After publishing a broad look at repurposed drugs in cancer, the follow-up questions came fast and focused. Nearly all of them asked some version of the same thing: what changes, if anything, when we are talking about brain cancer?
Glioblastoma (GBM) is one of the places where repurposed drug conversations surface most often, for understandable reasons. Standard options are limited, survival is often measured in months to a couple of years, and nearly everyone touched by this diagnosis starts looking for anything that might meaningfully tilt the odds.
The problem is that the internet moves faster than evidence.
The backbone of GBM treatment has not changed. Maximal safe surgery, radiotherapy, and temozolomide remain the foundation and offer the strongest proven benefit. At the same time, these core treatments can often be layered with adjunctive and supportive strategies, with the goal of improving tolerance, preserving neurologic function, and nudging outcomes.
In practice, this “layering” may include repurposed pharmaceuticals (the focus of this note), herbal and nutraceutical approaches (such as curcumin, boswellia, and cannabinoids), selected intravenous therapies, and lifestyle or dietary strategies like intermittent fasting or ketogenic approaches. More on those layers in future notes.
What counts as “repurposed” in GBM?
In this context, I am referring to medications originally developed for other conditions that are later explored in GBM because of biologically plausible mechanisms. These include radiosensitization, metabolic modulation, anti-angiogenesis, autophagy interference, or immune effects.
As with non-CNS solid tumors, some of these agents have human data in GBM. Most of that data is early-phase, mixed, or observational. That does not make it irrelevant, but it does mean these strategies are best understood as complements to standard care rather than replacements for it.
TL;DR: Quick repurposed drug takeaways in GBM
One reason GBM is so hard to treat is the blood–brain barrier, a protective filter that limits how much of many drugs can actually reach brain tissue. Even when a repurposed drug looks promising in the lab, the practical question is whether enough of it gets into the tumor at meaningful levels, and in GBM that access can be uneven because some areas are “leaky” while infiltrating cells at the margins may sit behind a more intact barrier.
Standard therapy still matters most. Surgery, radiotherapy, and temozolomide remain the backbone of GBM care. Repurposed drugs are best thought of as layers added around this core.
Valproic acid has some of the most consistent clinical signals among repurposed agents in GBM, particularly as a radiosensitizer. It is often used thoughtfully in integrative settings, most commonly in pre-treatment and during radiation.
Disulfiram (with copper) has a strong biologic rationale and reached phase II/III testing, but completed trials do not show a survival benefit in unselected GBM and higher doses increase toxicity. Any potential role appears limited to molecularly defined subgroups and remains unvalidated.
Metformin is widely discussed and biologically plausible, with mixed human data. It may make sense in selected patients, particularly those on chronic steroids or with metabolic indications, but it is not a proven survival-extending therapy in GBM.
Celecoxib and related anti-inflammatory agents are experimental chemosensitizers. Feasibility exists, but there is no clear survival advantage, and cardiovascular and GI risks must be considered.
Tadalafil is being explored as an immune-modulating add-on during chemoradiation. The interest here is less about direct tumor killing and more about shaping the immune environment during treatment.
Mebendazole has strong preclinical data and good tolerability, but randomized trials in recurrent GBM have not shown a survival benefit.
Chloroquine and hydroxychloroquine target autophagy and remain of interest, particularly in biologically defined contexts, though dosing and tolerability shape how they are used.
Low-dose naltrexone (LDN) appears neutral for tumor control but may still have a role in symptom-focused care, which I will cover separately.
Ivermectin currently remains a laboratory-driven story in GBM rather than a clinically established layer.
Proton pump inhibitors (PPIs) deserve special caution, as emerging data suggest they may interfere with chemoradiation efficacy, particularly in MGMT-methylated tumors.
ACE inhibitors, ARBs, statins, and beta-blockers should be continued when indicated for cardiovascular health. Their role in GBM is best viewed as supportive rather than disease-modifying.
The weeds: GBM-specific context for commonly discussed agents
One of the names that comes up again and again is valproic acid. It is an anti-seizure medication, and it also behaves like a histone deacetylase inhibitor, which is why it has been studied as a radiosensitizer and epigenetic modulator. The story here is complicated: some studies suggest improved outcomes when it is used during radiation, others are less convincing. Still, among repurposed drugs, it is one of the few with repeated clinical “signals.” Most often in integrative oncology we use valproic acid pretreatment.
Disulfiram is another recurring character. Most people know it as the alcohol-aversion drug, but it shows up in cancer biology because it can disrupt cellular detox pathways and proteasome-related processes. In glioblastoma, DSF is usually paired with copper and, on paper, it is one of the more biologically coherent repurposed candidates. It has made it into phase II/III testing.
Mechanistically, DSF can form copper complexes that induce proteotoxic stress, hitting targets like ALDH and proteasome-linked pathways (including p97–NPL4), and it has activity against GBM cells and stem-like cells in lab models. Completed clinical trials to date have not shown a general survival benefit in unselected GBM, and higher-dose regimens have been more toxic. Where DSF remains interesting is as a possible niche option in molecularly defined subgroups. Early studies have reported signals in tumors with alterations such as IDH1, BRAF, or NF1.
Metformin is probably the repurposed drug most people have heard about in oncology. In GBM, it is being explored for its effects on AMPK and mTOR signaling, and for broader metabolic impacts. The human data so far is mixed. Some observational series look positive, but much of that signal seems stronger in grade III gliomas than in pure grade IV glioblastoma. There are small trials showing it can be added, but we are still short on definitive proof that it meaningfully extends survival in GBM. I do use metformin in GBM, in addition to above potential benefit, most GBM patients are on continuous steroids and the additional blood glucose support is helpful.
Celecoxib is sometimes discussed in GBM because COX-2 signaling is often upregulated in high-grade gliomas, and blocking it may modestly shift inflammation, tumor signaling, and temozolomide sensitivity. In the lab, celecoxib can enhance temozolomide effects and may help reverse resistance in some models, which is why it keeps resurfacing in repurposing conversations. Clinically, though, the story is a bit more inconclusive: small phase I/II studies and nonrandomized combinations suggest feasibility, but there is no clear, practice-changing survival benefit. Celecoxib also has a significant cardiovascular and GI risk profile, making it not my top choice as an adjunct.
Tadalafil (a PDE5 inhibitor best known for erectile dysfunction) is being explored in GBM as an immune-modulating add-on, not as a direct anti-tumor drug. The main idea is that it may blunt treatment-related immune suppression, particularly by reducing myeloid-derived suppressor cells (MDSCs) and supporting T-cell activity during chemoradiation. Early phase data suggest it can be combined with radiation and temozolomide with acceptable tolerability and measurable immune effects.
Mebendazole is a well-tolerated antiparasitic with strong preclinical activity in glioblastoma and some early human safety data, but randomized studies so far have not shown a clear survival benefit. Mechanistically, it binds β-tubulin to disrupt microtubules, appears to have anti-angiogenic effects (including VEGF-related signaling), and can cross the blood–brain barrier, which is why it keeps showing up in GBM repurposing conversations. In orthotopic mouse glioma models, it prolonged survival substantially compared with controls, helping justify human trials.
In newly diagnosed high-grade glioma and GBM, a small phase 1 single-center study combined daily oral mebendazole with standard radiation and temozolomide primarily to assess safety and dosing. High doses (up to 4 g/day) were generally tolerated, and the survival numbers were encouraging, but they need to be interpreted cautiously given the small sample size, the mixed grade III and IV population, and the usual limitations of early-phase single-site studies.
In recurrent GBM, a randomized phase 2 trial tested adding mebendazole to lomustine (CCNU) or temozolomide compared with chemotherapy alone. Mebendazole was well tolerated, but it did not significantly improve overall survival, which is the key reason it has not moved into routine care for unselected recurrent GBM. In clinical trials, dosing is relatively high (often 1.5 to 4 g/day in divided doses), so liver function monitoring is essential.
Another category that comes up often is antimalarials like chloroquine and hydroxychloroquine, usually discussed as autophagy inhibitors. In the lab, both drugs enhance the effects of chemotherapy and radiation, and GBM models, particularly those with EGFRvIII expression, appear more dependent on autophagy. In human studies, however, results have been mixed. Trials adding hydroxychloroquine to chemoradiation have not shown a clear survival advantage in unselected patients, and higher doses are limited by hematologic toxicity, including neutropenia and thrombocytopenia.
Chloroquine has shown more intriguing signals in smaller studies, including one modern trial where EGFRvIII-positive tumors appeared to do better than EGFRvIII-negative disease, even though outcomes in broader, all-comer populations look similar to standard therapy. The way I think about chloroquine-class drugs in GBM is this: the rationale is real, the biology may matter most in certain subtypes, and the limiting factors are dosing and tolerability.
Two other drugs that come up frequently in GBM conversations are low-dose naltrexone (LDN) and ivermectin. In both cases, the evidence for direct anti-tumor effects in the CNS is much thinner than the online conversation suggests. LDN may have other potential benefits, which I will cover in future notes, but in glioma research it has been studied mostly as a quality-of-life intervention rather than a tumor-directed therapy. In a placebo-controlled randomized trial, adding nightly LDN during standard chemoradiation did not improve fatigue, mood, cognition, functional capacity, or overall survival compared with placebo. The upside is that it was generally safe and well tolerated, with one important limitation: it cannot be used with opioid medications.
Ivermectin sits earlier on the evidence spectrum. It has shown anti-glioma effects in cell lines and animal models, including impacts on mitochondrial function and signaling pathways like AKT/mTOR, which helps explain why it keeps resurfacing in repurposing discussions. But we do not yet have completed human GBM trials that clarify dosing, brain exposure, or clinical outcomes. For now, the most responsible way to think about ivermectin in GBM is as a research lead rather than a routine layer, especially given the real potential for neurotoxicity and drug–drug interactions with off-label use.
Not all commonly used non-oncology drugs behave the same way in glioblastoma. Proton pump inhibitors stand out as the exception: accumulating translational and real-world data suggest they may be harmful during chemoradiation. PPIs can upregulate ALDH1A1, a detoxification enzyme linked to resistance to temozolomide and radiation, and multiple large observational studies now associate PPI use with worse survival, particularly in MGMT-methylated tumors. As a result, many neuro-oncology teams now avoid routine or prophylactic PPIs during GBM treatment unless there is a strong gastrointestinal indication, favoring H2 blockers instead when gastric protection is needed.
By contrast, ACE inhibitors, ARBs, statins, and beta-blockers fall into a more neutral category. These drugs have plausible biologic rationales and some observational associations with outcomes in GBM, but results are inconsistent and no agent in these classes has demonstrated a clear survival benefit in controlled trials. Current thinking is not to start them for anti-tumor purposes, but also not to stop them when they are clearly indicated for cardiovascular health.
What about cocktail protocols?
One reason multi-drug “cocktail” approaches keep resurfacing in glioblastoma is that GBM is biologically redundant. It rarely depends on a single pathway for survival. When one route is blocked, others compensate. That reality has pushed some researchers to ask a different question: instead of aiming for one dominant target, what happens if you simultaneously weaken many of the tumor’s support systems?
CUSP9 (short for Coordinated Undermining of Survival Paths) is one of the best-known attempts to put that idea into practice. Designed primarily for recurrent GBM, it combines low-dose, continuous temozolomide with nine repurposed, non-cytotoxic drugs, each selected to interfere with a different growth-promoting or resistance pathway. The goal is not that any single drug “works,” but that together they create enough biologic pressure to make the tumor less adaptable and more vulnerable to standard treatment.
The commonly cited CUSP9 or CUSP9v3 regimen includes aprepitant, artesunate, auranofin, captopril, celecoxib, disulfiram, itraconazole, sertraline, and ritonavir. Each targets a distinct process, such as inflammatory signaling, oxidative-stress defenses, angiogenesis, efflux pumps, metabolic adaptation, or stress-response pathways. Conceptually, it is layered therapy taken to its logical extreme.
Early laboratory work suggested that CUSP9-style combinations can reduce GBM cell viability more than temozolomide alone and may impair migration and survival signaling. Translating that complexity into humans is the hard part. A phase Ib/IIa study of CUSP9v3 in recurrent GBM focused primarily on feasibility and safety. With careful monitoring, the full regimen could be delivered, which is meaningful given the number of interacting drugs. Signals such as disease stabilization and progression-free survival were observed, but the study was small and non-randomized, so the results are best read as proof of concept rather than a final verdict.
The larger takeaway is not that a “cocktail protocol” is “the answer,” but that it illustrates a different way of thinking about GBM: less as a single target to be hit, and more as a system to be constrained from multiple angles. This is also how I approach cases in practice. I think in layers, targeting inflammation, oxidative stress, angiogenesis, efflux pumps, metabolic health, and stress-response biology where it is relevant. Sometimes functional lab testing helps clarify where the biggest imbalances or vulnerabilities may be. The goal is not to do everything, but to choose fewer interventions more strategically, monitor closely, and reassess often.
Questions that lead to better conversations with your team
Instead of asking, “Should I add this protocol?” it is often more useful to ask:
What do we know about this approach in a case like mine?
Does my tumor biology (i.e. MGMT, IDH, EGFR, BRAF, NF1, PDGFRA) make me a candidate for any additional therapies?
Does my tumor biology or anything about my specific case make me a candidate for any clinical trials?
What is your personal experience with ____? What is the strongest level of evidence available for ____?
What are the alternatives with stronger evidence?
By layering ____ in what is the goal: tolerance, symptom control, slowing progression, or something else?
What are you most concerned about when thinking of adding ____ to my regime?
What would we monitor, and what would make us stop?
What is the tradeoff, including interactions or opportunity cost?
If this does not help, what is our next step?
Those questions tend to move conversations out of noise and into thoughtful, individualized care.
Educational content only. This is not medical advice and does not replace individualized care with a qualified neuro-oncology team.
For one-on-one consultations, visit drkseniamalarkey.com.
I like that nothing here is oversold like you see so much on Substack. Nice work!
I think I’ll stick with my croissant.