Stephanie Looijaard

The high prevalence of treatment-related complications and low survival rates in older individuals with cancer compared to younger individuals (12, 13) call for a different assessment and treatment approach in geriatric oncology. High expectations have been set on skeletal muscle status as a predictor of poor clinical outcomes and target for interventions to improve clinical outcomes in older individuals with cancer. In this thesis, the umbrella term (skeletal) muscle status was used to encompass measures of muscle mass, muscle quality, muscle strength and physical function and performance and poor muscle status was used to describe limitations or deficits in any of these muscle measures. Muscle status is of specific interest in geriatric oncology as its encompassing muscle measures can be perceived as reflections of overall health status and multiple organ systems (29, 30) and approximately 40% of individuals with cancer have a poor muscle status (31). The aims of this thesis were to elucidate the role of muscle status prior to cancer diagnosis (Part I), of possible underlying pathophysiological mechanisms between muscle status and clinical outcomes and targeted interventions (Part II) and the role of muscle status in the prediction of poor clinical outcomes (Part III), in older individuals with cancer.

The role of muscle status prior to cancer diagnosis
When muscle status falls below a clinically relevant threshold, this is considered a disease known as sarcopenia, which is defined as the age-related loss of muscle mass and muscle strength (38, 39). In Chapter two it was questioned whether older individuals at risk of sarcopenia could be identified by assessing a single muscle measure such as handgrip strength, balance or a physical function or performance measure in a cohort of geriatric outpatients. Facilitating the identification of older individuals with sarcopenia would also be highly useful in geriatric oncology. Several single muscle measures were associated with sarcopenia according to at least two common definitions but diagnostic accuracies were poor, indicating that a single muscle measure cannot identify older individuals at risk of sarcopenia.
The prevalence of poor muscle status is notably higher in older individuals with cancer compared to older individuals without cancer (31, 45-47). In the LASA cohort, we aimed to find out whether this could be the result of cancer activity before diagnosis (Chapter three). No differences between handgrip strength and physical function and performance measures were found between older individuals who were to be diagnosed with cancer and older individuals without cancer. Effects of cancer activity on muscle status before diagnosis could therefore not be confirmed in our study and might be dependent on the tumor burden and time period between cancer diagnosis and muscle status assessment.

The role of pathophysiological mechanisms underlying muscle status and poor clinical outcomes
In Chapter four, we provide an overview of potential pathophysiological mechanisms involved in sarcopenia and cachexia that might explain why low muscle mass would lead to a higher risk of poor clinical outcomes such as surgery- and chemotherapy-related complications and lower survival in older individuals with cancer. Altered myokine balance and pharmacokinetics of anticancer drugs were identified as mechanisms that could explain the link between low muscle mass and poor clinical outcomes. We consider glucose intolerance and circulating mitochondrial DNA to be topics of interest for future studies. However, due to reverse causation and the interplay with cancer it is impossible to determine whether these mechanisms are a direct consequence of low muscle mass. Physical exercise, neuro-muscular electrical stimulation and pharmacologic compounds reproducing the beneficial effects of exercise (exercise mimetics) may be targeted interventions to tackle the negative effects of these mechanisms and improve clinical outcomes in older individuals with cancer.
In the systematic review in Chapter five, prehabilitation programs consisting of preoperative physical and/or nutritional interventions were found to be ineffective in reducing postoperative complications, length of stay, readmission rates and mortality in older individuals with colorectal cancer. The lack of selection of individuals with physical limitations and/or nutritional deficits in the included studies is expected to have limited the effects of the interventions.

The role of muscle status in predicting poor clinical outcomes in older individuals with cancer
In the final part of this thesis, we were unable to demonstrate a role of muscle measures assessed by Comprehensive Geriatric Assessment (CGA) or total abdominal Computed Tomography (CT) scan analysis in predicting poor clinical outcomes in older individuals with respectively various solid cancers and colorectal cancer. In the prospective MOOP study, CGA was incorporated into oncological care for older individuals (Chapter six) to identify problems in the geriatric domains of comorbidity, polypharmacy, mood, cognition, nutrition and physical function and performance. The results of the MOOP study underline the relevance of conducting a CGA. CGA identified problems in more than one potentially modifiable geriatric domain in 70% of individuals and the medical oncologist had doubts about the expected tolerance of standard anticancer treatment in almost 50%. Poor clinical outcomes including chemotherapy toxicity, dose reduction, postponement of treatment, death before the start of treatment and early progression, occurred in 60% of individuals who received anticancer treatment but could not be predicted by problems in more than one CGA domain, problems in physical function and/or performance or by clinical doubts. It should be noted that this study is limited by the small cohort with heterogeneous cancer types and stages. The results highlight the challenges that remain to be faced in geriatric oncology and underscore the value of CGA to detect potentially modifiable geriatric problems that would otherwise have remained undiscovered. Based on the results of our study, the ability of CGA to predict poor clinical outcomes in a heterogeneous group of individuals with various cancer types and stages remains elusive.
The aim of the PREMUSCLE study was to clarify the role of preoperative total abdominal CT-based body composition measures in the prediction of poor clinical outcomes in older individuals with colorectal cancer (Chapter seven). In a large cohort of almost 400 older individuals with colorectal cancer, severe postoperative complications occurred in 13% of individuals, more than one-third died during five years of follow-up and 77% experienced dose-limiting chemotherapy toxicity. Continuously-measured CT-based body composition measures were not consistently associated with these poor clinical outcomes, and any associations disappeared after accounting for multiple testing. Thus, preoperative total abdominal CT-based body composition measures failed to predict poor clinical outcomes in older individuals with colorectal cancer in our study. A dedicated meta-analysis on continuous total abdominal CT-based muscle measures in relation to poor clinical outcomes in older individuals with colorectal cancer is warranted to draw a final conclusion on the predictive value of CT-based muscle measures in this population.
In Chapter eight we included older individuals with colon cancer from the PREMUSCLE cohort to investigate whether total abdominal muscle measures determined by CT scan analysis can be assumed representative of specific abdominal muscle groups. Large variety between muscle measures of specific abdominal muscle groups on a population level and within individuals was found, indicating that total abdominal muscle measures are not representative of specific abdominal muscle groups. Lower muscle density and a higher percentage of intermuscular adipose tissue of the lateral muscles were associated with severe postoperative complications, although the value of these results is limited due to the low number of individuals with complications. CT scan analysis of specific abdominal muscle groups may be of added value to predict poor clinical outcomes in older individuals with (colon) cancer but needs further underpinning by larger, longitudinal studies.

Clinical implications and future research
Anticancer management and treatment in older individuals call for a different approach from current oncological care, as highlighted by the high prevalence rates of treatment-related complications and low survival rates presented in this thesis. The high expectations of the role of muscle status to recognize cancer activity before diagnosis, of prehabilitation to improve clinical outcomes, and of muscle status assessed by CGA or total abdominal CT scan analysis in the prediction of poor clinical outcomes in older individuals with cancer could not be confirmed by the results presented in this thesis. Altered myokine balance, altered pharmacokinetics of anticancer drugs and other pathophysiological mechanisms can provide directions for interventions to improve clinical outcomes in older individuals with cancer and low muscle mass. CGA is still considered to be an essential element of oncological care for older individuals due to its value in identifying potentially modifiable geriatric problems. Finally, CT scan analysis of specific abdominal muscle groups might be of added value to predict poor clinical outcomes in older individuals with (colon) cancer but requires further substantiation in larger, longitudinal cohorts.
To conclude on the role of muscle status in geriatric oncology, comprehensive understanding of the interplay between muscle status, pathophysiology and cancer (outcomes) needs to be acquired. Moreover, a dedicated meta-analysis is warranted to draw a final conclusion on the value of preoperative continuous total abdominal CT-based muscle measures to predict poor clinical outcomes in older individuals with colorectal cancer. Future studies should focus on longitudinally measuring muscle mass and muscle quality, muscle strength and physical performance and function and put more emphasis on combining muscle measures and results of CGA to conclude on the role of muscle status in predicting and improving clinical outcomes in geriatric oncology. Potentially, analysis of specific abdominal muscle groups is of added value. Anticancer management and treatment in older individuals remain a challenge. Time will tell whether the longitudinal assessment of muscle status can make early recognition of cancer activity, prediction of poor clinical outcomes and improvement of clinical outcomes through targeted interventions possible in older individuals with cancer.