Population-Scale Identification of Cancer Cachexia and Weight Loss Trajectories

Researchers mapping Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) clinical sequencing data have quantified how cancer cachexia—a severe wasting syndrome—interacts with pan-cancer survival. The study examined multimodal clinical records from 52,236 adult patients across 589 distinct cancer histologies, tracking longitudinal body mass index trajectories, serology, and radiology-derived disease progression annotations.

In Plain English: The Clinical Takeaway

1. What cachexia means: It is not just ordinary weight loss from eating less; it is a complex metabolic wasting syndrome where the body rapidly loses both muscle and fat mass due to underlying disease processes.

2. How it was tracked: Researchers built an episode-resolved algorithm to spot consensus-defined weight loss—specifically losing five percent or more of body weight over a six-month window—across more than 1.7 million longitudinal measurements.

3. What the risk involves: Experiencing these weight loss episodes correlates with significantly lower overall survival rates across 17 out of 18 major cancer types studied, making early identification critical for patient monitoring.

Multimodal Data Reveals High Cachexia Prevalence Across Gastrointestinal Malignancies

Patients spent a mean of 13.7 percent of their observed follow-up time in a cachectic state, showing a mean weight loss magnitude of 10.45 percent. An independent validation cohort from the Dana-Farber Cancer Institute (DFCI), comprising 5,785 patients, confirmed these figures, finding that 55.2 percent met the consensus weight loss criteria at least once.

Consistent with established clinical literature, the prevalence of these episodes varied significantly by cancer type. Gastrointestinal cancers showed the highest rates of wasting. Pancreatic adenocarcinoma led the cohort with 72.7 percent prevalence, closely followed by stomach adenocarcinoma at 71.7 percent and colorectal adenocarcinoma at 49.6 percent. By contrast, prostate adenocarcinoma and chronic lymphocytic leukemia recorded lower initial prevalence rates at 26.1 percent and 19.9 percent, respectively.

Differences also emerged between histologically distinct tumor types arising from the same tissue. In non-small cell lung cancer, adenocarcinoma showed lower odds of meeting the consensus weight loss criterion than squamous cell carcinoma, with an odds ratio of 0.76. Uterine endometrioid carcinoma patients had substantially lower odds of cachexia than those with the more aggressive uterine serous carcinoma subtype, yielding an odds ratio of 0.33. Meanwhile, diffuse large B cell lymphoma patients exhibited higher odds of wasting compared to those with follicular lymphoma.

Survival Risks and Adipose Tissue Depletion During Wasting Episodes

To measure the physiological impact of these episodes, researchers performed survival analyses using time-independent and time-dependent Cox proportional hazards models. In fixed models, meeting the consensus weight loss criterion associated with a significantly increased mortality risk in 17 of 18 cancer types. For lung adenocarcinoma patients, the hazard ratio reached 2.68, while the pan-cancer hazard ratio stood at 3.02. Time-dependent models further demonstrated that episode onset intensified mortality risk across all 18 evaluated cancers, with the strongest associations appearing in prostate cancer, uterine endometrioid carcinoma, and clear cell renal cell carcinoma.

Compartment-specific body composition analysis of subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT) provided tissue-level context for these BMI-defined drops. Paired scans across 272 episodes in lung adenocarcinoma patients showed that mean SAT and VAT decreased by 23.5 percent and 27.2 percent, respectively. Post-episode trajectories showed marked heterogeneity; while 26.7 percent of episodes were followed by significant weight regain, tissue-level scans demonstrated that SAT increased in 64.1 percent of episodes and VAT increased in 58.1 percent of episodes among patients with available follow-up imaging.

Cancer Type	Cohort Prevalence (%)	Hazard Ratio (Time-Dependent)
Prostate Adenocarcinoma (PRAD)	26.1%	11.10
Uterine Endometrioid Carcinoma (UEC)	Data specific to subtype	9.27
Clear Cell Renal Cell Carcinoma (CCRCC)	Data specific to subtype	9.19
Pancreatic Adenocarcinoma (PAAD)	72.7%	Consistently elevated across models
Lung Adenocarcinoma (LUAD)	Subset analysis	2.68 (Fixed Model)

Funding Transparency and BioRxiv Repository Context

The underlying research analyzing multimodal clinical data to construct this cachexia atlas was shared via bioRxiv by researchers including Venise Jan Castillon, Amy X. Xie, Sonia Boscenco, Ethan Tse, Neil Ruthen, Jamie Wang, Ziad El Bakouny, Michele Waters, Chenlian Fu, Christopher Fong, Urvi A. Shah, A. Ari Hakimi, Justin Jee, Eileen White, Tobias Janowitz, Marcus D. Goncalves, and Ed Reznik. Repository disclosures indicate that the work integrates data from prospective clinical sequencing protocols to map the systemic impact of cancer-associated wasting syndromes.

Contraindications & When to Consult a Doctor

The development of this episode-resolved multimodal atlas establishes that consensus-defined weight loss is an acute clinical event tightly coupled with disease progression and heightened mortality risk across nearly all solid tumors and hematologic malignancies.

References

• Castillon, V. J., et al. (2025). A health-system-scale, episode-resolved multimodal atlas of cancer cachexia. bioRxiv.
• Memorial Sloan Kettering Cancer Center. MSK-IMPACT Clinical Sequencing Cohort Data.
• Dana-Farber Cancer Institute.