Paradox of protection: Re-examining cannabis use disorder in sepsis outcomes among cancer patients

Abstract

The intersection of cannabis use disorder (CUD) and critical illness outcomes in cancer patients represents a burgeoning area of research, particularly as cannabis legalization and therapeutic applications expand globally. Adjusted analyses of a retrospective cohort study by Sager et al revealed significantly lower odds of all-cause mortality (adjusted odds ratio (aOR) = 0.83) and respiratory failure (aOR = 0.8) in CUD-positive patients, alongside elevated hospitalization costs. These findings suggest the potential immunomodulatory and organ-protective effects of cannabinoids on sepsis. Future research must prioritize mechanistic studies, prospective clinical trials, and socioeconomic interventions to translate these findings into actionable clinical strategies, to align policy recommendations with guidelines, including those presented by the National Comprehensive Cancer Network.

Key Words: Cannabis/marijuana; Sepsis; Cannabis use disorder; Cardiovascular outcomes; Critical care; Cancer

Core Tip: This letter critiques a recent study revealing reduced sepsis mortality in cancer patients with cannabis use disorder. We highlight methodological limitations (e.g., coding bias and lack of dosage data) and propose future directions, including mechanistic studies on cannabinoid immunomodulation and policy reforms to address socioeconomic disparities. Key innovations include reconciling paradoxical findings through updated preclinical evidence, equitable policy frameworks to translate findings into clinical practice, and aligning policy recommendations with consensus and guidelines (e.g., National Comprehensive Cancer Network).

TO THE EDITOR

The intersection of substance use disorders and critical illness represents a growing challenge in oncology and critical care medicine. Cannabis use disorder (CUD), defined by compulsive use despite adverse consequences, has markedly increased in prevalence, driven by widespread legalization and evolving societal norms[1-3]. Among cancer patients, cannabis is increasingly utilized for symptom management, including chemotherapy-induced nausea, chronic pain, and anorexia[4,5].

However, its impact on critical illness outcomes, particularly sepsis, remains poorly understood. Sepsis, a life-threatening dysregulation of the host response to infection, disproportionately affects immunocompromised cancer patients, and mortality rate exceeds 25%[6-8].

A recent retrospective cohort study by Sager et al[3] investigated the association between CUD and outcomes of severe sepsis in cancer patients, revealing a paradoxical reduction in mortality and respiratory failure among patients with CUD, despite a higher burden of baseline comorbidities.

The present commentary aimed to critically evaluate the study’s methodology, interpret its paradoxical findings in terms of existing literature, and outline directions for future research to clarify the clinical implications of CUD regarding sepsis and cancer.

STRENGTHS

The study utilized the National Inpatient Sample (NIS), the largest publicly available all-payer inpatient database in the United States, encompassing 35 million weighted admissions. This provides robust epidemiological insights and ensures generalizability to the broader population. The use of validated ICD-9 and ICD-10 codes for CUD and severe sepsis enhances diagnostic accuracy, while multivariable regression modeling adjusts for a wide range of confounders, including demographics, comorbidities, and hospital-level characteristics. Furthermore, the inclusion of hospitalization costs and temporal trends from 2016 to 2020 adds both clinical and economic relevance to the analysis.

LIMITATIONS

Despite its strengths, the study has several limitations inherent to its retrospective design and reliance on administrative data. The first limitation is the diagnostic validity. Reliance on ICD coding for CUD may result in misclassification, with occasional or infrequent users potentially coded as having a formal disorder, thereby inflating prevalence rate. Sensitivity analysis excluding cases with mild or incidental CUD coding would help address this concern[5].

The second limitation is the Sepsis-3 alignment. The study defined "severe sepsis" using ICD codes rather than Sepsis-3 criteria (e.g., Sequential Organ Failure Assessment score), limiting comparability with contemporary cohorts that employ Sepsis-3 definitions[9].

The third limitation is the dose-response paradox. No mortality gradient was found across CUD severity tiers (mild vs moderate-to-severe), contradicting typical pharmacologic associations. This paradox may be attributed to threshold effects, where minimal cannabinoid exposure is sufficient to elicit immunomodulatory effects, or to unmeasured confounders, such as the ratio of tetrahydrocannabinol (THC) to cannabidiol (CBD)[5]. The NIS lacks detailed information on cannabis use patterns, including frequency, potency, route of administration, and product composition, restricting the ability to evaluate dose-response relationships. The NIS does not provide longitudinal data, precluding assessment of post-discharge outcomes or the long-term effects of CUD.

CLINICAL IMPLICATIONS

Notably, CUD-positive patients were younger (median age, 58 vs 69 years), predominantly male (67.9%), Black (23.7%), and Medicaid-dependent (35.2%)[3]. These demographic patterns reflect broader socioeconomic disparities in cannabis access and substance use prevalence[10,11]. Higher rates of depression (16.2% vs 9.8%) and polysubstance use (alcohol: 13.1%; cocaine: 6.9%) align with prior studies linking CUD to mental health disorders and adverse social determinants of health[1]. Paradoxically, CUD-positive patients exhibited fewer cardiovascular comorbidities (e.g., diabetes and hyperlipidemia), possibly due to younger age or the metabolic effects of cannabinoids (CBs)[12]. The most striking finding, a 17% reduction in mortality (aOR = 0.83) and a 20% lower risk of respiratory failure (aOR = 0.8), contrasts with prior studies associating cannabis use with adverse cardiopulmonary events[13,14]. The observed survival advantage may stem from CB-mediated immunomodulation. Recent preclinical studies suggested that CBD attenuates cytokine storm responses via inhibition of the TLR4/NF-κB signaling pathway in murine sepsis models[6]. CBs can reduce oxidative stress and organ injury in sepsis models[6,15]. CBs may also possess bronchodilatory properties, potentially mitigating the risk of respiratory failure[16]; however, clinical evidence in humans remains limited.

Despite comparable length of stay, hospitalization costs were higher among CUD-positive patients ($94574 vs $86615), likely reflecting the increased complexity of care, including the management of withdrawal syndromes and co-occurring psychiatric conditions. This underscores the necessity of integrated care models addressing both CUD and critical illness in socially vulnerable populations[12,17,18].

This study is the first one to examine CUD's association with sepsis outcomes in cancer patients, a population at high risk for both conditions[3]. By identifying a survival paradox, it challenges prevailing narratives of cannabis’s uniformly detrimental effects. The findings align with emerging evidence of CBs’ anti-inflammatory properties, while diverge from observational studies linking cannabis to arrhythmias and stroke[19,20]. This discrepancy may stem from heterogeneity in cannabis formulations (e.g., THC vs CBD dominance), dosage, and route of administration, which are factors that are not captured in administrative datasets.

FUTURE RESEARCH DIRECTIONS

Preclinical models of cancer-associated sepsis may help elucidate the immunomodulatory effects of cannabinoids under controlled conditions[13,21]. Prospective trials measuring plasma CB levels and biomarkers (e.g., interleukin-6 and tumor necrosis factor-α) are urgently needed to elucidate pathways underlying reduced mortality.

Randomized controlled trials (RCTs) evaluating CBs as an adjunctive therapy for sepsis are warranted. The feasibility of such trials is supported by early-phase evidence, including a Phase 1b study of nabiximols (a THC:CBD extract) in glioblastoma patients that demonstrated improved survival[22]. Longitudinal cohort studies assessing outcomes in cancer survivors with a history of CUD are also required to evaluate long-term effects on recurrence risk, functional status, and quality of life.

Socioeconomic and racial differences, particularly the overrepresentation of Black and Medicaid-insured patients in the CUD-positive cohort, should be explicitly addressed in the future research. Policy initiatives must expand access to CUD therapies while addressing stigma in healthcare settings.

POLICY RECOMMENDATIONS

The proposed tiered management framework requires refinement to align with existing guidelines, such as those from the National Comprehensive Cancer Network[23]. Suggested tiers include:

Tier 1 (High risk): Cannabis use should be contraindicated in patients receiving immune checkpoint inhibitors (e.g., programmed cell death protein 1/programmed death-ligand 1 therapies), due to concerns about potential immune dysregulation and toxicity.

Tier 2 (Moderate risk): For patients undergoing chemotherapy, THC intake should be limited to ≤ 10 mg/day, with monthly screening for signs of CUD.

Tier 3 (Palliative): It is essential to prioritize harm reduction, including recommending vaporizers over smoked cannabis to mitigate respiratory risk.

Global taxation framework should be established, and the elective policies need to be constantly improved and adjusted[4,6,13,17].

CONCLUSION

Sager et al[3] challenge assumptions about cannabis’s role in critical illness, and provide valuable insights into the paradoxical association between CUD and sepsis outcomes. Clinicians should balance potential benefits against socioeconomic differences in this cohort. Further clarification is needed on the interplay between confounders (e.g., socioeconomic variables) and biological mechanisms. Clinicians should screen for CUD in cancer patients, recognizing its dual role as a risk factor for substance use disorders and a potential modifier of critical illness. To translate findings into practice, future research must address coding biases, integrate Sepsis-3 criteria, explore CBs in RCTs, and disentangle CBs’ therapeutic potential from confounding socioeconomic variables, ensuring equitable translation of findings into practice. As cannabis legalization expands, rigorous scientific inquiry—unencumbered by regulatory barriers—is imperative to optimize care for this vulnerable population.