Background
Digestive cancers represent most of cancer spectrum in the world [
1] with cytotoxic drugs and targeted biotherapy mainly used. Cancer cachexia is a multifactorial syndrome defined by an ongoing loss of skeletal muscle (with or without loss of fat mass) that cannot be fully reversed by conventional nutritional support and leads to progressive functional impairment (loss of muscle strength) [
2]. Sarcopenia is defined as the age-related loss in skeletal muscle mass, and function (dynapenia [
3] or performance) [
4,
5]. Historically, calculation of anti-cancer treatment dose used body surface area (BSA) formula [
6] depending on patient’s weight and height. However, the objective at this time was not to develop a formula to dose anti-cancer drugs. Furthermore, body composition has emerged as an important predictor of anti-cancer drug efficacy and toxicity [
7‐
13]. In 2013, a systematic review led by Prado et al. has described a correlation between low lean tissue and severe toxicities of antineoplastic agents due to plasmatic surexposure [
14]. A growing literature suggests that the lean tissue compartment may be used to adjust dosage of drugs that are distributed in and metabolized by lean tissues [
14‐
17]. Loss of muscle mass occurs in 80% of patients with cancer [
18] but cannot be interpreted without its function, that is muscle strength.
Handgrip Strength (HGS) has been validated in geriatric and surgical studies. In oncology, HGS is associated with cancer-related fatigue, poor quality of life, loss of functional status in hospitalized patients, postoperative complications, length of hospital stay and short term survival [
19‐
23]. It has also been included in the consensual definition of sarcopenia, which associates muscle mass loss and weak strength [
5]. Muscle mass is one of the determinants of muscle strength [
24], but needs invasive and costly exams to be estimated (X-ray absorptiometry(DXA), anthropometry and bioelectrical impedance analysis (BIA), magnetic resonance imaging (MRI) or computerized tomography (CT)). HGS may be a low cost method to detect sarcopenia in clinical settings and by extrapolation to predict anticancer drug toxicity in patients with advanced cancers [
25]. Recently, Ordan et al. [
23] showed in FIGHTDIGO study the feasibility and acceptability of HGS measure with a Jamar dynamometer in routine. The present study focuses on the second aim of FIGHTDIGO study: the association between pre-therapeutic dynapenia and chemotherapy-induced Dose-Limiting Toxicities (DLT) in digestive cancer patients treated with ambulatory chemotherapy.
Methods
Study design and participants
This ancillary prospective monocentric study was conducted in the Ambulatory Cancer Unit (UMA-CH) of the Reims teaching hospital in France. The study population of FIGHTDIGO included patients older than 18 years-old, having a primary digestive cancer regardless of stage, and undergoing cytotoxic chemotherapy and/or biotherapy for digestive system cancers. All cytotoxic chemotherapy and/or biotherapy regimens according to guidelines were allowed. Patients who could not give their consent, did not understand the handgrip test, had history of neuro-muscular disorder and/or had appointed a health care proxy were excluded. The patients were recruited from May 18, 2016, to November 18, 2016, and were followed for 6 months. For this study on chemotherapy-induced DLT, in order to avoid an overestimation of toxicities in pre-treated patients, only the sub-group of chemotherapy-naïve patients were included.
Ethical approval
Informed written consent was obtained for each enrolled patient in the trial. The FIGHTDIGO study was approved by the ethics committee (Committee for the Protection of Person EST I DIJON, 25 March 2016) and was registered in Clinicaltrials.gov (
NCT02797197).
Outcome
The aim was to evaluate pre-therapeutic dynapenia as a predictive factor of chemotherapy-induced DLT and/or Dose-Limiting Neurotoxicity (DLN). Other potential predictive factors were analyzed (patient, tumoral, nutritional (as modified Glasgow Prognostic Score (mGPS)) and therapeutic characteristics).
Handgrip strength (HGS) measurement
HGS was measured, at first hospital stay, before chemotherapy administration, in non-dominant and dominant hand using a hydraulic Jamar Dynamometer. Position 2 was used among the 5 possible handle-positions. Patients performed the test while sitting comfortably (feet touching ground) with shoulder adducted, forearm neutrally rotated and elbow flexed to 90°. The other upper limb was placed alongside the body and relaxed. Patients were instructed (by physicians) to perform maximal isometric contraction within 3 s in both hand. Four measurements were taken. Two measurements were determined for each hand alternatively. A one-minute break was respected between two measurements with the same hand. The highest value (from the four measurements obtained) was chosen for final evaluation. According to the European Working Group on Sarcopenia, dynapenia was defined as HGS < 30 kg (men) and < 20 kg (women) [
5].
Chemotherapy-induced dose-limiting toxicities (DLT) and dose-limiting neurotoxicity (DLN)
Data were prospectively recorded in medical file at each hospital stay. Chemotherapy-induced DLT (all non-neurological toxicities) were defined as any toxicity leading to dose reduction (temporary or permanent), treatment delays or permanent treatment discontinuation. Chemotherapy-induced DLN was defined as permanent peripheral neuropathy (Levi scale grade 2 or 3) leading to dose reduction or permanent treatment discontinuation. DLN and progressive disease as the cause of treatment termination were not considered as DLT. Pre-therapeutic dose adaptation was defined as an initial dose reduction by individual clinical appreciation taking into account patient profile (age, ECOG Performance Status (PS), organ failure, malnutrition or mGPS). Malnutrition was defined as BMI < 21 kg/m
2 in patients aged more than 70 years old and BMI < 18.5 kg/m
2 in patients aged less than 70 years old [
26].
Statistical analysis
Data were described using mean ± standard deviation for quantitative variables and numbers (percentage) for qualitative variables.
The survival curves were established by the Kaplan-Meier method. Prognostic factors were identified by univariate analysis using the log rank tests and by multivariate analysis using a Cox proportional hazard model. Factors significant at the 0.20 level in univariate analysis were included in a stepwise regression multivariate analysis with entry and removal limits set at 0.20. Concerning analysis of factors associated with chemotherapy-induced DLN, neurotoxic chemotherapy was not included in multivariate analysis because of one strata has no event (convergence wasn’t satisfied) and BMI was not included in multivariate analysis because of multicollinearity. Comparisons between subgroups of patients were performed using Chi 2 test, Fisher’s exact test, Student’s t test or Wilcoxon test, as appropriate.
All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).
Discussion
To our knowledge, this prospective study is the first to show that pre-therapeutic dynapenia (low HGS testing with a Jamar dynamometer) could be predictive of chemotherapy-induced DLN in digestive cancer patients. Majority of participants received a neurotoxic regimen which enhanced these findings. Stomach, biliary tract and small intestine cancers were significantly associated with DLN in unadjusted and adjusted analysis because of receiving exclusively neurotoxic chemotherapy. There was no association of dynapenia with chemotherapy-induced DLT. This might be explained by several elements: efficacy of supportive cares, good safety profile of chemotherapy regimen in digestive cancers (as FOLFOX regimen [
27]), favorable prognostic profile of our population [
1] and a small sample size. Conversely, in a pilot study in 112 older newly diagnosed cancer patients with a minority of digestive cancer (
n = 20), low grip strength predicted toxicity at 3 months [
28]. In our study, as none significant association of chemotherapy-induced DLT could be detected with age, malnutrition, BMI, mGPS and stage, and ECOG PS 1 or 2 was found to be a protective factor of DLN, heterogeneity of population and pre-therapeutic dose adaptation of chemotherapy might be an explanation of discrepancies. Most of oncology trials excluded patients with ECOG PS > 1 [
29‐
31]. A bad condition at inclusion could be a bias for evaluation and limit the scope of potentially tested treatments. Modified GPS is an inflammation-based prognostic score. It has been shown to be correlated with mortality in colon cancer [
32] influenced by malnutrition and predicting toxicity [
33]. In present study, a small sample size with mGPS = 2 and/or ECOG PS 2 might be responsible of non-significant results.
Cancer cachexia is a recent concept that has been defined with the emergence of the body composition consideration in oncology [
18] [
34]. Prado et al. has reported the association of body composition with chemotherapy-induced DLT in stage II-III colon cancer patients treated with 5FU and leucovorin [
35]. These results had consequences on phase I studies with many works on relationship between sarcopenia/body composition and DLT [
9,
11]. Body composition and sarcopenia seem to be prognostic of enhanced chemotherapy and targeted therapy toxicity [
7,
8,
10,
11,
13], and more recently with immune checkpoint inhibitors such as anti-CTLA-4 (ipilimumab) [
12].
Lean Body Mass (LBM) was associated with chemotherapy-induced DLT and DLN in patients with colon cancers treated with FOLFOX regimen [
36]. Neurotoxicity affects 80% of patients and becomes chronic in 15–20% of cases, sometimes irreversibly [
27]. A recent systematic review [
37] has reported the impact of acute oxaliplatin-induced neuropathy and necessity of a large prospective study to established better preventive guidelines. In 2 cohorts of colorectal cancer patients treated with oxaliplatin, Ali et al. [
36] reported that overall DLT, and specifically oxaliplatin-induced neuropathy, occurred mostly in patients who receive > 3.09 mg/oxaliplatin/kg LBM. One study is ongoing to determine whether the oxaliplatin dose normalization based on the LBM index can prevent or reduce neurotoxicity associated with oxaliplatin, for colorectal cancer patients treated in adjuvant with FOLFOX regimen (LEANOX trial NCT03255434). Our findings might be an additional argument for this hypothesis. Dynapenia, which is the first step toward sarcopenia, may represent an interesting tool to estimate sarcopenia / LBM [
38]. It might be a more easily accessible predictive marker for oxaliplatin dose adaptation because patients with low muscle mass behave like patients “overdosed” with chemotherapy resulted in dose-limiting toxicities, independently of the patient’s weight.
In present study, there were 31% of overweight patients. Recently, it has been reported that overweight sarcopenic patients experienced more DLT [
12,
39,
40] than overweight patients with normal LBM. A potential explanation could be an altered volume distribution, metabolism or clearance in lean tissue of anticancer drugs in these patients [
41]. Dose of cytotoxic drugs has been shown to be correlated with LBM [
35]. These findings suggest a new paradigm in the future: body composition measurement leading to dose adaptation of antineoplastic agents.
Gray et al. have reported a prediction model used to identify sarcopenia based on parameters of functional fitness [
42]. Correlation between measure of dynapenia and skeletal muscle index by computed tomography in a cancer population are being evaluated in a further ancillary analysis of FIGHTDIGO study.
HGS measure is a useful low-cost method [
23] which could improve clinical practice by stratifying frailty patients who experienced more DLN. More relevant nutritional and adapted physical activity programs might be needed.
A major limitation of this ancillary study is the size of cohort with only 45 patients with various type of digestive cancers and heterogeneous chemotherapy regimens. Several aspects of using the Jamar dynamometer may require further investigation. Many studies are attempting to determine cut points for dynapenia with respect to age, sex, and ethnic group [
5,
43,
44]. The HGS dynamometer thresholds remain undetermined for cancer patients.