, normal weight, overweight and obesity was identified by BMI in 4%, 41%, 22% and 33% of pts respectively and 3 weeks later in 4%, 49%, 26% and 21% of pts respectively. Severe infections (Common Terminology Criteria for Adverse Events grade 3 to 5) occurred during the hospitalization in 68 pts. Fifteen pts died during hospitalization. Twelve pts failed to respond to induction chemotherapy regimen, all but one received a salvage regimen. Allogeneic bone marrow transplantation was performed in 15 pts. R/F occurred in 33 pts and 17 have died in remission after initial hospitalization. Finally, 62 pts have died at the stopping date. With a median follow-up of 64.6 months in alive pts, median overall survival was 29.3 months, and it was only negatively influenced by advanced age (p=0.003) and adverse molecular characteristics (p=0.01). BMI, GPS and albumin evaluated at treatment initiation and 3 weeks later did not influence the duration of hospitalization (p>0.16) nor the risk of severe infection during the initial hospitalization (p>0.14 in a logistic regression model). Among parameters listed above, only high BMI at 3 weeks after initiation of the induction regimen was significantly associated with an increased risk of R/F (p=0.04, hazard ratio 1.45, [1.01-2.12]), whereas none of these parameters retained a significant prognostic value for the risk of death without R/F.

Conclusion:

We confirmed the high incidence of overweight in adult pts with AML treated with intensive induction regimen and the absence of prognostic value of initial BMI (Am J Hematol 2016; 91: 199). However, high BMI at 3 weeks after treatment initiation was associated with an increased risk of subsequent R/F. Further studies are ongoing to assess the role of delay or dose reduction in treatment delivery and the potential statistical interactions between nutritional status and molecular characteristics of AML. Thus, our findings highlight the importance of taking any nutritional status into account during the management of AML pts with intensive induction chemotherapy.