Abstract
Background/Aim: No studies have examined the association between the Geriatric Nutritional Risk Index (GNRI) at the initiation of chemotherapy for malignant lymphoma and the occurrence of adverse events. Therefore, we investigated the impact of GNRI at treatment initiation on the occurrence of side effects and time to treatment failure (TTF) in patients with malignant lymphoma undergoing initial rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) therapy. Patients and Methods: This study included 131 patients who underwent initial R-CHOP therapy between March 2016 and October 2021. Patients were stratified into those with high (GNRI ≥92; n=56) or low (GNRI <92; n=75) GNRI status. Results: Comparing the High GNRI group and Low GNRI group, the incidence of febrile neutropenia (FN) and Grade ≥3 creatinine increase, alkaline phosphatase (ALP) increase, albumin decrease, hemoglobin decrease, neutropenia, and thrombocytopenia were significantly higher in the Low GNRI group. TTF in the High GNRI group was significantly longer than that in the Low GNRI group (p=0.045). Multivariate analysis showed that the factors influencing the duration of treatment were PS (≥2) at the start of treatment, serum albumin level, and GNRI. Conclusion: In patients undergoing R-CHOP therapy, GNRI <92 at regimen initiation increased the risks of developing FN and hematologic toxicity. Multivariate analysis revealed that performance status, albumin levels, and GNRI at regimen initiation were the factors influencing treatment duration. Nutritional status at treatment initiation may influence the development of hematologic toxicity and TTF.
Malignant lymphomas are histologically classified into Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL), and many patients in Japan have NHL. The male-to-female ratio is 3:2, with a peak incidence at the age between 70-80 years (1). Diffuse large B-cell lymphoma (DLBCL) accounts for more than 30% of all NHL cases in Japan (2). Rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) therapy is the first-line treatment of choice for DLBCL. In addition to hematologic toxicity, R-CHOP is highly emetogenic risk (3); therefore, patients are prone to nausea, vomiting, and anorexia, and 50%, 89%, 24%, and 19% of patients experience nausea and vomiting, leukopenia, thrombocytopenia, and anemia, respectively (4). Furthermore, nutritional disorders and weight loss occur in patients with malignant lymphoma undergoing cancer chemotherapy (5, 6), suggesting poor nutritional status and weight loss due to decreased food intake are associated with adverse events. Weight loss prior to the initiation of cancer chemotherapy is associated with a lower median survival (5) and nutritional status before and during treatment is associated with resistance to treatment, incidence of adverse events, and poor prognosis (5-7). Therefore, the nutritional status of patients undergoing chemotherapy must be assessed to ensure the safety and continuity of treatment, maintain the quality of life, and improve the prognosis. The Geriatric Nutritional Risk Index (GNRI) is a nutritional index that classifies “nutritional risk”, broadly defined as the risk of developing malnutrition and poor clinical outcomes when nutritional support is not provided, into four groups ranging from “severe nutritional risk” to “no nutritional risk” (8) for hospitalized older patients, and numerous prognostic factors have been reported to be predictive of outcomes of patients with cancer (9-11). Although there have been reports examining GNRI as a prognostic factor in the treatment of malignant lymphoma (12-16), the association between chemotherapy-induced adverse events and duration of treatment has not been examined. As patients affected by malignant lymphoma are often old, GNRI may be related to the occurrence of adverse events. We investigated the impact of GNRI at treatment initiation on the occurrence of side effects and time to treatment failure (TTF) in patients with malignant lymphoma undergoing initial R-CHOP therapy.
Patients and Methods
A total of 131 patients who underwent initial R-CHOP therapy at Kin-ikyo chuo Hospital between March 2016 and October 2021 were included. A retrospective survey was conducted on the background of the surveyed patients and the occurrence of adverse events based on the records of physicians, nurses, and pharmacists in the electronic medical record.
Patient characteristics including sex, age, weight, body mass index (BMI), number of treatment regimens, average relative dose intensity (ARDI), major laboratory values [lactate dehydrogenase (LDH), serum creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), total bilirubin, serum albumin, hemoglobin, and C-reactive protein (CRP) levels, and white blood cell, lymphocyte, neutrophil, and platelet counts] on the day of administration were investigated.
Adverse events included hematologic toxicity, insomnia, fatigue, nausea, constipation, dysgeusia, decreased appetite, stomatitis, alopecia, peripheral neuropathy, febrile neutropenia (FN), and infusion reaction, as defined by the Common Terminology Criteria for Adverse Events (CTCAE) ver. 5.0. The severity of each case was determined by the highest grade in the entire course.
This study was conducted in accordance with the “Ethical Guidelines for Medical Research Involving Human Subjects” and was approved by the Ethics Committee of Kin-ikyo chuo hospital.
Assessments. GNRI was calculated using the following formula:
GNRI=[14.89×serum albumin level (g/dl)]+41.7×[current weight (kg)/ideal weight (kg)]
Patients were stratified into the high GNRI (GNRI ≥92) and low GNRI groups (GNRI <92) based on the classification of Bouillanne et al. (8), for the comparative study.
TTF was defined as the period from the date of regimen initiation to the date of treatment discontinuation, the date of diagnosis of exacerbation or relapse after completion of treatment, or the date of death due to any reason.
Statistical analysis. The Mann-Whitney U-test was used to compare quantitative data between the two groups including patient background, severity of adverse events, and minimum hematologic toxicity, and Fisher exact probability test was used to compare qualitative data. The significance level for both was less than 5%. Cut-off values calculated from the receiver operating characteristics (ROC) curve were used for the continuous variable factors. TTF was analyzed using the Kaplan-Meier method. Cox proportional hazards regression analysis was used to obtain hazard ratios, 95% confidence intervals (95%CI), and p-values using TTF as the objective variable in the multivariate analysis. Excel Statistics ver. 3.22 (Social Information Service, Tokyo, Japan) was used for the statistical analysis.
Results
Patient characteristics. Patient characteristics and laboratory values at study initiation are shown in Table I. There were 131 patients, with 56 and 75 patients in the high and low GNRI groups, respectively; 4 of the 56 patients (7.1%) in the high GNRI group and 24 of the 75 patients (32%) in the low GNRI group had performance status (PS) ≥3; no patient in the high or low GNRI group who had PS ≥3 was treated with the full dose. The median serum albumin level was 3.0 g/dl for the entire study population, 3.7 g/dl for the high GNRI group and 2.5 g/dl for the low GNRI group. Among the two groups, the low GNRI group had significantly lower body weight, BMI, ARDI, serum albumin level, and hemoglobin level (p<0.05) and significantly higher PS; LDH, AST, ALP, and CRP levels; white blood cell and neutrophil counts at treatment initiation (p<0.05). In contrast, there were no significant differences in sex, age, serum creatinine level, ALT and total bilirubin levels, and lymphocyte and platelet counts.
Characteristics of study population.
Incidence of adverse events. As for hematologic toxicity, the low GNRI group had statistically significantly higher AST increase; hemoglobin decrease; thrombocytopenia of all grades; creatinine, ALP, albumin, and hemoglobin level decrease; neutropenia; and thrombocytopenia of Grades ≥3 (p<0.05) (Table II). The incidence of FN was statistically significantly higher in the low GNRI group than that in the high GNRI group (p<0.05) (Table II). There was no statistically significant difference in adverse events other than hematologic toxicity between the two groups for all grades and Grades ≥3 (Table II).
Adverse events in the patients.
Time to treatment failure. The median (interquartile range) TTF for all patients was 10.1 months (range=3.9-28.7 months). The medians (interquartile ranges) for the high GNRI and low GNRI groups were 16.8 (5.7-36.3) and 6.9 (3.0-20.5) months, respectively, with a significantly longer TTF in the high GNRI group (p=0.045) (Figure 1).
Comparison of time to treatment failure (TTF) between High Geriatric Nutritional Risk Index (GNRI) and Low GNRI groups. The median (interquartile range) TTF for all patients was 10.1 months (range=3.9-28.7 months). The medians (interquartile ranges) for the High GNRI and Low GNRI groups were 16.8 (5.7-36.3) and 6.9 (3.0-20.5) months, respectively, with a significantly longer TTF in the high GNRI group (p=0.045).
Univariate and multivariate COX regression proportional hazards analysis of TTF risk factors. The results of univariate and multivariate COX regression proportional hazards analysis of TTF risk factors are presented in Table III. In the univariate analysis, there were significant differences in BMI, PS, ARDI, serum albumin level, hemoglobin level, and GNRI at the initiation of the regimen. The areas under the ROC curves for BMI, PS, ARDI, serum albumin level, hemoglobin level, and GNRI were 0.557, 0.669, 0.639, 0.648 and 0.525, with cutoff values of 21.7 kg/m2, 2, 60.5%, 2.8 g/dl and 10.6 g/dl, respectively. The cutoff value for GNRI was set at 92. A multivariate Cox regression proportional hazards analysis was performed based on these results, and significant differences were found in PS (HR=2.34, 95%CI=1.09-5.50, p=0.028), serum albumin level (HR=0.54, 95%CI=0.31-0.93, p=0.029), and GNRI (HR=0.96, 95%CI=0.93-0.99, p=0.007). The factors affecting treatment duration were PS (≥2), serum albumin level, and GNRI.
Univariable and multivariable analysis with Cox proportional hazards model for time to treatment failure in malignant lymphoma patients treated with R-CHOP.
Discussion
We investigated the relationship between GNRI as an index of adverse events and TTF in patients with malignant lymphoma who received R-CHOP therapy. Although we defined the cutoff value of GNRI as 92, the cutoff value of GNRI was defined as 95.7 in another report (11) evaluating the relationship between OS and nutritional assessment indices in 615 patients with DLBCL and as 96.8 in another report (13) evaluating OS and PFS in 476 older patients with DLBCL. The cutoff values for the incidence of adverse drug reactions according to the GNRI, which was the objective of this study, and the cutoff value for predicting long-term prognosis in patients with DLBCL were evaluated at similar values. Although the GNRI has been developed for older patients, some reports (11, 13) have examined prognostic factors in patients with DLBCL, including young patients, and evaluation according to the GNRI is also applicable to young patients. The median age of the study subjects was 77 years, and the GNRI was effective in assessing the nutritional status and predicting prognosis and adverse effects in cancer types that tend to occur more frequently in the older population, such as malignant lymphoma (1).
In terms of patient background at regimen initiation, the low GNRI group had a significantly higher PS and LDH and CRP (p<0.01) levels and significantly lower BMI and ARDI (p=0.011) than the high GNRI group (Table I). This suggests that the low GNRI group had more disease progression at the start of the regimen than the high GNRI group and that the treatment intensity of the regimen was lower due to deterioration in the general condition and weight loss.
The incidence of FN was significantly higher in the low GNRI group compared to that in the high GNRI group (p<0.05). The ASCO and NCCN guidelines list several risk factors for the incidence of FN, including advanced age (>65 years), poor PS, and poor nutritional status (17, 18). In this study, the low GNRI group had normal renal and hepatic functions in the beginning, but the median age was 79 years, and the PS was significantly higher than that of the high GNRI group. The risk factors for the incidence of FN in the low GNRI group were advanced age, poor PS, and poor nutritional status. In addition, low body weight and hypoalbuminemia are risk factors for FN in several chemotherapy regimens (19-24). In this study, the low GNRI group also had significantly lower serum albumin levels and body weight at treatment initiation (Table I) and significantly more FN (Table II). FN is more likely to occur in patients in the low GNRI group due to their low albumin level and low body weight. Therefore, a GNRI <92 may be a risk factor for the incidence of FN.
In terms of hematologic toxicity, there were significant differences between the high GNRI and low GNRI groups in hemoglobin decrease, neutropenia, and thrombocytopenia in both the overall grades and Grade ≥3. The low GNRI group had a significantly lower hemoglobin level than the high GNRI group at treatment initiation, and the anemia may be even more severe under the effects of chemotherapy-induced bone marrow suppression and hemolysis. Multiple studies have reported that anemia causes tumor progression to malignancy and resistance to treatment and is an important prognostic factor for treatment efficacy and survival in patients with cancer (25-27). In this study, hemoglobin decrease was significantly more common in the low GNRI group, suggesting that GNRI may be a predictor of high risk for anemia. The incidence of neutropenia was significantly higher in the low GNRI group at Grade≥3, consistent with the high incidence of FN in the low GNRI group. In this study, the proportion of Grade≥3 thrombocytopenia was 34.7% in the low GNRI group, which was higher than the 7% reported in 604 patients in a study of patients with DLBCL (28). Cyclophosphamide and doxorubicin, which are included in the R-CHOP regimen, are known to cause thrombocytopenia; furthermore, thrombocytopenia may occur with disease progression; therefore, more attention should be paid when GNRI is <92.
Although this study focused on malignant lymphoma, undernutrition was correlated with an increased risk of serious hematologic toxicity in patients with advanced solid tumors (19). More severe side effects involving hematologic toxicity occur in patients with malignant lymphoma in the low GNRI group.
The multivariate analysis showed that PS, serum albumin level, and GNRI at regimen initiation were factors influencing treatment duration. An analysis of data from 3,047 patients enrolled in multiple chemotherapy protocols reported that prechemotherapy weight loss was associated with significantly shorter median survival of patients with weight loss (5). Weight loss leads to a decrease in BMI, which may affect treatment duration. The prognostic factor analysis of the R-CHOP regimen revealed that PS significantly affected TTF and OS (29), and a good PS at the start of treatment may lead to treatment that maintains ARDI and prolongs TTF. Hypoalbuminemia of <3.7 g/dl is an independent factor for survival in patients with DLBCL (30) and that of less than 3.2 g/dl affects TTF in the treatment of peripheral T-cell lymphomas (31). The serum albumin levels significantly differed at the beginning of treatment between the high and low GNRI groups. Furthermore, hypoalbuminemia Grade≥3 significantly differed, and the TTF was significantly longer in the high GNRI group than that in the low GNRI group, which corroborates a previous report (31) that revealed that hypoalbuminemia affects TTF.
Inflammatory cytokines and growth factors are released as part of the systemic inflammatory response to the tumor. Albumin production may be inhibited due to the production of cytokines such as IL-6, which regulates the production of albumin by hepatocytes, resulting in a decrease in serum albumin concentration (32). Hypoalbuminemia is known to be associated with inflammation. Hepatocytes may restore antioxidant activity by decreasing or resynthesizing albumin, and albumin synthesis and anti-inflammatory functions may occur depending on the severity of inflammation (33). However, albumin degradation also increases in inflammatory conditions, and systemic inflammation may lead to hypoalbuminemia (33). The American Society for Parenteral and Enteral Nutrition (ASPEN) stated in an October 2020 position paper (34) that low serum albumin is an inflammatory marker associated with nutritional risk, broadly defined as the risk of developing malnutrition and/or poor clinical outcomes when nutritional support is not provided. The results of this study indicate that GNRI is a prognostic factor affecting TTF for the primary treatment of malignant lymphoma. The formula for calculating the GNRI includes serum albumin level (8), and it is thought that low albumin level results in a lower GNRI, which indicates severe or moderate nutritional risk. When such nutritional risks are present, appropriate interventions such as the aggressive use of oral nutritional supplements or enteral feeding are necessary (35).
One limitation of this study is that it was conducted at a single institution, and the number of cases is limited. Therefore, it is necessary to increase the number of cases to extract more data and examine associations more accurately.
Conclusion
In patients undergoing R-CHOP therapy, GNRI <92 at regimen initiation increased the risk of developing FN and hematologic toxicity. PS, albumin level, and GNRI at the start of the regimen were factors affecting treatment duration. Nutritional status at treatment initiation may influence the development of hematologic toxicity and TTF.
Footnotes
Authors’ Contributions
Ken Kikuchi conceived the idea of the study and drafted the original manuscript. Momoka Sawada made significant contributions to the design of the study and the interpretation of data. Genki Takeda, Taiki Watanabe, and Takafumi Souma contributed to the interpretation of the results. Hideki Sato supervised this study. Hideki Sato and the other authors substantially contributed to the revision of manuscript drafts. All Authors have approved the submitted version of the manuscript and agreed to be accountable for any part of the study.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received March 1, 2023.
- Revision received March 13, 2023.
- Accepted March 14, 2023.
- Copyright © 2023 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.