For decades, the treatment of glioblastoma multiforme (GBM), the most common and aggressive primary brain cancer in adults, has been a story of incremental progress against a relentless foe. The standard of care—surgery, radiation, and the chemotherapy temozolomide—offers only a modest survival benefit, with median overall survival stubbornly hovering around 10 to 14 months.
One of the most formidable obstacles is not the cancer cell itself, but a biological fortress protecting it: the blood-brain barrier (BBB). This selective shield, designed to keep toxins out of the brain, also excludes approximately 98% of all small-molecule drugs and nearly 100% of large-molecule therapeutics. The central challenge in neuro-oncology has become as much about delivery as it is about discovery.
This intractable problem is the focus of CNS Pharmaceuticals, a clinical-stage biotechnology company employing a “back to the future” strategy. Rather than seeking entirely novel compounds, CNS is chemically re-engineering established, potent chemotherapeutic classes, anthracyclines and taxanes, to transform them from systemic agents into brain-penetrant weapons.
To understand this approach and the evolving landscape of brain cancer therapy, Drug and Device World spoke with Dr. Sandra Silberman, MD, PhD, the Chief Medical Officer of CNS Pharmaceuticals. A seasoned oncologist who helped pioneer early targeted therapies like Novartis’ Gleevec (imatinib), Dr. Silberman brings a historical perspective and pragmatic urgency to the fight against glioblastoma. In a wide-ranging interview, she detailed the company’s pipeline, the lessons learned from past failures, and the cautious optimism surrounding two investigational drugs: Berubicin and TPI 287.
From Poisons to Precision (and Back Again)
Dr. Silberman’s career provides a living timeline of oncology’s evolution, a journey that contextualizes CNS’s current mission. She began her fellowship when chemotherapy was largely about managing the toxicity of broadly cytotoxic “poisons.” The first major paradigm shift was the advent of targeted therapy, a revolution in which she participated directly.
“I actually did the Phase I and Phase II studies for Pfizer for Tarceva (erlotinib), and put Novartis’ Tasigna (nilotinib) into the clinic,” she recounted. The greater breakthrough came with Gleevec. “It really did hit a target that drove the tumor… We were getting long-term responders… this was a true turning point for us in oncology.”
This experience cemented a core principle: identify the driver and hit it precisely. She and her team later executed what she calls “the first basket protocol,” seeking out rare tumors driven by imatinib’s known targets, leading to new FDA approvals. However, she observed that this lesson was often ignored in the subsequent rush. “A lot of companies started employing this strategy of ‘throwing the spaghetti on the wall’… where they just use these targeted therapies, without fully answering the questions like – But do they work in patients? Have we explored enough to see whether these are drivers of the tumor?”
The immunotherapy era followed, bringing its own triumphs and lessons. While checkpoint inhibitors like MSD’s Keytruda (pembrolizumab) transformed outcomes in several cancers, their application in GBM has been disappointing. “We didn’t do our homework again,” Dr. Silberman noted, emphasizing that not all tumors are immunogenic. “Which brings me to what I’m working on today with CNS. However, glioblastoma is again unique as it has another problem… that’s the blood-brain barrier.”
This history underscores a critical point: for glioblastoma, a potent, effective mechanism of action is necessary but insufficient. It must first be delivered to the battlefield. This is the linchpin of CNS Pharmaceuticals’ strategy.
Overcoming the Blood-Brain Barrier Wall
Overcoming the BBB is not a new ambition, but past attempts have faltered. Dr. Silberman critiqued several approaches. Liposomal encapsulation, like in Doxil, had “very limited success.” More recent techniques using ultrasound and microbubbles to temporarily open the BBB also raise concerns. “Unfortunately, not allowing the drug to get in sufficiently to have any meaningful efficacy and also sometimes disrupting the blood-brain barrier… to have poor results later on for the patient,” she explained.
The barrier’s defense is multifaceted. Tight junctions between endothelial cells prevent passive diffusion. More actively, efflux pumps like P-glycoprotein (P-gp), part of the multidrug resistance (MDR) system, recognize and eject many foreign molecules, including most chemotherapies. “They have to go through the cell and the cell has these MDR proteins that are extruding all of these unknown things, mostly chemotherapy out of the brain, so they don’t get in,” said Dr. Silberman.
CNS’s approach is fundamentally different: chemically redesign the drug so that it is no longer a substrate for these efflux pumps and can passively or actively cross the barrier. “What CNS has done is looked at molecules… that have been chemically engineered to be able to cross the blood-brain barrier, they are not substrates for the MDR and extrusion pathways,” Dr. Silberman stated.
Berubicin: An Anthracycline Reborn
The company’s lead candidate, Berubicin, is a novel anthracycline discovered by Professor Waldemar Priebe at MD Anderson Cancer Center. It is structurally modified from doxorubicin, a highly effective chemotherapeutic whose use in brain cancer has been futile due to the BBB.
Preclinical data has been promising. Studies showed Berubicin achieved concentrations in the brain several-fold higher than doxorubicin and demonstrated potent activity against glioma cell lines and in animal models. This translated into a recent Phase II clinical trial (NCT04762069) comparing Berubicin to lomustine in patients with recurrent or progressive GBM after first-line therapy.
In March 2024, CNS Pharmaceuticals announced the primary analysis from this study. While the trial did not meet its primary endpoint of a statistically significant improvement in overall survival (OS) versus lomustine, the data revealed encouraging signals. The median OS was 7.8 months for the Berubicin arm versus 7.0 months for the lomustine arm. Notably, one patient on Berubicin remained on treatment for 2.5 years at the time of the analysis.
The safety profile was distinct, with neutropenia as the major concern for Berubicin, compared to thrombocytopenia for lomustine.
Dr. Silberman provided a nuanced context for these results. “We just completed a study showing that it was as effective as lomustine, but it had different side effects. It was well tolerated… we now have a patient still on that study, still alive 2.5 years later.” She emphasized the drug’s different mechanism (topoisomerase II inhibition) from alkylating agents like temozolomide or lomustine, making it a candidate for combination therapy. “The need to do with that compound is look at combinations.”
Perhaps most compelling is Berubicin’s potential in pediatric brain cancers, specifically diffuse intrinsic pontine glioma (DIPG) and diffuse midline glioma (DMG), where options are tragically limited. “We have with Berubicin pediatric… PDX models. We’re working with a group in Switzerland, and they showed that the drug has incredible activity based on comparison with either temozolomide or lomustine, exponentially better,” Dr. Silberman revealed. The company is actively planning clinical development in this desperate patient population.
A Next-Generation Taxane for the Brain
The second asset in CNS’s pipeline is TPI 287, a third-generation taxane derivative specifically designed to cross the BBB and evade efflux pumps. Like Berubicin, it is a modification of a proven chemotherapeutic class, in this case, paclitaxel. Dr. Silberman noted it may have “better binding affinity to the microtubule apparatus than paclitaxel.”
Pharmacokinetic data is a key part of its promise. “One of the pieces of data is showing that there was at least 10 times more drug in the brain than there was in the plasma and it was sustained in the brain longer than it was in the plasma,” Dr. Silberman cited. This high and sustained brain concentration is the holy grail of neuro-oncology drug delivery.
The most intriguing clinical data for TPI 287 comes from earlier studies combining it with Roche’s Avastin (bevacizumab), an anti-angiogenic antibody used in recurrent GBM. In a Phase I dose-escalation study in heavily pre-treated patients (median of 5 prior therapies), the combination showed remarkable activity. “It showed that there were three complete responses and 9 partial responses… they were unequivocal,” Dr. Silberman reported. The reported median overall survival for this refractory population was 13.4 months, which compares favorably to historical controls of 8-10 months with lomustine.
However, Dr. Silberman was careful to temper excitement with scientific rigor. These were investigator-assessed responses without central radiology review, a necessity for registration trials. Furthermore, the Phase I trial “did not achieve a maximum tolerated dose because there were no dose limiting toxicities,” raising questions about the optimal dose and schedule.
CNS Pharmaceuticals is now designing a meticulous development path under FDA guidance. “We realized we really needed to look at the drug, the drug itself. The FDA has asked us to do this and then the drug in combination with bevacizumab to get any pharmacodynamic interaction… then actually look at whether we can reproduce some of these results,” Dr. Silberman explained. A Phase II study will investigate pharmacokinetics, drug-drug interactions, and confirm the dose before proceeding to a Phase III trial.
A significant strategic question for the Phase III design is the control arm. The FDA has indicated that a randomized control of TPI 287 + bevacizumab versus bevacizumab alone is required. Dr. Silberman questioned whether a synthetic control arm using real-world evidence for bevacizumab monotherapy could be a viable, ethical alternative to randomizing patients to a known ineffective single agent. “Can you absolutely compare that to every single patient based on demographics, prior therapy, age, sex…?” she pondered.
The Path Forward
Looking ahead, Dr. Silberman sees the future in rational combinations. Both Berubicin and TPI 287 are being studied preclinically with other BBB-penetrant agents. “We’re looking at combinations with drugs that do cross the blood-brain barrier in orthotopic models… if we [see] significant differences, these are drugs that do play a significant role… moving forward with those combinations,” she said.
This work embodies what she calls “back to the future.” “These are two drug classes, taxanes and anthracyclins, that have been used for over 50 years, and they’re still used. Why are they still used? Because they’re effective. Why are they not used in glioblastoma? Because they don’t cross the blood-brain barrier.” The “future” is the sophisticated chemical engineering that finally allows these proven warriors to reach their target.
When asked if a universal BBB delivery technology would be preferable to drug-specific re-engineering, Dr. Silberman was clear. “No, no, it doesn’t… The most compelling argument is to take the drugs that are effective and be able to engineer them so that they get across.” She expressed skepticism towards disruptive methods like ultrasound, favoring the elegance of a molecule intrinsically designed for the task. “It’s unique. You know, none of the other companies are really looking at this.”
The journey of CNS Pharmaceuticals highlights the complex, iterative nature of cancer drug development. It is a story that connects the foundational lessons of targeted therapy, know your target and your drug’s mechanism, with the singular physical challenge of glioblastoma. The data on Berubicin and TPI 287 are not yet the definitive breakthrough the field desperately needs, but they represent a strategically sound and biologically plausible path forward.
Update: The article was updated on 26 December to clarify comment and the study.
