Pre-Transplant Nutritional Risk Index Shows a Lack of Predictive Accuracy in AML


Findings from a study examining the prognostic value of Nutritional Risk Index prior to hematopoietic stem cell transplant in patients with acute myeloid leukemia highlighted a need to consider other tools for assessing nutritional status among those undergoing transplant.

Nutritional Risk Index (NRI) prior to hematopoietic cell transplant (HCT) did not provide additional predictive value over serum albumin alone in patients with acute myeloid leukemia (AML), highlighting a need to examine other tools for assessing nutritional status, according to results from a study published in Transplant and Cellular Therapy.

Investigators reported that low NRI at the time of conditioning for HCT was associated with an increased non-relapse mortality (HR, 0.97; 95% CI, 0.96-0.98) and risk of relapse (HR, 0.98; 95% CI, 0.96-0.98; P <.001), as well as a decrease in relapse-free survival (RFS; HR, 0.97; 95% CI, 0.96-0.98) and overall survival (OS; HR, 0.97; 95% CI, 0.96-0.98; P <.001); similar associations were identified for pre-HCT albumin levels.

A total of 1011 patients received transplant for AML, 979 of whom remained after exclusions and 970 were included in the analysis. At diagnosis, 1% of patients were underweight and 3% were underweight at relapse. Those with any level of malnutrition as evaluated by NRI appeared to be older (P = .001) with a shorter time between last remission and transplant (P = .023). Increasing degree of malnutrition—defined as decreased NRI—was associated with a worse ECOG performance status (P <.001), higher treatment-related mortality scores (P <.001), and a higher HCT comorbidity index (P = .003). Severely malnourished patients by NRI criteria were more likely to have a higher body mass index at diagnosis or relapse (P <.001), but not prior to HCT (P = .14).

There was a significant correlation between usual body weight and ideal body weight, however, the correlation coefficient was low (r = .42; P <.001). Moreover, the correlation coefficient was low when NRI was determined using usual body weight or ideal body weight (r = .37; P <.001). When the usual body weight was used, 18% of patients were classified as severely malnourished compared with when the ideal body weight was used.

At a median follow-up of 5.13 years following HCT, a total of 308 patients relapsed, 460 died, and 196 non-relapse mortality events took place. Investigators also reported that cumulative incidence of non-relapse mortality was elevated in patients who were severely malnourished by NRI criteria at 33% (95% CI, 17%-50%) compared with 19% (95% CI, 15%-23%) in those with mild malnutrition, 15% (95% CI, 9%-21%) with borderline malnutrition, and 13% (95% CI, 10%-17%) with no malnutrition (P = .004). Risk of relapse was elevated in patients with malnutrition vs no malnutrition. The risk of relapsed was 33% (95% CI, 17%-50%) for severe malnutrition, 33% (95% CI, 28%-38%) for mild malnutrition, 39% (95% CI, 30%-47%) for borderline malnutrition, and 25% (95 CI, 21%-29%) for no malnutrition (P = .002).

This translated to a 3-year RFS rate of 33% (95% CI, 21%-54%) for severe malnutrition, 48% (95% CI, 43%-53%) for mild malnutrition, 46% (95% CI, 39%-55%) for borderline malnutrition, and 62% (95% CI, 57%-67%) for no malnutrition (P <.001). The OS rate at 3-years was 36% (95% CI, 23%-57%) for severe malnutrition, 54% (95% CI, 49%-60%) for mild malnutrition, 52% (95% CI, 44%-61%) for borderline malnutrition, and 67% (95% CI, 62%-72%) for no malnutrition (P <.001).

The univariable analysis highlighted statistically significant associations between NRI and non-relapse mortality (HR, 0.89; 95% CI, 0.83-0.97; P = .005), relapse (HR, 0.92; 95% CI, 0.86-0.99; P = .022), RFS (HR, 0.91; 95% CI, 0.86-0.96; P <.001), and OS (HR, 0.92; 95% CI, 0.87-0.97; P = .002). The predictive value of NRI was similar regardless of whether it was used as a continuous or categorical variable. However, if ideal body weight was used to determine NRI, investigators reported a significant association with relapse (HR, 0.89; 95% CI, 0.81-0.97; P = .010) but not non-relapse mortality (HR, 1.05; 95% CI, 0.94-1.16; P = .4), RFS (HR, 0.95; 95% CI, 0.89-1.02; P = .13), or OS (HR, 0.97; 95% CI, 0.90-1.04; P = .4).

After assessing post-HCT outcomes using NRI, investigators set out to determine the relative important of individual components of the index such as weight loss and serum albumin. Decreasing albumin levels prior to HCT was associated with non-relapse mortality (HR, 0.36; 95% CI, 0.25-0.53; P <.001), relapse (HR, 0.62; 95% CI, 0.46-0.85; P = .002), RFS (HR, 0.51; 95% CI. 0.40-0.65; P <.001), and OS (HR, 0.51; 95% CI, 0.39-0.65; P <.001). NRI’s prognostic value was similar to that of pre-HCT serum albumin levels alone in terms of non-relapse mortality, PFS, and OS..


Orvain C, Byelykh M, Othus M, et al. Relationship between pre-transplant nutritional status and outcomes of adults with acute myeloid leukemia undergoing allogeneic hematopoietic cell transplantation. Transplant Cell Ther. Published online September 27, 2022. doi:10.1016/j.jtct.2022.09.023

Related Videos
Ghayas C. Issa, MD, an expert on acute myeloid leukemia
Ghayas C. Issa, MD, an expert on acute myeloid leukemia
Mikkael A. Sekeres, MD, discusses how data from the phase 3 QuANTUM-First trial may advance the treatment field for patients with newly diagnosed FLT3-ITD–positive acute myeloid leukemia.
An expert from Duke Health says that patients with NPM1-mutatant relapsed/refractory acute myeloid leukemia did not experience any significant safety signals following treatment with ziftomenib.
At the recommended 600 mg dose, ziftomenib interestingly produced complete remissions in patients relapsed/refractory acute myeloid leukemia who harbored NPM1 mutations .
An expert from Duke Health reviews the design of the phase 1/2 trial of the KOMET-001 study in heavily pretreated patients with relapsed/refractory acute myeloid leukemia.
Related Content