**1. Introduction**

Malignant peripheral nerve sheath tumor (MPNSTs) is the sixth most common soft tissue sarcoma [1] and has an incidence rate of 0.1–0.2 per 100,000 persons per year [2]. MPNSTs are often associated with neurofibromatosis type 1 (NF1). The incidence rate of MPNSTs in patients with NF1 is much higher than that of the general population, estimated to be 1.6 per 1000 per year, or a lifetime risk of 8–13% [3]. Approximately 50% of MPNSTs occur in patients with neurofibromatosis [4–7], and the other 50% of MPNSTs occur sporadically or in the setting of previous radiation therapy [4,6]. In the setting of NF1, MPNSTs often arise within a pre-existing benign nerve sheath tumor (plexiform neurofibroma) [4,7].

Prognosis remains poor for patients with MPNST despite multi-modality therapy [2,5–10]. In the setting of metastatic disease, treatment is limited to cytotoxic chemotherapy, typically consisting of single agen<sup>t</sup> doxorubicin or a combination of doxorubicin and ifosfamide [11–13].

A number of di fferent genes have been implicated in the development of MPNSTs. One of the most commonly used models for preclinical testing was developed by Cichowski et al. and Vogel et al; they demonstrated that mice with germline variants in *Nf1* and *Tp53* develop MPNSTs, supporting a cooperative and causal role for these tumor suppressors in the context of MPNST formation [14,15]. Other groups have found a reduction in expression of *PTEN*, a tumor suppressor in the *PI3K*/*AKT*/*mTOR* pathway, in MPNSTs compared to benign nerve sheath tumors in a manner that is not regulated by *NF1* [16]. Keng et al. went on to demonstrate the cooperative roles of *Pten* and *Nf1* in the tumorigenesis of MPNSTs in vivo with transgenic mouse models [17]. Gregorian et al. further elucidated the cooperative relationship between *k-ras* activation and *Pten* deletion, showing that both variants in combination led to 100% penetrable development of MPNSTs [18]. Another gene implicated in MPNST pathogenesis is *INK4A*, a tumor suppressor encoding both *p16* and *p19*. Deletions in this gene have been identified in MPNSTs but not in benign neurofibromas [19]. Lu et al. demonstrated a di fference in aberrant expression of *ATRX*, a DNA helicase that plays a role in chromatin regulation and maintenance of telomeres, between MPNSTs and benign neurofibromas [20]. Additionally, variants in *EED* and *SUZ12* have been observed in MPNST. These genes code for components of the PRC2 complex which is involved in transcriptional repression. Lee et al. showed loss-of-function somatic alterations of PRC2 components in 92% of sporadic, 70% of NF1-associated and 90% of radiotherapy-associated MPNSTs. Further, introduction of the lost PRC2 component in a PRC2-deficient MPNST cell line decreased cell growth [21]. Others have found alterations such as structural alterations of *PDGFRA* (platelet-derived growth factorα) in 26% of MPNST samples [22]; increased expression of *EGF-R* (epidermal growth factor receptor) by immunohistochemistry in MPNSTs [23]; and *IGFR1* gene amplification in 24% of MPNSTs [24].

Despite all of this research, no e ffective therapies have emerged from recent clinical studies based on this genomic data and subsequent preclinical studies. Intra-tumoral heterogeneity is a possible reason for these shortcomings. Prior studies have relied on a single sample from these tumors. All the subclones within a tumor may not be captured by this approach. Our aim in this study is to investigate intra-tumoral heterogeneity more thoroughly through analysis of samples taken from multiple sites of the same MPNST.

#### **2. Materials and Methods**

## *2.1. Study Approvals*

Blood and tumor were obtained from an individual diagnosed with NF1 according to established criteria [25] and treated for a MPNST at Washington University/St. Louis Children's Hospital NF Clinical Program (St. Louis, MO, USA). The human tumor samples were collected under an approved IRB protocol (#201203042) at Washington University, and the patient was appropriately consented.

## *2.2. Sample Collection*

Samples were taken from three distinct areas within a single tumor from a patient with an NF1-MPNST immediately after surgical resection with guidance from a pathologist (SD). While area "1" represented solid, tan homogeneous tumor lacking hemorrhage and/or necrosis, areas "2" and "3" of the tumor grossly appeared necrotic and hemorrhagic respectively. 20 g of tissue was taken from each area. Each area was then divided to be used for RNA extraction, DNA extraction, and slide preparation to analyze the histology. A gross image of the tumor was taken at this time and is shown as Figure 1.

**Figure 1.** Malignant peripheral nerve sheath tumor (MPNST) sampled areas. Area 1 shows an area centrally located in MPNST, Area 2 an area of hemorrhage, and Area 3 an area of necrosis.
