Highlights
In a nationwide multicenter cohort of 166,663 patients treated for non-small cell lung cancer (NSCLC) in Japan, 1,346 had pre-existing schizophrenia spectrum disorder (SSD).
Patients with SSD were more often diagnosed with stage IV disease than patients without psychiatric disorders (45.0% vs 31.4%).
After multivariable adjustment, SSD was associated with lower odds of receiving surgery, adjuvant chemotherapy for pathological stage II/IIIA disease, and systemic therapy for stage IV disease.
No statistically significant difference was observed in receipt of concurrent chemoradiotherapy among patients with clinical stage III disease.
Background
People living with serious mental illness, including schizophrenia spectrum disorders, experience substantially worse physical health outcomes than the general population. Cancer is a major contributor to this excess mortality. While some of the survival gap reflects smoking exposure, cardiometabolic comorbidity, delayed diagnosis, and social disadvantage, disparities in cancer treatment are increasingly recognized as an important and potentially modifiable mechanism.
Lung cancer is especially relevant in this context. Tobacco use is more common in patients with schizophrenia, and lung cancer remains the leading cause of cancer death worldwide. For NSCLC, survival depends heavily on whether patients receive stage-appropriate treatment. Surgical resection offers the best chance for cure in early-stage disease, while adjuvant chemotherapy improves outcomes in resected stage II and selected stage IIIA disease. Concurrent chemoradiotherapy is a standard curative-intent approach for unresectable stage III disease, and systemic therapy is central to the management of metastatic stage IV disease.
Despite this well-defined therapeutic framework, patients with severe mental illness may face barriers at multiple points in the care pathway: symptom recognition, access to screening and diagnostic work-up, informed consent, perioperative assessment, treatment adherence concerns, fragmented communication between oncology and psychiatry, and implicit bias in clinical decision-making. However, high-quality contemporary evidence quantifying these treatment gaps in NSCLC has been limited, particularly in Asian healthcare systems.
The study by Yamada and colleagues addresses this gap by examining whether pre-existing SSD is associated with lower receipt of key NSCLC treatments in a large national Japanese cohort.
Study Design
Design and data sources
This was a retrospective cohort study using linked national hospital-based cancer registry data and administrative claims data in Japan. The linkage allowed the investigators to identify cancer characteristics, psychiatric diagnoses, comorbidities, functional status, and delivered treatments.
Population
The study included patients who received initial treatment for NSCLC between 2018 and 2021. The final cohort consisted of 166,663 individuals, including 1,346 patients with pre-existing SSD. SSD was defined by an International Classification of Diseases, Tenth Revision diagnosis code in the F20-F29 range.
The comparator group was patients with NSCLC without psychiatric disorders.
Exposure
The principal exposure was the presence of pre-existing schizophrenia spectrum disorder before or at the time of lung cancer treatment initiation.
Endpoints
The primary outcome was receipt of surgery for NSCLC. Secondary outcomes were stage-specific and clinically meaningful: receipt of adjuvant chemotherapy within 180 days among patients with pathological stage II/IIIA disease, receipt of concurrent chemoradiotherapy among patients with clinical stage III disease, and receipt of systemic therapy among patients with stage IV disease.
Statistical approach
The investigators used multivariable logistic regression to estimate adjusted odds ratios for treatment receipt. The models accounted for age, sex, clinical stage, comorbidities, and functional status. This adjustment is important because patients with SSD often have older physiologic age, higher medical complexity, and reduced performance status, all of which could independently influence treatment selection.
Key Findings
More advanced stage at diagnosis
One of the most clinically important findings was stage distribution. Patients with SSD were substantially more likely to present with stage IV disease than those without psychiatric disorders, 45.0% versus 31.4%. This result suggests that disparities begin before treatment selection, likely during symptom appraisal, help-seeking, diagnostic access, or referral. Advanced-stage presentation narrows therapeutic options and strongly influences prognosis.
Lower likelihood of surgery
Surgery was performed in 31.5% of patients with SSD compared with 49.9% of those without psychiatric disorders. After adjustment for measured confounders, SSD remained associated with significantly lower odds of surgery, with an adjusted odds ratio of 0.70 and a 95% confidence interval of 0.57 to 0.85.
This is a clinically meaningful difference. In resectable NSCLC, failure to receive surgery can represent a lost opportunity for cure. Although some of the gap may reflect factors not fully captured in registry data, such as pulmonary reserve, smoking-related frailty, social support, or patient preference, the persistence of the association after adjustment points to a probable treatment inequity rather than case-mix alone.
Marked disparity in adjuvant chemotherapy
Among patients with pathological stage II/IIIA disease, those with SSD were much less likely to receive adjuvant chemotherapy within 180 days. The adjusted odds ratio was 0.31, with a 95% confidence interval of 0.17 to 0.57.
This was one of the strongest associations in the study. Adjuvant platinum-based chemotherapy confers a well-established survival benefit in appropriately selected patients after resection of stage II and selected IIIA NSCLC. A reduction of this magnitude raises concern that patients with SSD may not be consistently receiving postoperative oncologic evaluation, multidisciplinary review, or supportive care needed to complete potentially beneficial therapy.
No significant difference in concurrent chemoradiotherapy for stage III disease
For patients with clinical stage III disease, receipt of concurrent chemoradiotherapy did not significantly differ between groups. The adjusted odds ratio was 0.99, with a 95% confidence interval of 0.73 to 1.34.
This neutral result is noteworthy. It suggests that once patients with SSD enter a treatment pathway for locally advanced disease, access to intensive combined-modality therapy may be more comparable. One possible explanation is that stage III management often occurs in specialized centers with established multidisciplinary protocols, potentially reducing discretionary variation. Another possibility is that the eligible subset reaching this treatment decision is more selected and functionally preserved.
Lower use of systemic therapy in stage IV disease
Among patients with stage IV NSCLC, SSD was associated with a lower likelihood of receiving systemic therapy, with an adjusted odds ratio of 0.54 and a 95% confidence interval of 0.45 to 0.65.
In the contemporary era, systemic therapy encompasses cytotoxic chemotherapy, immune checkpoint inhibitors, targeted therapy, or combinations thereof, depending on tumor biomarkers and patient fitness. Reduced treatment in metastatic disease may translate directly into shorter survival and poorer symptom control. This finding also raises the question of whether disparities extend beyond first-line treatment selection to molecular testing, immunotherapy access, and palliative oncology integration.
Clinical Interpretation
This study adds robust population-level evidence that patients with schizophrenia spectrum disorder are less likely to receive several cornerstone NSCLC treatments, even after adjustment for important clinical variables. The pattern of findings suggests disparities across the continuum of care rather than at a single decision point.
First, the excess of stage IV presentation indicates delayed diagnosis or delayed entry into oncologic care. In lung cancer, even modest delays can shift management from potentially curative to palliative. Second, lower rates of surgery and adjuvant chemotherapy indicate that treatment intensity falls short even when disease stage would ordinarily support standard-of-care therapy. Third, lower systemic therapy use in stage IV disease suggests that inequity persists in advanced cancer, where treatment goals include both survival prolongation and symptom relief.
The absence of a significant disparity in concurrent chemoradiotherapy should not be overinterpreted as proof of equity. Rather, it may identify a setting in which protocolized multidisciplinary care partially mitigates treatment differences. This observation could help inform intervention design: the more structured the pathway, the smaller the disparity may become.
Why Might These Disparities Occur?
Registry studies cannot fully determine mechanisms, but several explanations are plausible and likely overlapping.
Diagnostic delay is one. Patients with SSD may have reduced access to primary care, cancer screening pathways, or timely specialist evaluation. Somatic symptoms may also be under-recognized by patients, caregivers, or clinicians.
Medical complexity is another. Schizophrenia is associated with smoking, chronic obstructive pulmonary disease, cardiovascular disease, metabolic disorders, and poorer functional status, all of which can complicate surgical candidacy and systemic treatment tolerance. Although the study adjusted for comorbidity and functional status, residual confounding almost certainly remains.
Care fragmentation may be especially important. Oncology and psychiatry often operate in parallel rather than integrated systems. Problems with appointment coordination, transportation, adherence support, and medication reconciliation can influence whether a treatment plan is actually delivered.
Clinician-level factors may also contribute. Teams may worry about informed consent capacity, treatment adherence, perioperative behavioral issues, or psychiatric destabilization. Some concerns are valid and require planning, but they can also lead to therapeutic nihilism if not addressed through structured support. Importantly, schizophrenia alone should not be treated as a proxy for inability to benefit from cancer therapy.
Social determinants of health matter as well. Patients with SSD often have lower income, social isolation, unstable housing, or reliance on caregivers or institutions, all of which may affect cancer care logistics and decision-making.
Strengths of the Study
This investigation has several notable strengths. The cohort was very large and nationally derived, improving precision and reducing center-specific idiosyncrasy. Use of linked cancer registry and administrative data allowed more detailed treatment ascertainment than either source alone. The endpoints were clinically relevant, stage-specific, and directly interpretable by practicing clinicians. In addition, the analysis adjusted for major confounders including age, sex, stage, comorbidities, and functional status.
Another strength is the focus on real-world care rather than trial populations. Patients with severe mental illness are often underrepresented in clinical trials, so registry-based evidence is especially valuable for health services planning and policy development.
Limitations
As with all retrospective observational studies, causal inference is limited. Residual confounding is likely. Important variables may not have been fully captured, such as smoking burden, pulmonary function, frailty, social support, detailed performance status, tumor genomics, patient preference, hospital resources, or severity and current activity of psychiatric illness.
The exposure definition relied on coded diagnoses of F20-F29 disorders. Misclassification is possible, and the study may not distinguish between chronic stable illness and acute psychiatric instability. The comparator excluded psychiatric disorders, but other mental health conditions could influence treatment in different ways.
Treatment receipt was analyzed, but the study did not report cancer-specific survival, overall survival, treatment completion rates, perioperative complications, or quality-of-life outcomes in the abstract provided. Those endpoints would be important for determining the clinical consequences of lower treatment intensity.
Generalizability beyond Japan should be considered cautiously. Japan has a distinct healthcare system, psychiatric care structure, and hospital-based cancer registry environment. Nonetheless, the broad direction of effect is consistent with international concerns about cancer inequity in serious mental illness.
Implications for Practice and Policy
The practical message is clear: oncology systems should not assume that equal availability of services produces equal care. Patients with SSD appear to need proactive, structured support to ensure access to diagnosis and stage-appropriate treatment.
Potential interventions include integrated psycho-oncology and consultation-liaison psychiatry, routine navigation services, standardized preoperative assessment pathways that avoid blanket exclusion, and multidisciplinary case review for patients with serious mental illness. Early involvement of family members, caregivers, community mental health teams, and social workers may help address consent, logistics, and adherence barriers.
At the system level, quality metrics could include equity-sensitive indicators such as stage at diagnosis and receipt of guideline-concordant therapy by serious mental illness status. Education for thoracic oncology teams should emphasize that psychiatric diagnosis alone should not preclude curative or palliative cancer treatment. Conversely, psychiatric clinicians should be equipped to recognize cancer symptoms, facilitate referral, and support treatment continuity.
Future research should evaluate where the largest drop-off occurs: diagnosis, referral, treatment recommendation, treatment initiation, or treatment completion. Studies linking treatment disparities to survival outcomes, biomarker testing, immunotherapy use, and patient-reported outcomes would be particularly informative. Qualitative work involving patients, caregivers, oncologists, surgeons, and psychiatrists could also identify modifiable barriers invisible to administrative data.
Conclusion
This nationwide Japanese cohort study shows that patients with pre-existing schizophrenia spectrum disorder and NSCLC are more likely to present with metastatic disease and less likely to receive several key stage-appropriate treatments, including surgery, adjuvant chemotherapy, and systemic therapy. The findings are clinically important because these treatments are central to both cure and disease control in NSCLC.
The study does more than document disparity; it points to a care gap that health systems can address. Better integration of oncology and mental health services, closer navigation through diagnostic and treatment pathways, and explicit equity-focused quality improvement may help reduce avoidable excess mortality in this vulnerable population.
Funding and Trial Registration
Clinical trial registration: Not applicable.
Funding information was not provided in the abstract cited here; readers should consult the full Chest article for complete funding and disclosure details.
Citation
Yamada Y, Fujiwara M, Ishii T, Watanabe T, Fujimori M, Nakaya N, Kawamura T, Otsuki K, Ichihara E, Shimazu T, Hinotsu S, Uchitomi Y, Inagaki M. Impact of pre-existing schizophrenia spectrum disorder on the receipt of surgery and other treatments for non-small cell lung cancer: A multicenter nationwide cohort study in Japan. Chest. 2026-05-20. PMID: 42167588. URL: https://pubmed.ncbi.nlm.nih.gov/42167588/
Selected References
National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Non-Small Cell Lung Cancer. Current version available at NCCN.org.
Pignon JP, Tribodet H, Scagliotti GV, et al. Lung adjuvant cisplatin evaluation: a pooled analysis by the LACE Collaborative Group. J Clin Oncol. 2008;26(21):3552-3559.
Auperin A, Le Pechoux C, Rolland E, et al. Meta-analysis of concomitant versus sequential radiochemotherapy in locally advanced non-small-cell lung cancer. J Clin Oncol. 2010;28(13):2181-2190.
Kisely S, Crowe E, Lawrence D. Cancer-related mortality in people with mental illness. JAMA Psychiatry. 2013;70(2):209-217.
Zhuo C, Tao R, Jiang R, et al. Cancer mortality in patients with schizophrenia: systematic review and meta-analysis. Br J Psychiatry. 2017;211(1):7-13.
