Cancer cachexia may be detectable before major weight loss becomes obvious, according to a large Nature Medicine study that analyzed longitudinal clinical and genomic data from 52,236 adults with cancer. The researchers found that cachectic weight-loss episodes clustered around disease progression, carried a consistent blood-test signature across many cancer types and could be modeled in advance in lung, colorectal and prostate cancer.
Cancer cachexia is more than ordinary weight loss
Cachexia is a complex wasting syndrome seen in many people with advanced cancer. It can involve loss of body weight, skeletal muscle and fat, along with changes in appetite, inflammation and metabolism.
Unlike weight loss caused simply by eating fewer calories, cancer cachexia is driven by systemic changes linked to the tumor and the body’s response to it. That makes it difficult to reverse with nutrition alone.
Cachexia is clinically important because it can weaken patients, reduce treatment tolerance and worsen quality of life. It is also associated with higher mortality across many cancers.
The study followed more than 52,000 adults with cancer
The researchers repurposed data collected through the MSK-IMPACT prospective tumor-sequencing program at Memorial Sloan Kettering Cancer Center between 2010 and 2024.
After applying the study’s eligibility criteria, the final analysis included 52,236 adults across 589 distinct cancer histologies. Their tumor-genomic data were linked with longitudinal body-mass measurements, treatment histories and routine laboratory tests such as complete blood counts and comprehensive metabolic panels.
This design allowed the team to study cachexia as a dynamic process rather than as a single measurement taken at one point in time.
Cachectic episodes clustered around disease progression
The researchers identified episodes of clinically meaningful weight loss using consensus-based cachexia criteria and then examined when those episodes occurred relative to the course of each patient’s cancer.
Across many cancer types, cachectic episodes were enriched around periods of disease progression.
They were also associated with inferior survival in nearly all of the cancer types examined.
That does not mean every episode of weight loss is caused directly by tumor progression. Treatment toxicity, appetite changes, gastrointestinal symptoms and other illnesses can also contribute. But the population-scale pattern suggests that cachexia often intensifies as the disease itself becomes more active.
Routine blood tests showed a consistent cachexia signature
One of the most clinically relevant findings was that cachectic weight-loss episodes were accompanied by a reproducible pattern in routine laboratory tests.
Across diverse malignancies, patients experiencing cachectic episodes tended to have lower albumin and lower hemoglobin levels, along with higher levels of liver enzymes.
Albumin is a blood protein influenced by nutrition, inflammation, liver function and illness severity. Hemoglobin reflects the blood’s oxygen-carrying capacity and can fall for many reasons in cancer, including inflammation, bleeding, bone-marrow effects and treatment.
The study does not show that these laboratory changes are specific to cachexia. Instead, it suggests that routine blood tests may contain useful information about the systemic state that accompanies clinically important wasting.
Tumor genetics were also linked to cachexia risk
The investigators then asked whether particular tumor genotypes were associated with a higher likelihood of developing cachectic weight loss.
They identified numerous somatic alterations associated with more frequent cachexia onset. One recurring signal was mutation of TP53, which was associated with increased cachexia risk across several cancer types.
TP53 is one of the most commonly altered tumor-suppressor genes in cancer. Its association with cachexia does not mean the mutation directly causes wasting, but it raises the possibility that tumor biology helps shape the systemic metabolic and inflammatory response of the patient.
The authors emphasize that cachexia is unlikely to have a single molecular cause. Different cancers and genotypes may reach a similar wasting phenotype through multiple pathways.
The researchers built models to predict impending weight loss
Using the clinical, laboratory and genomic patterns identified in the dataset, the team developed multivariable models to estimate the future risk of clinically meaningful weight loss.
They focused these prediction models on lung adenocarcinoma, colorectal adenocarcinoma and prostate adenocarcinoma.
The models incorporated routinely available features such as cancer stage, baseline body mass index and selected tumor-genomic alterations.
For lung adenocarcinoma, the discovery cohort included 4,176 patients in the held-out model evaluation, while the broader lung cohort used for external validation included 4,813 patients.
The externally tested lung model used variables that were available in both institutions, including stage, baseline BMI, TP53 status and STK11 status.
The lung cancer model was tested in an independent cohort
A major strength of the study was external validation.
The researchers applied their framework to an independent Dana-Farber Cancer Institute dataset containing 5,785 patients across 12 cancer types.
For the lung model specifically, external validation was performed in 1,159 people with non-small cell lung cancer, including 516 weight-loss events.
The model was applied without refitting to that external cohort.
External validation does not prove that the model is ready for clinical use, but it reduces the chance that the prediction signal was simply an artifact of one hospital’s patient population or data system.
The study does not establish a bedside screening test yet
The phrase “predictable cachexia” needs to be interpreted carefully.
The researchers demonstrated that impending cachectic weight loss could be estimated statistically from existing clinical and genomic data. They did not establish a universal clinical cutoff that doctors can use today to determine which individual patient will develop cachexia.
The study also did not test whether acting on a high-risk prediction improves outcomes.
For a prediction model to become clinically useful, future studies would need to show that it identifies patients early enough for an intervention to make a meaningful difference.
Earlier identification could matter because cachexia is difficult to reverse
Cachexia is often recognized after substantial weight and muscle loss has already occurred.
By that stage, patients may already be weaker and less able to tolerate chemotherapy, radiation, surgery or other treatments.
An earlier warning system could potentially allow clinicians to intensify nutritional support, assess reversible contributors to weight loss, monitor muscle mass more closely or enroll patients in trials of anti-cachexia therapies.
But the current study does not show that any particular intervention prevents cachexia when started earlier.
Weight alone does not capture the full syndrome
Another important limitation is that the study centered on episodes of weight loss because body weight is one of the few physiological measurements collected repeatedly in routine cancer care at very large scale.
Cachexia, however, also involves changes in skeletal muscle, fat mass, inflammation, appetite and function.
Two patients can lose the same amount of weight but differ substantially in how much muscle they lose and how severely their physical function is affected.
The authors therefore describe the work as an atlas of cachectic weight-loss episodes rather than a complete molecular map of every component of cachexia.
The data came from patients receiving care at major cancer centers
The primary cohort came from Memorial Sloan Kettering, and the external validation cohort came from Dana-Farber.
Both are large academic cancer centers with extensive genomic testing and longitudinal electronic health records.
That depth of data is a strength, but it also means the study population may not perfectly represent patients treated in community oncology settings or health systems where genomic profiling is less common.
Future validation in broader and more diverse care settings will be important before such models are used routinely.
The findings link tumor biology with whole-body physiology
One of the broader implications of the study is that cancer genomics may help explain not only how a tumor grows but also how the disease affects the rest of the body.
Most precision-oncology studies focus on tumor mutations because they can predict response to targeted therapy. This work uses the same type of genomic information to investigate a systemic complication of cancer.
That approach could help researchers identify why certain tumors are more likely to drive wasting, inflammation or metabolic dysfunction.
It also reinforces the idea that cancer should not be viewed only as a local mass of malignant cells. Advanced disease can alter physiology throughout the body.
Routine clinical data may become a tool for cachexia research
The investigators argue that large clinicogenomic datasets can help study cancer-associated syndromes that are difficult to capture in traditional clinical trials.
Weight measurements, laboratory tests, imaging, treatment records and tumor sequencing are already collected as part of routine care. When linked longitudinally, those data can reveal patterns that would be difficult to observe in smaller prospective studies.
The challenge is turning those patterns into tools that improve care rather than simply describe risk.
The study is observational and cannot prove causation
The analysis identified associations among disease progression, laboratory abnormalities, tumor genotypes and cachectic weight loss.
It cannot prove that a particular mutation or blood-test change causes cachexia.
Confounding is also possible. Patients with more aggressive disease may simultaneously have more tumor progression, more abnormal laboratory tests and greater weight loss.
Mechanistic experiments and prospective clinical studies will be needed to determine which signals are causal and which are markers of worsening disease.
The next step is to test whether prediction can change outcomes
The study provides a framework for identifying patients at increased risk before severe wasting becomes obvious.
The clinically important question now is whether that warning can be acted on.
Future trials could test whether risk-guided interventions preserve muscle, improve treatment tolerance, reduce hospitalization or extend survival.
Until those studies are done, the new models should be viewed as research tools rather than validated clinical decision systems.
The findings make cachexia easier to study at scale
The biggest contribution of the work is scale.
By integrating longitudinal data from more than 52,000 patients, the researchers showed that cancer-associated wasting follows measurable patterns across tumor types, blood tests and tumor genotypes.
Those patterns may eventually allow clinicians to recognize cachexia earlier, but the study stops short of showing that early prediction improves outcomes.
For now, the results provide a detailed map of when cachectic weight loss occurs, what systemic changes accompany it and which patients may be at higher risk.
References
- Castillon VJ, Xie AX, Boscenco S, et al. A health-system-scale, episode-resolved multimodal atlas of cancer cachexia. Nature Medicine. 2026. DOI: 10.1038/s41591-026-04689-9.
- Nature Medicine. Characterizing the onset and progression of cancer cachexia using routine clinical data. October 13, 2026. Research Briefing.