Papillary craniopharyngioma is a World Health Organization Grade 1 tumor arising from remnants of Rathke’s pouch, an embryonic structure that helps form the pituitary gland. Although histologically benign, the tumor sits in a crowded anatomical neighborhood near the optic chiasm, hypothalamus, and pituitary stalk, so growth translates directly into progressive visual loss and devastating neuroendocrine dysfunction. Unlike its pediatric counterpart, the adamantinomatous subtype, papillary craniopharyngioma is defined by a single, highly conserved mutation: BRAF V600E. That molecular signature, first identified through whole exome sequencing a decade ago, provided an obvious therapeutic hook, because the same mutation drives a subset of melanomas, colorectal cancers, and lung cancers for which BRAF and MEK inhibitors are already approved. Tumor exome analyses showed that papillary craniopharyngiomas are genomically stable and carry few mutations beyond BRAF, making them, in principle, ideal targets for small-molecule inhibitors.

The new analysis, published in the Journal of Neuro-Oncology, retrospectively reviewed all adult patients who underwent surgery at a RAPID center and subsequently received BRAF and/or MEK inhibitor therapy between January 2017 and December 2024. Of 124 papillary craniopharyngioma patients across the consortium, only 19 had received targeted therapy, a figure that underscores both the rarity of the tumor and how unevenly the new approach has been adopted. The cohort had a median age of 52.5 years at surgery, and no patient achieved a gross total resection; 18 underwent subtotal or partial removal while one received only a biopsy. Tumors were large, with a median preoperative volume of roughly 5,700 cubic millimeters, and therapy was typically started about three months after surgery.

The results were strikingly consistent with earlier clinical trial data. Among 18 patients with sufficient imaging follow-up, graded according to the Response Assessment in Neuro-Oncology 2.0 criteria, two achieved a complete radiographic response and 11 achieved a partial response, meaning 72.2 percent of patients derived measurable tumor shrinkage. Five patients had stable disease, and, critically, not a single patient progressed while on therapy during a median follow-up of 19 months from surgery. Median tumor volume fell by approximately 1,450 cubic millimeters from the start to the end of medical therapy. Treatment was generally well tolerated: 79 percent of patients experienced either no adverse events or only Grade 1 or 2 toxicity on the Common Terminology Criteria for Adverse Events scale, with just one Grade 4 event and no treatment-related deaths. Nevertheless, nearly half of the patients, nine of 19, eventually discontinued therapy because of side effects, a discontinuation rate higher than that reported in the pivotal phase I/II Alliance trial of vemurafenib and cobimetinib, likely reflecting more flexible stopping rules outside a strict protocol.

Perhaps the most consequential finding concerns radiation. The standard modern paradigm for adult craniopharyngioma is maximal safe resection followed by radiotherapy when resection is incomplete, a strategy that provides good local control but exacts a heavy toll on the surrounding hypothalamic-pituitary axis. Previous studies have shown that radiation produces new or worsening endocrine dysfunction in the vast majority of patients within five years, and around 11 percent of adults suffer measurable vision deterioration or symptomatic memory impairment after treatment. In the RAPID cohort, only four patients received radiation, two before starting targeted drugs and two after stopping them without ever having progressed. The 14 patients managed with medication alone achieved similar overall response rates to those who were irradiated, and remarkably, they finished therapy with far smaller residual tumor volumes, a median of just 23.7 cubic millimeters compared with 612.4 cubic millimeters in the radiation group, a difference that reached statistical significance despite the small comparison groups. The authors caution that selection bias likely influenced these numbers, since clinicians may have reserved radiation-sparing strategies for patients with smaller or less symptomatic tumors, but the data provide the strongest multi-institutional support to date for the idea that targeted therapy could replace radiotherapy in selected patients.

The consortium also examined whether the two-drug approach borrowed from melanoma oncology is truly necessary. Fourteen patients received combined BRAF and MEK inhibition using agents such as vemurafenib, dabrafenib, or encorafenib paired with trametinib, cobimetinib, or binimetinib, while four received BRAF inhibitor monotherapy. In melanoma, the MEK inhibitor is added to suppress compensatory reactivation of the ERK signaling pathway, a resistance mechanism driven by loss of negative feedback in rapidly dividing malignant cells. But papillary craniopharyngioma is genetically quiet, and the researchers reasoned that this resistance machinery may be less relevant in a benign tumor with few alternative drivers. Their clinical data bore this out: monotherapy produced comparable radiographic responses, achieved complete responses in half of the monotherapy patients versus none of the dual-therapy patients, left no measurable residual tumor on median after treatment, and produced no Grade 3 or 4 adverse events, compared with a 28.5 percent rate of serious toxicity in the dual-therapy group. One patient treated with MEK inhibitor monotherapy alone also achieved a partial response with only mild toxicity, though single-patient experiences cannot support firm conclusions.

Treatment duration emerged as the third major variable dividing practice across centers. Using the median of four cycles, and the four-cycle design of the phase I/II Alliance trial, as a natural dividing line, the researchers compared short-course regimens of four or fewer cycles with longer courses exceeding four cycles. Patients on longer courses, some continuing for as many as 26 cycles of three-to-four-week treatment, achieved partial radiographic responses far more often, 81.8 percent versus 28.6 percent, a statistically significant difference, without any accompanying increase in adverse event severity or discontinuation. This finding complicates earlier reports suggesting that most shrinkage occurs within the first month of therapy and raises the question of whether many patients are currently stopping treatment too early. It also highlights how little of the dosing logic in this field has been optimized for the brain; the regimens in use descend directly from melanoma dose-escalation studies that never specifically titrated for central nervous system penetration.

When the targeted therapy cohort was compared against a contemporaneous control group of 57 surgically treated papillary craniopharyngioma patients who did not receive the drugs, median progression-free survival was not reached in the treated group because no patient progressed, compared with five months in the control group, although this difference did not reach statistical significance, likely owing to the small treated sample and short follow-up. No patients in either group died during the study period. The authors are careful to frame all of their subgroup analyses as hypothesis-generating rather than definitive; the retrospective design introduces treatment selection bias, adverse event capture is imperfect, the median post-therapy follow-up of roughly 12 months limits claims about long-term durability, and each comparison arm was small. Detailed neuroendocrine outcomes, arguably the most important measure of success for tumors in this location, were also not systematically captured and remain a priority for future work.

What the study does establish, its authors argue, is a map of the decision landscape. Radiation-sparing protocols, BRAF monotherapy, and prolonged treatment courses all appear feasible and safe in this disease, and each deserves prospective evaluation. Only 19 of 124 RAPID patients with papillary craniopharyngioma ever received targeted drugs, and the cases came from fewer than half of the consortium’s active sites, showing that despite dramatic early trial results, molecular therapy for this tumor remains far from routine. Given the diagnosis’s extreme rarity, which makes single-institution randomized trials essentially impossible, the researchers point to multi-institutional consortia like RAPID as the only realistic vehicle for determining, in a rigorously controlled way, which patients can be spared radiation, which need dual inhibition, and how long the drugs should be given. For now, the message for clinicians is that a tumor once managed exclusively with scalpel and radiation beam has become, for a growing number of patients, a disease that a daily pill regimen can shrink, and in some cases seemingly eliminate.

Leptomeningeal metastasis—the invasion of the brain and spinal cord’s protective membranes by cancer cells—remains one of the most formidable complications of advanced malignant diseases. This clinical condition, characterized by diffuse dissemination of tumor cells within the cerebrospinal fluid, precipitates rapid neurological decline. Conventional treatments, predominantly radiotherapy and systemic chemotherapy, have shown limited efficacy, with survival rarely extending beyond months. The challenge multiplies when tumor cells acquire resistance to EGFR-TKIs, rendering traditional targeted therapies ineffective.

Furmonertinib, a third-generation EGFR-TKI, has distinguished itself by targeting mutant EGFR variants with increased potency and selectivity, while sparing wild-type receptors, thus minimizing off-target toxicity. This molecule has demonstrated formidable blood-brain barrier penetration—a critical factor for central nervous system (CNS) malignancies and metastases. However, resistance mechanisms inevitably evolve, diminishing the monotherapeutic impact of this agent. Addressing this concern, the integration of bevacizumab, which antagonizes vascular endothelial growth factor (VEGF), disrupting tumor angiogenesis, promises a synergistic attack on tumor biology by simultaneously inhibiting proliferative signaling and vascular nourishment.

The study meticulously tracked molecular alterations in CSF ctDNA, a liquid biopsy surrogate of tumor burden and molecular landscape within the CNS. ctDNA analysis offers unprecedented noninvasive insight into tumor dynamics, enabling real-time assessment of therapeutic efficacy at a molecular resolution. The researchers demonstrated that the addition of bevacizumab potentiated furmonertinib’s effectiveness, as evidenced by significant molecular response rates in the CSF ctDNA, characterized by reduction or clearance of mutant EGFR alleles.

This molecular response corresponded with meaningful improvements in median overall survival and progression-free survival metrics. These outcomes not only corroborate the clinical benefit but also validate the approach of combinational targeted therapy guided by precise molecular monitoring. The integration of ctDNA analysis advances the paradigm of adaptive treatment modulation, wherein therapeutic decisions are continually refined according to evolving tumor genomics rather than solely relying on radiographic or symptomatic changes.

The implications of these findings extend beyond immediate survival benefits. They suggest a route to circumvent established resistance mechanisms such as T790M mutations or alternative pathway activations that frequently undermine EGFR-TKI monotherapies. By interrupting angiogenic support alongside mutant EGFR signaling, the cancer microenvironment becomes less hospitable for resistant clones, potentially delaying or preventing the emergence of treatment refractoriness.

Clinically, this study redefines the management algorithm for patients afflicted by EGFR-mutant leptomeningeal metastases resistant to frontline TKIs. The intervention combines molecular precision with biologic rationale, offering an evidence-based pathway to enhance CNS disease control. Treatment protocols incorporating furmonertinib and bevacizumab could soon become standard of care, pending validation in larger, multi-center trials. For oncologists, neuro-oncologists, and molecular pathologists, these insights underscore the necessity of integrating molecular diagnostics with therapeutic selection.

Moreover, this research highlights the transformative power of CSF ctDNA as a biomarker platform. Beyond diagnostic utility, serial CSF ctDNA evaluations enable clinicians to detect molecular relapse before clinical deterioration and adjust regimens proactively. In an era where personalized medicine thrives, this represents a quantum leap toward truly dynamic, patient-tailored oncology care.

The study also opens avenues for exploring other combinational regimens targeting parallel resistance pathways—immune checkpoint inhibitors, alternative angiogenesis inhibitors, or novel small molecules—in synergy with furmonertinib. The layered molecular approach may be the key to sustained remissions in leptomeningeal metastasis, a realm long constrained by therapeutic nihilism.

While promising, this combinational therapy requires vigilant evaluation of potential adverse effects, including hypertension, proteinuria from bevacizumab, and off-target toxicities from intensive EGFR inhibition. The balance of risks versus benefits necessitates robust clinical monitoring frameworks and patient selection criteria. Future research should elucidate biomarkers predictive of both therapeutic success and toxicity to optimize individual outcomes.

Equally transformative is the study’s methodology which utilized next-generation sequencing platforms to quantify and characterize ctDNA mutations with high sensitivity and specificity. This technological precision allows discrimination between subclonal variants contributing to resistance, facilitating preemptive treatment adjustments. Such advances in molecular diagnostics are pivotal for managing the heterogeneous and rapidly evolving landscape of metastatic CNS cancers.

This pioneering research by Wang and colleagues situates itself at the intersection of molecular oncology, neuro-oncology, and targeted therapeutics, embodying a multidisciplinary approach essential for tackling leptomeningeal metastases. By leveraging novel agents and cutting-edge diagnostics, the clinical community edges closer to converting a once universally fatal complication into a manageable, chronic condition.

As the global oncology field embraces these innovations, patients burdened by leptomeningeal metastasis might anticipate new standards of care that not only extend survival but also preserve neurological function and quality of life. This progress underscores the enduring value of translational research bridging laboratory findings to clinical applications.

In sum, the combinational regimen of furmonertinib plus bevacizumab established by this study offers a potent and promising therapeutic avenue to overcome EGFR-TKI resistance in leptomeningeal metastasis. The intricate molecular insights gained through CSF ctDNA analysis provide an exemplar for personalized treatment strategies targeting intracranial tumor genotypes. These advancements collectively push the frontier toward precision neuro-oncology in the battle against metastatic brain disease.

Subject of Research: Treatment strategies for EGFR-TKI-resistant leptomeningeal metastasis using furmonertinib and bevacizumab, with CSF ctDNA analysis

Article Title: Furmonertinib combined with bevacizumab in EGFR-TKI-resistant leptomeningeal metastasis: analysis of the CSF ctDNA molecular response and survival outcomes

Article References:
Wang, X., Xie, Y., Hu, J. et al. Furmonertinib combined with bevacizumab in EGFR-TKI-resistant leptomeningeal metastasis: analysis of the CSF ctDNA molecular response and survival outcomes. Br J Cancer (2026). https://doi.org/10.1038/s41416-026-03407-z

Image Credits: AI Generated

DOI: 06 April 2026