Highlights
Pretreatment Risk Marker
Baseline fatigue is not merely a common symptom but a potent predictor of severe (grade 3+), life-threatening (grade 4+), and fatal (grade 5) adverse events in patients undergoing systemic cancer therapy.
Dose-Response Relationship
There is a clear dose-response correlation; patients reporting severe baseline fatigue face an approximately five-fold increased risk of fatal toxic effects compared to those with minimal or no fatigue.
Consistency Across Domains
The association remains robust across various toxic effect categories, including symptomatic, hematologic, and non-hematologic adverse events, suggesting a systemic underlying vulnerability.
The Clinical Context of Fatigue in Oncology
Fatigue is perhaps the most pervasive symptom reported by patients with cancer, often persisting from diagnosis through survivorship or end-of-life care. Traditionally, clinicians have viewed baseline fatigue as a quality-of-life issue or a secondary consequence of the underlying malignancy. However, as precision medicine evolves, there is an increasing need to identify clinical markers that can predict how a patient will tolerate systemic therapy, including chemotherapy, immunotherapy, and targeted agents.
While performance status (such as the ECOG or Karnofsky scales) has long been the gold standard for assessing treatment eligibility, these clinician-rated tools often fail to capture the nuanced physiological reserves of a patient. Patient-reported outcomes (PROs) have gained traction for their ability to provide a more accurate reflection of the patient’s lived experience. The study by Unger et al. (2025) published in JAMA Oncology addresses a critical knowledge gap: whether a patient’s self-reported fatigue before starting therapy can serve as an early warning sign for severe biological toxicity.
Study Design and Methodology
This comprehensive cohort study and pooled analysis utilized data from 17 SWOG (formerly the Southwest Oncology Group) phase 2 and 3 clinical trials conducted between 1990 and 2022. The analysis included 7,086 patients across a diverse range of malignancies, including prostate, lung, colorectal, breast, ovarian, and pancreatic cancers, as well as melanoma and lymphoma.
Exposure Assessment
Baseline fatigue was measured using a 5-point Likert scale. Researchers analyzed this data through binary thresholds (comparing those with some or more fatigue against those with no or minimal fatigue) and across the full spectrum of fatigue severity.
Outcome Measures
Adverse events (AEs) were meticulously categorized using the Common Terminology Criteria for Adverse Events (CTCAE). To maintain consistency across three decades of data, various CTCAE versions were mapped to version 4.0. The study distinguished between symptomatic AEs (such as nausea or neuropathy) and objective toxic effects (such as hematologic or laboratory-based abnormalities). The primary endpoints were the incidence of grade 3 or higher, grade 4 or higher, and grade 5 toxicities.
Statistical Rigor
To ensure the findings were not confounded by other variables, the researchers used generalized estimating equations clustered by trial. They adjusted for age, sex, race, and obesity, providing a high degree of confidence in the independent predictive value of fatigue.
Key Findings: The Predictive Power of Fatigue
Among the 7,086 participants (mean age 62.1 years), nearly 40% reported at least some fatigue at baseline. The study recorded a staggering 103,738 adverse events in total, allowing for a robust statistical analysis of risk.
Severe and Life-Threatening Risks
Patients entering treatment with some or more fatigue had a significantly higher risk of experiencing severe (Grade 3+) toxic effects, with an odds ratio (OR) of 2.09 (95% CI, 1.58-2.78; P < .001). The risk for life-threatening or fatal toxic effects (Grade 4 or 5) was similarly elevated (OR, 1.96; 95% CI, 1.36-2.82; P < .001).
The Fatal Correlation
One of the most striking findings was the association between baseline fatigue and treatment-related mortality. Patients reporting baseline fatigue had a 2.35-fold higher risk of fatal toxic effects (Grade 5). When looking at the highest levels of fatigue—those reporting quite a lot or very much fatigue—the risk of a fatal adverse event skyrocketed to an OR of 4.99 (95% CI, 1.84-13.51; P = .002).
Symptomatic vs. Objective Toxicities
Interestingly, baseline fatigue was not only predictive of symptomatic toxicities (which might be expected given the subjective nature of fatigue) but also of objective hematologic and non-hematologic toxicities. This suggests that baseline fatigue is a marker of diminished physiological reserve or a high systemic inflammatory state that predisposes patients to bone marrow suppression and organ dysfunction.
Expert Commentary and Biological Plausibility
From a clinical perspective, these findings suggest that fatigue may be a functional manifestation of the frailty phenotype. In many patients, fatigue is driven by a pro-inflammatory milieu characterized by elevated levels of cytokines such as interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-alpha). These same inflammatory pathways are often implicated in the pathogenesis of chemotherapy-induced toxicities and the cytokine release syndromes seen in newer immunotherapies.
Furthermore, fatigue often correlates with subclinical sarcopenia (muscle wasting) and mitochondrial dysfunction. A patient who is fatigued at baseline may have less metabolic flexibility to handle the stress of systemic therapy, leading to the severe and sometimes fatal outcomes observed in this study. This research underscores the importance of integrating PROs into standard clinical workflows. While a clinician might see a patient with a decent ECOG performance status, the patient’s own report of fatigue might reveal a vulnerability that is otherwise invisible.
Clinical Implications and Future Directions
These findings have immediate implications for individualized treatment planning and clinical trial design:
Risk Stratification
Fatigue assessment should be a standard part of the pretreatment workup. Patients reporting high levels of fatigue may require more frequent monitoring, prophylactic support, or even dose modifications at the outset of therapy.
Pre-habilitation
Identifying fatigued patients early provides an opportunity for pre-habilitation. Interventions such as structured exercise, nutritional support, and psychological counseling may reduce baseline fatigue and, potentially, the subsequent risk of severe toxicities.
Informing Shared Decision-Making
This data provides clinicians with concrete numbers to share with patients during the consenting process. Knowing that high baseline fatigue carries a five-fold risk of fatal toxicity might significantly alter a patient’s or clinician’s perspective on the benefit-risk ratio of aggressive therapy.
Limitations
While the study is large and multi-centered, it is a retrospective pooled analysis of clinical trial participants, who generally have better performance status than the general oncology population. Additionally, while the study controlled for many factors, other unmeasured variables like psychological distress or undiagnosed comorbidities could contribute to both baseline fatigue and treatment toxicity.
Conclusion
Baseline patient-reported fatigue is a powerful, independent predictor of treatment-related toxicity in cancer patients. By recognizing fatigue as a clinical marker of risk rather than just a subjective complaint, oncology teams can better identify vulnerable patients, personalize treatment strategies, and ultimately improve the safety of systemic therapy. As we move toward more patient-centered care, the voice of the patient—expressed through simple fatigue scales—may prove to be as valuable as the most sophisticated laboratory biomarker.
References
Unger JM, Fisch MJ, Jones SMW, Henry NL, Hershman DL. Baseline Fatigue and Severe Toxic Effects in Patients With Cancer Receiving Systemic Therapy. JAMA Oncol. 2025 Dec 26:e255549. doi: 10.1001/jamaoncol.2025.5549. Epub ahead of print. PMID: 41452615; PMCID: PMC12743311.
