What Happens with Glioblastoma After Radiation and Chemo?

Glioblastoma almost always recurs after the standard combination of radiation and temozolomide chemotherapy, typically within about seven months to a year. The landmark trial that established this treatment as the standard of care found a median survival of roughly 14.6 months, with only about one in four patients alive at two years. What happens between the end of treatment and that eventual recurrence involves a complex interplay of tumor biology, treatment side effects, and diagnostic uncertainty that shapes every decision patients and their care teams face going forward.

What the Standard Treatment Actually Achieves

The combination of radiation therapy with concurrent and adjuvant temozolomide, often called the Stupp protocol after the researcher who led the pivotal trial, became the standard of care for newly diagnosed glioblastoma in the mid-2000s. That trial showed adding temozolomide to radiation improved median survival from about 12 months to roughly 14.6 months, and the two-year survival rate more than doubled, from about 10% to roughly 27%.¹ Those numbers sound modest, and they are. But they represent the best proven first-line approach available, and everything that follows in a patient’s course is measured against this baseline.

After finishing six weeks of daily radiation with concurrent temozolomide, patients typically continue taking temozolomide alone for six additional monthly cycles. During this maintenance phase, MRI scans are performed every two to three months to monitor the tumor. This is when things start to get complicated, because what the scans show in those first months after treatment can be genuinely misleading.

The Pseudoprogression Problem

Within the first few months after completing chemoradiation, up to about 20% of patients develop what looks like tumor growth on MRI but is actually a treatment effect called pseudoprogression. These enlarging lesions can account for roughly half of all cases where imaging shows a growing mass after treatment.² The treatment itself causes tumor-cell death, blood vessel damage, swelling, and abnormal vessel leakiness in the area where the tumor was, and all of that can look almost identical to actual tumor regrowth on a standard MRI.

Pseudoprogression is counterintuitive and anxiety-inducing, but it often signals that treatment worked well. The lesions tend to stabilize or shrink on their own without any additional therapy, and many patients remain symptom-free throughout the episode. The clinical challenge is that distinguishing pseudoprogression from true tumor progression on conventional imaging is extremely difficult. Standard MRI features overlap heavily between the two. Getting this distinction wrong in either direction has real consequences: misidentifying pseudoprogression as real growth could lead to unnecessarily switching a patient off an effective treatment, while missing true progression delays intervention.

Advanced imaging techniques have been explored to solve this problem. Hybrid PET-MRI scans using an amino acid tracer called FET have shown promise for telling the two apart, outperforming standard MRI alone in some case series.³ However, timing matters. When PET imaging is done within the first six months after treatment, its accuracy at telling recurrence from pseudoprogression drops considerably; one systematic review found diagnostic accuracy in that early window was only around 60-74%, compared to roughly 90% when performed after six months.⁴ In practice, clinicians often take a “watch and wait” approach during the first three months after treatment, repeating scans at short intervals before concluding that the tumor is actually progressing.

Where and When Recurrence Happens

When glioblastoma does recur, it overwhelmingly comes back right where it started. About 75-80% of recurrences develop locally, within or immediately adjacent to the original tumor site and within the area that received the highest radiation dose.⁵ Central recurrence, meaning regrowth of the main tumor mass, occurs in the majority of patients at a median of about seven months from diagnosis. New lesions that pop up farther away from the original site are less common but do happen, appearing at a median of roughly 11 to 18 months depending on their distance from the treated area.⁶

There is an interesting wrinkle involving proximity to neural stem cell regions in the brain. Glioblastomas that initially contacted these zones were roughly three times more likely to recur outside the radiation field compared to tumors that were farther away.⁷ This pattern hints at the possibility that stem-like tumor cells near these niches may be capable of migrating before treatment and seeding new tumors at a distance, though the biology behind this remains under active study.

The fact that most recurrences happen right where the tumor was originally treated, in tissue that received the full radiation dose, tells you something important about glioblastoma: it is remarkably good at surviving the very treatments aimed at destroying it.

Why the Tumor Survives Treatment

Several overlapping biological mechanisms explain how glioblastoma cells persist through radiation and chemotherapy. Understanding these helps clarify why recurrence is the norm rather than the exception.

One key player is a subpopulation of cells often called glioblastoma stem cells or cancer stem cells. These cells are especially good at repairing the DNA damage caused by radiation, thanks to highly efficient internal repair machinery.⁸ While radiation kills the bulk of rapidly dividing tumor cells, these stem-like cells can ride out the damage, survive, and eventually repopulate the tumor.

Temozolomide resistance has its own biology. The drug works by attaching chemical groups to DNA in a way that causes lethal errors during cell division. A protein called MGMT can directly undo this damage, so tumors where the gene for MGMT is silenced (through a process called promoter methylation) are generally more sensitive to temozolomide. But even in those favorable cases, resistance can develop. The tumor’s internal mismatch repair system, which normally helps temozolomide kill cells by recognizing and amplifying the drug’s damage, can acquire defects over time. Mutations in mismatch repair genes have been found even in treatment-naïve glioblastomas, meaning some tumors arrive pre-equipped with temozolomide resistance before treatment ever begins.⁹

Perhaps most strikingly, temozolomide itself can drive new mutations. A large-scale analysis of over 10,000 gliomas identified a common pathway to hypermutation, where tumors exposed to temozolomide develop defects in their mismatch repair machinery and then accumulate a surge of new mutations. This temozolomide-driven hypermutation creates enormous genetic diversity within the recurrent tumor.¹⁰ Ironically, while hypermutation in other cancers sometimes makes tumors more vulnerable to immunotherapy, these hypermutated gliomas showed poor T-cell infiltration and low response rates to immune checkpoint drugs. The chemotherapy, in other words, can push the tumor to evolve in ways that close off treatment options rather than open them.

How Treatment Reshapes the Brain Environment

Radiation does not just affect tumor cells. It fundamentally alters the surrounding brain tissue in ways that can actually help a future tumor thrive. The brain’s immune cells, blood vessels, and supporting cells are all changed by radiation, and these changes tend to create a more tumor-friendly environment when recurrence eventually happens.

Radiation has lasting effects on microglia (the brain’s resident immune cells) and on monocytes that infiltrate from the bloodstream, altering their behavior and the chemical signals they produce in ways that can promote tumor growth rather than fight it.¹¹ Radiation-induced changes to blood vessels in the tumor area also contribute to therapeutic resistance.¹²

One particularly revealing finding involves astrocytes, the star-shaped cells that provide structural and metabolic support throughout the brain. Cranial radiation pushes astrocytes into a state of cellular senescence, where they stop dividing but become highly active secretors of inflammatory molecules. These senescent astrocytes were found to promote tumor regrowth by recruiting inflammatory immune cells to the brain and changing the chemical signals that glioblastoma cells produce.¹³ This research has raised interest in senolytic drugs, agents that selectively kill senescent cells, as a possible addition to radiation therapy to reduce the tumor-promoting effect of irradiated brain tissue.

Recurrent Tumors Often Behave Differently

The glioblastoma that comes back is frequently not the same disease that was treated the first time. Under the selective pressure of radiation and temozolomide, tumors undergo a shift in their molecular character. One well-documented change is the proneural-to-mesenchymal transition, where tumor cells adopt a more aggressive, invasive, and treatment-resistant identity.¹⁴ Radiation has been shown to drive this transition directly: in both mouse and human tumor cells, irradiation induced a sustained shift toward the mesenchymal phenotype, which came with increased invasiveness and greater resistance to temozolomide.¹⁵

This mesenchymal shift is connected to broad patterns of multitherapy resistance. Resistant tumor-initiating cells showed enrichment for processes related to tissue invasion, drug transport pumps, and drug metabolism, while sensitivity was linked to active cell division, a trait that ironically makes cells more vulnerable to treatments that target dividing cells.¹⁶ The slowly dividing, invasive cells are the ones that survive treatment and drive the recurrence.

Metabolic reprogramming is another layer of this adaptation. Irradiated glioblastoma cells have been observed to shift their metabolism toward accumulating fats, particularly unsaturated fatty acids, which they store in lipid droplets to protect against cellular stress. This radiation-driven shift toward a fat-storing metabolic program appears to promote survival in the post-treatment environment.¹⁷ Treatment-resistant glioblastoma stem cells also show reduced dependence on glucose, instead relying more on fatty acid burning for energy and activating stress-response pathways that help them weather the metabolic strain of the post-treatment environment.¹⁸

Long-Term Side Effects in Survivors

For the minority of patients who survive well beyond the median, the cumulative effects of treatment on healthy brain tissue become a major concern. Radiation-induced leukoencephalopathy, a form of white matter damage caused by radiation, is one of the most significant long-term complications. Among long-term glioblastoma survivors, roughly 85% had at least one significant neurological deficit, and about 28% had clinically significant leukoencephalopathy. Radiation necrosis (death of healthy tissue caused by radiation) affected about 23%, and treatment-related strokes occurred in a similar proportion. These complications appeared a median of about 2.7 years after diagnosis.¹⁹

The progression of radiation-induced leukoencephalopathy typically begins with subtle cognitive changes, such as shortened attention span, mental slowing, and memory trouble. Over time, this can worsen to gait problems, urinary dysfunction, and in some patients overt dementia. In one study of patients who developed leukoencephalopathy, about two-thirds showed neurological symptoms, and cognitive decline progressed to dementia in a substantial proportion. Cognitive symptoms appeared at a median of about 33 months after completing radiation, with gait and urinary problems following later.²⁰ There is also growing recognition that early-onset leukoencephalopathy, appearing within six months of treatment, occurs in some patients, though its risk factors are less well understood.²¹

Distinguishing radiation necrosis from tumor recurrence on MRI is its own challenge, separate from the pseudoprogression question that arises early on. Both radiation necrosis and recurrent tumor can show similar patterns of enhancement on standard imaging.²² At the tissue level, recurrent tumor is packed with dense cancer cells, while radiation necrosis shows dead tissue, dilated blood vessels, and scarring, but imaging often cannot reliably tell them apart without advanced techniques or biopsy.²³

Treatment Options When Glioblastoma Recurs

There is no standard-of-care second-line treatment for recurrent glioblastoma that commands the same level of evidence as the initial Stupp protocol. Options include repeat surgery, further chemotherapy, bevacizumab (an anti-blood-vessel drug), and enrollment in clinical trials. Decisions are highly individualized, depending on the patient’s functional status, location of recurrence, time since initial treatment, and molecular features of the tumor.

Repeat surgery can extend survival for selected patients, particularly those who are in good functional condition. A meta-analysis found a statistically significant survival advantage for reoperation, though this finding carries a strong caveat: the patients selected for repeat surgery tend to be healthier and have more accessible tumors, so the survival benefit may partly reflect selection rather than the surgery itself.²⁴ Permanent complications from repeat resection occur in a notable proportion of patients, approaching 30% in one series.²⁵

Among drug combinations, lomustine plus bevacizumab has shown the most consistent signal. A phase 2 trial found that nine-month survival in the combination group was about 59-87% depending on dose, compared to roughly 43% for lomustine alone and 38% for bevacizumab alone.²⁶ A meta-analysis confirmed an improvement in both overall and progression-free survival for the combination.²⁷ In practice, median progression-free survival with this combination is still only about five to six months, and overall survival from the start of second-line treatment runs roughly eight to nine months.²⁸ These are sobering numbers, and blood-count toxicity, particularly low platelet counts, can be a dose-limiting problem.

Tumor-Treating Fields and Newer Approaches

Tumor-treating fields, or TTFields, represent one of the few additions to the glioblastoma treatment toolkit in recent years. The technology uses a portable device worn on the scalp that delivers low-intensity alternating electric fields to the brain. These fields interfere with cell division and DNA repair in tumor cells. Randomized trial data and subsequent observational studies have shown a meaningful survival benefit when TTFields are added to standard maintenance temozolomide after chemoradiation, and the treatment has been described as the fourth established modality for glioblastoma alongside surgery, radiation, and chemotherapy.²⁹ Side effects are mainly skin irritation at the electrode sites, which is a markedly different toxicity profile from systemic chemotherapy.

On the immunotherapy front, CAR-T cell therapy has moved past the proof-of-concept stage for glioblastoma, though it remains experimental. Engineered T cells targeting several glioblastoma surface markers have shown biological activity, and researchers are developing next-generation platforms designed to overcome the major barriers: the blood-brain barrier, the tumor’s immunosuppressive local environment, and the heterogeneity of surface markers across different cells within the same tumor.³⁰ Direct delivery into the brain rather than intravenous infusion appears to improve tumor penetration, and dual-target designs aim to reduce the risk that the tumor simply sheds the target and escapes.

One practical complication for any immunotherapy approach in glioblastoma is that patients frequently take dexamethasone, a potent steroid used to control brain swelling. Dexamethasone has been shown to reduce both the number and function of T cells and other immune cells, undermining the very immune responses that checkpoint inhibitors and CAR-T cells depend on.³¹ Managing swelling without suppressing immune function is one of the practical challenges that makes immunotherapy in brain tumors harder than in many other cancers.

Liquid Biopsy and Monitoring Without Surgery

Repeat brain surgery just to learn what the tumor looks like molecularly is not always feasible or safe. This has driven growing interest in liquid biopsy approaches, particularly analyzing tumor DNA shed into cerebrospinal fluid. Cerebrospinal fluid contains fragments of cell-free DNA from the tumor, and sequencing this DNA can potentially identify molecular changes at recurrence, track how the tumor’s genetics evolve over time, and detect early recurrence before it becomes visible on imaging.³² Early clinical work has shown that cerebrospinal fluid DNA sequencing can often distinguish progressive disease from stable cases, though the technology is not yet reliable enough for routine clinical use.³³

The Role of Palliative Care

Given the limited survival gains from second-line treatments, palliative care plays a central role in the management of recurrent glioblastoma, and there is increasing evidence that it should begin well before the end of life. Early integrated palliative care addresses symptom burden, home-based support, and advance care planning from the time of recurrence rather than only when treatment options are exhausted.³⁴ A randomized trial has formally evaluated early palliative care starting at the time of diagnosis or recurrence in glioblastoma patients.³⁵ This represents a shift from treating palliative care as something that happens when oncology is “done” to recognizing it as a parallel track that improves quality of life throughout the disease course. Glioblastoma’s unique neurological symptoms, including cognitive decline, personality changes, seizures, and progressive weakness, make this parallel approach especially important, because these are symptoms that affect not just the patient but everyone around them, and they benefit from expert management well before the final weeks of illness.