# Brain Bleeding Risks with MAPK-Targeted Cancer Drugs: What Patients and Families Need to Know

Drugs that block the MAPK signaling pathway are a key driver of several brain tumors. These drugs carry a consistent but variable risk of bleeding inside the brain or inside the tumor itself. This finding comes from a systematic review of 10 data sources. Reported rates ranged from under 1% for symptomatic intracranial hemorrhage with dabrafenib in adults with melanoma that had spread to the brain. Reported rates went up to 9% for intratumoral hemorrhage in the pooled safety population for the pediatric drug tovorafenib. The highest-risk combination identified was a BRAF inhibitor plus stereotactic radiosurgery, which tripled the odds of intracranial bleeding compared with radiosurgery alone (odds ratio 3.16; 95% confidence interval 1.43–6.96). The authors call for standardized definitions and prospective imaging surveillance to pin down the true incidence of this complication.

# Intratumoral and intracranial hemorrhage associated with MAPK-pathway targeted therapy: a systematic review and mechanistic synthesis.

## Table of Contents

- Key Points
- Background: Why This Research Matters
- How the Researchers Conducted This Review
- How Many Studies Were Included
- Dabrafenib Findings: Melanoma That Has Spread to the Brain
- A Detailed Patient Case: Bleeding After Dabrafenib Plus Trametinib
- Tovorafenib Findings: Pediatric Low-Grade Glioma
- Other MAPK Inhibitors in Brain Tumors and NF1
- Combining MAPK Inhibitors With Brain Radiation
- How Might These Drugs Cause Bleeding?
- What This Means for Patients
- Limitations: What This Review Could Not Prove
- Recommendations and Practical Advice
- Frequently Asked Questions
- Source Information

## Key Points

- MAPK-pathway inhibitors carry a consistent but variable risk of bleeding inside the brain or tumor, ranging from under 1% to 9% depending on the drug and patient group.
- In a pediatric tovorafenib safety population of 140 patients, intratumoral hemorrhage occurred in 9%; most bleeding events were minor nosebleeds.
- Combining a BRAF inhibitor with stereotactic radiosurgery tripled the odds of intracranial hemorrhage compared with radiosurgery alone (odds ratio 3.16).
- Serious bleeding can occur months after starting treatment or even after the drug is stopped because of disease progression.
- Report new or severe headaches, seizures, sudden weakness, confusion, or vision changes immediately, and tell your team about all medications and supplements.

## Background: Why This Research Matters

The **MAPK pathway** (mitogen-activated protein kinase pathway) is a chain of chemical signals inside cells that tells them when to grow and divide. In many brain tumors, this pathway is switched on permanently, driving the tumor's growth. Researchers call this "constitutive activation."

Three specific genetic changes account for most of these cases. They are **BRAF fusions** (abnormal joining of the BRAF gene to another gene), **BRAF V600E mutations** (a specific spelling error in the BRAF gene), and **NF1 loss** (loss of a gene that normally puts the brakes on growth). All three converge on the same signaling chain, known as RAS–RAF–MEK–ERK. Together, these alterations account for the majority of molecular changes found in **pediatric low-grade glioma (pLGG)** — the most common childhood brain tumor.

Drugs that block this pathway have transformed treatment. This class includes BRAF inhibitors, MEK inhibitors, combinations of the two, and more recently **type II RAF inhibitors**. These drugs are now used to treat pLGG, pediatric BRAF V600–mutant high-grade glioma, **NF1** (neurofibromatosis type 1)–associated plexiform neurofibromas (nerve sheath tumors), and BRAF V600–mutant melanoma that has spread to the brain.

One of the newest agents is **tovorafenib** (brand name Ojemda). It is a once-weekly oral type II RAF inhibitor that penetrates the central nervous system (CNS) and works against both BRAF fusions and BRAF V600 alterations. It received accelerated FDA approval in April 2024 for relapsed or refractory BRAF-altered pLGG.

The approval was based on the phase 2 FIREFLY-1 trial. In its initial report, the overall response rate (ORR — the proportion of patients whose tumors shrank) was **67%** using prespecified RANO-HGG criteria, with a median duration of response of **16.6 months**. Using the more disease-appropriate RAPNO-LGG criteria in the registrational population of 76 patients, the FDA approval summary reported an ORR of **51%** and a median duration of response of **13.8 months**.

These results highlight how valuable sustained MAPK-pathway inhibition can be. But they also make it critical to understand the side effects, because these drugs are moving into earlier lines of therapy and broader groups of patients with CNS tumors.

One such concern is bleeding. **Intratumoral hemorrhage (ITH)** means bleeding inside the tumor itself. **Intracranial hemorrhage (ICH)** means bleeding anywhere inside the skull. Intratumoral hemorrhage is already a known complication of certain primary and metastatic brain tumors, particularly melanoma. But a distinct signal has emerged specifically with MAPK-pathway inhibition itself.

Hemorrhage is a labeled warning for selumetinib and tovorafenib. Fatal CNS bleeding has been reported with dabrafenib-based regimens in melanoma brain metastases. However, these observations are scattered across different tumor types, drugs, and reporting systems. Classification of CNS versus non-CNS bleeding is inconsistent, use of imaging surveillance varies, and integration with data on other exposures such as radiation is limited. The authors performed this systematic review to pull that evidence together.

## How the Researchers Conducted This Review

The team searched **PubMed/MEDLINE** and **Embase** — two major medical research databases — from each database's beginning through **May 29, 2026**. They followed **PRISMA** (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) principles, a standard set of rules for conducting and reporting reviews.

The search combined bleeding-related terms ("intratumoral hemorrhage," "intracranial hemorrhage," "tumor hemorrhage," "bleeding") with drug and drug-class terms. Those drug terms included "MEK inhibitor," "BRAF inhibitor," "RAF inhibitor," and the specific names selumetinib, mirdametinib, trametinib, cobimetinib, binimetinib, dabrafenib, vemurafenib, encorafenib, and tovorafenib.

The search was supplemented in three ways: FDA approval packages, FDA prescribing information, and hand-searching the reference lists of included articles and relevant narrative reviews.

To be included, a study had to be in humans. The study also had to report intratumoral or intracranial hemorrhage occurring in a timeframe linked to exposure to a MAPK-pathway targeted agent, in any tumor type. Non-CNS bleeding events, such as nosebleeds (epistaxis), were recorded only when the source reported them alongside CNS events, to provide context rather than to be counted as CNS hemorrhage.

Melanoma studies were included when they reported outcomes in patients with brain metastases. The authors note an important caveat here: some melanoma safety datasets pool patients with and without documented intracranial disease. That means the denominators in those datasets may not correspond only to patients actually at risk for CNS hemorrhage. The authors interpreted those numbers with that limitation in mind.

Eligible study designs included:

- Prospective and retrospective clinical trials
- Regulatory safety summaries
- Observational cohorts
- Case series
- Case reports

Narrative reviews without original patient-level or trial-level data were excluded from data extraction but were screened for citation chaining (following their references to find additional studies). The team also excluded non-human studies, preclinical reports, conference abstracts without enough denominator information (that is, without knowing how many patients were studied), and non-English publications without an available translation.

The researchers extracted a detailed set of variables for each study: design and enrollment period, patient demographics, tumor type, tumor location, molecular alterations, targeted agents and regimens, cumulative exposure or time on therapy, timing of hemorrhage relative to starting therapy, hemorrhage phenotype (intratumoral versus intracranial), **CTCAE grade** (Common Terminology Criteria for Adverse Events — a standard severity scale), prior or concurrent radiation therapy including stereotactic radiosurgery, use of blood thinners or antiplatelet drugs, management approach, and clinical outcome including death.

Because reporting was so inconsistent and the events were rare, the team did **not** perform a statistical meta-analysis (pooling numbers mathematically). Instead, they synthesized the results narratively and in summary tables. When denominators were unclear, they relied on the regulatory or primary trial publication as the preferred source.

Two researchers (S.D. and A.C.) screened titles and abstracts independently. Full-text review was done by S.D. with verification by A.C., and disagreements were resolved by consensus with a third author (M.S.A.). No automation tools were used, and no data were sought from the original study investigators.

Where possible, hemorrhage events are expressed as incidence proportions (n/N). For the combined analysis of BRAF inhibitors plus stereotactic radiosurgery, the reported odds ratio with 95% confidence interval is cited. The authors did not perform a formal risk-of-bias assessment using a single validated instrument, because the evidence base spanned too many different designs for one appraisal tool to apply uniformly. The full search strings for each database are provided in a supplementary appendix, and the study selection process is summarized in a PRISMA flow diagram.

## How Many Studies Were Included

The database search identified **580 records** — 312 from PubMed/MEDLINE and 268 from Embase. An additional **25 records** came from FDA approval packages, prescribing information, and reference-list hand-searching.

After removing **206 duplicates**, the team screened **399 records** at the title and abstract level. Of these, **344 were excluded**. The remaining **55 reports** were assessed in full text, and **45 were excluded** for specific reasons:

- Non-human or preclinical report: 14
- Conference abstract without sufficient denominator data: 10
- Non-English-language publication without available translation: 7
- Narrative review without original patient- or trial-level data: 14

This left **10 data sources** in the final synthesis. The evidence base consisted of prospective clinical trials and their pooled safety updates. The evidence base also included regulatory safety reviews and FDA prescribing information, post-marketing pharmacovigilance analyses (safety monitoring after a drug reaches the market), small observational cohorts, and individual case reports.

Notably, **no prospective study was specifically designed to evaluate CNS hemorrhage risk as a primary endpoint** in patients treated with MAPK-pathway drugs. Across all sources, intratumoral and intracranial hemorrhage were infrequent but reproducibly reported. When they were severe, they represented dose-limiting or fatal toxicities.

## Dabrafenib Findings: Melanoma That Has Spread to the Brain

Dabrafenib-based regimens in melanoma brain metastases have produced the most mature adult CNS safety dataset available. The results show a low absolute rate of severe CNS bleeding — but events do occur, and they are occasionally fatal.

In the phase 2 **BREAK-MB** trial of dabrafenib used alone (172 patients), intracranial hemorrhage was reported in **10 patients, or 6%**. No hemorrhagic event was fatal. The investigators attributed these bleeds to the underlying melanoma metastases rather than to a direct drug effect.

In the phase 2 **COMBI-MB** trial of dabrafenib plus trametinib (125 patients), a single fatal intracranial tumor hemorrhage was reported — **1 of 125 patients, or 0.8%**. Other serious bleeding events were uncommon.

In pooled clinical-trial safety data for dabrafenib-containing regimens across all indications (1,087 patients), intracranial hemorrhage occurred in **7 patients (0.6%)**, including **5 fatal events (0.5%)**. The pooled population included patients with melanoma, non-small cell lung cancer, thyroid cancer, and glioma.

On the reassuring side, five-year follow-up of the COMBI-d and COMBI-v trials did **not** identify any new late-emerging CNS hemorrhage signals with long-term dabrafenib/trametinib exposure. In other words, the bleeding risk did not appear to grow over years of treatment.

Putting these numbers in plain terms: for adults with melanoma brain metastases treated with dabrafenib-based therapy, roughly **6 in 100** experienced an intracranial hemorrhage in the monotherapy trial, and about **1 in 125** experienced a fatal intracranial hemorrhage in the combination trial. In the broad pooled dataset, about **6 in 1,000** had an intracranial hemorrhage.

## A Detailed Patient Case: Bleeding After Dabrafenib Plus Trametinib

A published case report provides a close-up view of how this complication can unfold. Lee and colleagues described a 48-year-old man with metastatic melanoma of unknown primary origin.

His disease involved the right collarbone (clavicle), liver, adrenal gland, and a lesion in the left frontotemporal region of the brain. He underwent gross total resection (complete surgical removal) of the brain lesion, followed by adjuvant CyberKnife stereotactic irradiation — a precise, high-dose form of radiation.

After his disease progressed through two prior lines of therapy, he started treatment with dabrafenib plus trametinib. **Four months after starting therapy**, he developed a severe headache and had a seizure.

Brain MRI showed extensive edema (swelling) in the left frontal lobe with midline shift — meaning the swelling was pushing brain structures out of their normal position, a dangerous sign. An emergency craniotomy (surgical opening of the skull) was performed, and an intracranial hemorrhage was identified during the operation.

When pathologists examined the removed tissue, they found organizing hemorrhage and necrosis (dead tissue) with surrounding gliosis (scarring of brain tissue). Critically, there were **no viable tumor cells** present.

## Tovorafenib Findings: Pediatric Low-Grade Glioma

In children, tovorafenib has produced the most striking hemorrhage signal identified to date. The numbers require careful reading, because different analyses use different patient groups.

In the **FIREFLY-1** trial (137 patients across arms 1 and 2), hemorrhagic events of any type occurred in **58 patients (42%)**. Most of these were minor. Grade 3–4 events (moderate to severe) were reported in **7 patients (5%)**. Separately, a single **grade 5 (fatal) tumor hemorrhage** occurred — **1 of 137 patients, or 0.7%** — which was not counted among the seven grade 3–4 events.

That fatal event occurred in a patient with a disseminated leptomeningeal mixed glial-neuronal tumor. It developed **21 days after the final dose** of tovorafenib, which had previously been discontinued because of disease progression.

Among the four patients in FIREFLY-1 with serious treatment-related tumor hemorrhage (as described in the trial report), two had a documented history of intratumoral hemorrhage before starting tovorafenib. All four of these serious events resolved. Together with the fatal event — which occurred after the drug was stopped, at the time of tumor progression — these observations suggest that **prior bleeding and the tumor's natural history are important contributors** to the uncommon symptomatic CNS hemorrhage seen with tovorafenib.

A separate pooled safety population informed the FDA label. In that group of **140 patients**, hemorrhagic events occurred in **37%** of patients overall. Specifically:

- Intratumoral hemorrhage: **9%** (about 9 in 100 patients)
- Epistaxis (nosebleeds): **26%**

"Hemorrhage" is listed among the most common (occurring in 30% or more of patients) adverse reactions in the label, and it is included as a labeled Warning and Precaution. This labeling language requires baseline assessment of bleeding risk, monitoring during therapy, and dose interruption or discontinuation after significant hemorrhagic events.

**Three different tovorafenib denominators recur in the literature, and they should not be confused or added together:**

1. The registrational efficacy population (**n = 76**), used for the RAPNO-LGG response analysis
1. The FIREFLY-1 safety population (**n = 137**)
1. The pooled safety population informing FDA labeling (**N = 140**)

The 42% any-site bleeding figure comes from the 137-patient safety population. The 9% intratumoral hemorrhage figure comes from the 140-patient pooled population. These groups overlap substantially and represent the same clinical program, so the two figures must not be summed.

Because the trial grouped bleeding events across all body sites, the 42% figure is dominated by minor events — most commonly nosebleeds. CNS tumor hemorrhage represented the severe tail of that distribution.

## Other MAPK Inhibitors in Brain Tumors and NF1

Reports in primary CNS tumors outside of pLGG are limited. In a phase 2 trial of dabrafenib plus trametinib in BRAF V600–mutant pediatric high-grade glioma (41 patients), hemorrhagic events were not highlighted as a major toxicity. However, the small number of patients limits how precise any conclusion can be.

In the **TADPOLE** trial — the largest prospective pediatric study of first-line dabrafenib plus trametinib in BRAF V600–mutant pLGG — nosebleeds and other bleeding events combined were reported in **25%** of patients receiving the targeted combination, compared with the carboplatin/vincristine arm. Isolated intracranial hemorrhage events were described, but they were not systematically quantified, because CNS-specific hemorrhage data were not a prespecified endpoint. No grade 3 or higher intratumoral or intracranial hemorrhage was attributed to dabrafenib plus trametinib as a treatment-related event in that population. The trial enrolled 73 patients in the relevant comparison.

In the phase 1/2 pediatric trametinib program, bleeding events were infrequent and primarily affected the skin or mucous membranes.

For **selumetinib** in NF1-associated plexiform neurofibromas, the **SPRINT** trial (50 patients) did not identify excess clinically significant bleeding in the pediatric cohort. However, the FDA label does include an increased risk of bleeding as a Warning and Precaution. This is attributed in part to the **vitamin E (α-tocopheryl) excipient** — a filler ingredient — in the selumetinib capsule formulation, which may impair platelet function and add to antithrombotic effects.

The multicenter phase 2 **PBTC-029B** trial studied selumetinib in children with recurrent or progressive low-grade glioma (130 patients across strata). One of two patients in stratum 6 discontinued treatment because of a **grade 2 intratumoral hemorrhage** accompanied by a **grade 3 headache**; both were considered possibly related to the study drug. Stratum 6 included patients with non-NF1-associated recurrent or progressive pLGG whose molecular characterization could not be completed. Molecular characterization could not be completed because of insufficient tumor tissue or because the laboratory test failed. These patients otherwise met all eligibility criteria. No other clinically significant intratumoral or intracranial hemorrhages were reported as treatment-related toxicity; bleeding was otherwise limited to minor skin and mucous membrane events.

The absence of a broader CNS hemorrhage signal in that cohort may reflect the lower intrinsic bleeding tendency of the tumor types enrolled. Those tumor types were predominantly pilocytic astrocytoma. This contrasts with melanoma brain metastases or MAPK-altered pLGG subgroups with more aggressive histology.

## Combining MAPK Inhibitors With Brain Radiation

A consistent secondary signal involves combining MAPK-pathway inhibitors with radiation directed at the CNS.

In a systematic review and meta-analysis comparing BRAF inhibitors plus **stereotactic radiosurgery (SRS)** — a highly focused radiation technique — against SRS alone in melanoma brain metastases, combination therapy was associated with **higher odds of intracranial hemorrhage**. The odds ratio was **3.16** (95% CI 1.43–6.96; p = 0.004).

In plain terms, the odds of bleeding were about three times higher with the combination. This was true even though the combination improved survival and local tumor control. The pooled cohort across eight studies included **976 patients** for the survival and local-control analyses, though the intracranial hemorrhage comparison came from a smaller subset of those studies.

## How Might These Drugs Cause Bleeding?

The authors propose a mechanistic framework linking MAPK-pathway inhibition to blood vessel problems in the brain. The explanation has several interlocking parts.

Under normal conditions, growth factor binding to a **receptor tyrosine kinase (RTK)** — a docking protein on the cell surface — recruits an adaptor complex called **SHC–GRB2–SOS** (growth factor receptor–bound protein 2 and son of sevenless). This activates **RAS**, which drives sequential phosphorylation of **RAF**, **MEK** (mitogen-activated protein kinase kinase), and **ERK** (extracellular signal-regulated kinase). Once in the nucleus, ERK regulates the transcription of genes governing angiogenesis (new blood vessel formation), endothelial proliferation, and cell survival.

Drugs in this class block the pathway at three different points:

-
**Type I BRAF V600E inhibitors**: dabrafenib, encorafenib, vemurafenib
-
**Type II RAF inhibitor**: tovorafenib, which is active against both BRAF fusions and V600 alterations
-
**MEK inhibitors**: trametinib, binimetinib, cobimetinib, selumetinib, mirdametinib

Here is the key problem. In endothelial cells (the cells lining blood vessels), normal **VEGF receptor-2 signaling** (vascular endothelial growth factor receptor-2) converges on this very same axis to maintain the integrity of the vessel wall. When drugs block the pathway, they compromise endothelial homeostasis — the blood vessels' ability to keep themselves stable — increasing vascular permeability (leakiness) and fragility.

At the same time, MAPK inhibition reduces the expression of **tissue factor** in tumor cells. Tissue factor is a protein involved in blood clotting. Additionally, the rapid shrinking of the tumor mass destabilizes the disorganized, abnormal blood vessels that had formed to feed the tumor.

Together, these three effects provide a plausible mechanistic explanation for the intratumoral and intracranial hemorrhage observed across BRAF, MEK, and type II RAF inhibitor exposure. The three effects are leaky and fragile vessel walls, reduced clotting capacity, and destabilized tumor vasculature.

## What This Means for Patients

The overarching message is that MAPK-pathway inhibitors are associated with a **consistent but heterogeneous** CNS hemorrhage signal. In other words, the bleeding risk is real and shows up across multiple drugs and studies. But the reported rates vary a lot depending on the drug, the patient population, and how carefully researchers looked for bleeding.

That last point matters enormously. The authors highlight a key distinction between two ways of measuring hemorrhage:

-
**Active imaging-based surveillance** — regularly scanning patients' brains, which finds more bleeds, including silent ones with no symptoms
-
**Clinical reporting** — counting only bleeds that caused symptoms and were noticed by doctors or patients, which finds fewer

This difference explains why reported incidence ranges so widely: from **less than 1% symptomatic intracranial hemorrhage** with dabrafenib in adult melanoma, to **9% intratumoral hemorrhage** in the pooled pediatric tovorafenib safety population.

For patients and families, the practical takeaways are these. The most common bleeding symptom is a nosebleed, which is usually minor. Serious bleeding into the brain or tumor is uncommon. But when bleeding happens, it can be life-threatening. Bleeding may occur months after starting treatment. Bleeding may even occur after the drug has been stopped because of disease progression, as happened with the fatal tovorafenib case.

It is also worth noting that a history of previous tumor bleeding appears to increase the risk of serious bleeding on tovorafenib. Two of the four FIREFLY-1 patients with serious treatment-related tumor hemorrhage had documented prior intratumoral hemorrhage before starting the drug.

## Limitations: What This Review Could Not Prove

The authors are candid about the limits of this evidence base, and patients and clinicians should weigh them when interpreting the findings.

-
**No standardized CNS-specific bleeding definitions.** Different trials defined and categorized bleeding in different ways, making direct comparisons difficult.
-
**Rare events, heterogeneous designs.** Because the events are uncommon and the studies vary widely in design, the team could not mathematically pool the data into a single unified incidence estimate.
-
**Susceptibility to selective reporting.** Studies may report bleeding more or less thoroughly depending on what they were designed to examine.
-
**Ascertainment bias.** How hard a study looked for bleeding strongly influenced what it found.
-
**Pooled melanoma datasets.** Some safety datasets combine patients with and without documented intracranial disease, so their denominators do not map exactly onto patients truly at risk for CNS hemorrhage.
-
**No single risk-of-bias tool applied.** The evidence spanned too many study designs for one validated appraisal instrument to apply uniformly.
-
**No prospective study with CNS hemorrhage as the primary endpoint** has ever been conducted in MAPK-pathway–treated patients.

One additional transparency point: this systematic review was **not** prospectively registered in PROSPERO, the international database for registering systematic reviews in advance. Registration helps reduce bias by locking in the study plan before results are known. The authors note that all database search strings are available in a supplementary appendix to support reproducibility.

## Recommendations and Practical Advice

The authors conclude with a clear call to action for the research community, and their findings carry practical implications for patients currently on these drugs.

For the field, three steps are needed to define the true incidence and risk of CNS bleeding as MAPK-directed therapy expands in neuro-oncology:

1.
**Standardized definitions** of intratumoral versus intracranial hemorrhage across all trials
1.
**CNS-specific CTCAE capture** — recording brain bleeding events using the standard severity scale in a way that specifically distinguishes them from bleeding elsewhere in the body
1.
**Prospective imaging surveillance** — scanning patients on a schedule, rather than relying only on symptoms, to find out how often bleeding actually occurs

Existing FDA labeling already sets expectations for clinical practice. For tovorafenib, the label requires baseline assessment of bleeding risk, monitoring during therapy, and dose interruption or discontinuation after significant hemorrhagic events.

For patients and caregivers, the actionable points are:

- Ask your care team about your specific bleeding risk before starting a MAPK-pathway inhibitor, especially if you have had any prior bleeding in a brain tumor.
- Tell your team about all medications you take, including blood thinners, antiplatelet drugs, and supplements — particularly vitamin E, which is also present as an ingredient in the selumetinib capsule.
- Report new or severe headaches, seizures, sudden weakness or numbness, confusion, or vision changes immediately. In the published case, a severe headache and seizure were the first signs of a serious bleed four months into treatment.
- Report nosebleeds that are frequent, prolonged, or hard to stop.
- If you are receiving stereotactic radiosurgery alongside a BRAF inhibitor, discuss the roughly three-fold higher odds of intracranial hemorrhage with your team. This risk must be balanced against the improved tumor control and survival this combination provides.
- Remember that bleeding events can occur even after the drug has been stopped because of disease progression, so continued vigilance matters.

It is also worth keeping the benefits in perspective. In FIREFLY-1, 67% of patients responded by RANO-HGG criteria with a median response duration of 16.6 months. For most patients, the therapeutic value of these drugs continues to outweigh the bleeding risk — but that risk deserves informed, ongoing attention.

## Frequently Asked Questions

### What is the risk of bleeding in the brain from MAPK-targeted drugs?

Reported rates vary widely. In adults with melanoma spread to the brain, symptomatic intracranial hemorrhage with dabrafenib was under 1% in pooled data. In a pediatric tovorafenib safety population of 140 patients, intratumoral hemorrhage was 9%. Most bleeding events are minor nosebleeds, but serious brain bleeding can occur and may be life-threatening.

### What symptoms of brain bleeding should I watch for?

Report new or severe headaches, seizures, sudden weakness or numbness, confusion, or vision changes immediately. In one published case, a severe headache and seizure were the first signs of a serious bleed four months after starting dabrafenib plus trametinib. Also report nosebleeds that are frequent, prolonged, or hard to stop.

### Does combining a BRAF inhibitor with stereotactic radiosurgery increase bleeding risk?

Yes. A systematic review and meta-analysis found that combining a BRAF inhibitor with stereotactic radiosurgery for melanoma brain metastases tripled the odds of intracranial hemorrhage compared with radiosurgery alone. The odds ratio was 3.16, with a 95% confidence interval of 1.43–6.96. This was true even though the combination improved survival and local tumor control.

### Can bleeding happen after stopping the drug?

Yes. In the FIREFLY-1 trial of tovorafenib, one fatal tumor hemorrhage occurred 21 days after the final dose, in a patient whose drug had been stopped because of disease progression. This means continued vigilance is important even after treatment ends, especially if the tumor is progressing.

### What should I tell my care team before starting a MAPK inhibitor?

Ask about your specific bleeding risk, especially if you have had any prior bleeding in a brain tumor. Tell your team about all medications and supplements, including blood thinners, antiplatelet drugs, and vitamin E, which is also an ingredient in the selumetinib capsule. For tovorafenib, the label requires baseline bleeding risk assessment and monitoring during therapy.

### How common is serious bleeding with tovorafenib in children?

In the FIREFLY-1 trial of 137 patients, hemorrhagic events of any type occurred in 42%, but most were minor, such as nosebleeds. Grade 3–4 events occurred in 5%, and one fatal tumor hemorrhage occurred in 0.7%. In a separate pooled safety population of 140 patients, intratumoral hemorrhage was 9%.

### Why do these drugs cause bleeding in the brain?

The authors propose that blocking the MAPK pathway affects blood vessels in several ways. It can make vessel walls leaky and fragile, reduce tissue factor (a clotting protein) in tumor cells, and destabilize the abnormal tumor blood vessels as the tumor shrinks. Together, these effects may explain the bleeding seen with BRAF, MEK, and type II RAF inhibitors.

### When should a patient with a brain tumor starting a MAPK-targeted drug like dabrafenib, trametinib, or tovorafenib seek a second opinion?

A second opinion is worth considering before starting a MAPK-pathway inhibitor. This is especially important if you have had prior bleeding in a brain tumor. That history appears to raise the risk of serious bleeding on tovorafenib. It also helps if stereotactic radiosurgery is planned alongside a BRAF inhibitor, because that combination carries roughly three times higher odds of intracranial hemorrhage. A review can clarify your bleeding risk, medication interactions, and monitoring plan. Diagnostic Detectives Network provides independent expert second opinions.

## Source Information

**Original article title:** Intratumoral and intracranial hemorrhage associated with MAPK-pathway targeted therapy: a systematic review and mechanistic synthesis.

**Publication details:** Journal of Neuro-Oncology (2026) 178:103. DOI: 10.1007/s11060-026-05714-0. Received 23 June 2026; accepted 13 July 2026; published online 15 July 2026. The article is open access under a Creative Commons license.

**Article type:** Review (systematic review and mechanistic synthesis), not prospectively registered in PROSPERO.

*This patient-friendly article is based on peer-reviewed research.*


---
Publisher: Diagnostic Detectives Network (https://diagnosticdetectives.com) — independent multi-expert medical second opinions, worldwide, private-pay.
Author byline: Anton Titov, MD, PhD. Contact: https://diagnosticdetectives.com/pages/contact
Canonical page: https://diagnosticdetectives.com/products/brain-bleeding-risks-with-mapk-targeted-cancer-drugs-what-patients-and-families-need-to-know