Yes, immunotherapy can make cancer worse in a small but significant number of patients. The phenomenon is called hyperprogressive disease, and it occurs in roughly 12% of patients treated with immune checkpoint inhibitors. Rather than slowing tumor growth, these patients experience an acceleration of it, with tumors growing faster than they were before treatment began. This is distinct from side effects or treatment failure. It represents a paradox where the immune system’s response to the drug actively fuels the cancer.
What Hyperprogressive Disease Looks Like
Hyperprogressive disease (HPD) isn’t simply immunotherapy not working. It’s the cancer growing at a measurably faster rate than it was before treatment started. Oncologists identify it by comparing tumor growth rates on imaging before and after immunotherapy begins, looking for a sharp uptick that can’t be explained by the cancer’s natural trajectory.
The overall incidence sits around 12.4% across cancer types, but the rate varies considerably. Stomach cancer has the highest reported rate at roughly 22.5%. Liver cancer and non-small cell lung cancer each come in around 14%. Kidney cancer sits at the low end, closer to 2%. These numbers come from a meta-analysis spanning dozens of observational studies, so the real-world range for any individual study can stretch from near zero to over 36%, depending on the cancer type and how strictly HPD is defined.
Why Immunotherapy Can Backfire
Checkpoint inhibitors work by releasing the brakes on your immune system so it can attack cancer cells. The drugs block a protein called PD-1 (or its partner PD-L1) that tumors exploit to hide from immune detection. In most patients, this unleashes an effective anti-tumor response. In HPD, three things can go wrong.
The first involves regulatory T cells, a type of immune cell whose job is to prevent the immune system from overreacting. Blocking PD-1 on these cells can accidentally supercharge them, making them proliferate faster and suppress the very immune cells that should be fighting the tumor. Multiple studies on HPD patients have found a significant increase in regulatory T cell infiltration inside tumors.
The second mechanism centers on immune cells called macrophages that live inside tumors. The antibody used to block PD-1 has a structural tail called an Fc domain. When this tail interacts with certain receptors on tumor-associated macrophages, it can reprogram them from a neutral or anti-tumor state into one that actively promotes tumor growth. Experiments have shown that when researchers used a version of the antibody with the Fc tail removed, the tumor-boosting effect disappeared entirely.
The third pathway involves T cell exhaustion. Cancer-fighting T cells in HPD patients are already heavily worn down from prolonged exposure to the tumor. When PD-1 is blocked, these exhausted cells don’t recover. Instead, they compensate by switching on other inhibitory receptors, entering what researchers describe as “secondary exhaustion.” The net result is an immune environment that’s even more suppressed than before treatment.
Who Is at Higher Risk
Certain clinical profiles are associated with a greater chance of HPD. In a study of advanced melanoma patients, the strongest independent risk factors were having three or more sites where cancer had spread and poor overall physical condition at the start of treatment. Liver metastases were also significantly linked to HPD, as were elevated levels of lactate dehydrogenase, an enzyme that rises when there’s widespread tissue damage or high tumor burden.
On the genetic side, amplification of genes called MDM2 and MDM4 has been flagged as a potential risk marker. These genetic changes are relatively common in certain cancers, appearing in up to 20% of high-grade brain tumors, for instance. Mutations that activate the EGFR pathway have also been implicated. These genetic factors may prime the tumor microenvironment in ways that make it more likely to exploit the immune changes triggered by checkpoint inhibitors.
Interestingly, older age does not appear to increase risk. One melanoma study found patients over 65 actually trended toward a lower rate of HPD, though the result didn’t quite reach statistical significance.
HPD vs. Pseudoprogression
Not every case where tumors appear to grow during immunotherapy is cause for alarm. Pseudoprogression, where tumors temporarily swell before shrinking, occurs because immune cells flood into the tumor and make it look larger on imaging. Biopsies of these swollen tumors reveal dense clusters of cancer-killing immune cells rather than expanding cancer tissue.
The key difference is how the patient feels. Pseudoprogression is generally accompanied by stable or improving overall health, while true progression or HPD comes with worsening symptoms. Specialized imaging can also help distinguish the two. PET scans show low metabolic activity in pseudoprogressing tumors but intense activity in truly progressing ones. Ultrasound can detect decreased blood flow inside tumors that are merely swollen with immune cells rather than actively growing.
The distinction matters enormously because stopping immunotherapy during pseudoprogression means abandoning a treatment that’s actually working, while continuing it during HPD means prolonging a treatment that’s causing harm.
Which Drugs Are Involved
HPD has been reported with all major classes of checkpoint inhibitors. PD-1 inhibitors, PD-L1 inhibitors, and CTLA-4 inhibitors have all been linked to accelerated tumor growth. Combination therapy using both PD-1/PD-L1 and CTLA-4 inhibitors does not appear to reduce the risk. One meta-analysis found similar HPD rates whether patients received checkpoint inhibitors alone or in combination with other treatments.
How Doctors Catch It Early
Because HPD can progress rapidly, oncologists increasingly recommend an early CT scan around three to four weeks after starting checkpoint inhibitor therapy, rather than waiting for the standard assessment window. If that scan shows signs of progression, the next step is typically a tumor biopsy or a blood test for circulating tumor DNA to understand what’s driving the growth at a molecular level.
When HPD is suspected, the general approach is to stop immunotherapy promptly. Waiting for a confirmatory scan a month later, as is standard practice for ruling out pseudoprogression, is not recommended when the patient’s condition is clearly deteriorating. An early switch to conventional chemotherapy may help counteract the rapid growth, though specific evidence in HPD patients remains limited. The molecular analysis from the biopsy, which typically takes about three weeks, can guide the choice of the next treatment and often coincides with a follow-up scan that confirms the trajectory.
Cytokine Release Syndrome
Beyond HPD, immunotherapy can cause a separate acute crisis called cytokine release syndrome, most commonly associated with a type of immunotherapy called CAR-T cell therapy rather than checkpoint inhibitors. This happens when activated immune cells release a massive flood of inflammatory signaling molecules all at once. Mild cases cause fever and fatigue. Severe cases can lead to dangerous drops in blood pressure, fluid buildup in the lungs, kidney and liver damage, and in the most extreme situations, multi-organ failure. The severity is graded on a scale from 1 to 4, with grade 4 requiring ventilator support. Unlike HPD, cytokine release syndrome doesn’t accelerate the cancer itself, but it can be life-threatening on its own and represents another way immunotherapy can make a patient significantly worse before it makes them better.