HDL proteomic signatures as potential biomarkers of cachexia and sarcopenia in gastrointestinal cancer

Cancer cachexia is an inflammatory condition characterised by rapid weight loss (skeletal muscle and/or fat mass) that affects up to 80% of patients with cancer. It negatively impacts patients quality of life, treatment tolerance and survival. Despite its prevalence, the condition is severely underdiagnosed, partially due to a lack of awareness and consensus definition, but also due to difficulties with screening, such as reliance on weight history and the cost and time required for muscle quantification. Therefore, a blood based biomarker that will make the screening of cachexia, easy, routine and actionable is sorely needed. However, despite the large number of candidate markers identified in the literature, none have made it to clinic. In this thesis, we propose a novel biomarker in the form of the High Density Lipoprotein (HDL) proteome. HDL are protein rich particles and previous research has shown they are affected in inflammatory and metabolic disorders such as cardiovascular disease and diabetes. In study 1, we investigated the HDL proteome in a subset of the Nutrimal study which investigated the effects of Leucine (Leu) ± long chain n-3 polyunsaturated fatty acids (LC n-3 PUFA) on skeletal muscle mass and strength in older, healthy adults. It should be noted that this study was not within the original research scope of this thesis, however, due to delays caused by COVID, we utilised a well-characterised cohort available to us for proof of concept. Therefore, we used this subset to identify biomarkers of sarcopenia risk (moderate- and high-risk) and also correlate the HDL proteome with muscle strength (handgrip and leg) and skeletal muscle index (SMI). Using this approach we identified several biomarkers associated with high-sarcopenia risk on small (S) and (L) HDL including CP, F2, ACTB1 and PON-1, and markers of strength (handgrip and leg) and SMI. We also investigated the impact of supplementation on the HDL proteome. We showed that the HDL proteome is affected by even small changes in diet (Leu ± LC n-3 PUFA) and may reflect positive changes in patients that are not yet seen in more physical measures such as muscle mass and strength. In study 2, we investigated the utility of the HDL proteome as a biomarker of cachexia and sarcopenia (low muscle mass) in patients with gastrointestinal cancer. In patients with cachexia, we identified n=16 biomarkers including VTN, CNDP1, APOA2 and CP, when comparing patients to those without cachexia. A score was created using all the identified proteins and its performance was evaluated using receiver operating characteristic (ROC) area under the curve (AUC). The score had an AUC of 0.810, which is considered clinically relevant, and preformed better than individual proteins or the commonly proposed cachexia screening tool, the modified Glasgow Prognostic Score (AUC=0.536). A similar approach was used to identify a sarcopenia score in this group. We identified n=13 biomarkers associated with sarcopenia including CNDP1, SUN3, SEPPP and A2M. Significant proteins were combined into a score and its AUC was 0.861. In study 3, we validated our findings in a C26 model of cachexia. The HDL of mice with cachexia (C26) was significantly different from mice without cachexia (NC) and control mice (CT), while there were few differences between the NC and CT mice. Similarly, the livers of C26 mice were significantly different from NC and CT mice, with upregulated pathways related to protein synthesis, downregulated pathways related to xenobiotic metabolism and activation of the acute phase response. We also validated several HDL biomarkers identified in the human cohort (ApoB, CP, HGFAC, HSPA5, IGFBP3, PROC, SERPINA3). To conclude, we have shown that the HDL proteome is a potential biomarker of cancer cachexia, however more validation is needed. We have also shown that the liver of mice with cachexia is severely affected, contributing to the scarce research on the subject.