Abstract
Purpose
Malignant peripheral nerve sheath tumor (MPNST) is a rare, aggressive soft tissue sarcoma with highest incidence in patients with neurofibromatosis type 1 (NF1). Data on outcomes and second malignancies following radiation therapy (RT) are limited, particularly in patients with NF1. This study evaluated overall survival (OS), progression-free survival (PFS), local failure, and secondary malignancies in patients with localized resectable MPNST.
Methods and Materials
Patients diagnosed with localized resectable MPNST between 2010 and 2024 were retrospectively identified. Survival was estimated using the Kaplan–Meier method, multivariable analysis was conducted using Cox proportional hazards modeling, and cumulative incidence was estimated using Fine and Gray model.
Results
Sixty-eight patients met inclusion criteria; median age was 34 years, and 75% had NF1. Median follow-up was 43.5 months. RT was delivered in 61.8% of patients, primarily preoperatively (71.1%). RT was associated with improved 3-year OS (87.2% vs 51.5%, P = .039) and 3-year PFS (55.8% vs 41.5%, P = .017). On multivariable analysis, RT remained independently associated with improved PFS (HR, 0.34, P = .003) and OS (HR, 0.4, P = .034). Three-year local failure was for RT versus no RT was 26.3% versus 35.3% (P = .240). One secondary malignancy arose within a prior RT field 9.5 years after initial RT, and another at the RT field edge 8.6 years later.
Conclusions
RT was independently associated with improved PFS and OS in patients with localized resected MPNST. These hypothesis-generating findings need validation in a larger cohort. The risk of RT-associated secondary malignancy was low, occurring beyond the median disease progression timeframe.
Introduction
Malignant peripheral nerve sheath tumors (MPNSTs) are aggressive soft tissue sarcomas that arise from peripheral nerve tissue.1 MPNSTs can occur sporadically with an annual incidence of 0.001%,2 but approximately half of all diagnoses occur in people with neurofibromatosis type 1 (NF1).1,2 Patients with NF1 have an 8% to 13% lifetime risk of developing MPNSTs, which most commonly arises from pre-existing plexiform neurofibromas.3
Negative-margin surgical resection is associated with improved overall survival (OS) and is recommended as definitive treatment of localized disease.2,4, 5, 6, 7, 8, 9, 10, 11, 12, 13 However, despite negative-margin resection, the rate of local recurrence and distant metastasis is high5,9,11,14, 15, 16, 17, 18, 19, 20 with a 5-year OS ranging from 34% to 66% for all patients.2, 3, 4, 5, 6,9, 10, 11, 12, 13,16, 17, 18, 19, 20, 21, 22 These poor outcomes emphasize the need to investigate additional treatment modalities.
Although radiation therapy (RT) reduces local recurrence in other soft tissue sarcomas,23,24 its role in the treatment of MPNST is not well established. Studies investigating RT in MPNST are limited, with most studies largely composed of patients treated with outdated techniques prior to 2000. Furthermore, results regarding the efficacy of RT for both local control (LC) and survival are mixed, and most studies have not shown significant association between RT and improved OS and/or disease specific survival.2,4,10,13,16,17,20, 21, 22,25,26 Several studies have also failed to show an association between RT and LC17,22 and/or progression-free survival (PFS).4,10,13,21 Although 2 recent studies have found association between RT and LC (but not OS),26,27 only 1 series has found an association between RT and both LC and OS among 251 patients with localized disease diagnosed between 1960 and 2016.19
Additionally, there are concerns about developing radiation-associated secondary malignancies, especially in patients with NF1 where plexiform neurofibromas may be at risk for malignant transformation.2,3,28 MPNSTs account for 5% of RT-associated soft tissue sarcomas with a median latency time of approximately 12 to 17 years after RT.2,15,19,28, 29, 30 The estimated risk of developing a MPNST after RT for all patients is 0.06%,30 and for children is 0.1% at 30 years.31 These data, coupled with the lack of robust data to guide the use of RT in MPNSTs, can lead to the avoidance of RT in patients with NF1.
The purpose of this study is to characterize the OS, PFS, and local failure (LF) in patients with localized resected MPNST treated with and without RT, as well as the incidence of secondary malignancies after RT and their relation to the RT field.
Methods and Materials
Patients
This retrospective study was approved by the institutional review board of the authors’ institution. Patients diagnosed with localized MPNST between 2010 and 2024 were retrospectively identified from our institutional sarcoma database. Patients were excluded if they had radiographically detectable metastatic disease at diagnosis, fewer than 3 months of follow-up, incomplete RT data, and/or treated with noncurative intent therapy. Medical records were reviewed to obtain data on patient demographics, tumor characteristics, treatment details, secondary malignancies, and outcomes.
Clinicopathologic and treatment variables
Patient demographic and tumor characteristic data included sex, age at diagnosis, germline NF1 mutation status (or a clinical diagnosis of NF1 without genetic confirmation), tumor anatomic location, tumor size as determined radiographically using restricted diffusion magnetic resonance imaging sequences when available,32 tumor grade (high/intermediate grade included Fédération Nationale des Centres de Lutte Contre le Cancer grades 2 and 3, low grade included Fédération Nationale des Centres de Lutte Contre le Cancer grade 1), histologic subtype, surgical margin status, exposure to chemotherapy, neoadjuvant or adjuvant RT, and RT dose and fractionation. When tumor grade was not provided on biopsy specimen, grade was labeled as indeterminate. Tumor location was categorized as involving the brachial plexus, extremity (upper or lower), head and neck, paraspinal region, pelvis (pelvis or sacrum), or trunk (chest, abdomen, retroperitoneum). Surgical margins were defined as negative if noted as negative (R0) on the initial surgical pathology report or if an initially positive margin was converted to a negative margin after reresection. Microscopic (R1) and macroscopic (R2) margins that were not cleared on a reresection were defined as positive margins. RT dose was converted to an equieffective dose using equivalent dose in 2-Gy fractions using α/β = 4 as the mean α/β ratio used for soft tissue sarcomas.33
Selection of therapy
All patients were discussed at our institutional multidisciplinary tumor board to determine therapy selection and sequence. For patients with localized MPNST, en bloc maximal safe resection to negative margins was generally recommended. Neoadjuvant chemotherapy was offered at the physician’s discretion to determine radiographic response to therapy and continued in the postoperative setting.34
Radiation therapy
All patients underwent CT simulation with or without magnetic resonance simulation. If magnetic resonance simulation was not performed, a diagnostic magnetic resonance imaging was fused to the CT simulation. The clinical target volume (CTV) was defined as the gross target volume (GTV) expanded by 1.5 cm radially and 3 to 4 cm superiorly and inferiorly, respecting anatomic boundaries. In children <18 years of age, at the physician’s discretion, the superior-inferior CTV expansion was reduced to no less than 1.5 cm, following guidelines from ARST0332.35 Planning target volume margin was 0.3 to 0.5 cm and trimmed 3 to 5 mm from skin for deep-seated tumors to reduce skin dose. For postoperative cases, the GTV was defined as the resection bed and any residual tumor. The CTV was defined as the GTV expanded by 1 to 1.5 cm radially and 1.5 to 2 cm superiorly and inferiorly, respecting anatomic boundaries. If a boost was used, the CTV boost was defined as a 1 to 2 cm expansion of the GTV and region of positive margin. Both 3-dimensional conformal RT and intensity-modulated RT techniques were used.
Definition of endpoints
Primary endpoints were OS and PFS. OS was defined as time from MPNST diagnosis to death or last patient contact. PFS was defined as time from MPNST diagnosis to disease progression (clinical suspicion of progression confirmed radiographically with or without biopsy confirmation), death, or last patient contact. Secondary endpoints included LF, distant failure (DF), and secondary malignancy. LF was defined as a recurrence within the primary site with or without a DF. DF was defined as a failure outside of the primary site without a LF. Secondary malignancy was defined as any pathologically confirmed malignancy diagnosed after the initial MPNST diagnosis. A subsequent MPNST diagnosis was considered a secondary malignancy if arising within a separate plexiform neurofibroma or if occurring after 5 years without evidence of disease.36 Secondary malignancies were assessed for relationship to a prior RT field.
Statistical analysis
All statistical analyses were completed using Stata18.0. Fisher’s exact test and Wilcoxon rank sum were used to compare categorical and continuous variables, respectively. The Kaplan–Meier method was used to estimate PFS, OS, LC, and distant control (DC). Effect of confounding variables was assessed using Cox proportional hazards. Variables with a P value <.10 on univariable analysis were included in the multivariable analysis in a stepwise fashion. For variables with P values ≥.10 but ≤0.105, a nested model comparison was performed: the reduced model included only variables with P values <.10, and the extended model additionally incorporated variables with P values ≤.105. Model fit was evaluated using the log-likelihood ratio test, with the model demonstrating superior fit selected as the final multivariable model. Patients with indeterminate tumor grade (n = 11) were excluded from univariable and multivariable analyses incorporating grade as a covariate. All other analyses included the full cohort of 68 patients unless otherwise specified. Multivariable analyses of both PFS and OS with a priori inclusion of size and margin status, along with RT receipt, were planned for evaluation in case a P value-driven analysis did not identify size and margin status as included variables, based on prior literature establishing the prognostic value of these variables.2,4, 5, 6,9, 10, 11, 12, 13,15, 16, 17, 18, 19,21,22 Cumulative incidence of LF was estimated using the Fine-Gray model with death and DF as competing risks. Cumulative incidence of DF was also estimated using the Fine-Gray model with death and LF as competing risks. A preplanned subgroup analysis for patients with NF1 was completed to determine the effect of RT on OS and PFS in this population.
Results
Sixty-eight patients met eligibility criteria and were included. Table 1 shows detailed patient, tumor, and treatment characteristics. The median age at diagnosis was 34 years. Fifty-one patients (75.0%) had an identified germline NF1 mutation or were clinically diagnosed with NF1. The most common tumor location was the extremity (29.4%). Median tumor size was 7.5 cm. Most patients (72.0%) had high/intermediate grade tumors. Eleven patients had indeterminate tumor grade based on diagnostic biopsy specimen, of whom 6 received preoperative RT, 4 received postoperative RT, and 1 underwent surgery alone. All patients underwent surgical resection, of whom 50 (73.5%) had a negative-margin resection. Thirty-eight patients (55.9%) underwent chemotherapy, of whom 27 (71.1%) received neoadjuvant chemotherapy. Ifosfamide and doxorubicin are first-line systemic therapy agents at our institution and were used in 28 of 38 patients (73.7%) receiving systemic therapy. An additional 7 patients (18.4%) received a regimen that included mesna, doxorubicin, and ifosfamide, and 3 (7.9%) received other systemic therapy. Forty-two patients (61.8%) underwent RT, of whom 27 (64.3%) underwent preoperative RT. Photon intensity-modulated RT was used in 36 patients (85.7%), intensity-modulated proton therapy was used in 5 patients (11.9%), and 1 patient (2.4%) received 3-dimensional conformal RT. Median equivalent dose in 2-Gy fractions was 50 Gy. Compared to those not treated with RT, patients treated with RT had fewer low-grade tumors (4.8% vs 23.1%, P = .047). There were no significant differences in any other patient or tumor characteristics between those who received RT and those who did not (all P > .05).
Table 1.
Patient and tumor characteristics
| Characteristic | All patients (n = 68) | RT (n = 42) | No RT (n = 26) |
|---|---|---|---|
| Sex | |||
| Female | 29 (42.6%) | 18 (42.9%) | 11 (42.3%) |
| Male | 39 (57.4%) | 24 (57.1%) | 15 (57.7%) |
| Age at diagnosis, median (range) | 34 (5-75) | 35 (5-75) | 29 (12-74) |
| NF-1 mutated | |||
| No | 17 (25.0%) | 12 (28.6%) | 5 (19.2%) |
| Yes | 51 (75.0%) | 30 (71.4%) | 21 (80.8%) |
| Tumor location | |||
| Brachial plexus | 4 (5.9%) | 2 (4.8%) | 2 (7.7%) |
| Head and neck | 6 (8.8%) | 3 (7.1%) | 3 (11.5%) |
| Paraspinal | 11 (16.2%) | 7 (16.7%) | 4 (15.4%) |
| Pelvis | 12 (17.6%) | 8 (19.0%) | 4 (15.4%) |
| Trunk | 15 (22.1%) | 6 (14.3%) | 9 (34.6%) |
| Extremity | 20 (29.4%) | 16 (38.1%) | 4 (15.4%) |
| Tumor size (cm), median (range) | 7.5 (1.4-16) | 8.2 (3.5-16) | 6.9 (1.4-14.5) |
| Tumor grade | |||
| Low grade | 8 (11.8%) | 2 (4.8%) | 6 (23.1%) |
| High/Intermediate grade | 49 (72.0%) | 30 (71.4%) | 19 (73.1%) |
| Indeterminate | 11 (16.2%) | 10 (23.8%) | 1 (3.9%) |
| Histologic subtype | |||
| Classic | 56 (82.4%) | 32 (76.2%) | 24 (92.3%) |
| Epithelioid | 4 (5.9%) | 3 (7.1%) | 1 (3.9%) |
| Triton | 6 (8.8%) | 5 (11.9%) | 1 (3.9%) |
| Other* | 2 (2.9%) | 2 (4.8%) | 0 |
| Surgical margin status | |||
| Negative | 50 (73.5%) | 34 (81.0%) | 16 (61.5%) |
| Positive margin | 18 (26.5%) | 8 (19.0%) | 10 (38.5%) |
| Chemotherapy | |||
| No | 30 (44.1%) | 15 (35.7%) | 15 (57.7%) |
| Yes | 38 (55.9%) | 27 (64.3%) | 11 (42.3%) |
| Chemotherapy sequence | |||
| Preoperative† | 27 (71.1%) | 21 (77.8%) | 6 (54.5%) |
| Postoperative only | 11 (28.9%) | 6 (22.2%) | 5 (45.5%) |
| Radiation | |||
| No | 26 (38.2%) | 0 | 26 (100%) |
| Yes | 42 (61.8%) | 42 (100%) | 0 |
| Median EQD2 (range) | 50 Gy (34.8-97.5 Gy) | ||
| Radiation sequence | |||
| Preoperative | 27 (64.3%) | ||
| Postoperative | 15 (35.7%) |
Abbreviation: EQD2 = equivalent dose in 2 Gy fractions.
Other histologic subtypes included 1 malignant peripheral nerve sheath tumor with neuroblastoma features and 1 malignant peripheral nerve sheath tumor with divergent differentiation with chondromyxoid changes.
Preoperative chemotherapy included those who received preoperative chemotherapy with or without postoperative chemotherapy.
Table E1 shows detailed patient, tumor, and treatment characteristics for patients with NF1-associated MPNST (n = 51) versus those with sporadic MPNST (n = 17).
Median follow-up for all patients was 43.5 months (range, 5.4-172.9 months). The 3-year and 5-year OS for all patients was 73.4% and 61.1%, respectively (Fig. 1A). RT was associated with improved OS (3-year OS: 87.2% vs 51.5%, P = .039; Fig. 1B). Improved OS was also associated with negative compared to positive-margin status (3-year OS: 77.9% vs 61.1%, P = .009, Fig. 1C). On multivariable analysis, both receipt of RT and margin status remained associated with OS, in addition to tumor size (Table 2). When comparing this model to a more restricted model excluding tumor size, the unrestricted model demonstrated a better fit (P = .012). RT was associated with a reduction in the risk of death (HR, 0.4; 95% CI, 0.17-0.93; P = .034). Positive-margin status was associated with an increased risk of death (HR, 2.62; 95% CI, 1.18-5.78; P = .017), and increased tumor size was associated with an increase in risk of death (HR, 1.19; 95% CI, 1.04-1.36; P = .010).
Figure 1.
(A) Overall survival (OS), (B) OS stratified by receipt of RT, (C) OS stratified by margin status, (D) Progression-free survival (PFS), (E) Progression-free survival stratified by receipt of RT, (F) PFS stratified by margin status. The 3-year and 5-year OS for all patients were 73.4% and 61.1%, respectively, while the 3-year and 5-year PFS for all patients were 50.4% and 44.7%, respectively. Receipt of RT and negative-margin status were associated with both improved OS and PFS.
Abbreviation: RT = radiation therapy.
Table 2.
Variables associated with overall survival
| Overall survival |
||||||
|---|---|---|---|---|---|---|
| Univariate |
Multivariate |
|||||
| Variable | Variable count | Death events | HR (95% CI) | P value | HR (95% CI) | P value |
| Age | 68 | 27 | 1.01 (0.99-1.03) | .539 | ||
| Sex | ||||||
| Female | 29 | 14 | 1 | |||
| Male | 39 | 13 | 0.80 (0.37-1.70) | .560 | ||
| NF1 status | ||||||
| NF1 wildtype | 17 | 7 | 1 | |||
| NF1 mutant | 51 | 20 | 0.95 (0.40-2.26) | .907 | ||
| Size | 68 | 27 | 1.11 (0.98-1.25) | .104 | 1.19 (1.04-1.36) | .010 |
| Subtype | ||||||
| Nontriton | 62 | 25 | 1 | |||
| Triton | 6 | 2 | 0.69 (0.16-2.93) | .612 | ||
| Grade | ||||||
| Low grade | 8 | 2 | 1 | |||
| High grade | 49 | 18 | 0.91 (0.41-2.04) | .827 | ||
| Location | ||||||
| Nonextremity | 48 | 22 | 1 | |||
| Extremity | 20 | 5 | 0.46 (0.17-1.21) | .116 | ||
| Margin status | ||||||
| Negative/Close margin | 50 | 14 | 1 | |||
| Positive margin | 18 | 13 | 2.66 (1.24-5.69) | .012 | 2.62 (1.18-5.78) | .017 |
| Chemotherapy status | ||||||
| No chemotherapy | 30 | 14 | 1 | |||
| Chemotherapy | 38 | 13 | 0.76 (0.36-1.62) | .473 | ||
| Chemotherapy sequence | ||||||
| Adjuvant only | 11 | 6 | 1 | |||
| Neoadjuvant ± adjuvant | 27 | 7 | 0.41 (0.14-1.23) | .112 | ||
| Radiation status | ||||||
| No radiation | 26 | 13 | 1 | |||
| Radiation | 42 | 14 | 0.46 (0.21-0.98) | .044 | 0.40 (0.17-0.93) | .034 |
| Radiation sequence | ||||||
| Postoperative | 15 | 8 | 1 | |||
| Preoperative | 27 | 6 | 0.46 (0.16-1.33) | .151 | ||
Abbreviation: HR = hazard ratio.
The 3-year and 5-year PFS for all patients were 50.4% and 44.7%, respectively (Fig. 1D). RT was associated with improved PFS (3-year PFS: 55.8% vs 41.5%, P = .017; Fig. 1E). Negative compared to positive-margin status was also associated with improved PFS (3-year PFS: 55% vs 38.9%, P = .021; Fig. 1F). On multivariable analysis, increasing tumor size (HR, 1.18; 95% CI, 1.06-1.32; P = .002) and positive surgical margin (HR, 2.08; 95% CI, 1.08-4.03; P = .030) were associated with worse PFS, whereas receipt of RT (HR, 0.34; 95% CI, 0.16-0.69; P = .003) was associated with improved PFS (Table 3).
Table 3.
Variables associated with progression-free survival
| Progression-free survival |
||||||
|---|---|---|---|---|---|---|
| Univariate |
Multivariate |
|||||
| Variable | Variable count | Progression events | HR (95% CI) | P value | HR (95% CI) | P value |
| Age | 68 | 39 | 1.01 (0.99-1.02) | .461 | ||
| Sex | ||||||
| Female | 29 | 18 | 1 | |||
| Male | 39 | 21 | 0.99 (0.52-1.85) | .963 | ||
| NF1 status | ||||||
| NF1 wildtype | 17 | 12 | 1 | |||
| NF1 mutant | 51 | 27 | 0.74 (0.38-1.48) | .397 | ||
| Size | 68 | 39 | 1.11 (1.00-1.22) | .047 | 1.18 (1.06-1.32) | .002 |
| Subtype | ||||||
| Nontriton | 62 | 37 | 1 | |||
| Triton | 6 | 2 | 0.47 (0.11-1.97) | .303 | ||
| Grade | ||||||
| Low grade | 8 | 3 | 1 | |||
| High grade | 49 | 28 | 1.07 (0.53-2.16) | .840 | ||
| Location | ||||||
| Nonextremity | 48 | 30 | 1 | |||
| Extremity | 20 | 9 | 0.54 (0.25-1.14) | .105 | ||
| Margin status | ||||||
| Negative/Close margin | 50 | 24 | 1 | |||
| Positive Margin | 18 | 15 | 2.10 (1.10-4.02) | .024 | 2.08 (1.08-4.03) | .030 |
| Chemotherapy status | ||||||
| No chemotherapy | 30 | 20 | 1 | |||
| Chemotherapy | 38 | 19 | 0.65 (0.35-1.22) | .176 | ||
| Chemotherapy sequence | ||||||
| Adjuvant | 11 | 8 | 1 | |||
| Neoadjuvant ± adjuvant | 27 | 11 | 0.48 (0.19-1.21) | .119 | ||
| Radiation status | ||||||
| No radiation | 26 | 18 | 1 | |||
| Radiation | 42 | 21 | 0.46 (0.24-0.88) | .019 | 0.34 (0.16-0.69) | .003 |
| Radiation sequence | ||||||
| Postoperative | 15 | 9 | 1 | |||
| Preoperative | 27 | 12 | 0.82 (0.34-1.95) | .652 | ||
The 3-year cumulative incidence of LF for patients treated with and without RT was 25% versus 37% (P = .240), respectively (Fig. 2A). The 3-year cumulative incidence of DF for patients treated with and without RT was 15.9% versus 22% (P = .516), respectively (Fig. 2B). When DF was not considered as a competing event, RT was associated with improved LC (3-year LC: 66.3% vs 40.4%, P = .017; Fig. 2C). When LF was not considered as a competing event, RT was not associated with improved DC (3-year DC: 72.1% vs 50.0%, P = .055; Fig. 2D). The most common site of DF was the lungs. Table E2 characterizes progression events, broken down by local, distant, or local and DF.
Figure 2.
(A) Cumulative incidence of local failure for patients treated with or without RT with distant failure as a competing event, (B) Cumulative incidence distant failure for patients treated with or without RT with local failure as a competing event, (C) Kaplan–Meier curve of local control stratified by receipt of radiation, (D) Kaplan–Meier curve of distant control stratified by receipt of radiation. Receipt of RT was significantly associated with improved local control versus no RT based on Kaplan–Meier curves (P = .017), whereas the cumulative incidence of local failure, when accounting for distant failure as a competing event, was not significantly different between those receiving RT versus no RT.
Abbreviation: RT = radiation therapy.
In a subgroup analysis of patients with NF1, the 3-year and 5-year OS were 72.2% and 61.3%, respectively (Figure E1). Patients with NF1 who received RT had significantly improved OS compared to patients with NF1 who did not receive RT (3-year OS: 89.1% vs 48.8%, P = .031, Fig. E2). For patients with NF1, the 3-year and 5-year PFS were both 51.9% (Fig. E3), and receipt of RT was associated with improved PFS (3-year PFS: 62.4% vs 36.7%, P = .011, Fig. E4). On multivariable analyses, receipt of RT, margin status, and tumor size were all associated with both OS and PFS in patients with NF1 (Tables E3, E4).
Six patients developed 7 total secondary malignancies after their initial MPNST diagnosis, all occurring in patients with NF1, and included 4 MPNSTs (1 patient developed 2 subsequent MPNSTs), 1 astrocytoma, 1 gastrointestinal stromal tumor, and 1 case of chemotherapy-related acute myeloid leukemia. Table E5 summarizes all secondary malignancy characteristics, initial treatment details, and relation to RT field, if applicable. Two patients had no prior RT exposure (cases 2-3 in Table E5). Three patients who received RT for their initial MPNST subsequently developed solid secondary malignancies (cases 4-6 in Table E5). Two of these patients developed subsequent MPNSTs either within or at the edge of the prior RT field. One patient was initially diagnosed with a left neck MPNST, involving C4 to C5 nerve roots, and underwent positive-margin resection with rapid recurrence, prompting a subsequent reresection with adjuvant chemotherapy and adjuvant proton RT to 60 Cobalt Gray Equivalent. This patient then developed a second MPNST outside of the RT field 2 years later and a third MPNST 9.5 years after initial RT, directly within the prior RT field. This patient’s second and third MPNST arose from plexiform neurofibromas and were not thought to represent metastatic disease by the clinical team. Another patient developed a second MPNST at the edge of the prior RT field in the setting of an initial left inguinal MPNST which was treated with preoperative RT to 50 Gy in 25 fractions with subsequent margin-negative resection (Fig. 3A). This patient’s second MPNST was in the left medial thigh compartment and was diagnosed 8.6 years after the initial MPNST treatment and was thought to arise from a separate plexiform neurofibroma. The isocenter of the second MPNST was outside of the prior RT field’s 12.5% isodose line (Fig. 3B, C).
Figure 3.
(A) RT plan for initial left inguinal MPNST with isodose lines. (B) Initial left inguinal RT plan fused to new CT simulation for second MPNST in the left medial thigh. GTV of second MPNST is shown in red. Yellow dot marks the isocenter of the second MPNST. (C) Second MPNST GTV (red), CTV (orange), and PTV (light blue) in the left medial thigh. For the patient with a secondary MPNST at the edge of a prior RT field, the center of the new MPNST falls outside of the prior 12.5% isodose line, and it is difficult to distinguish based on anatomic changes over time, whether this secondary MPNST arises from a new anatomic compartment altogether. Nevertheless, the new secondary MPNST could be considered possibly RT-associated given its overlap with the prior RT field.
Abbreviations: CT = computed tomography; CTV = clinical target volume; GTV = gross target volume; MPNST = malignant peripheral nerve sheath tumor; PTV = planning target volume; RT = radiation therapy.
Discussion
In this contemporary retrospective series of patients with localized resected MPNST treated at a single institution with a large population of patients with NF1 and a large multidisciplinary sarcoma practice, we found that RT was independently associated with improved PFS and OS. All secondary malignancies arose in patients with NF1. The percent of second malignancies arising within or at the edge of the RT field was 4.8% (n = 2), whereas 7.7% of those who did not receive RT also developed a secondary malignancy (n = 3).
The 5-year OS for our entire cohort was 61.1%, which is among the highest reported in comparison to prior studies (range, 34%-66%).2, 3, 4, 5, 6,9, 10, 11, 12, 13,16, 17, 18, 19, 20, 21, 22 Our relatively high survival outcomes may reflect a modern cohort of patients benefiting from improvements in diagnostic imaging, modern RT techniques, evolving systemic therapy options, and regular surveillance through a dedicated NF1 clinic. Additionally, 75% of patients in our cohort had NF1, which is the highest reported proportion among similar literature investigating outcomes (range, 11.7%-56%).2, 3, 4, 5, 6,9,10,12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22,26 Patients with NF1 generally have inferior outcomes compared to those with sporadic MPNSTs, as reported in 2 recent meta-analyses.11,37 In our series, NF1 was not significantly associated with differences in survival outcomes, but within this context, our OS rate may be considered particularly high. It should additionally be noted that our institution has a dedicated NF1 clinic with regular surveillance of this patient population, which may further contribute to the relatively favorable outcomes seen in this population. Overall, our series enhances the understanding of treatment outcomes for localized resectable MPNST, particularly for patients with NF1 where prior data are limited.
Consistent with prior literature, tumor size and surgical margin status were significantly associated with both OS and PFS, affirming that these variables are critical prognostic factors for MPNST.2,4, 5, 6,9, 10, 11, 12, 13,15, 16, 17, 18, 19,21,22 Tumor size reached significance in the multivariable model (P = .010) after adjustment for margin status and RT receipt. This likely reflects the interrelated nature of these variables in clinical practice, as larger tumors are more challenging to resect with negative margins and more commonly receive adjuvant RT. After adjusting for tumor size and margin status, RT remained significantly associated with improved OS and PFS. These findings are notable and in contrast to most prior studies which have not found significant associations between RT and survival outcomes on multivariable analyses.2,4,5,9,10,12,13,16,17,20, 21, 22,25
Two recent series found RT to be associated with improved LC in patients with MPNST, but did not find a significant association between RT and improved OS.26,27 Otherwise, a 2019 series of 251 patients with localized MPNST treated between 1960 and 2016 by Miao et al19 is the only other study to find a significant association between RT and both LC and OS. Although most prior literature is limited to small retrospective series, a recent meta-analysis by Cai et al11 consisting of patients with MPNST from 28 studies published between 1966 and 2020 found RT to be a significantly protective factor for OS, consistent with our findings.
Although the 3-year cumulative incidence of LF did not differ significantly between patients treated with or without RT (25% vs 37%, P = .240), this analysis accounts for DF as a competing event. When DF was not treated as a competing event, RT was associated with significantly improved LC on Kaplan–Meier analysis (3-year LC: 66.3% vs 40.4%, P = .017). This discrepancy reflects the limitations of competing-risks analyses in cohorts with high rates of competing events. In a population with frequent competing events (eg, DF), competing-risks models have reduced power to detect LF differences compared to OS analyses. The Kaplan–Meier analysis isolates the LC effect of RT and suggests a meaningful local benefit. Taken together, the improvement in LC on Kaplan–Meier analysis, combined with the numerically lower cumulative incidence of LF among patients treated with RT, could explain the improvement in PFS and OS observed in our series. Additionally, in contrast to most prior studies, our series focuses entirely on patients treated in a modern era with modern RT techniques (97.6% of patients treated with RT received intensity-modulated techniques), which may account for our findings of significant positive association between RT and clinical outcomes.
In our cohort, only patients with NF1 developed a secondary malignancy. Two patients developed a secondary MPNST within or adjacent to a prior RT field with latency times of 9.5 years and 8.6 years. It is not possible to know whether these tumors would have developed without RT exposure in the setting of NF1. Overall, the percent of secondary malignancies potentially associated with RT was similar to the percent of secondary malignancies observed in patients with NF1 who did not receive RT. Prior studies report median latency periods for development of RT-associated MPNST from approximately 12 to 17 years, but these estimates are not specific to patients whose initial malignancy was also an MPNST.2,15,19,28, 29, 30 Given the 5-year OS in our cohort of 61.1%, many patients do not survive long enough to develop a secondary RT-associated MPNST.
Furthermore, improvements in RT technique and a shift toward preoperative RT allow for smaller volumes to be irradiated to lesser doses, potentially further reducing secondary malignancy risk. At our institution, all MPNST cases are discussed in a multidisciplinary fashion, and surgical goals include the resection of the entire irradiated surrounding plexiform neurofibroma (when feasible). This practice may contribute to our relatively low rates of secondary malignancy development within RT fields. Although secondary malignancy risk should be acknowledged, in a modern treatment era, the survival benefit of RT likely outweighs the long-term risk of RT-associated secondary malignancy development, particularly when there is close multidisciplinary coordination between radiation oncology and surgical teams.
There are several limitations to our analysis. Despite being the largest cohort of similar patients published to date, our cohort size limits the statistical power. As a retrospective study, there is an inherent selection bias, especially regarding treatment decisions. Although we controlled for confounding variables such as tumor size and margin status, there may have been other variables that influenced whether a patient was treated with RT that were not captured in our dataset. Aside from tumor grade, there were no differences between the treatment groups, however, residual confounding by indication remains an important limitation. Although we adjusted for known prognostic factors (tumor size, margin status), unmeasured confounders such as performance status, comorbidities, patient preferences, and physician judgment regarding suitability for aggressive local therapy may influence treatment selection and outcomes.
Additionally, limitations in the multivariable analyses warrant consideration, and the association between RT and improved OS and PFS should be interpreted with caution and viewed as hypothesis-generating observations in need of validation in larger, prospective cohorts. Our multivariable analyses utilized a P value-driven stepwise variable selection approach which carries recognized limitations. This method may overestimate effect sizes and does not reliably protect against overfitting, even when the number of included variables is modest relative to the number of events. Additionally, this model may not capture all clinically relevant relationships. The significance of tumor size in multivariable but not univariable analyses of OS reflects confounding relationships between size, margin status, and RT use—variables that are inherently interrelated in clinical decision-making. Although our event-per-variable ratios are acceptable, the modest sample size limits our ability to fully disentangle these relationships or conduct extensive sensitivity analyses. An alternate method using a priori variable selection was considered and would have included tumor size and margin status, along with RT as our variable of interest, as tumor size and margin status are the only variables consistently identified in prior studies as having prognostic value in MPNST clinical outcomes.2,4, 5, 6,9, 10, 11, 12, 13,15, 16, 17, 18, 19,21,22 The convergence between the P value-driven selection method and these literature-informed prognostic variables is reassuring that the final multivariable models were not an artifact of stepwise selection, but still warrant cautious interpretation.
Furthermore, the inability to assign grade in 11 cases (16%), predominantly due to small biopsy specimens, represents a limitation common to retrospective MPNST series. Although these cases were distributed across treatment groups, and the missing data pattern does not appear systematically biased, we cannot fully exclude the possibility that grade misclassification contributes to residual confounding in our models. Taken altogether, the observed associations between RT and improved survival should be interpreted as hypothesis-generating rather than establishing causation. Prospective studies or propensity-matched analyses in larger cohorts are needed to further delineate treatment effects from selection bias.
Conclusions
RT was independently associated with improved PFS and OS in patients with localized resectable MPNST. Secondary malignancies arising within the irradiated field were rare and occurred beyond the median survival timeframe for this patient population. These findings of RT’s association with improved survival are hypothesis-generating and should be interpreted cautiously given the retrospective design and modest cohort size. Prospective studies and larger cohort sizes are needed to validate the role of RT in MPNST. Further research is warranted to better define the role of preoperative RT and to explore effective adjuvant systemic therapy strategies to guide optimal treatment decision-making.
Disclosures
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
We would like to acknowledge Ljubitca Fadic, MS, for her administrative assistance in preparing and submitting this manuscript. We would also like to acknowledge Victoria Doss, MD and Sahaja Acharya, MD, who were responsible for the statistical analysis.
Footnotes
This work had no specific funding.
Research data are stored in an institutional repository and will be shared upon request to the corresponding author.
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.adro.2026.102131.
Appendix. Supplementary materials
References
- 1.Wu J.M., Montgomery E. Classification and pathology. Surg Clin North Am. 2008;88:483. doi: 10.1016/j.suc.2008.03.007. [DOI] [PubMed] [Google Scholar]
- 2.Ducatman B.S., Scheithauer B.W., Piepgras D.G., Reiman H.M., Ilstrup D.M. Malignant peripheral nerve sheath tumors. A clinicopathologic study of 120 cases. Cancer. 1986;57:2006–2021. doi: 10.1002/1097-0142(19860515)57:10<2006::aid-cncr2820571022>3.0.co;2-6. [DOI] [PubMed] [Google Scholar]
- 3.Evans D.G.R., Baser M.E., Mcgaughran J., Sharif S., Howard E., Moran A. Malignant peripheral nerve sheath tumours in neurofibromatosis 1. J Med Genet. 2002;39:311–314. doi: 10.1136/jmg.39.5.311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Carli M., Ferrari A., Mattke A., et al. Pediatric malignant peripheral nerve sheath tumor: The Italian and German soft tissue sarcoma cooperative group. J Clin Oncol. 2005;23:8422–8430. doi: 10.1200/JCO.2005.01.4886. [DOI] [PubMed] [Google Scholar]
- 5.Wong W.W., Hirose T., Scheithauer B.W., Schild S.E., Gunderson L.L. Malignant peripheral nerve sheath tumor: Analysis of treatment outcome. Int J Radiat Oncol Biol Phys. 1998;42:351–360. doi: 10.1016/s0360-3016(98)00223-5. [DOI] [PubMed] [Google Scholar]
- 6.Anghileri M., Miceli R., Fiore M., et al. Malignant peripheral nerve sheath tumors: Prognostic factors and survival in a series of patients treated at a single institution. Cancer. 2006;107:1065–1074. doi: 10.1002/cncr.22098. [DOI] [PubMed] [Google Scholar]
- 7.Gupta G., Mammis A., Maniker A. Malignant peripheral nerve sheath tumors. Neurosurg Clin N Am. 2008;19:533–543. doi: 10.1016/j.nec.2008.07.004. [DOI] [PubMed] [Google Scholar]
- 8.Dunn G.P., Spiliopoulos K., Plotkin S.R., et al. Role of resection of malignant peripheral nerve sheath tumors in patients with neurofibromatosis Type 1. J Neurosurg. 2013;118:142–148. doi: 10.3171/2012.9.JNS101610. [DOI] [PubMed] [Google Scholar]
- 9.Bishop A.J., Zagars G.K., Torres K.E., Bird J.E., Feig B.W., Guadagnolo B.A. Malignant peripheral nerve sheath tumors: A single institution’s experience using combined surgery and radiation therapy. Am J Clin Oncol. 2018;41:465–470. doi: 10.1097/COC.0000000000000303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Longhi A., Errani C., Magagnoli G., et al. High grade malignant peripheral nerve sheath tumors: Outcome of 62 patients with localized disease and review of the literature. J Chemother. 2010;22:413–418. doi: 10.1179/joc.2010.22.6.413. [DOI] [PubMed] [Google Scholar]
- 11.Cai Z., Tang X., Liang H., Yang R., Yan T., Guo W. Prognosis and risk factors for malignant peripheral nerve sheath tumor: A systematic review and meta-analysis. World J Surg Oncol. 2020;18:257. doi: 10.1186/s12957-020-02036-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Martin E., Coert J.H., Flucke U.E., et al. A nationwide cohort study on treatment and survival in patients with malignant peripheral nerve sheath tumours. Eur J Cancer. 2020;124:77–87. doi: 10.1016/j.ejca.2019.10.014. [DOI] [PubMed] [Google Scholar]
- 13.Sobczuk P., Teterycz P., Czarnecka A.M., et al. Malignant peripheral nerve sheath tumors - outcomes and prognostic factors based on the reference center experience. Surg Oncol. 2020;35:276–284. doi: 10.1016/j.suronc.2020.09.011. [DOI] [PubMed] [Google Scholar]
- 14.Doorn P.F., Molenaar W.M., Buter J., Hoekstra H.J. Malignant peripheral nerve sheath tumors in patients with and without neurofibromatosis. Eur J Surg Oncol. 1995;21:78–82. doi: 10.1016/s0748-7983(05)80073-3. [DOI] [PubMed] [Google Scholar]
- 15.Lafemina J., Qin L.X., Moraco N.H., et al. Oncologic outcomes of sporadic, neurofibromatosis-associated, and radiation-induced malignant peripheral nerve sheath tumors. Ann Surg Oncol. 2013;20:66–72. doi: 10.1245/s10434-012-2573-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Stucky C.C., Johnson K.N., Gray R.J., et al. Malignant peripheral nerve sheath tumors (MPNST): The mayo clinic experience. Ann Surg Oncol. 2012;19:878–885. doi: 10.1245/s10434-011-1978-7. [DOI] [PubMed] [Google Scholar]
- 17.Kahn J., Gillespie A., Tsokos M., et al. Radiation therapy in management of sporadic and neurofibromatosis type 1-associated malignant peripheral nerve sheath tumors. Front Oncol. 2014;4:324. doi: 10.3389/fonc.2014.00324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ferrari A., Bergamaschi L., Chiaravalli S., et al. Malignant peripheral nerve sheath tumor in children and adolescents: Local treatment in a retrospective single-center experience. Pediatr Blood Cancer. 2025;72 doi: 10.1002/pbc.31813. [DOI] [PubMed] [Google Scholar]
- 19.Miao R., Wang H., Jacobson A., et al. Radiation-induced and neurofibromatosis-associated malignant peripheral nerve sheath tumors (MPNST) have worse outcomes than sporadic MPNST. Radiother Oncol. 2019;137:61–70. doi: 10.1016/j.radonc.2019.03.015. [DOI] [PubMed] [Google Scholar]
- 20.Yuan Z., Xu L., Zhao Z., et al. Clinicopathological features and prognosis of malignant peripheral nerve sheath tumor: A retrospective study of 159 cases from 1999 to 2016. Oncotarget. 2017;8:104785–104795. doi: 10.18632/oncotarget.18975. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Valentin T., Le Cesne A., Ray-Coquard I., et al. Management and prognosis of malignant peripheral nerve sheath tumors: The experience of the French Sarcoma Group (GSF-GETO) Eur J Cancer. 2016;56:77–84. doi: 10.1016/j.ejca.2015.12.015. [DOI] [PubMed] [Google Scholar]
- 22.Zou C., Smith K.D., Liu J., et al. Clinical, pathological, and molecular variables predictive of malignant peripheral nerve sheath tumor outcome. Ann Surg. 2009;249:1014–1022. doi: 10.1097/SLA.0b013e3181a77e9a. [DOI] [PubMed] [Google Scholar]
- 23.Kaushal A., Citrin D. The role of radiation therapy in the management of sarcomas. Surg Clin North Am. 2008;88 doi: 10.1016/j.suc.2008.03.005. 629-vii. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.O'Sullivan B., Davis A.M., Turcotte R., et al. Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: A randomised trial. Lancet. 2002;359:2235–2241. doi: 10.1016/S0140-6736(02)09292-9. [DOI] [PubMed] [Google Scholar]
- 25.Kolberg M., Høland M., Agesen T.H., et al. Survival meta-analyses for >1800 malignant peripheral nerve sheath tumor patients with and without neurofibromatosis type 1. Neuro Oncol. 2013;15:135–147. doi: 10.1093/neuonc/nos287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jansma CYMN, Grünhagen D.J., Flucke U.E., et al. The role of radiotherapy in MPNST and the impact of NF1 status on outcomes: Insights from a multicenter cohort study. Neuro Oncol. 2025;27:3214–3223. doi: 10.1093/neuonc/noaf186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Roohani S., Claßen N.M., Ehret F., et al. The role of radiotherapy in the management of malignant peripheral nerve sheath tumors: A single-center retrospective cohort study. J Cancer Res Clin Oncol. 2023;149:17739–17747. doi: 10.1007/s00432-023-05449-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yamanaka R., Hayano A. Radiation-induced malignant peripheral nerve sheath tumors: A systematic review. World Neurosurg. 2017;105:961–970.e8. doi: 10.1016/j.wneu.2017.06.010. [DOI] [PubMed] [Google Scholar]
- 29.Riad S., Biau D., Holt G.E., et al. The clinical and functional outcome for patients with radiation-induced soft tissue sarcoma. Cancer. 2012;118:2682–2692. doi: 10.1002/cncr.26543. [DOI] [PubMed] [Google Scholar]
- 30.Mavrogenis A.F., Pala E., Guerra G., Ruggieri P. Post-radiation sarcomas. Clinical outcome of 52 patients. J Surg Oncol. 2012;105:570–576. doi: 10.1002/jso.22122. [DOI] [PubMed] [Google Scholar]
- 31.Bright C.J., Hawkins M.M., Winter D.L., et al. Risk of soft-tissue sarcoma among 69 460 five-year survivors of childhood cancer in Europe. J Natl Cancer Inst. 2018;110:649–660. doi: 10.1093/jnci/djx235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ahlawat S., Blakeley J.O., Rodriguez F.J., Fayad L.M. Imaging biomarkers for malignant peripheral nerve sheath tumors in neurofibromatosis type 1. Neurology. 2019;93:e1076–e1084. doi: 10.1212/WNL.0000000000008092. [DOI] [PubMed] [Google Scholar]
- 33.Naghavi A.O., Yang G.Q., Latifi K., Gillies R., Mcleod H., Harrison L.B. The future of radiation oncology in soft tissue sarcoma. Cancer Control. 2018;25 [Google Scholar]
- 34.Higham C.S., Steinberg S.M., Dombi E., et al. SARC006: Phase II trial of chemotherapy in sporadic and neurofibromatosis type 1 associated chemotherapy-naive malignant peripheral nerve sheath tumors. Sarcoma. 2017;2017 doi: 10.1155/2017/8685638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Spunt S.L., Million L., Chi Y.Y., et al. A risk-based treatment strategy for non-rhabdomyosarcoma soft-tissue sarcomas in patients younger than 30 years (ARST0332): A Children's Oncology Group prospective study. Lancet Oncol. 2020;21:145–161. doi: 10.1016/S1470-2045(19)30672-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cahan W.G., Woodard H.Q., Higinbotham N.L., Stewart F.W., Coley B.L. Sarcoma arising in irradiated bone: Report of eleven cases. Cancer. 1998;82:8–34. doi: 10.1002/(sici)1097-0142(19980101)82:1<8::aid-cncr3>3.0.co;2-w. [DOI] [PubMed] [Google Scholar]
- 37.Lim Z., Gu T.Y., Tai B.C., Puhaindran M.E. Survival outcomes of malignant peripheral nerve sheath tumors (MPNSTs) with and without neurofibromatosis type I (NF1): A meta-analysis. World J Surg Oncol. 2024;22:14. doi: 10.1186/s12957-023-03296-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.