Beyond body surface area: CT-derived body composition as a predictor of chemotherapy toxicity and survival in pancreatic ductal adenocarcinoma: a systematic review.

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) is characterised by a high prevalence of severe muscle wasting (sarcopenia) and fatty muscle infiltration (myosteatosis), yet chemotherapy dosing still relies on body surface area (BSA), a metric that does not reflect individual patients’ lean body mass (LBM) or muscle quality. A growing body of evidence from oncology meta-analyses demonstrates that low skeletal muscle mass (sarcopenia) independently predicts chemotherapy toxicity across multiple cancer types, and that myosteatosis is associated with significantly increased mortality risk.

We performed a systematic review to determine whether CT-based body composition metrics better predict chemotherapy toxicity and survival outcomes in PDAC than conventional BSA-based dosing.

METHODS

We searched PubMed and EMBASE (up to 10 April 2026) according to PRISMA 2020 guidelines. Of 340 identified records, a total of 16 were included after screening and eligibility assessment: 14 primary studies/abstracts (10 retrospective cohorts, 1 prospective study, 3 conference abstracts); 2 prior systematic reviews were appraised qualitatively for contextual background.

Methodological quality was assessed using the Newcastle-Ottawa Scale (NOS) for cohort studies and AMSTAR-2 for systematic reviews. Conference abstracts were assessed qualitatively; their inclusion and associated limitations are transparently acknowledged.

Data were extracted on severe (grade ≥ 3) toxicities graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE), dose-limiting toxicities (DLT), treatment modifications, and overall survival. Body composition measures assessed included skeletal muscle index (SMI) and skeletal muscle density (SMD) (surrogates for muscle quantity and quality, respectively), visceral adipose tissue (VAT), subcutaneous adipose tissue (SAT), and longitudinal changes in these parameters.

RESULTS

Muscle quality (low SMD/myosteatosis) was as predictive of severe toxicity as muscle mass (SMI) in multiple studies, and when low SMI and SMD co-occurred, patients had significantly higher odds of grade ≥ 3 toxicity (odds ratio ∼ 1.7 in the largest cohort of 636 patients).

Patients receiving high chemotherapy doses relative to LBM (e.g. >5.8 mg of nab-paclitaxel per kg LBM) were significantly more likely to experience DLT (p = 0.028), whereas standard BSA-normalised dosing did not discriminate risk. Early skeletal muscle loss (≥7.9% SMI decline within 2 months of FOLFIRINOX) was linked to a fourfold higher risk of mortality (HR 4.02; 95% CI 1.54-10.5).

Overall, CT-derived body composition measures consistently outperformed BSA for toxicity and outcome prediction, although evidence remains largely retrospective and heterogeneous. Automated CT body composition analysis was demonstrated to be feasible, supporting integration into routine PDAC care.

CONCLUSIONS

CT-derived body composition metrics, particularly LBM-normalised dosing parameters and serial skeletal muscle measurements, appear to provide a more patient-tailored, physiologically relevant approach to chemotherapy dosing in PDAC than exclusive reliance on BSA.

Given the predominantly retrospective evidence base and limited prospective validation to date, these findings should be interpreted as hypothesis-generating rather than practice-changing. Prospective trials of LBM-adapted dosing protocols are needed to validate their impact on toxicity and survival.