Cancer prevalence has increased due to the expanding elderly population in many countries (1), and cancer is now the leading cause of death worldwide (2). Cancers are sometimes detected at incurable stages, and palliative chemotherapy is administered to prolong survival (3,4). However, progressive cancer causes systemic declines and decreasing chemotherapy tolerability, and leads to a poor prognosis (5). Cachexia, which is characterized by a decrease in skeletal muscle and adipose tissue, impairs physical activity and quality of life (6). Moreover, anti-cancer treatment can cause adverse events that impair systemic conditions (7). These situations lead to a decline in both physical activity and nutrition in patients with cancer.

While significant efforts have been made to advance anti-cancer therapy, such as with surgery, radiotherapy, chemo-therapy, and immunotherapy, the importance of supportive care should also be emphasized. The field of supportive care is broad, involving many types of healthcare providers, such as physicians, nurses, physical therapists, nutritionists, and others. Rehabilitation contributes to maintaining muscle strength and the physical condition of advanced cancer patients (8-10). Additionally, malnutrition, which is common in advanced cancer patients, could negatively affect the outcome of anti-cancer treatment (11,12). Several guidelines suggest that nutritional assessment should be provided with rehabilitation (13-15).

Although the results of two phase II trials have supported the efficacy of a multimodal intervention that involves physical and nutritional evaluation and corresponding intervention (16,17), each trial included only approximately 40 patients. There are still limited numbers of studies of combined nutritional and physical care. Moreover, multimodal assessment and intervention require efforts from patients and medical staff, and may involve burden for patients and financial constraints (18). Therefore, it might be difficult for all palliative cancer patients to receive such intensive care Furthermore, there are still a limited number of studies regarding which population of advanced cancer patients is more likely to receive benefits from nutritional care combined with rehabilitation. Therefore, it is important to identify the population of advanced cancer patients who are more likely to benefit from intensive nutritional care.

This retrospective study evaluated whether the efficacy of rehabilitation differed between patients with sufficient caloric intake and those with insufficient intake. We also evaluated patients whose physical status improved after rehabilitation and determined whether certain nutritional indices improved concurrently.

Study design/Patients. This was a single-center, retrospective observational study. We examined patients diagnosed with advanced cancer who underwent chemotherapy and received cancer rehabilitation in the Medical Oncology Department of Sasebo Kyosai Hospital from April 2017 to July 2021. Rehabilitation was provided only to hospitalized patients because medical insurance in Japan covers cancer rehabilitation only during hospitalization. Rehabilitation was initiated in accordance with physicians’ orders and terminated by orders or discharge. When patients received physical rehabilitation several times in their clinical course, we examined the first rehabilitation and its results. Patients who died during hospitalization without discharge after receiving rehabilitation were excluded. In this study, the term “rehabilitation” meant exercise training 1) ordered by physicians, 2) managed and provided by physiologists, and 3) covered under the Japanese medical insurance system.

Analysis 1. Patients were categorized into two groups on the basis of their nutritional status: a Well group, comprising those whose caloric intake was above or equal to basal metabolism calculated by the Harris–Benedict equation (19), and a Poor group, consisting of individuals whose caloric intake was below their basal energy expenditure. In these defined criteria, caloric intake referred to the dietary calories on the day rehabilitation was initiated, as documented in the medical record.

We assessed the Eastern Cooperative Oncology Group Performance Status (ECOG PS) (20) and Barthel Index (BI) (21) of patients before and after rehabilitation as physical markers. ECOG PS is a scale to measure patients’ systemic condition and ranges from Grade 0 (fully active) to Grade 5 (dead). The Barthel index (BI) assesses a patient’s activities of daily living (ADL), and a full score is 100 points, with points deducted in 5-point increments by ADL. Changes in ECOG PS or BI were evaluated by comparing these indices before and after rehabilitation and were categorized as improved, stable, or worsened. In this study, we also defined the term “maintained”, to signify that a patient’s ECOG PS or BI was either improved or stable. The study endpoint was the rate of patients whose ECOG PS was maintained, and the rate of patients whose BI was maintained.

The change in nutritional status was analyzed in patients whose nutritional data after rehabilitation were available. We evaluated body mass index (BMI), serum albumin, and the geriatric nutritional risk index (GNRI) as nutritional markers. GNRI is a tool to evaluate the nutritional risk for older people and is calculated as follows (22):

GNRI = 1.489 × serum albumin (g/l) + 41.7 × body weight (kg) ÷ ideal body weight (kg)

Analysis 2. While we focused mainly on the maintenance of physical status, we performed an additional analysis to identify the factors associated with improvement in physical status. We restratified the cohort into Responders and Non-responders groups. Responders were defined as patients whose ECOG PS or BI improved and whose ECOG PS and BI did not worsen, during rehabilitation; Non-responders comprised the remaining patients. The endpoint was a comparison of nutritional status, using serum albumin, BMI, and GNRI. Caloric intake was also evaluated, as described earlier.

Statistical analysis. Regarding the patients’ characteristics, all p-value evaluations were two-tailed. We used the χ² test for categorical data and the Mann–Whitney U-test for continuous data. All data were analyzed with R, version 4.2.1. (R Foundation for Statistical Computing, Vienna, Austria).

Ethics approval. This study was approved by the Institutional Review Board of the Sasebo Kyosai Hospital. All data acquisition and analysis were conducted after the approval. We complied with the Declaration of Helsinki in this study. This observational study complies with The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.

Informed consent. Because this was a retrospective study, written informed consent was not required. We disclosed explanatory documents available to patients and their families on our hospital’s website and excluded any patients who requested to be excluded from this study. We also did not obtain consent for publication from individuals because of the retrospective design of this study.

Analysis 1. In our department, 101 patients received chemotherapy and rehabilitation by physical therapists during the designated period. After excluding 17 patients who died during hospitalization without being discharged after receiving rehabilitation, 84 patients were evaluated for their physical and nutritional statuses (Figure 1).

The patients’ background characteristics are shown in Table I. Of the total population, 51 patients comprised the Well group and 33 patients comprised the Poor group. Compared with the Poor group, the Well group had a higher frequency of previous surgeries, shorter duration of rehabilitation days, and lower median serum albumin. There was no significant difference in median BMI and GNRI between the groups. Background ECOG PS tended to be better in the Well group compared with the Poor group; background BI was similar between the groups.

In the Well group, ECOG PS was maintained in 50 patients and worsened in one, while BI was maintained in all 51 patients. In the Poor group, ECOG PS was maintained in 30 patients and worsened in three patients; BI was maintained in 29 patients and worsened in four patients. The PS-maintained rate was 98% and 91% (p=0.29) (Table II), and the BI-maintained rate was 100% and 88% (p=0.02) in the Well and Poor groups, respectively (Table III).

Analysis 2. To analyze the nutritional status, we excluded 12 patients, namely five patients from the Well group and seven patients from the Poor group, whose nutritional data after rehabilitation were unavailable. As a result, we evaluated nutritional change after rehabilitation in 72 patients, namely 46 patients in the Well group and 26 patients in the Poor group (Figure 1).

We subsequently restratified the 72 patients who received a nutritional evaluation. There were 17 Responders and 55 Non-responders in the cohort (Table IV). Data showed that the Responders had a longer duration of rehabilitation days. ECOG PS and BI were significantly worse in the Responders group compared with the Non-responders group, while BMI was similar between the groups. Serum albumin and GNRI before rehabilitation were significantly lower in the Responders compared with the Non-responders groups.

Data showed that caloric intake before rehabilitation was sustained in the Responders group (Table IV). Serum albumin and GNRI before rehabilitation were worse in the Responders compared with the Non-responders groups; BMI before rehabilitation was similar in both groups. Changes in the nutritional markers after rehabilitation between the Responders and Non-responders groups are presented in the boxplots in Figure 2. The medians and p-values for the comparison of the markers for the Responders and Non-responders groups, respectively, were as follows: 181.7 kcal and 0 kcal (p=0.28) for caloric intake; 0.4 g/dl vs. 0 g/dl (p=0.15) for serum albumin; 0.64 kg/m² vs. 0.52 kg/m² (p=0.94) for BMI; and 4.98 vs. −1.69 (p=0.12) for GNRI (Figure 2). These results indicated no statistically significant differences between the groups.

To achieve better quality of life and reduce toxicity in cancer treatment, it is important to maintain the systemic condition of cancer patients. Rehabilitation provides important supportive care to maintain the physical ability of cancer patients and is recommended during chemotherapy (8). Nonetheless, the nutritional status of cancer patients tends to decline (23). This led to our hypothesis that insufficient caloric intake could impair rehabilitation and maintenance of physical activities in advanced cancer patients. We categorized advanced cancer patients into two groups according to their caloric intake to assess their physical status during rehabilitation.

In this study, we evaluated ECOG PS and BI as markers of physical status. ECOG PS is a widely used marker to indicate systemic conditions and tolerability for chemotherapy in cancer patients. BI is a standard tool to assess the ADL of patients and ranges from 0 to 100 points with 5-point increments. BI evaluates abilities, such as eating, moving, and walking. Physical training has been shown to maintain BI in cancer patients and has a prognostic value in patients with brain tumors (24,25). In this study, we were able to collect both ECOG PS and BI data from electronic medical records, which enabled the evaluation of changes in physical status using different indices.

We stratified eligible patients on the basis of caloric intake on the first day of rehabilitation. Caloric intake is lower in cancer patients compared with healthy people (16,17,26). The Harris–Benedict equation was used to estimate basal energy expenditure (BEE) (19), which was the cutoff point in this study. Although total energy expenditure is calculated by the product of BEE, an activity factor, and a stress factor (27), we did not evaluate these factors. Setting BEE as a cutoff point was considered reasonable to prevent an overly large Poor group. As a result, the caloric intake in the Poor group was determined to be below BEE for each patient.

Our results showed that the Poor group benefited less from rehabilitation compared with the Well group. Moreover, caloric intake appeared to be lower in the Non-responders group compared with the Responders group, even though other nutritional background data tended to be better in the Non-responders group. This supports our initial hypothesis that low caloric intake might negatively impact the physical benefits of rehabilitation. While previous studies proposed potential benefits of multimodal intervention (16,17,28), there were problems with the adherence rate for in-home therapy, burdens on patients and social finances, and the workload of healthcare providers. Our results suggest that caloric intake could be a predictive marker to identify patients who would benefit from intensive nutritional care during rehabilitation.

Study limitations. First, this was a retrospective study, and the patients’ backgrounds differed in the Well and Poor groups. Factors other than caloric intake could have been associated with the worse results in the Poor group compared with the Well group. Second, food intake and physical activity were affected by both cachexia and its treatment. As appetite can decrease in response to various factors, such as cachexia and adverse events associated with chemotherapy, caloric intake may not be the best marker to identify a population that needs intensive supportive care. Finally, patients received only short-term exercise intervention, and the median duration of rehabilitation days was longer in the Poor group and the Responders group compared with the other groups. This is likely because patients in the Poor group had poorer nutritional statuses and required longer hospitalization, and because longer duration of rehabilitation might have contributed to improvement in physical status in the Responders group compared with the Non-responders group. Of note, patients in the Well group, whose caloric intake was high, did not show worse outcomes in the maintenance rate for both ECOG PS and BI, despite shorter durations of exercise and observation.

In summary, we examined the impact of nutritional status on the efficacy of rehabilitation in advanced cancer patients. Prospective verification of our results is necessary to evaluate the advanced cancer patient population in more detail.

Adequate caloric intake, accompanied by rehabilitation, might contribute to the maintenance of physical status for advanced cancer patients.

E.B. received a personal fee from Chugai and Daiichi-Sankyo. The other Authors have no competing interests or financial disclosures to declare in relation to this study.

All Authors contributed to the study’s conception and design. TI, YO, and HU performed data collection. TI and TS analyzed data. TI wrote the first draft of the manuscript. All Authors checked and commented on previous versions of the manuscript. All Authors read and approved the final manuscript.

The Authors appreciate all the patients, their families, and medical workers in Sasebo Kyosai Hospital who contributed to this study. The Authors thank Jane Charbonneau, DVM, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

The Authors declare that no funds, grants, or other support were received during the preparation of this manuscript.