Disease Stabilization with MEK Inhibitors in NF1-Associated Plexiform Neurofibromas: A Systematic Review and Meta-Analysis with Subgroup Analyses by Age and Study Design

Abstract

Background

Plexiform neurofibromas (PN) represent a significant cause of morbidity among patients diagnosed with neurofibromatosis type 1 (NF1). MEK inhibitors continue to be developed as targeted therapies by inhibiting the mitogen-activated protein kinase pathway to treat PN; nonetheless to this day, therapeutic responses have varied across different patient populations and clinical contexts, and the overall efficacy and tolerability of these agents remain incompletely characterized.

Objective

We aimed to systematically evaluate the efficacy and safety of MEK inhibitor therapy in patients with NF1-associated PN and to evaluate differences among key subgroups based on the most contemporary metadata.

Methods

A comprehensive search was performed across electronic databases to identify studies that reported outcomes related to MEK inhibitor therapy in NF1-associated PN. Pooled proportions were calculated using a random-effects meta-analysis with logit transformation. Outcomes assessed included objective response rate, disease control rate, disease progression rate, and grade ≥ 3 adverse events.

Results

A total of 23 studies comprising 769 patients were included. The pooled objective response rate was estimated at 56% (95% confidence interval [CI] 46–65; I² = 79.2%), with significantly higher response rates observed in clinical trials (61%) compared with real-world cohorts (44%) [p = 0.035], while no statistically significant difference was observed between pediatric (58%) versus adult populations (51%) [p = 0.407]. The pooled disease control rate was 96% (95% CI 91–98; I² = 17%) and the pooled disease progression rate was estimated at 2% (95% CI 1–5; I² = 0%), both reflecting on-treatment outcomes. Grade ≥ 3 adverse events occurred in 13% of patients (95% CI 6–25; I² = 51.9%). Subgroup analyses revealed comparable disease control across study settings, with moderate variability in response and toxicity estimates.

Conclusions

The use of MEK inhibitors is associated with high rates of disease control and minimal tumor progression in patients with NF1-related PN, with consistent effects observed across clinical trial and real-world environments. Although tumor reduction occurs in some patients, the predominant therapeutic benefit appears to be sustained disease stabilization, with response variability noted among different age groups and study designs.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40263-026-01316-6.

Key Points

MEK inhibitors help control tumor growth in most patients with plexiform neurofibromas related to neurofibromatosis type 1.

About half of patients experience measurable tumor shrinkage, but the main benefit of treatment is preventing tumors from growing during therapy.

Treatment effects appear similar across different patient groups, including children and adults, and in both clinical trials and real-world settings.

Serious side effects are relatively uncommon, but less severe side effects may still affect long-term treatment use and patient quality of life.

Differences in how tumor response is measured and the natural behavior of these tumors make it difficult to directly compare results across studies.

Introduction

Neurofibromatosis type 1 (NF1) is one of the most common inherited tumor predisposition syndromes, affecting about 1 in 3000 people worldwide [1]. It results from pathogenic variants in the NF1 gene, which encodes neurofibromin, a negative regulator of the RAS signaling pathway [2, 3]. The loss of neurofibromin causes continuous activation of the RAS–RAF–MEK–ERK signaling cascade, which promotes tumor growth and cell proliferation [2]. One of the most clinically important features of NF1 is the presence of plexiform neurofibromas (PN), which are complex tumors of the peripheral nerve sheath that develop in up to 30–50% of patients with NF1 and are usually diagnosed during childhood [4–6]. These tumors can lead to significant health problems through progressive growth, pain, neurological deficits, disfigurement, airway issues, and functional impairments, greatly impacting quality of life [4–6].

Historically, the management of PN has been challenging. Surgical resection remains the only potentially curative option; however, complete removal is often not feasible because of the infiltrative growth pattern of these tumors and their close association with critical neurovascular structures [5, 7]. As a result, many patients experience disease progression with limited therapeutic options. Importantly, prospective natural history studies conducted by the National Cancer Institute have demonstrated that PN frequently exhibit progressive growth over time and rarely undergo spontaneous regression, providing a benchmark for which treatment responses can be interpreted. These observations underscored the need for effective systemic therapies [7–9]. The identification of dysregulated RAS/mitogen-activated protein kinase (MAPK) pathway signaling in NF1-associated tumors has offered a valuable therapeutic target, leading to the development of MEK inhibitors as targeted therapies for PN [3, 6, 10] (Fig. 1).

Fig. 1.

Pathophysiological basis of neurofibromatosis type 1 (NF1)-associated plexiform neurofibromas (PN) and mechanism of action of MEK inhibitors. Schematic of the RAS–RAF–MEK–ERK signaling pathway under normal conditions, in NF1-associated PN, and following MEK inhibitor therapy. In healthy cells (panel A), neurofibromin (NF1) functions as a GTPase-activating protein that facilitates the conversion of active RAS-guanosine triphosphate (GTP) to its inactive RAS-guanosine diphosphate (GDP) form, thereby regulating downstream mitogen-activated protein kinase (MAPK) signaling and ensuring controlled cellular proliferation, growth, and survival. In NF1-associated PN (panel B), loss-of-function mutations in the NF1 gene impair neurofibromin activity, resulting in persistent activation of RAS (RAS-GTP). This leads to continuous downstream signaling through RAF, MEK, and ERK, which promotes unregulated cell proliferation, growth, and tumor formation. During MEK inhibitor therapy (panel C), pharmacologic inhibition of MEK diminishes downstream MAPK signaling despite ongoing upstream RAS activation. This suppression of ERK-mediated signaling reduces tumor cell proliferation and contributes to disease stabilization rather than complete tumor regression. GRB2 growth factor receptor-bound protein 2, MEK i MEK inhibitor, SOS son of sevenless homolog

Over the past decade, several clinical trials have demonstrated the promising activity of MEK inhibitors, particularly selumetinib, which showed significant tumor reduction and clinical improvement in pediatric patients with inoperable PN in a pivotal phase II trial by Dombi et al. Subsequent studies have corroborated these findings, reporting objective response rates (ORRs) exceeding 60% in some cohorts and durable tumor control in many patients [6, 8, 9, 11–20]. These outcomes ultimately culminated in the regulatory approval of selumetinib for pediatric patients with symptomatic inoperable PN. These results led to the regulatory approval of selumetinib for pediatric patients with symptomatic inoperable PN, with more recent expansion of its clinical use. More recently, additional MEK inhibitors such as mirdametinib, trametinib, cobimetinib, and tunlametinib have been evaluated through clinical trials and observational studies, thereby expanding the available therapeutic options for NF1-associated PN [11, 12, 14–16, 21].

Despite these advances, the current evidence remains varied. Published studies comprise a combination of prospective phase I–III clinical trials and emerging real-world observational cohorts, which differ in patient selection, imaging techniques, and how outcomes are reported [6, 8, 9, 11–28]. Furthermore, most early trials primarily enrolled pediatric populations, while recent studies are increasingly including adult patients with NF1, a group where treatment responses and toxicity profiles may differ [6, 8, 9, 11–28]. As a result, the comparative effectiveness of MEK inhibitors across adult and pediatric populations, as well as between clinical trial and real-world settings, is still not fully understood.

To address these gaps, we conducted a systematic review and meta-analysis of studies evaluating MEK inhibitors for NF1-associated PN. Our goal was to synthesize the available evidence regarding treatment effectiveness and safety to assess differences in outcomes across study settings and age groups. By combining evidence from various study designs and patient populations, this analysis aims to provide a comprehensive evaluation of how effective MEK inhibitors are in managing NF1-associated PN.

Methods

This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. The protocol was prospectively registered with PROSPERO (Registration Number: 1346241). Data supporting the findings of this study are available from the corresponding author upon reasonable request.

Search Strategy and Eligibility Criteria

A comprehensive systematic literature search was conducted to identify studies assessing MEK inhibitor therapy for NF1-associated PN. The following electronic databases were searched from inception through March 2026: PubMed (MEDLINE), Embase, Scopus, and Web of Science. The search strategy used both controlled vocabulary and free-text terms related to neurofibromatosis type 1, plexiform neurofibromas, and MEK inhibition. Search terms included combinations of the following keywords: “neurofibromatosis type 1,” “NF1,” “plexiform neurofibroma,” “plexiform neurofibromatosis,” “MEK inhibitor,” “selumetinib,” “mirdametinib,” “trametinib,” “cobimetinib,” and “tunlametinib.” Boolean operators were employed to combine search terms across databases, maximizing the sensitivity and specificity of the search. The complete search strategies for each database are provided in Table S1 of the Electronic Supplementary Material (ESM).

Eligible studies met the following criteria: (1) included patients diagnosed with NF1-associated PN; (2) evaluated treatment with a MEK inhibitor, including selumetinib, mirdametinib, trametinib, cobimetinib, or tunlametinib; (3) reported clinical outcomes related to tumor response, disease control, or disease progression; and (4) were prospective clinical trials (phase I–III), prospective observational studies, or retrospective cohort studies. Studies were excluded if they were case reports, small case series with fewer than ten patients, review articles, editorials, conference abstracts lacking full outcome data, preclinical studies, or animal studies. Publications lacking extractable response outcome data were also excluded.

When multiple publications reported data from overlapping patient populations, the most comprehensive or methodologically robust dataset was selected to prevent patient-level data duplication. Detailed inclusion and exclusion criteria are summarized in Table S2 of the ESM.

Study Selection and Data Collection

Study screening and selection were carried out following PRISMA 2020 guidelines. All records identified through database searches were imported into EndNote X9 for initial handling, where duplicate entries were both automatically and manually removed. The cleaned dataset was then transferred to Mendeley and exported in RIS format for systematic screening using the Rayyan platform (https://www.rayyan.ai/). Two independent reviewers (SS and ANS) reviewed study titles and abstracts to identify potentially eligible studies. Articles that met preliminary eligibility criteria underwent a full-text review to confirm their inclusion. Any discrepancies at any stage of the screening process were resolved through consensus discussions, with senior investigators acting as adjudicators when needed.

To ensure completeness of the literature search, the reference lists of all included studies and relevant review articles were manually screened for additional potentially eligible studies not captured by the electronic database search. The overall study selection process is summarized in the PRISMA flow diagram (Fig. 1 of the ESM).

Data Extraction

Data extraction was carried out independently by two investigators using a standardized data extraction form developed before the review process. The extracted information included study details, patient demographics, treatment specifics, imaging methods, and clinical outcomes. Study-level variables covered first author, publication year, country, study design, enrollment period, and total sample size. Patient characteristics included age and sex distribution, tumor location, and baseline disease features when available.

Treatment-related variables encompassed the type of MEK inhibitor used, dosing schedule, therapy duration, and follow-up time. Imaging-related details recorded included the imaging modality, intervals between scans, tumor measurement technique (volumetric vs non-volumetric), and response evaluation criteria. Outcome data gathered from each study included ORR, disease control rate (DCR), disease progression rate (DPR), and grade ≥ 3 adverse events. When needed, response outcomes were calculated from reported response categories using study-specific criteria. Any disagreements in data extraction were addressed through joint review and consensus among the investigators.

Outcome Definitions

Objective Response Rate (ORR)

The ORR was defined as the proportion of patients who achieved either a complete response or partial response during treatment, based on study-specific response criteria. In most of the included studies, partial response was defined as a reduction of at least 20% in tumor volume compared with baseline, usually measured with volumetric magnetic resonance imaging (MRI) of the target plexiform neurofibroma. Complete response was defined as the total disappearance of the target lesion when reported.

Disease Control Rate (DCR)

Disease control rate was defined as the proportion of patients achieving complete response, partial response, or stable disease during treatment. Stable disease was generally defined as a tumor volume change that did not meet the criteria for either response or progression, typically representing less than a 20% decrease in tumor volume compared with baseline measurements.

Disease Progression Rate (DPR)

Disease progression rate was defined as the proportion of patients demonstrating progressive disease during treatment, most commonly defined as a ≥ 20% increase in tumor volume relative to baseline measurements. In studies using alternative assessment methods, progression was defined according to the respective study-specific imaging or clinical response criteria.

Grade ≥ 3 Adverse Events

Safety outcomes included the incidence of grade ≥ 3 adverse events, as reported in the included studies, and reflected all events of grade ≥ 3 severity. Distinction between treatment-related and overall adverse events was not consistently reported across studies. Grade ≥ 3 adverse events were defined based on the Common Terminology Criteria for Adverse Events (CTCAE), representing severe or medically significant toxicities that may require hospitalization or clinical intervention and may limit self-care activities of daily living.

Risk of Bias and Certainty Assessment

The methodological quality of the included studies was assessed following the Cochrane Handbook for Systematic Reviews of Interventions [29]. Two independent reviewers (SS and ANS) applied the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) tool for the non-randomized studies [30]. Discrepancies were resolved through consensus. The confidence in the evidence supporting pooled estimates was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework, which considers risk of bias, consistency, directness, precision, and publication bias [29].

Statistical Analysis

Descriptive statistics summarized the characteristics of included studies and patient populations. Meta-analysis of pooled proportions was performed to estimate overall response and disease control outcomes across studies. Pooled estimates for ORR, DCR, and DPR were calculated using random-effects models to address expected clinical and methodological differences among studies. Proportions were stabilized with the Freeman–Tukey double arcsine transformation, a common method in meta-analyses of proportions.

Statistical heterogeneity was assessed using Cochran’s Q test, the I² statistic, and the estimated between-study variance (τ²). I² values exceeding 75% were considered indicative of significant heterogeneity. Predefined subgroup analyses were conducted to identify potential sources of heterogeneity, including comparisons between clinical trial populations and real-world observational cohorts, as well as adult versus pediatric patient groups. Differences between subgroups were assessed using a test for subgroup differences based on the χ² (Chi-square) statistic as implemented in the meta-analysis model. Sensitivity analyses involved sequentially removing individual studies to evaluate the robustness of pooled estimates. All statistical tests were two-sided with a significance level of P < 0.05. All analyses, including forest plot creation and heterogeneity evaluation, were carried out using R statistical software (Version 4.3.2).

Results

Study Characteristics

Following systematic screening of 1057 records, a total of 28 full-text articles were assessed for eligibility. After applying the predefined inclusion and exclusion criteria, 21 studies [6, 8, 9, 11–28] comprising 731 patients [347 (47.5%) female] were included in the final systematic review and meta-analysis (Fig. 1 of the ESM). Two studies reported separate outcomes for adult and pediatric cohorts, which were analyzed independently to preserve population-specific estimates, yielding 23 analytic cohorts included in the quantitative synthesis. Across the included literature, 13 studies (61.9%) were prospective clinical trials, including phase I, phase II, and phase IIb investigations [6, 8, 9, 11–20], while eight studies (38.1%) were observational studies [21–28], consisting primarily of retrospective or prospective real-world cohorts. The studies were conducted across multiple geographic regions, including the USA (8, 38.1%) [6, 8, 9, 11, 12, 16, 18, 19], Europe (7, 33.3%) [21–25, 27, 28], and Asia (4, 19%) [14, 15, 17, 20], one from Canada [26], and one was a multinational trial [13] reflecting the global clinical experience with MEK inhibitor therapy for NF1-associated PN.

The most frequently studied MEK inhibitor was selumetinib (15; 71.4%) [6, 8, 9, 13, 17–20, 22–28], evaluated in most included studies. Other agents assessed included mirdametinib, trametinib, cobimetinib, tunlametinib, and FCN-159, highlighting the growing range of MEK inhibitors used in NF1-related tumors [11, 12, 14–16, 21]. Treatment protocols varied across studies but primarily involved oral administration in 28-day cycles, with selumetinib typically given at 20–25 mg/m² twice-daily or similar dosing schedules for other MEK inhibitors. Radiologic response assessment was mainly conducted using MRI. Most studies used three-dimensional volumetric MRI analysis of target PN, enabling quantitative evaluation of tumor burden over time. Imaging intervals typically ranged from every 3–6 months, although several trials used cycle-based imaging schedules. Some studies included centralized or blinded independent radiologic review to ensure standardized response evaluation. A comprehensive summary of study characteristics was provided in Table 1.

Table 1.

Characteristics of included studies evaluating MEK inhibitors for NF1-associated PN

Author, year [ref.]	Country	Study design	Enrollment period	Total patients (n)	Female (n)	Age range (years)	Median age (years)	Age group	Tumor location	Follow-up period	MEK inhibitor	Dose regimen	Progressive disease at enrollment	Tumor-related morbidity
Weiss et al., 2021 [11]	USA	Multi-center phase II trial	2014-06 to 2018-08-01	19	8	> 16	24	Mixed	Head (face), neck, combined head, neck, and chest, combined neck and chest, trunk, extremity, combined trunk and extremity	Study evaluations at enrollment and after courses 4, 8, and 12 months, and then after courses 18 and 24 months for those who continued therapy	Miradametenib	2 mg/m2/dose BID (maximum dose of 4 mg BID; capsules swallowed whole) in a 3-week on/1-week off sequence	2	17
Passos et al., 2024 [22]	Portugal	Single-center Retrospective study	April 2018 to 2023	54	23	4.5–58.0	16.4	Mixed	NA	5 years	Selumetinib	5 mg/m2 of body surface area, rounded to the nearest achievable 5 mg or 10 mg dose (between 30 and 100 mg per day taken over 2 oral doses)	NA	NA
Dombi et al., 2016 [9]	USA	Multi-center phase I trial	9.21.2011–2.27.2014	24	11	3.0–18.5	10.9	Pediatric	Face, head and neck, neck and chest, trunk, trunk and extremity, whole body	Median 30 cycles (range 6–56 cycles), 1 cycle = 28 days	Selumetinib	20 (n = 12), 25 (n = 6), 30 (n = 6) BID	9	8
Trippett et al., 2022 [12]	USA	Multi-center phase I and II trials	May 2016 and November 2018	12	NA	NA	NA	Mixed	NA	NA	Cobimetinib	Suspension or 20-mg tablets, taken orally, days 1–21 of each 28-day cycle, with/without food. Patients received a cumulative weekly dose (rounded to the nearest 20 mg) based on body weight	NA	NA
Chen et al., 2025 [13]	Multinational	Multi-center phase III trial	11.19.2021 to ongoing	145 (71)	38	24–40	31	Adult	Neck-trunk, trunk extremity, head and neck, head, extremity, body, trunk, other	16 cycles (approximately 15 months)	Selumetinib	Oral selumetinib 25 mg/m2 (capped at 50 mg BID for a body surface area ≥1.9 m2)	NA	71
Wang et al., 2026 [14]	China	Single-center retrospective study	10.18.2021 - ongoing	15	5	18–45	27	Adult	Head, neck and trunk, trunk, thoracic, pelvis, extremity (upper or lower)	Median 22.8 months (range 5.1–33.5 months)	Tunlametinib	9 mg BID	NA	NA
Gross et al., 2023 [36]	USA	Multi-center phase II trial	9.21.2011 to 2.27.2021	50	20	3.5–17.4	10.2	Pediatric	Face, head and neck, neck and chest, trunk, trunk and extremity, whole body	Median 55.5 cycles (range 1–73 cycles), 1 cycle = 28 days	Selumetinib	20 (n = 12), 25 (n = 6), 30 (n = 6) BID	21	50
Hu et al., 2023 [15]	China	Multi-center phase I trial	3.26.2021 to 3.30.20205	19	8	20–57	26	Adult	Face, orbit, anterior neck/upper airway, posterior neck (cervical paraspinal), thoracic/paraspinal/chest wall, posterior abdomen/pelvis (lumbosacral plexus), forearm, thigh/upper leg, foot, others	2 years	FCN-159	3 in 4 mg, 4 in 6 mg, 8 in 8 mg, and 4 in 12 mg	NA	NA
Santoro et al., 2025 [23]	Italy	Multi-center retrospective study	NA	70	31	2.6–12.6	7.9	Pediatric	Head, neck, lower abdomen, mediastinum/thorax, lower legs, upper arm, upper abdomen, thigh, hand, foot, forearm, ankle	Median 49 cycles (7.5–79), 1 cycle = 28 days	Selumetinib	25 mg/m2 BID for ≥ 6 cycles	NA	NA
Moertel et al., 2024 (adult) [16]	USA	Multi-center phase IIb trial	9.29.2019 to 09.20.2023	58	37	18–69	34	Adult	Head and neck, lower extremities, paraspinal, chest wall, mesentery and pelvis, upper extremities, others	24-cycle treatment	Mirdametinib	Administered as a capsule or tablet for oral suspension, at a dose of 2 mg/m2 (maximum dose, 4 mg) orally BID in 28-day cycles	31	58
Moertel et al., 2024 (pediatric) [16]	USA	Multi-center phase IIb trial	9.29.2019 to 09.20.2023	56	30	(2–17)	10	Adult	Head and neck, lower extremities, paraspinal, chest wall, mesentery and pelvis, upper extremities, abdominal wall, others other	24-cycle treatment	Mirdametinib	Administered as a capsule or tablet for oral suspension, at a dose of 2 mg/m2 (maximum dose, 4 mg) orally BID in 28-day cycles	35	56
Cacchione et al., 2022 [24]	Italy	Single-center retrospective study	1.1.2019 to 12.31.2021	13	6	0.25–17	3	Pediatric	Face and neck, chest and mediastinum, abdomen, limbs, spine (intrarachid and paravertebral)	12-month period	Selumetinib	25 mg/m2 BID	6	13
Kim et al., 2024 (adult) [17]	South Korea	Multi-center phase I trial	May 2019 to December 2021	30	10	(19.0–47.0)	26.5	Adult	Abdomen, ankle, pelvis, pelvis, thigh, shoulder, back, pelvis, buttock, chest, chest wall, chest, leg, eye, face, face and chest, face and paraspinal, face and pelvis, foot, leg, liver, neck, neck and pelvis, neck and abdomen, neck and back, neck and chest, neck and face, neck and pelvis, orbit, paraspinal, paraspinal and chest, pelvis and chest, thoracic and paraspinal, thoracic paraspinal and leg	2 years; median cycles at analysis 21 (4–26)	Selumetinib	20 or 25 mg/m2 every 12 h	NA	NA
Kim et al., 2024 (pediatric) [17]	South Korea	Multi-center phase I trial	May 2019 to December 2021	60	25	(3–18)	8	Pediatric	Abdomen, ankle, pelvis, pelvis, thigh, shoulder, back, pelvis, buttock, chest, chest wall, chest, leg, eye, face, face and chest, face and paraspinal, face and pelvis, foot, leg, liver, neck, neck and pelvis, neck and abdomen , neck and back, neck and chest, neck and face, neck and pelvis, orbit, paraspinal, paraspinal and chest, pelvis and chest, thoracic and paraspinal, thoracic paraspinal and leg	2 years; median cycles at analysis 26 (16–26)	Selumetinib	25 mg/m2 or 50 mg every 12 hours	NA	NA
Santo et al., 2020 [25]	Portugal	Single-center retrospective study	4.2018 to 4.2019	19	4	(3–19)	13	Pediatric	Head and neck, chest, pelvis, upper and lower limbs	223 days (range 35–420 days)	Selumetinib	25 mg/m2 BID	8	19
Coltin et al., 2022 [26]	Canada	Multi-center retrospective study	2016 to 2021	19	10	(2.5–18)	11	Pediatric	Head, neck, trunk, limbs, buttock, face, upper trunk	19.6 months (range 3–51 months)	Selumetinib		10	19
Gross et al., 2025 [18]	USA	Multi-center phase II trial	1.7.2016 to ongoing	33	11	(18.3–60.2)	35.6	Adult	Trunk and limbs, neck, trunk and limbs, limbs only, neck and limbs, trunk only, neck only, head and neck, neck and trunk	28 cycles completed (range 1–78, 28 days per cycle)	Selumetinib	50 mg/m2/dose, BID	4	33
Gross et al., 2022 [8]	USA	Multi-center phase II trial	11.12.2015 to 8.6.2018	25	9	4.5–18.1	12.3	Pediatric	Head only, head/neck, neck/trunk, trunk only, trunk/extremity, extremity only	41 cycles (IQR 34–49)	Selumetinib	25 mg/m2/dose, BID	11	25
Suenobu [20] et al., 2023	Japan	Multi-center phase I trial	8.31.2020 to 12.8.2021	12	9	7.5–18.2	13.3	Pediatric	Neck or trunk, trunk, head, extremity, head and neck	12-month period	Selumetinib	25 mg/m2 BID	NA	12
Boitez et al., 2026 [27]	France	Single-center prospective study	NA	10	2	(4–16)	12.5	Pediatric	Face, neck, torso, limbs	19.5 months (± 6.6 months), 20 cycles	Selumetinib	25 mg/m2/day administered in one or two doses	NA	10
Jackson et al., 2020 [19]	USA	Single-center phase II trial	8.1.2015 to 22.10.2018	24	6	6.2–60.3	16.9	Mixed	Spinal region (cervical, thoracic, lumbosacral or any combination of these)	36 cycles (8–54)	Selumetinib	25 mg/m2 BID (pediatric), 50 mg/m2 BID (adult)	NA	NA
Styczyńska et al., 2025 [28]	Poland	Multi-center prospective study	1.1.2024 to 2.28.20205	79	28	(3–18)	12	Pediatric	Head and neck, chest and abdomen and limb	4 months	Selumetinib	25 mg/m2 BID	NA	NA
Noordhoek et al., 2026 [21]	Netherlands	Single-center phase II trial	7.17.2020 to 5.10.2023	30	16	(19–67)	45	Adult	Head/neck, thorax/abdomen, upper extremities, lower extremities	24 months (range 2–46 months)	Trametinib	Trametinib 2 mg once per day orally	NA	30

Data represent study characteristics of clinical trials and observational cohorts included in this systematic review and meta-analysis evaluating MEK inhibitors for NF1-associated PN. Progressive disease at baseline refers to radiographic tumor progression prior to treatment initiation when reported. Tumor-related morbidity includes clinical symptoms attributable to PN such as pain, neurologic deficits, airway compromise, or functional impairment

(A) adult cohort, BID twice daily, (C) pediatric cohort, IQR interquartile range, NA not available, NF1 neurofibromatosis type 1, PN plexiform neurofibroma

Across studies, tumor response criteria were mostly consistent, with most defining partial response as a ≥ 20% reduction in tumor volume from baseline, progressive disease as a ≥ 20% increase in tumor volume, and stable disease as tumor volume change within ± 20% of baseline. These definitions broadly align with the Response Evaluation in Neurofibromatosis and Schwannomatosis (REiNS) criteria, which are commonly used in clinical trials assessing PN. However, a small number of real-world studies used alternative evaluation methods, including two-dimensional diameter measurements or qualitative radiologic assessments (Table 3 of the ESM). Reporting of treatment-related adverse events varied across studies. In total, 18 (85.7%) studies reported grade ≥ 3 adverse events, of which 11 were clinical trials, and 7 were observational cohorts. The number of grade ≥ 3 adverse events reported per study ranged from 0 to 23 events (Table 4 of the ESM).

ORR

The pooled analysis of ORR across the included cohorts demonstrated an overall response rate of 56% (95% CI 45–66; I² = 79.2%) [Fig. 2a], indicating that a substantial proportion of patients with NF1-associated PN experienced measurable tumor reduction following MEK inhibitor therapy.

Fig. 2.

Objective response rate (ORR) following MEK inhibitor therapy in neurofibromatosis type 1-associated plexiform neurofibromas, with subgroup analyses. Forest plots demonstrating the pooled ORR following MEK inhibitor therapy in patients with neurofibromatosis type 1-associated plexiform neurofibromas. Pooled estimates were calculated using a random-effects model with a logit transformation. Squares represent individual study estimates, with sizes proportional to study weight, and horizontal lines indicate 95% confidence intervals (CIs); diamonds represent pooled estimates. (a) Overall pooled ORR across all included studies, (b) ORR in clinical trials, (c) ORR in real-world observational cohorts, (d) ORR in pediatric populations, and (e) ORR in adult populations

Subgroup analyses demonstrated statistically significant difference response rates between clinical trials (ORR = 61%; [95% CI 49–73]; I² = 82.8%) and real-world observational cohorts (ORR = 44%; [95% CI 34–54]; I² = 56.2%) [test of subgroup difference p = 0.035] (Fig. 2b, c, and Fig. 3a of the ESM). When stratified by age group, pediatric patients exhibited a higher pooled response rate (ORR = 58%; [95% CI 45–70]; I² = 78.1%) compared with adult cohorts (ORR = 51%; [95% CI 34–67]; I² = 83.7%) with no significant between both groups (test of subgroup difference p = 0.407) (Fig. 2d, e and Fig. 3b of the ESM).

DCR

The pooled DCR across all included studies was 96% (95% CI 91–98; I² = 17%) [Fig. 3a], indicating that the majority of patients achieved either tumor response or stable disease stabilization during MEK inhibitor therapy. Subgroup analyses demonstrated similar DCR estimates between clinical trials (DCR = 96%; [95% CI 88–99]; I² = 10.4%) and real-world cohorts DCR = 96%; [95% CI 89–98]; I² = 0%) [test of subgroup difference p = 0.82) [Fig. 3b, c, and Fig. 4a of the ESM]. When stratified by age group, pediatric populations showed a slightly higher pooled DCR 96% (95% CI 88–99]; I² = 0%) compared with adult cohorts (DCR = 95%; [95% CI 78–99]; I² = 6.1%), [test of subgroup difference p = 0.77], although both groups demonstrated consistently high rates of disease control (Fig. 3d, e, and Fig. 4b of the ESM).

Fig. 3.

Disease control rate (DCR) following MEK inhibitor therapy in neurofibromatosis type 1- associated plexiform neurofibromas, with subgroup analyses. Forest plots demonstrating the pooled DCR following MEK inhibitor therapy in patients with neurofibromatosis type 1- associated plexiform neurofibromas. Pooled estimates were calculated using a random-effects model with a logit transformation. Squares represent individual study estimates, with sizes proportional to study weight, and horizontal lines indicate 95% confidence intervals (CIs); diamonds represent pooled estimates. (a) Overall pooled DCR across all included studies, (b) DCR in clinical trials, (c) DCR in real-world observational cohorts, (d) DCR in pediatric populations, and (e) DCR in adult populations

DPR

The pooled DPR across studies was 2% (95% CI 1–5; I² = 0%) [Fig. 4a], indicating that tumor progression during MEK inhibitor therapy was relatively uncommon.

Fig. 4.

Disease progression rate following MEK inhibitor therapy in neurofibromatosis type 1-associated plexiform neurofibromas, with subgroup analyses. Forest plots demonstrating the pooled disease control rate (DCR) following MEK inhibitor therapy in patients with neurofibromatosis type 1-associated plexiform neurofibromas. Pooled estimates were calculated using a random-effects model with a logit transformation. Squares represent individual study estimates, with sizes proportional to study weight, and horizontal lines indicate 95% confidence intervals (CIs); diamonds represent pooled estimates. (a) Overall pooled DCR across all included studies, (b) DCR in clinical trials, (c) DCR in real-world observational cohorts, (d) DCR in pediatric populations, and (e) DCR in adult populations

Subgroup analyses demonstrated slightly higher progression rates in real-world cohorts (DPR = 4% [95% CI 1–11]; I² = 0%) compared with clinical trials (DPR = 4% [95% CI 0–5]; I² = 2.9%) with no significant difference between both groups (test of subgroup difference p = 0.22) [Fig. 4b, c, and Fig. 5a of the ESM). When stratified by age group, pediatric cohorts exhibited a marginally higher DPR (DPR = 3% [95% CI 1–10]; I² = 0%) than adult populations (DPR = 1% [95% CI 0–4]; I² = 0%) [test of subgroup difference p = 0.509], although progression remained infrequent in both groups (Fig. 4d, 4e, and Fig. 5b of the ESM).

Grade ≥3 Adverse Events

The pooled rate of grade ≥ 3 adverse events across included studies was 13% (95% CI 6–25; I² = 51.9%) [Fig. 5a], indicating that severe treatment-related toxicity occurred in a minority of patients receiving MEK inhibitor therapy. Subgroup analyses demonstrated comparable rates between study designs, with a pooled incidence of 14% (95% CI 6–31; I² = 48.8%) in clinical trials and 10% (95% CI 1–48; I² = 62.7%) in real-world cohorts (test of subgroup difference p = 0.705) [Fig. 5b, c, and Fig. 6a of the ESM). When stratified by age group, adult cohorts exhibited a higher pooled rate of grade ≥ 3 adverse events (23%, [95% CI 4–66]; I² = 81%) compared with pediatric populations (10%, [95% CI 3–26]; I² = 0%) [test of subgroup difference p = 0.275] (Fig. 5d, e, and Fig. 6b of the ESM). Despite this difference, the overall incidence of severe adverse events remained relatively low across both groups.

Fig. 5.

Grade > 3 adverse events (AEs) following MEK inhibitor therapy in neurofibromatosis type 1-associated plexiform neurofibromas, with subgroup analyses. Forest plots demonstrating the pooled grade > 3 AEs following MEK inhibitor therapy in patients with neurofibromatosis type 1-associated plexiform neurofibromas. Pooled estimates were calculated using a random-effects model with a logit transformation. Squares represent individual study estimates, with sizes proportional to study weight, and horizontal lines indicate 95% confidence intervals (CIs); diamonds represent pooled estimates. (a) Overall pooled grade > 3 AEs across all included studies, (b) grade > 3 AEs in clinical trials, (c) grade > 3 AEs in real-world observational cohorts, (d) grade > 3 AEs in pediatric populations, and (e) grade > 3 AEs in adult populations

Risk of Bias and Certainty Assessment

Risk of bias was assessed using the ROBINS-I tool, with most included studies judged to be at moderate risk of bias and several retrospective or real-world cohorts at serious risk, primarily due to confounding and participant selection. Bias related to outcome measurement was generally low, reflecting the use of standardized volumetric MRI-based assessments, although deviations from intended interventions and incomplete data contributed to overall limitations (Fig. 2 of the ESM).

Using the GRADE framework, the certainty of evidence was rated as low for ORR and grade ≥3 adverse events, driven by risk of bias and heterogeneity, and moderate for DCR and DPR, reflecting consistent findings with low statistical heterogeneity and stable effect estimates (Table 5 of the ESM). Subgroup analyses were considered exploratory, and certainty of evidence was not formally assessed for these comparisons.

Discussion

In this systematic review and meta-analysis of MEK inhibitor therapy for NF1-associated PN, we show that treatment has high rates of disease control (96%) and very low rates of progression (2%), indicating a strong and consistent effect on tumor stabilization across various clinical settings. Importantly, these findings should be interpreted in the context of prospective natural history studies conducted by the National Cancer Institute, which have demonstrated that untreated PN frequently exhibit continued growth and rarely undergo spontaneous regression. Although objective tumor responses were seen in about half of the patients (56%), these findings suggest that the main benefit of MEK inhibition is in maintaining disease stability rather than in radiographic tumor shrinkage alone. Importantly, these effects were consistent across both clinical trials and real-world cohorts, as well as across different age groups, supporting the broad applicability of MEK inhibitor therapy. The occurrence of grade ≥3 adverse events remained relatively low (13%), though variability across studies emphasizes the need for ongoing evaluation of long-term safety. Overall, these findings position MEK inhibitors as an effective approach for managing tumor growth in patients with PN, addressing a crucial unmet need in this population.

The therapeutic effectiveness of MEK inhibitors in PN closely relates to the underlying dysregulation of the RAS–RAF–MEK–ERK signaling pathway in NF1 [2, 10, 31]. While this mechanistic rationale is well established, our findings provide clinical evidence that pharmacologic inhibition of this pathway primarily results in suppression of tumor growth during treatment rather than consistent tumor shrinkage [6, 7, 9]. This is biologically plausible, as MEK inhibition primarily reduces proliferative signals and stabilizes tumor size rather than causing widespread tumor cell death [10, 31, 32]. These results strengthen the idea that targeted MAPK pathway inhibition in NF1 is a disease-modifying approach, aiming for long-term control of tumor progression rather than just tumor shrinkage.

Notably, our findings suggest that the primary clinical benefit of MEK inhibition lies in disease stabilization rather than tumor shrinkage alone. Although ORRs were moderate, the consistently high DCRs and extremely low progression rates across studies indicate that MEK inhibitors effectively halt tumor growth over time [6, 8, 9, 11–28]. This distinction is clinically meaningful, as PN are often characterized by progressive enlargement and cumulative morbidity rather than rapid malignant transformation [4, 5, 7]. While biological differences between pediatric and adult tumors have been described, our analysis did not demonstrate statistically significant differences in response between these groups [5–7, 9]. These findings suggest that age-related differences in tumor biology may influence therapeutic responsiveness to MEK inhibition.

Importantly, the consistency of outcomes seen between clinical trial populations and real-world groups supports the external validity and generalizability of MEK inhibitor therapy. While clinical trials usually involve highly selective patient groups under controlled conditions [6, 8, 9, 11–20], real-world studies include a broader diversity of patients, such as variations in age, tumor size, and other health conditions [21–28]. The similar DCRs across both settings suggest that the effects of MEK inhibitors are strong and reliable in everyday clinical practice. However, the slightly lower response rates and greater variability seen in real-world groups may be due to differences in adherence, dose adjustments, and imaging assessment methods [21–28]. In particular, the use of volumetric MRI-based assessment in clinical trials may allow for more sensitive and reproducible quantification of tumor burden compared with conventional measurements, potentially contributing to differences in reported response rates across studies [21–28].

The safety profile of MEK inhibitors observed in this analysis was generally consistent with prior reports, with grade ≥ 3 adverse events occurring in a minority of patients [6, 9]. Although the incidence of severe toxicity varied across studies, these events were typically manageable with dose modification or temporary treatment interruption [6, 8, 9, 11–28]. Importantly, the favorable safety profile, together with sustained disease control, supports the role of MEK inhibitors as a disease-modifying therapeutic approach, rather than a short-course cytotoxic intervention [6, 9]. This paradigm shift is particularly relevant in NF1, where treatment goals often prioritize symptom control, preservation of function, and prevention of tumor progression over complete radiographic remission.

Beyond radiographic outcomes, emerging evidence suggests that MEK inhibitor therapy may also confer meaningful improvements in patient-reported outcomes, including pain, functional impairment, and health-related quality of life [6, 8, 9, 11–20]. Plexiform neurofibromas are frequently associated with a substantial symptom burden that is not fully captured by volumetric tumor measurements alone [6, 33, 34]. Prior studies have demonstrated improvements in patient-reported domains even in the absence of substantial tumor shrinkage, highlighting a potential dissociation between radiographic response and clinical benefit [34, 35]. These findings further support the concept that MEK inhibitors function as disease-modifying therapies, whereby stabilization of tumor growth may translate into clinically meaningful improvements in symptoms and functional outcomes [6, 9]. Incorporation of standardized patient-reported outcome measures in future studies will be essential to more comprehensively define treatment benefit in NF1-associated PN.

Limitations

Several limitations should be considered when interpreting these findings. First, the majority of included studies were non-randomized or single-arm trials, contributing to an overall moderate-to-serious risk of bias and limiting causal inference. Second, substantial heterogeneity was observed in some analyses, particularly for ORRs and adverse events, likely reflecting differences in study design, patient populations, and outcome assessment methods. Third, variability in response criteria and imaging methodologies across studies may have influenced outcome estimates. In particular, while many clinical trials employed a volumetric MRI-based assessment, some studies relied on two-dimensional or linear measurements, which are less sensitive for detecting changes in plexiform neurofibroma burden. This heterogeneity in measurement approaches limits comparability of response rates across studies. Furthermore, multiple MEK inhibitors were included across studies, and although they share a common mechanism of action, differences in drug-specific efficacy and toxicity profiles may have contributed to heterogeneity in the pooled estimates. The limited number of studies available for several individual agents precluded meaningful drug-specific subgroup analyses, and therefore, pooled estimates should be interpreted as reflecting the overall therapeutic class effect of MEK inhibition rather than the efficacy or safety profile of any single agent.

Fourth, important clinical variables, including tumor size, tumor subtype (e.g., classic plexiform neurofibroma vs diffuse nodular lesion), durability of response, and treatment discontinuation rates, were inconsistently reported and could not be systematically analyzed. Most studies reported outcomes during treatment over relatively limited follow-up durations, precluding a robust assessment of long-term durability. In addition, discontinuation of therapy because of adverse events, particularly low-grade but intolerable toxicities, was not consistently captured, despite its potential impact on treatment adherence and real-world effectiveness. Adverse event reporting was heterogeneous, with inconsistent distinction between treatment-related and overall events, and limited reporting of lower-grade toxicities that may significantly affect patient quality of life. Additionally, this analysis focused on grade ≥ 3 adverse events because they were the most consistently reported safety outcome; lower-grade toxicities, particularly dermatologic and gastrointestinal adverse effects that may affect treatment adherence and quality of life, were reported inconsistently and could not be pooled quantitatively.

The interpretation of disease control and progression rates is further limited by the absence of standardized baseline growth data and the heterogeneous natural history of PN, which may remain stable without treatment, particularly in older patients. As a result, stable disease may not necessarily reflect a treatment effect. In addition, differences in reporting of progression events and a lack of uniform follow-up intervals across studies further limit the reliability of these measures. Finally, the lack of patient-level data further precluded adjustment for important clinical covariates, including tumor burden, prior treatments, and genotype-phenotype correlations.

Conclusions

MEK inhibitors are associated with high rates of disease control and minimal progression in patients with NF1-associated PN, supporting their role as an effective therapeutic strategy in this population. While objective tumor shrinkage is observed in a subset of patients, the predominant clinical benefit appears to be durable stabilization of tumor growth, consistent with a disease-modifying effect rather than cytoreductive activity alone. This distinction is particularly relevant in a condition in which many PN demonstrate progressive growth and contribute to cumulative morbidity. Despite these promising findings, the certainty of evidence remains limited by the predominance of non-randomized and single-arm studies. Future prospective and comparative investigations incorporating standardized imaging and patient-reported outcomes are needed to more fully define the long-term efficacy and clinical benefit of MEK inhibition in NF1.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

No external funding was used in the preparation of this article.

Declarations

Conflict of interest

Sai Sanikommu, Alejandro N. Santos, Bashar M. Dawoud, Adam S. Levy, Ricardo J. Komotar, and Victor M. Lu have no conflicts of interest that are directly relevant to the content of this article.

Ethics approval

This study does not require ethics approval.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and material

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Coda availability

Not applicable.

Author contributions

Concept and design: SS, VML, RJK. Literature review: SS, BD, ANS, ASL. Data collection: SS, BD, ANS, ASL. Analysis: SS. Writing and editing: SS, VML, RJK, ANS. Reviewing: all authors.