Nearly 4 in 10 Young Lung Cancer Patients Also Smoked Cannabis, and Their Cancers Were More Aggressive

A French prospective cohort study finds dual cannabis and tobacco smoking is common in patients under 60 with lung cancer, and is linked to more emphysema, rarer tumor types, and earlier symptom onset, though universal tobacco co-use prevents isolating cannabis as an independent cause.

Why This Matters

Lung cancer in patients under 60 represents a distinct clinical challenge. These patients are often diagnosed at stages that demand aggressive treatment, and they face decades of potential life-years lost. Understanding which exposures drive cancer risk in this younger population is essential for both prevention strategies and clinical surveillance. Cannabis smoking, which involves deep inhalation and prolonged breath-holding of unfiltered combustion products, has long been hypothesized to carry independent pulmonary carcinogenic risk beyond what tobacco alone confers. The endocannabinoid system itself modulates inflammatory signaling, immune surveillance, and cell proliferation in ways that could theoretically either promote or suppress tumor development, making the relationship between cannabis exposure and cancer biology far more complex than a simple “smoke equals harm” narrative. As cannabis legalization expands globally and use rises dramatically among younger populations, the clinical urgency of understanding whether and how cannabis smoking contributes to lung cancer has never been greater, yet high-quality prospective data with objective exposure confirmation have been remarkably scarce.

Clinical Summary

Lung cancer in patients younger than 60 accounts for a meaningful minority of new diagnoses and tends to present with distinct biological features compared to older cohorts. These younger patients are more likely to harbor driver mutations, may have different exposure histories, and face treatment decisions with profound long-term consequences. Despite decades of epidemiologic inquiry, the specific contribution of cannabis smoking to lung cancer risk remains poorly characterized. The near-universal co-use of cannabis and tobacco in most populations has been the central obstacle to disentangling their respective effects. Prior studies have relied heavily on retrospective designs and self-reported exposure data, both of which introduce substantial bias. The result is a landscape where clinical suspicion about cannabis and lung cancer runs ahead of the evidence base.

This prospective multicenter cohort study, conducted across three French tertiary hospitals (Gustave Roussy, Paris-Saint Joseph, and Marie Lannelongue) and published in the Journal of Thoracic Oncology, enrolled consecutive lung cancer patients aged 18 to 60 between 2021 and 2023. The study’s mechanistic rationale rests on the recognition that cannabis combustion generates many of the same carcinogens found in tobacco smoke, including polycyclic aromatic hydrocarbons and benzopyrene, but that cannabis smoking patterns differ markedly from tobacco use. Cannabis smokers typically inhale more deeply, hold smoke longer in the airways, and smoke unfiltered material, all of which increase particulate deposition in the lung parenchyma. Additionally, cannabinoid receptors CB1 and CB2 are expressed throughout the respiratory epithelium and immune cells resident in the lung, and chronic exposure to phytocannabinoids may alter local immune surveillance, inflammatory tone, and epithelial repair mechanisms in ways that could modify cancer risk independently of the combustion products themselves.

Of 148 patients analyzed, 39% were classified as cannabis and tobacco smokers (CTSs), 52% as tobacco-only smokers (TSs), and 9% as never-smokers. Cannabis use in the CTS group was substantial: a median of 26 years of use at a median of 4 joints per day. A critical methodological strength was the use of hair analysis by liquid chromatography-tandem mass spectrometry to objectively confirm THC, CBD, nicotine, and cotinine levels. Strikingly, tobacco exposure was statistically equivalent between the CTS and TS groups by both self-reported pack-years and objective hair cotinine concentrations, meaning that cannabis was the primary distinguishing variable. CTSs were significantly younger at diagnosis (median 53 versus 56 years, P=.006), more frequently presented with chest pain (22% versus 8%, P=.03), had markedly higher rates of emphysema on imaging (64% versus 38%, P=.003), lower diffusing capacity for carbon monoxide (DLCO 63% versus 70% predicted, P=.004), and strikingly higher prevalence of rare and aggressive tumor histological types (17% versus 4%, P=.007).

Despite these compelling associations, the study carries fundamental limitations that prevent causal conclusions. Most importantly, every cannabis user in the cohort also smoked tobacco; no cannabis-only group existed, making it impossible to attribute observed differences solely to cannabis. The sample size of 148 patients, while reasonable for a prospective study in a restricted age range, limits statistical power for subgroup analyses and multivariable adjustment for potential confounders such as socioeconomic status, occupational exposures, diet, and genetic susceptibility. No formal confounder adjustment through multivariable regression is described in the available data. The never-smoker group was too small (approximately 13 patients) to serve as a robust reference. The authors themselves conclude that larger, multicentric studies with formal causal analytic methods are needed before cannabis smoking can be declared an independent risk factor for more aggressive lung cancer in young patients.

Dr. Caplan’s Take

This study makes a genuinely important contribution by doing something that most prior work in this space has failed to do: objectively confirming exposure through hair biomarker analysis rather than relying entirely on self-report. The use of LC-MS/MS to measure THC, CBD, nicotine, and cotinine in hair samples provides a level of exposure verification that meaningfully strengthens the associational findings. The convergence of earlier diagnosis, worse pulmonary physiology, and more aggressive histology in dual smokers, despite equivalent tobacco burden, creates a biologically coherent signal that deserves serious attention.

In my clinical practice, I see patients who use cannabis regularly and want to know whether their use puts them at risk for lung cancer. This is one of the most common and most difficult questions in cannabinoid medicine. Patients are not asking abstractly. They are asking because they smoke cannabis daily, often have for decades, and want a clinician to tell them honestly whether they should be worried. What honest clinical communication requires in this moment is acknowledging that combustion of any plant material delivers carcinogens to the lungs, that the patterns of cannabis inhalation may worsen that delivery, and that studies like this one raise a credible signal of harm, while also being clear that we still cannot say definitively that cannabis itself, independent of tobacco, causes lung cancer.

The reason we cannot close this question is precisely the limitation this study illustrates. Without a cannabis-only cohort, and without rigorous multivariable adjustment for the constellation of lifestyle and socioeconomic factors that cluster with heavy cannabis use, we are left with associations that are suggestive but not dispositive. The difference between “unproven” and “disproven” is critical here. We have not disproven that cannabis smoking causes or accelerates lung cancer. We simply have not yet proven it to a standard that would support a definitive clinical recommendation. The rarity of exclusive cannabis smoking in most populations makes the ideal study enormously difficult to conduct.

What I do in practice with patients who smoke cannabis is straightforward. I counsel strongly in favor of non-combustion routes of administration, including vaporization of regulated products, oral preparations, and sublingual formulations. For patients who have smoked cannabis heavily for decades, I discuss the findings from studies like this one honestly, noting that they suggest an added risk beyond tobacco alone without claiming certainty. I pay close attention to pulmonary symptoms, maintain a lower threshold for chest imaging in heavy smokers of any substance, and document cannabis exposure history with the same rigor I would apply to tobacco. This study reinforces that cannabis smoking history should be a standard part of any pulmonary or oncologic intake, not an afterthought.

Clinical Perspective

This study sits at an important juncture in the slowly accumulating evidence base on cannabis and lung cancer. Prior retrospective and case-control studies have yielded contradictory results, with some suggesting increased risk from cannabis smoking and others finding no association after adjusting for tobacco. The prospective design, multicenter recruitment, and objective biomarker confirmation of this French cohort represent a meaningful methodological step forward. However, the study does not resolve the central question of whether cannabis is an independent pulmonary carcinogen. It sharpens the hypothesis and provides stronger associational data than most predecessors, but it does not cross the threshold into causal evidence.

Clinicians discussing cannabis and lung cancer with patients should frame the evidence precisely. The preclinical and toxicological data on cannabis combustion products are genuinely concerning: cannabis smoke contains many of the same carcinogens as tobacco smoke, and in vitro studies demonstrate mutagenic potential. This study adds epidemiologic signal by showing that dual smokers fare worse than tobacco-only smokers despite equivalent tobacco exposure. What the evidence does not yet support is telling a patient that cannabis smoking definitively caused their cancer or that cannabis abstinence would have prevented it. The honest communication framework involves acknowledging a credible and growing concern, recommending risk reduction through non-combustion routes, and being transparent about the uncertainty that remains.

Specific pharmacological and safety considerations are relevant here. THC and CBD interact with cytochrome P450 enzymes, particularly CYP3A4 and CYP2C9, which metabolize many chemotherapeutic agents and targeted therapies commonly used in lung cancer. Patients actively using cannabis during cancer treatment may experience altered drug levels, potentially affecting both efficacy and toxicity of their oncologic regimens. The emphysema and reduced DLCO observed in dual smokers also have direct implications for surgical candidacy, tolerance of pneumonectomy or lobectomy, and perioperative risk stratification. Clinicians managing these patients should assess cannabis exposure as part of preoperative pulmonary evaluation and factor it into functional reserve calculations.

The actionable recommendation for clinical practice is to incorporate detailed cannabis exposure history, including years of use, frequency, and route of administration, into standard oncology and pulmonology intake documentation. For patients under 60 presenting with lung cancer, explicitly asking about cannabis use is now supported by data showing its high prevalence in this population. Clinicians should document this exposure with the same granularity applied to tobacco pack-years. Where available, referral to prospective registries or clinical studies examining cannabis and cancer outcomes should be offered to interested patients, as building the evidence base depends on systematic data collection across diverse populations.

Study at a Glance

Study Design

Prospective multicenter observational cohort

Enrollment Period

2021 to 2023

Setting

Three French tertiary hospitals (Gustave Roussy, Paris-Saint Joseph, Marie Lannelongue)

Population

Consecutive lung cancer patients aged 18 to 60 years

Total Analyzed

148 patients

Exposure Groups

Cannabis plus tobacco smokers (39%, approximately 58 patients), tobacco-only smokers (52%, approximately 77 patients), never-smokers (9%, approximately 13 patients)

Exposure Confirmation

Hair analysis via liquid chromatography-tandem mass spectrometry for THC, CBD, nicotine, and cotinine

Cannabis Use in CTS Group

Median 26 years of use; median 4 joints per day

Follow-Up

12 months

Primary Outcomes

Clinical features, tumor characteristics, lung function, and surgical outcomes

Journal

Journal of Thoracic Oncology

Funding Source

Not specified in available data

What Kind of Evidence Is This

This is a prospective observational cohort study, meaning researchers enrolled patients consecutively as they were diagnosed and followed them forward in time, comparing groups defined by their smoking exposure history. This design sits in the middle tier of the evidence hierarchy: it is stronger than retrospective chart reviews or cross-sectional surveys because it captures data prospectively and can verify exposures objectively, but it is fundamentally weaker than randomized controlled trials for establishing cause and effect.