Today, only two boron delivery agents are approved for clinical use, the amino acid analog boronophenylalanine, known as BPA, and sodium borocaptate, known as BSH. Both have enabled meaningful advances, particularly in head and neck cancers and melanoma, but both carry well-recognized weaknesses. Tumor uptake can be heterogeneous, boron retention inside cells is often brief, and both agents depend on biological transport mechanisms that vary widely between patients and tumor types. A research team spanning institutions in Argentina and Uruguay has now reported a candidate designed to attack these limitations from a different angle — a single molecule that not only ferries boron into tumor cells but also attacks the cancer through an entirely separate pharmacological mechanism.

The compound, referred to in the study as Cmpd. 1, is a hybrid molecule built on the quinazoline scaffold of erlotinib, a widely used inhibitor of the epidermal growth factor receptor, or EGFR, a signaling protein that drives proliferation and survival in many lung cancers and, to a more variable degree, in melanoma. Fused to this kinase-inhibitor core is a carborane, a bulky, three-dimensional cluster rich in boron atoms. Earlier work by the same groups had shown that the compound inhibits wild-type EGFR and retains measurable activity against the T790M resistance mutation, displays favorable drug-like properties, and showed no mutagenicity or relevant acute oral toxicity in preclinical testing. In glioma models, it promoted selective boron accumulation and improved survival in mice. The new study, published in Frontiers in Chemical Biology, extends the platform to two new tumor contexts: cutaneous melanoma and non-small cell lung cancer.

The team began by profiling the two tumor models at the molecular level. Using quantitative real-time PCR, they found that the human lung cancer cell line NCI-H125 expressed markedly higher levels of both EGFR and LAT1 — the amino acid transporter responsible for shuttling BPA into cells — than the melanoma line Mel J. This biological asymmetry set up a natural experiment: a tumor heavily dependent on EGFR signaling and LAT1-mediated transport versus one with more modest expression of both. When the researchers measured sensitivity to the compound, the lung cancer cells proved more vulnerable, with a 24-hour IC50 of 39.73 micromolar compared with 60.79 micromolar in melanoma, a pattern consistent with the compound’s EGFR-directed design.

The boron delivery results were striking. Cells were incubated with the compound at a total boron concentration of 2.7 parts per million, roughly ninefold lower than the 25 parts per million supplied with BPA. Yet after 24 hours, the compound produced greater intracellular boron accumulation in both cell lines than BPA did. When uptake was normalized to the amount of boron actually added to the culture medium — a measure of delivery efficiency rather than sheer dose — the compound outperformed BPA by a factor of 56.65 in melanoma cells and 24.06 in lung cancer cells. Neutron autoradiography, performed with UV-C sensitization on cells grown on Lexan polycarbonate, confirmed that detectable particle tracks were generated in cells treated with the compound even at just 0.5 parts per million of boron-10, one-twentieth of the concentration used for BPA, supporting efficient intracellular delivery.

To test whether this translated into therapeutic effect, the researchers irradiated treated cells with a thermal neutron beam at the RA-3 nuclear reactor facility at Argentina’s Centro Atómico Ezeiza, calibrated to deliver a total physical dose of 2 gray. In vitro BNCT significantly reduced cell viability in both tumor models seven days after irradiation. In melanoma cells, BPA-mediated and compound-mediated BNCT produced comparable reductions. In lung cancer cells, however, the response was more pronounced, particularly with the new compound, suggesting that the EGFR-targeted approach may offer an edge in EGFR-dependent tumors. Notably, the compound alone also modestly but significantly reduced lung cancer cell viability even without irradiation, underscoring its dual nature.

Cell death analysis 48 hours after irradiation revealed mechanistic differences between the two tumor types. Fluorescence microscopy using Hoechst 33258, fluorescein diacetate and propidium iodide showed that BNCT predominantly induced necrotic death in melanoma cells, with necrosis reaching roughly 50 percent — an expected outcome given the high-linear-energy-transfer nature of the alpha and lithium particles, which shred cellular structures beyond repair. The lung cancer cells, by contrast, displayed a mixed response: compound-mediated BNCT produced approximately 35 percent apoptosis alongside 35 percent necrosis, a larger apoptotic fraction than BPA-mediated BNCT induced. The researchers suggest this apoptotic component may reflect suppression of EGFR-driven survival signaling through the PI3K/AKT and RAS/RAF/MEK/ERK pathways, effectively a pharmacological assist to the physical radiation damage.

The in vivo experiments strengthened the case considerably. Athymic nude mice bearing either melanoma or lung cancer xenografts received the compound intraperitoneally at 1 milligram per kilogram three times per week over an 18-day period, without any neutron irradiation. In both models, treatment significantly delayed tumor growth relative to vehicle controls, demonstrating that the compound’s intrinsic antitumor activity survives in living animals. Body weight remained stable throughout the treatment period in both models, with no significant differences between treated and control animals, suggesting the regimen was well tolerated. Boron biodistribution was then assessed at the end of the treatment schedule, one hour after a final administration. In the lung cancer model, co-administration of the compound with BPA yielded a tumor boron concentration of approximately 23 micrograms per gram of tissue — a tumor-to-blood ratio of 6.3, a tumor-to-surrounding-skin ratio of about 2.2 and a tumor-to-lung ratio of roughly 3.7, the last metric included as a disease-relevant parameter since normal lung would be a critical tissue in any future lung-directed application. In the melanoma model, the combination achieved about 16 micrograms per gram with a tumor-to-blood ratio near 4.7.

The authors are careful to note the boundaries of what these experiments show. Because biodistribution was measured at the end of a repeated-dose protocol, residual compound from prior injections may have contributed to the measured boron concentrations, and the design does not establish the compound’s independent biodistribution profile or allow a direct head-to-head comparison of tumor selectivity with BPA. The choice to evaluate the compound alongside BPA was informed by earlier experience with a structurally related boron-containing lapatinib derivative, which alone produced low tumor boron levels at a comparable intraperitoneal dose. Future work, the team writes, should optimize the administration route — potentially intravenous or intratumoral — before attempting an independent biodistribution assessment. Even so, the study establishes that an EGFR-targeted carborane-quinazoline can be incorporated into a BPA-based regimen without compromising preferential tumor boron accumulation, while simultaneously contributing antitumor activity of its own. As the search for next-generation boron carriers intensifies, molecules that fuse targeted kinase pharmacology with dense boron payloads offer a compelling blueprint: a therapy that softens up the tumor chemically before the neutrons finish the job.

Subject of Research: Development of an EGFR-targeted carborane-quinazoline compound that combines intrinsic antitumor activity with boron delivery for boron neutron capture therapy in melanoma and non-small cell lung cancer models.

Article Title: A bifunctional carborane-quinazoline combines tumor growth inhibition and boron delivery for boron neutron capture therapy in melanoma and non-small cell lung cancer models

Article References: Dagrosa, M. A., Carpano, M., Pastini, A., Pavlik, E., Bechis, A., Olivera, M. S., Vallecorsa, P., Dávila, B., Parodi, L., Salem, E. Y., Espain, M. S., Portu, A. M., Todaro, L., Urtreger, A., Menendez, P., Cerecetto, H., & Couto, M. (2026). A bifunctional carborane-quinazoline combines tumor growth inhibition and boron delivery for boron neutron capture therapy in melanoma and non-small cell lung cancer models. Frontiers in Chemical Biology, 5, Article 1945455. https://doi.org/10.3389/fchbi.2026.1945455

Image Credits: AI Generated

DOI: 10.3389/fchbi.2026.1945455

Keywords: boron neutron capture therapy, BNCT, carborane, quinazoline, EGFR, erlotinib, melanoma, non-small cell lung cancer, BPA, LAT1, boron delivery agents, targeted cancer therapy