Study Finds Nearly Half of Metastatic Lung Cancer Patients Receive No Treatment

More than 200,000 Americans will be diagnosed with lung cancer this year. About 100,000 of those cases will already be metastatic — meaning the cancer has spread beyond the lungs — at the time of diagnosis. And roughly half of those patients will never receive any systemic treatment: no chemotherapy, no immunotherapy, no targeted oral pills [1][2].

That is the central finding of a study published May 7, 2026, in JAMA Oncology by Adam Fox, M.D., and Gerard Silvestri, M.D., of the MUSC Hollings Cancer Center, on behalf of the American Cancer Society National Lung Cancer Roundtable. The researchers analyzed more than 250,000 older adults with metastatic non-small cell lung cancer (NSCLC) in the SEER-Medicare database from 2006 to 2021 — a period that spans the introduction of immunotherapy, the expansion of targeted therapies, and a surge in published research on lung cancer treatment [1][2][3].

The increase in treatment rates over that 15-year period was, in the researchers' own words, "unimpressive" [2].

Source: Fox et al., JAMA Oncology 2026; SEER-Medicare Data

Data as of May 7, 2026CSV

The 90-Day Window That Closes Before Treatment Begins

One of the study's starkest findings: 40% of patients were dead within 90 days of their metastatic diagnosis [1][2]. That timeline matters because it suggests a substantial portion of untreated patients may have had no realistic window in which treatment could have been initiated, selected, and delivered.

For many advanced lung cancer patients, the median time from diagnosis to death without treatment is measured in weeks, not months. Previous propensity-matched analyses have found median overall survival of just 2.0 months for untreated stage IV NSCLC patients, compared with 9.3 months for those who received chemotherapy [4]. In the immunotherapy era, the gap has grown wider: 5-year survival for metastatic NSCLC nearly doubled from 6.8% in 2010–2014 to 10.7% in 2015–2020, according to a SEER database analysis of nearly 192,000 patients [5].

Source: Wang et al., Cancer 2025; SEER Database

Data as of Jan 1, 2025CSV

Those numbers mean that the patients who are missing treatment are missing increasingly effective treatment.

Who Goes Untreated: Demographics and Disparities

The JAMA Oncology study identified several variables independently associated with whether a patient received treatment: age, race and ethnicity, marital status, histologic findings, and receipt of biomarker testing [1][2][3].

Oncologists who see patients in their clinics estimated that 80–90% of their patients receive treatment [2]. But that figure reflects a survivorship bias — patients who make it to an oncologist's office are already a filtered population. The study's finding that the actual treatment rate is closer to 50% implies a large group of patients never makes it through the door.

Race and Ethnicity

The disparities literature on lung cancer is extensive and consistent. Black patients are less likely than white patients to receive lung cancer screening, genetic testing for mutations, systemic treatments, and surgical intervention [6][7]. A SEER-based analysis found that five-year overall survival was lower for Black patients, single patients, uninsured patients, and those on Medicare or Medicaid, with Black and uninsured individuals specifically less likely to receive stage-appropriate treatment [6].

However, the racial disparity appears to be driven primarily by access rather than biology. When patients of different racial backgrounds receive the same treatments — particularly immune checkpoint inhibitors — their survival outcomes converge, with no significant racial disparity [8]. This suggests the gap is a systems problem, not a biological one.

Insurance and Income

Insurance type is among the strongest predictors of treatment access. Medicaid patients are 40% less likely than commercially insured patients to receive biomarker testing and 30% less likely to receive targeted therapies after testing [9]. In one Medicaid database study, biomarker testing rates were 14.1% among Black patients, 26.2% among white patients, and 32.8% among Asian and other patients — with Medicaid-eligible patients tested at a rate of 20.6% compared to 28.4% for those without Medicaid [10].

Source: PMC / Medicaid Database Studies

Data as of Mar 1, 2024CSV

Geography

Rural residents face compounding disadvantages. They have higher lung cancer incidence and mortality than urban residents — rural women had 10% higher incidence by 2021, and rural men maintained roughly 15% higher incidence [11]. Rural patients are also less likely to live within 30 miles or a 30-minute drive of a designated lung cancer screening center, with approximately 15% and 28% of rural adults lacking such access, respectively [11].

Clinical Reasons vs. System Failure: Parsing the "Untreated"

Not all untreated patients represent a failure of the system. Some are genuinely too sick to benefit from therapy. Some choose comfort-focused care after fully informed conversations with their physicians. Distinguishing between appropriate non-treatment and preventable non-treatment is the central interpretive challenge of this data.

Performance Status and Rapid Decline

Performance status — a clinical measure of how well a patient can carry out daily activities — is one of the strongest prognostic factors in metastatic lung cancer. Patients with poor performance status (ECOG 2 or greater, meaning they are in bed or a chair more than 50% of waking hours) have substantially shorter survival regardless of treatment [4]. The 40% mortality rate within 90 days suggests many patients were diagnosed late, with advanced disease burden and poor functional status that made treatment infeasible.

In prior studies, mean survival has been longer for patients treated with targeted therapy than untreated patients among those with adenocarcinoma and performance status of 1 or less [4]. But for patients with very poor performance status, the survival gap between treated and untreated narrows — raising the question of whether "no treatment" in those cases is independently harmful or simply reflects patients who were already dying.

Patient Choice and Comfort Care

Research on end-of-life preferences in advanced cancer has found that 72% of patients with advanced cancer preferred treatment focused on comfort over life extension when given the choice [12]. Hospice care — designed for patients with a life expectancy of less than six months who have decided to forgo curative treatment — represents a legitimate and often preferred care pathway [12].

The question is how many of the study's "untreated" patients made an informed choice versus how many were never offered the choice at all. The study found that many untreated patients were never referred to an oncologist [2][3], which means they could not have participated in shared decision-making about treatment versus comfort care. A patient who declines treatment after a thorough conversation with an oncologist about risks, benefits, and alternatives is in a fundamentally different situation from a patient who was never told treatment was an option.

Biomarker Testing: The Bottleneck Before the Bottleneck

The study identified receipt of biomarker testing as one of the variables most strongly associated with receiving treatment [1][2]. This is significant because more than 50% of metastatic NSCLC patients have either a targetable mutation (EGFR, ALK, ROS1, BRAF, and others) or high PD-L1 expression that would make them eligible for immunotherapy [2][3].

EGFR mutations, present in 10–15% of U.S. NSCLC cases, are treated with oral tyrosine kinase inhibitors that have extended survival by years in some patients [13]. ALK rearrangements, found in about 5% of cases, are similarly treatable with oral agents [13]. These are not toxic intravenous chemotherapy regimens — they are pills with manageable side-effect profiles, and they work.

Yet testing is where the pipeline breaks. International guidelines from the NCCN, ESMO, CAP, IASLC, and AMP recommend that all patients with advanced NSCLC undergo testing for EGFR, ALK, ROS1, BRAF, and PD-L1 [13]. In practice, only 57% of Medicaid patients with metastatic lung cancer had evidence of any biomarker testing [10]. Less than half of community oncologists reported using biomarker testing to guide patient discussions, compared with 73% of academic clinicians [9].

The reimbursement landscape compounds the problem. Next-generation sequencing (NGS), the most comprehensive form of biomarker testing, costs an estimated $4,932 per patient, and many healthcare systems provide only partial coverage [14]. Reflex testing — where the pathologist automatically orders biomarker tests at diagnosis rather than waiting for an oncology referral — could bypass much of the delay, but it is not standard operating procedure at most institutions [13].

Source: OpenAlex

Data as of Jan 1, 2026CSV

The explosion of research into lung cancer immunotherapy — more than 271,000 papers published to date, peaking at over 45,000 in 2024 [15] — has not yet translated into universal access to the basic molecular tests that determine which therapies are appropriate.

International Comparison: Is This an American Problem?

Data from the United Kingdom suggests the untreated rate for stage IV NSCLC is comparable. A REAL-Oncology database analysis of 3,739 patients diagnosed between 2007 and 2017 at a large NHS teaching hospital found that treatment rates declined steeply with disease stage, from 78.2% for stage I to 49.8% for stage IV [16]. That is, roughly half of UK patients with stage IV disease also went untreated in the pre-immunotherapy period, mirroring the U.S. pattern.

Across five European countries, the adoption of immunotherapy in the first-line setting transformed treatment between 2018 and 2021: the use of platinum-etoposide chemotherapy alone fell from 91.8% to 42.3%, while combined chemo-immunotherapy rose from 0% to 41.2% [17]. But these figures describe patients who received treatment — the proportion who went entirely untreated is less well documented in European registries.

The similarity between U.S. and UK untreated rates is notable because the UK has a universal-coverage, single-payer system. This suggests that while insurance barriers amplify disparities in the U.S., the fundamental challenge of late-stage diagnosis and rapid disease progression is not unique to the American system. Poor performance status and late presentation are universal clinical realities in advanced lung cancer.

What Has — and Hasn't — Worked to Close the Gap

Several systemic barriers correlate with untreated status: oncology appointment wait times, biomarker testing turnaround, lack of patient navigators, and transportation barriers [2][3][9].

Some interventions have shown measurable results:

Patient navigation programs have been proposed as a way to guide patients through the diagnostic and referral process. The LUNGevity Foundation's LungMATCH program connects patients with navigators who help them understand biomarker testing results and treatment options [9]. The American Cancer Society's National Lung Cancer Roundtable, which funded the JAMA Oncology study, is focused on closing implementation gaps [3].

Reflex biomarker testing — where pathologists automatically order molecular testing at the time of NSCLC diagnosis rather than waiting for an oncologist to request it — has been adopted at some academic centers and can eliminate days to weeks of delay [13].

State-level legislation requiring insurance coverage for biomarker testing has gained momentum, with multiple states enacting laws to mandate coverage by commercial and Medicaid plans [9].

MUSC Hollings Cancer Center, where the study's authors are based, received an American Cancer Society grant specifically to improve lung cancer biomarker testing at community hospitals — an acknowledgment that the problem is concentrated in under-resourced settings [18].

As Fox put it: "If you make it to a medical oncologist... your odds of getting treated are much higher" [2]. The challenge is everything that happens — or fails to happen — before that appointment.

The Steelman Case for Non-Treatment

There is a legitimate argument that some portion of the "untreated" population received appropriate care — that non-treatment, in their cases, was not a system failure but a reasonable clinical or personal decision.

Patients with very poor performance status (ECOG 3–4) have limited expected benefit from systemic therapy and face real risks of treatment-related toxicity. For these patients, hospice enrollment or symptom-focused palliative care may offer better quality of remaining life than aggressive treatment [12].

Framing all non-treatment as a problem risks pathologizing end-of-life choices and creating pressure toward aggressive care that patients do not want. The 72% of advanced cancer patients who prefer comfort-focused care [12] deserve to have that preference respected rather than counted as a treatment failure.

The counter-argument is equally strong: without an oncology consultation, patients cannot make an informed choice. The study's finding that many untreated patients were never referred to an oncologist [2][3] means they were denied the opportunity for shared decision-making. A system that produces 50,000 untreated patients per year — many of whom never saw a specialist — cannot claim that non-treatment uniformly reflects patient preference.

The Core Tension

Silvestri, the study's senior author, framed the imperative in direct terms: "The onus is on the medical community to diagnose disease early, refer to an oncologist quickly and recognize and communicate to patients that there are available treatment options that can improve both quality of life and survival" [1].

The data supports that urgency. Newer therapies have extended survival by almost seven years in some cases [2]. Five-year survival has nearly doubled since immunotherapy became available [5]. Oral targeted agents for EGFR and ALK mutations offer manageable treatment with significant benefit [13].

Yet 15 years of therapeutic progress have moved the treatment rate by only a few percentage points [1][2]. The therapies exist. The evidence base is enormous. The patients are dying without ever learning they had options.

The question is no longer whether the treatments work. The question is why the system that is supposed to deliver them keeps failing half of the people who need them.