When to Stop Immunotherapy and What Happens Next?

Most immunotherapy treatment plans revolve around the two-year mark as a reasonable stopping point, though the right moment depends on how well the cancer has responded, what side effects have appeared, and what type of cancer is being treated. In landmark lung cancer trials, patients who stopped after two years had survival rates nearly identical to those who kept going indefinitely. But the decision is rarely that clean-cut, and what happens after you stop involves its own set of risks, monitoring needs, and options if the cancer comes back.

Why Two Years Became the Default Benchmark

The two-year treatment cap did not emerge from rigorous dose-finding studies. It started as a practical ceiling in the pivotal clinical trials of drugs like pembrolizumab and nivolumab, where patients were treated for up to two years and then observed. Because those trials showed durable responses well beyond the last infusion, two years became a widely adopted stopping point in real-world practice. An analysis of advanced non-small cell lung cancer patients found that stopping after two years yielded an overall survival rate of about 79%, compared with 81% for patients who continued treatment indefinitely, a difference that was not statistically meaningful.¹

That said, the picture has some wrinkles. The real-world I-STOP study tracked 173 advanced lung cancer patients and confirmed no overall survival difference between those who stopped at two years and those who continued. But the patients who stopped did have a higher risk of their cancer progressing, with a median progression-free survival of about 36 months compared with a median that had not yet been reached in the continuation group.² In other words, more patients who stopped saw their cancer grow back, but this did not translate into dying sooner. That gap is reassuring, partly because many of those patients responded well when immunotherapy was restarted.

A European analysis of anti-PD-1 therapy in advanced melanoma adds a useful detail: among patients who achieved a complete response, stopping treatment before six months was linked to a much higher chance of the cancer returning, while those treated for longer than six months fared considerably better.³ So the duration of treatment before stopping matters, and very short courses may not give the immune system enough time to build a lasting response.

The Timeline Varies by Cancer Type

Two years works as a general guideline, but different cancers behave differently. In advanced melanoma, an emulated trial analysis found that stopping immunotherapy too early was risky. Patients who discontinued treatment before accumulating meaningful time on therapy had substantially lower survival compared with those who continued for at least three more months. But by the two-year mark, stopping actually appeared slightly favorable, with an absolute 48-month survival rate that was roughly 10 percentage points higher among those who stopped at two years than those who kept going.⁴ That counterintuitive finding may reflect that patients still on treatment at two years included those who needed to stay on because they had not responded as strongly.

For head and neck cancers, the dynamics are different. A study of patients with recurrent or metastatic head and neck squamous cell carcinoma treated with nivolumab found that long-term responses were rare overall, but for those who did respond well, the progression-free survival curve plateaued at around three years, suggesting that may be a better stopping point for this cancer type.⁵ No single number works for every tumor.

Tests That Help Predict Whether Stopping Is Safe

One of the hardest parts of the decision is figuring out whether the cancer is truly beaten or merely suppressed. Standard CT scans can show that a tumor has shrunk, but they struggle to distinguish scar tissue from residual living cancer. This is where metabolic imaging with PET scans becomes valuable. A study of patients discontinuing checkpoint inhibitors found that those with a complete metabolic response on PET had a two-year progression-free survival of 94%, compared with 62% for those who still had metabolic activity.⁶ In the statistical analysis, metabolic response was the only factor that reliably predicted whether the cancer would stay away.

A separate study in melanoma patients echoed these findings. Among patients who still had residual disease visible on CT but showed a complete metabolic response on PET, the time to progression was dramatically longer than in those whose PET scans still lit up. In the non-complete-metabolic-response group, the median time to progression was just under 13 months; in the complete-response group, it had not yet been reached at the time of reporting.⁷ PET scans essentially reclassified patients whose CT results looked ambiguous, helping distinguish who could safely stop from who probably should not.

Liquid biopsies, which look for circulating tumor DNA in blood samples, are another promising tool. Prospective clinical trials are actively investigating whether ctDNA levels can guide decisions about stopping or continuing treatment in both early-stage and advanced cancers.⁸ The idea is appealing because a blood draw is far less burdensome than imaging, and it could be repeated frequently. But this technology is not yet standard practice for immunotherapy stopping decisions.

The I-STOP study also identified specific patient characteristics that predicted a higher risk of progression after stopping at two years. Patients with lower PD-L1 expression, brain metastases, poorer performance status, or KRAS mutations were more likely to see their cancer return.² These factors can help oncologists tailor the stopping decision to each individual rather than applying a blanket two-year rule.

Long-Term Outcomes After Stopping

The evidence on what happens after discontinuation is encouraging for many patients, particularly those who stopped electively rather than because of side effects. A meta-analysis of melanoma patients who stopped immunotherapy found that 86% were alive at one year after their last dose and that figure held at 86% at three years as well. Progression-free survival was 86% at one year and 71% at three years.⁹ Patients who stopped electively (because they were doing well) had a one-year progression-free survival of 91%, compared with 79% for those who stopped due to toxicity. Longer treatment duration before stopping was also associated with better outcomes.

These numbers reflect the nature of checkpoint inhibitors: they do not kill cancer cells directly. Instead, they release the brakes on the immune system, which then mounts its own attack on the tumor. Once the immune system learns to recognize the cancer, that memory can persist long after the drug itself has cleared the body. Pharmacokinetic studies show that these antibodies have half-lives of roughly two weeks, meaning the drug itself is substantially gone within a couple of months.¹⁰ The durability of the response comes from the immune system’s memory, not from lingering drug levels.

When Side Effects Force an Early Stop

Not everyone gets to choose when to stop. Immune-related adverse events can be serious enough to require immediate discontinuation. These side effects arise because the same immune activation that fights cancer can also attack healthy tissues. Colitis, liver inflammation, lung inflammation, and hormonal gland disruption are among the most common reasons for ICU admission in patients on checkpoint inhibitors.¹¹

A registry dedicated to tracking immunotherapy side effects found that half of all reported immune-related adverse events were graded as severe or life-threatening, and nearly a quarter of affected patients needed additional therapy beyond steroids because their initial treatment for the side effect was not working.¹² This creates a difficult situation: the treatment is working against the cancer, but the patient cannot tolerate it.

Stopping early for toxicity does not necessarily mean worse cancer outcomes. A UK study of lung cancer patients who developed severe immune-related side effects found that those who had received more than four cycles of immunotherapy before the side effect appeared had better overall and progression-free survival than those who stopped very early.¹³ There is even a provocative strand of research suggesting that more intense side effects may correlate with better anti-tumor immune responses, though this remains an area of active debate.¹⁴ It complicates the picture: the immune system’s overenthusiasm may be both the problem and the solution.

Side Effects That Appear After You Have Already Stopped

One of the more unsettling aspects of immunotherapy is that side effects can show up months or even more than a year after the last infusion. These delayed immune-related events, sometimes called DIRE, are a recognized phenomenon. A review identified 23 cases in the literature where new autoimmune problems were diagnosed at least 90 days after immunotherapy ended. The median time to diagnosis was six months, but the range extended to 28 months. The affected organs included hormonal glands, skin, the nervous system, lungs, heart, and gut.¹⁵ Over half of these cases occurred after a brief treatment course of four or fewer doses, which means even patients who stop early are not exempt.

Case reports illustrate how late these events can strike. One patient developed colitis more than a year after stopping nivolumab and ipilimumab.¹⁶ In another pair of cases, patients with renal cell cancer experienced liver inflammation 142 and 436 days after their last nivolumab dose, the latter case emerging well over a year later.¹⁷ Both patients recovered with immunosuppressive treatment, but the delay in recognition can be dangerous if neither the patient nor their doctor connects the symptoms to prior immunotherapy.

This means that monitoring should not end when treatment does. Thyroid function, liver enzymes, and basic inflammatory markers warrant periodic checking for at least a year after the last dose, and both patients and general practitioners need to know that new autoimmune symptoms could be a delayed drug effect. Endocrine side effects like thyroid dysfunction or adrenal insufficiency are often permanent and require lifelong hormone replacement, regardless of when they first appear.

Restarting Immunotherapy If Cancer Returns

One of the strongest arguments for the two-year-stop approach is that immunotherapy can often be restarted successfully if the cancer comes back. In the I-STOP study, patients who progressed after stopping at two years and were rechallenged with immunotherapy achieved a response rate of about 63%.² That is a high number and suggests that the cancer often remains sensitive to the same class of drugs after a treatment break.

A study focused on melanoma patients who relapsed after achieving a complete response and then electively stopping found that retreatment produced an initial complete response in 20% of cases. With the addition of individualized salvage therapies, another 40% were converted to a second remission. All of those patients were eventually able to stop therapy again.¹⁸

Rechallenge does carry risks. A review of the literature concluded that patients who previously tolerated immunotherapy well are generally good candidates for restarting it, but those who experienced severe (grade 3 or higher) side effects the first time around need careful evaluation before trying again.¹⁹ Even after severe initial toxicity, rechallenge is considered feasible for most patients if managed at centers with multidisciplinary experience in handling immune-related side effects. New or recurrent toxicities do occur but tend to be manageable.²⁰ The interval between the first and second course of treatment appears to influence how well rechallenge works, though the optimal gap has not been established.

Treatment Options If Immunotherapy Stops Working

Not every recurrence can be treated by restarting the same immunotherapy. When cancer progresses on or after checkpoint inhibitors, several pathways exist. A real-world study of advanced lung cancer patients who progressed after initial immunotherapy found that combining immunotherapy with anti-angiogenic drugs (which cut off blood supply to tumors) outperformed combining immunotherapy with chemotherapy. The immunotherapy-plus-anti-angiogenic group had a 12-month overall survival of about 78%, compared with roughly 62% for the immunotherapy-plus-chemotherapy group.²¹

For squamous cell lung cancer specifically, targeted therapies offer another route. A study of afatinib, which targets specific growth receptors, showed a disease control rate of about 60% in patients who had already failed both chemotherapy and immunotherapy.²² Molecular profiling of the tumor to look for targetable mutations is increasingly guiding these later-line decisions. The post-immunotherapy treatment landscape is expanding, and failing immunotherapy does not mean running out of options.

The Financial Weight of Continued Treatment

The question of when to stop immunotherapy is not purely medical. These drugs are expensive, and continuing treatment indefinitely carries a real financial toll. A study of cancer survivors on Medicare found that high-cost immunotherapy significantly increased the likelihood of financial hardship. Blood cancer survivors were hit particularly hard, with a roughly 24-percentage-point increase in being unable to afford medical care and a 43-percentage-point increase in taking fewer medications than prescribed due to cost.²³

When two years of treatment offers similar survival to indefinite treatment, the financial argument for stopping becomes substantial. Each additional month of checkpoint inhibitor therapy adds thousands of dollars in drug costs, infusion fees, and monitoring visits. For patients who are responding well and have strong predictive markers favoring a safe stop, the calculus often tips clearly toward discontinuation. For those with ambiguous scans, borderline biomarkers, or aggressive tumor biology, the cost of continuing may feel worthwhile as insurance against relapse. These are conversations where oncologists, patients, and sometimes financial counselors all need to be in the room together.