An investigation of computed tomography-derived skeletal muscle measurements in clinical cancer care

McGovern, Josh (2025) An investigation of computed tomography-derived skeletal muscle measurements in clinical cancer care. PhD thesis, University of Glasgow.

Abstract

Cancer is a leading cause of death globally, responsible for nearly 10 million deaths annually. In the UK, it is responsible for one in four deaths. Despite the significant mortality rates, cancer survival continues to improve. Such advances are considered multifactorial and attributable to the evolution of new anti-cancer therapies and the identification of novel biomarkers for the optimization of anti-cancer therapy.

Determining which patients will derive benefit from anti-cancer therapy, and when in their cancer journey, remains an area of interest in oncology. At present, such decisions are informed by tumour and host factors. With reference to the tumour, cancer stage and grade are commonly utilised by clinicians for the determination of treatment intent and modality. With reference to the host, age and performance status are routinely considered when determining the appropriateness of anti-cancer therapy. Whilst performance status has historically been considered a robust determinant of likely outcome to anti-cancer therapy, a lack of granularity in the measures of performance status has meant that there is continued interest in the identification of tools that can objectively determine functional status in cancer patients.

Computed tomography (CT)- derived skeletal muscle measurements, skeletal muscle index (SMI) and density (SMD), are considered surrogate markers of muscle quantity and quality, respectively. Readily quantified from the analysis of CT images obtained during routine clinical cancer care, SMI and SMD have been reported to be associated with functional status in patients with cancer. Moreover, are considered to provide a global assessment of the cancer patient, that also inform of nutritional and frailty status. The work presented in this thesis examines how CT-derived measurements of skeletal muscle may be utilised in clinical cancer care.

The prevalence and determinants of CT-derived skeletal muscle measurements, SMI and SMD, are examined in Chapter 4. The results of Chapter 4 reported that a low SMI and SMD had a percentage prevalence of between 30-60% in a substantial cohort of patients with cancer and that this was similar irrespective of threshold values used to define a low SMI/SMD. Moreover, reported that a low SMI and SMD are endemic across a range of cancer subtypes and disease stages, suggesting the poor muscle status is largely constitutional and not the result of the cancer per se.

Given their respective associations with skeletal muscle mass and function, the combination of SMI and SMD, may provide an objective measure by which sarcopenia can be characterized. Chapters 5 and 6 examined the relationships between the CT-derived sarcopenia score (CT SS), a score that combines SMI and SMD, and physical function, malnutrition, systemic inflammation and survival in patients with potentially curative disease. Chapter 5 reported that the CT-SS was significantly associated with malnutrition, systemic inflammation and poorer survival in 1,002 patients with primary operable colorectal cancer. Chapter 6 reported that CT-SS was associated with cardio-pulmonary exercise testing (CPET) performance, an assessment of cardiopulmonary fitness likely to inform on the patient’s baseline functional status, systemic inflammation and survival in 232 oesophagogastric cancer patients with good performance status who underwent neoadjuvant chemotherapy with a view to potentially curative surgical resection. However, that the prognostic value of CT-SS to survival was not maintained when adjusted for systemic inflammation.

The relationship between CT-derived skeletal muscle measurements, systemic inflammation and survival in patients with cancer remains unclear. This relationship was further examined in Chapter 7, that reported systemic inflammation, but not the CT-SS, was significantly associated with survival in 307 good performance status patients with advanced cancer. Taken collectively, the results of Chapter 6 and 7 support the hypothesis that systemic inflammation dominates the prognostic value of CT-derived skeletal muscle measurements in patients CT derived skeletal muscle measurements, systemic inflammation and survival in patients with cancer is required to determine if CT-derived skeletal muscle measurements have independent prognostic value to clinical outcomes and are a useful adjunct for the prediction of likely outcome in patients with cancer.

Sarcopenia is considered a cause of frailty in older adults with cancer. However, the relationship between the CT-SS, frailty and clinical outcomes in patients with cancer is unclear. Specifically, if CT-derived skeletal muscle measurements capture the prognostic value of frailty in patients with cancer. Chapter 8 examined the prevalence and prognostic value of frailty screening tools in patients with colorectal cancer, reporting that frailty was prevalent and had prognostic value to both short- and long-term clinical outcomes. Chapter 9 examined the relationship between frailty and malnutrition, CT-derived skeletal muscle measurements, systemic inflammation and short-term clinical outcomes in 1,002 patients undergoing potentially curative surgery for colorectal cancer. The results reported that frailty was associated with CT-derived skeletal muscle measurements. However, remained independently associated with short-term clinical outcomes (post-operative complications) when adjusted for CT-derived skeletal muscle measurements. The results suggest that whilst sarcopenia and frailty are closely associated in patients with cancer, CT-derived muscle measurements do not completely capture the prognostic significance of frailty. Nevertheless, the results suggest that the CT-SS may be a useful adjunct to frailty screening tools/measures in patients with cancer.

Cancer cachexia is a complex metabolic syndrome that is associated with dysregulated glucose metabolism. However, there is currently a paucity of studies examining the relationship between an elevated serum lactate dehydrogenase (LDH), an early biomarker of dysregulated glucose metabolism, and a low skeletal muscle mass, considered the defining feature of cachexia. Chapter 10 reported that an elevated LDH was significantly associated with performance status, systemic inflammation and survival in 436 patients with advanced cancer. However, also reported that there was no significant association between an elevated LDH and a low SMI. Whilst the results of Chapter 10 do not suggest that the loss of skeletal muscle mass is directly related to dysregulated glucose metabolism in patients with cancer, further study is required.

Whilst skeletal muscle mass is considered to reduce with cancer progression, liver mass is thought to be preserved. CT is considered a reliable modality for the quantification of both skeletal muscle and liver mass. However, the quantification of liver mass is significantly more time-consuming and laborious compared with that of skeletal muscle. The current gold standard methodology requires the measurement of the total liver volume, calculated by manual segmentation of sequential axial CT images. As such, there is a paucity of studies examining the relationship between skeletal muscle and liver mass, quantified using CT, in patients with cancer. We hypothesized that the maximal cross-sectional liver area on an axial CT slice, determined using manual segmentation, may be an easily quantified surrogate measure of liver mass, analogous to how skeletal muscle mass is quantified using CT. Chapter 11 reported that the maximal cross-sectional liver area was strongly correlated with the total liver volume in patients undergoing potentially curative surgery for colonic cancer, suggesting that it was a reliable surrogate marker. Chapter 12 reported that CT-derived liver mass, quantified using the novel proposed methodology, was significantly associated with SMI in 385 patients undergoing potentially curative surgery for colonic cancer, suggesting that a higher skeletal muscle mass is associated with a higher liver mass in patients with early stage disease. The results are informative and provide a foundation for future work examining the relationship between skeletal muscle and liver mass in patients with cancer.

In summary, a low SMI and SMD appear to be constitutional and not the result of cancer per se. The combination of SMI and SMD would appear to objectively characterize sarcopenia and is closely associated with malnutrition, physical function, frailty and systemic inflammation in patients with cancer. However, it remains unclear if CT-derived skeletal muscle measurements have independent prognostic value to clinical outcomes and therefore questions their utility as prognostic tools in patients with cancer. Whilst a low skeletal muscle mass was not found to be significantly associated with biomarkers of dysregulated metabolism, it was significantly associated with CT-derived liver mass. The present work reports a reliable methodology for future examination of this relationship in patients with cancer.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: Q Science > QP Physiology; Q Science > QR Microbiology; R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer); R Medicine > RD Surgery
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
• Supervisor's Name: Dolan, Dr. Ross and McMillan, Professor Donald
• Date of Award: 2025
• Depositing User: Theses Team
• Unique ID: glathesis:2025-85395
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 19 Aug 2025 07:36
• Last Modified: 19 Aug 2025 08:43
• Thesis DOI: 10.5525/gla.thesis.85395
• URI: https://theses.gla.ac.uk/id/eprint/85395

Related URLs

• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI
• Article DOI