Pediatric Soft Tissue Sarcoma in Limb-Girdle Muscular Dystrophy: Molecular Findings and Clinical Implications

Abstract

Patient: Male, 17-year-old

Final Diagnosis: Limb-girdle muscular dystrophy recessive 1 • desmoplastic small round cell tumor

Symptoms: Shoulder and pelvic girdle muscle weakness • progressive pain in the right thigh, radiating to knee and hip

Clinical Procedure: —

Specialty: Genetics

Objective:

Rare coexistence of disease or pathology

Background:

Limb-girdle muscular dystrophy recessive 1 (LGMDR1) is an autosomal recessive degenerative muscle disorder characterized by progressive muscular weakness caused by pathogenic variants in the CAPN3 gene. Desmoplastic small round cell tumors (DSRCT) are ultra-rare and aggressive soft tissue sarcomas usually in the abdominal cavity, molecularly characterized by the presence of a EWSR1::WT1 fusion transcript. Mouse models of muscular dystrophy, including LGMDR1, present an increased risk of soft tissue sarcomas. However, the DSRCT risk and general cancer risk in patients with LGMD is unknown. Here, we delineate the clinical, molecular, and genetic findings of a patient with LGMDR1 who developed a DSRCT.

Case Report:

The patient was a boy who was diagnosed at the age of 9 years with LGMDR1, caused by the biallelic pathogenic variants NP_000061.1:p.(Arg448Cys) and NP_000061.1:p.(Thr184ArgfsTer36) in CAPN3. At 17 years of age, a pathologic soft tissue mass was found in the right pelvis. Immunostaining was positive for Desmin and negative for Myogenin and MyoD1, and RNA sequencing showed a EWSR1::WT1 fusion transcript, confirming the diagnosis of DSRCT. The patient relapsed after 1 year and, following a second relapse, he was started on palliative treatment. No germline variants in childhood cancer predisposition genes were detected by whole genome sequencing.

Conclusions:

We describe a patient with LGMDR1 who developed a DSRCT. Since associations between LGMD and pediatric cancer are hitherto unknown, further studies are warranted, as little information is currently published about the pediatric cancer risk in this patient group.

Introduction

Limb-girdle muscular dystrophy recessive 1 (LGMDR1, MIM #253600), previously known as LGMD2A, is an autosomal recessive disorder characterized by progressive weakness of the proximal leg and shoulder girdle muscles [1]. Based on public sequencing databases, the disease has an estimated prevalence of 0.02 to 27.0 per million individuals, highly dependent on the geographical location [2]. The debut of symptoms is very variable, ranging from early childhood to over 40 years of age [3]. The disease-causing gene, CAPN3, encodes a calcium-dependent cysteine protease with roles in muscle formation, turnover, and repair and calcium release/uptake in the sarcomere [4]. Loss of CAPN3 activity in dystrophic muscle leads to calcium imbalances, oxidative damage, mitochondrial alterations, sarcomere disorganization, impaired muscle adaptation, and abnormal muscle regeneration [5].

Soft tissue sarcomas (STS) are rare malignant tumors arising from soft tissue. Pediatric presentations have an incidence of 12 in 100 000 children, accounting for 6.7% of all childhood malignancies [6]. STS are a heterogeneous group of tumors, with distinct clinical behavior, histology, and tumor biology [7]. Although there is no clear cause, different factors associated with the development of STS have been suggested, including genetic predisposition, oncogenic viruses, immunodeficiency, chemical carcinogens, and chronic inflammation [8].

Desmoplastic small round cell tumors (DSRCT) are aggressive malignant STS, which often develop in the abdominal cavity, commonly metastasizing to the peritoneum, liver, and lymphatic system [9,10]. DSRCT are extremely rare. An epidemiological study in the United States calculated an age-adjusted incidence of 0.3 cases per million with a 3.6 male to female ratio, and a peak age at diagnosis of 20 to 24 years [11]. Molecularly, DSRCT are characterized by the EWSR1::WT1 gene fusion, usually due to the recurrent chromosomal translocation t(11;22)(p13;q12), which combines the transactivation region of EWSR1 and the DNA binding domain of WT1 [10,12]. The chimeric protein acts as an oncogenic transcription factor, regulating the expression of several growth and transcription factors involved in carcinogenesis, including, for instance, PDGFRα, IGF1R, EGFR, MYC, PAX2, and WT1 [10]. DSRCT diagnosis includes a positive dot-like Desmin staining, often positive cytokeratin, and negative staining for skeletal muscle-associated markers Myogenin and MyoD1. Detection of the EWSR1::WT1 fusion transcript is desirable [13]. The current management for DSRCT is multimodal, including a combination of chemotherapy, radiotherapy, and cytoreductive surgery. The prognosis remains poor, as recurrences are common, and the survival rate is low [10].

Independent studies have demonstrated that mouse models of different muscular dystrophies (MD), including LGMDR1, are susceptible to STS, with varied penetrance and age-at-onset depending on the muscular disease, mouse strain, and genetic variant modelled [14–19]. The study from Schmidt et al included a small cohort of Capn3 −/− mice, of which 1/19 (5%) developed STS. The STS rate was increased in Dmd −/− Capn3 −/− double-knockout mice, with 24/55 (44%) mice affected [14]. Seven reports of cancer in individuals with LGMD have been published in the literature [20–22]. Only one of them had an STS; a patient with concomitant liposarcoma and LGMD2B [21].

Here, we delineate the clinical, molecular, and genetic characteristics of a 17-year-old male patient with LGMDR1 who developed a DSRCT and discuss possible pediatric cancer associations.

Case Report

The patient was born after an uneventful pregnancy, to parents with unremarkable family history. He was born by cesarean delivery, as the umbilical cord was wrapped around the shoulder, with a birth weight, length, and head circumference within the reference range. Postnatally, a hemodynamically insignificant muscular ventricular septum defect that closed spontaneously was diagnosed. His growth and psychomotor development were normal.

At 9 years of age, the patient was referred to an orthopedic specialist with concern for persistent tiptoe walking. General physical and neurological examinations were unremarkable, apart from symmetric Achilles tendon contracture and calf hypertrophy. Additional investigations revealed an elevated serum creatine phosphokinase concentration, which motivated further MD workup. Initial Multiplex ligation-dependent probe amplification (MLPA) analysis for the detection of deletions and duplications within the DMD gene was normal. Histopathologic evaluation of a muscle biopsy, complemented by calpain-3 immunoblot testing, showed reduced expression of calpain-3 in muscle tissue, suggesting a calpainopathy. Subsequent MLPA analyses and sequencing of the entire coding region of the CAPN3 gene revealed 2 heterozygous pathogenic variants, a CAPN3 frameshift variant in exon 4 (NM_000070.3: c.550del (NP_000061.1:p.(Thr184ArgfsTer36) and a missense variant in exon 10 (NM_000070.3:c.1342C>T NP_000061.1:p.(Arg448Cys) in the patient, confirming the diagnosis of LGMDR1. The variants have been reported as pathogenic in ClinVar [23]. Both parents were asymptomatic carriers of the respective variants, while a younger brother also has a diagnosis of LGMDR1 (See Table 1 for a summary of the case report).

Table 1.

Summary of clinical characteristics of the patient.

Sex	Male
Family history	Unremarkable
Clinical history: Limb-girdle muscular dystrophy
Age at diagnosis	9 years of age
CAPN3 variants	NM_000070.3:c.550del (NP_000061.1:p.(Thr184ArgfsTer36) and NM_000070.3:c.1342C>T (NP_000061.1:p.(Arg448Cys)
Muscular symptoms	At 9 years of age: Persistent tiptoe walking, symmetric Achilles tendon contracture and calf hypertrophy At 13 years of age: Shoulder and pelvic girdle muscle weakness
Clinical history: Primary desmoplastic small round cell tumor
Age at diagnosis	17 years of age
Location	Right pelvis with multiple skeletal metastases.
Tumor histology	Desmin-positive staining. CD99, NKX2-2, CD45, BCOR, Myogenin, MyoD1 and CKMNF-negative cells.
Tumor genetics	– Near triploid karyotype with multiple numerical and segmental chromosomal changes – EWSR1::WT1 (t(11;22)(p13;q12)) fusion transcript
Germline genetics	No variants found in 189 childhood cancer predisposition genes [26]
Treatment regime	Upfront treatment: – Local radiotherapy – Systemic courses of ifosfamide, vincristine, adriamycin, ifosfamide and actinomycin-D; followed by actinomycin-D Maintenance therapy: – Six months cyclophosphamide and vinorelbine
Clinical history: Relapse tumors
Age at diagnosis	18 years of age
Location	First relapse: Left hip Second relapse: Scalp
Histology of the second relapse tumor	Desmin, vimentin and NTRK (Pan-Trk antibody)-positive staining. Pan cytokeratin-negative cells.
Genetics of the second relapse tumor	– Near triploid karyotype with multiple numerical and segmental chromosomal changes – EWSR1::WT1 (t(11;22)(p13;q12)) fusion transcript
Treatment regime	After first relapse: – Systemic 6-month therapy with topotecan, etoposide, carboplatin, and cyclophosphamide After second relapse: – Palliative pazopanib and radiotherapy treatment. Good response

At the time of LGMDR1 diagnosis, the patient was asymptomatic, apart from toe walking. The first symptoms of proximal muscle weakness with diminishing muscle strength appeared at the age of 13 years. Initially, he noticed performance changes in sport competitions, because of his shoulder girdle weakness. Later, he struggled to “keep up” with his peers while running or playing football, due to progressive weakness of the pelvic girdle. The patient had regular appointments with the neuromuscular team, reporting only episodes of symptomatic hypoglycemia during fasting, which was attributed to his MD. The muscle weakness progressed slowly, but the patient continued to be very active physically.

At the age of 17 years, the patient presented with a 4-week-long history of progressive pain in the right thigh, radiating to the knee and the hip, often becoming more severe at night. Diagnostic imaging including magnetic resonance imaging and positron emission tomography scan revealed a pathologic soft tissue mass also invading the skeleton (9×12×9 cm) in the right pelvis, located between the acetabulum, tuber ischiadicum, and ramus superior, with multiple skeletal metastases.

At primary tumor diagnosis, core biopsy material and fine-needle aspiration from the soft tissue component of the tumor in the right os ileum of the patient, performed under guidance of ultrasound, were collected in parallel. The fine-needle aspiration material was evaluated using May-Grünwald-Giemsa staining and the core biopsy with hematoxylin and eosin staining. The exchange was cellular with rounded, hyperchromatic, fragile cells with scant cytoplasm. Focally, there was a tendency of rosette formations. Apoptotic bodies and mitoses were seen (Figure 1A). Tumor cells were positive for Desmin (Figure 1B) and negative for CD99, NKX2-2, CD45, BCOR, Myogenin, MyoD1, and CKMNF. Results were confirmed by evaluation of core biopsies, which showed a malignant small round cell tumor. As part of routine diagnostic workup, reverse transcription polymerase chain reaction was performed, showing an EWSR1::WT1 gene fusion through a translocation (t(11;22)(p13;q12)). The combined findings led to a diagnosis of DSRCT.

Figure 1.

Diagnostic findings in the primary and relapse tumors. (A, B) Histology of core biopsies in the primary tumor. (A) Hematoxylin and eosin (HE) staining showing malignant small blue round cell tumor with hyperchromatic, immature, rounded cells with scant cytoplasm growing in a partly solid and partly trabecular pattern in a desmoplastic stroma. (B) Positive perinuclear cytoplasmic and focally dotlike Desmin staining. (C, D) Histology of open biopsies in the second relapse, scalp tumor. (C) HE staining showing a malignant small blue round cell tumor with oval to rounded cells with hyperchromatic nuclei and scant cytoplasm in a desmoplastic stroma. (D) Desmin immunostaining with perinuclear cytoplasmic pattern. (E) Visualization of RNA sequencing results from the soft tissue sarcoma, showing the EWSR1::WT1 fusion transcript. The upper part of the figure displays the fusion partners with their chromosomal localization (GRCh38) and orientation, while the lower part visualizes the predicted fusion transcript and the retained exons. All histology figures are presented at a magnification of 40×.

The metastatic disease and the complicated location of the primary tumor made radical surgery impossible. The patient was treated locally with definite radiotherapy to the pelvic tumor, and systemically with 6 courses of ifosfamide, vincristine, adriamycin, ifosfamide, vincristine, actinomycin-D and adriamycin, followed by 3 courses of actinomycin-D alone, with an interval of 3 weeks between courses, for a total of 25 weeks. The cumulative doses/m² for therapy were 19.5 mg vincristine, 7.5 mg actinomycin D, 54 g ifosfamide, and 320 mg doxorubicin. A good response was observed. After upfront treatment, the patient was put on maintenance therapy for 6 months, with cyclophosphamide and vinorelbine. The 28-day treatment cycle comprised intravenous vinorelbine (25 mg/m²) over 5–10 min on days 1, 8, and 15; and daily orally administrated cyclophosphamide (25 mg/m²).

One year after the initial DSRCT diagnosis, a relapse in his left hip was found, and the patient was started on a 6-month intravenous treatment with a cumulative dose/m² of topotecan (16 mg), etoposide (1200 mg), carboplatin (2400 mg), and cyclophosphamide (2000 mg). Partial response was observed, but eventually he developed a second relapse. Evaluation of the open biopsies from the second relapse tumor by hematoxylin and eosin staining showed a small blue round cell tumor with rounded cells, hyperchromatic nuclei, and scant cytoplasm in a desmoplastic stroma (Figure 1C). Immunohistochemical staining was positive for Desmin (Figure 1D) and vimentin (not shown), while pan cytokeratin was negative. Of note, the biopsies for the primary and relapsed tumors presented arti-factual changes, which made the acquisition of high-quality images difficult. However, the histological and immunohistological findings together with the molecular results confirmed the diagnosis of DSRCT.

Although the second relapse presented immunohistochemical NTRK overexpression (Pan-Trk antibody, not shown), it was judged after careful examination that treatment with NTRK inhibitor was not relevant in the absence of a fusion gene. Instead, the patient was put on palliative treatment with 600–800 mg per day of pazopanib, a protein kinase inhibitor, in combination with palliative radiotherapy to the scalp metastasis, resulting in a remarkable response, with good quality of life.

Tumor and Germline Whole Genome Sequencing Results

Upon diagnosis of the primary DSRCT, the patient was enrolled in the national Genomic Medicine Sweden pediatric cancer project, as part of which matched genomic DNA from peripheral blood and DNA and RNA from fresh frozen tumor tissue were extracted at Clinical Genomics, Stockholm. Whole genome sequencing was then performed via paired Illumina sequencing with 30× depth for germline DNA and 90× for tumor DNA, as previously described [24–26], and RNA sequencing with the Illumina stranded paired-end mRNA method, as previously described [24].

Manual inspection of germline whole genome sequencing results did not identify any pathogenic variants in 189 known childhood cancer predisposition genes [26]. RNA sequencing analyses of both the primary and relapsed tumor on the scalp confirmed the EWSR1::WT1 (t(11;22)(p13;q12) fusion transcript (Figure 1E). Furthermore, both the primary and relapsed tumors exhibited a near triploid karyotype and displayed multiple shared numerical and segmental chromosomal aberrations, indicative of a clonal relationship. Nonetheless, new rearrangements emerged in distinct chromosomal regions, suggesting genomic instability during tumor evolution (Figure 2). No potentially damaging single nucleotide variants or insertions/deletions were discovered in either tumor.

Figure 2.

Genomic profiles of the primary and relapse tumors. Copy-number profiles, ploidy, and tumor cell ratio from whole genome sequencing of tumor material generated by ASCAT [43]. For each figure, the upper panel shows the allele-specific copy number across the genome (copy number on y axis vs genomic location on the x axis). The allele with the lowest copy number is shown in green, while that with highest copy number, in red. For illustrative purposes, both lines are slightly shifted (red, down; green, up) such that they do not overlap. Only germline heterozygous probes are shown. In the middle and lower panels, the normalized log transform of read depth (LogR) and B-allele frequencies (BAF) values overlaid with segmented LogR and BAF across the genome are shown, respectively, representing the relative presence of each allele. (A) Near triploid desmoplastic small round-cell tumor showing multiple numerical and segmental chromosomal aberrations, including 1q gain (5 copies) and loss of heterozygosity in chromosomes 6 and 16. (B) The relapsed scalp tumor was also near triploid. Multiple aberrations were shared with the primary tumor, while new rearrangements emerged in specific chromosomal regions, such as 1p loss, 4q gain, loss of heterozygosity in chromosomes 11 and 13, and 15q rearrangements, which may suggest the involvement of genomic instability in the evolution of the tumor.

Discussion

We present the clinical and genetic characteristics of a patient with LGMDR1 who developed a DSRCT. To the best of our knowledge, this is the first reported case of DSRCT or any type of pediatric cancer in LGMD. The patient’s LGMDR1 was caused by compound heterozygous variants in the CAPN3 gene, namely NP_000061.1: p.(Thr184ArgfsTer36) and NP_000061.1:p.(Arg448Cys), while the tumor was diagnosed as a DSRCT with an EWSR1::WT1 (t(11;22)(p13;q12)) fusion transcript. We suggest a possible association between the development of DSRCT and LGMD. However, we cannot rule out the possibility that carcinogenesis in this patient was fortuitous.

The patient described in this case report developed DSRCT, a very rare STS [11]. The diagnosis was confirmed by positive dot-like Desmin staining, negative staining of the skeletal muscle markers Myogenin and MyoD1, and the presence of a EWSR1::WT1 fusion transcript, detected by RT-PCR in the primary tumor. In addition, RNA sequencing in the primary and relapse tumors confirmed the fusion transcript. Although the WT1 C-terminal antibody, generally used to diagnose DSRCT, was not available in our laboratory, 3 independent assays confirmed the chromosomal translocation t(11;22)(p13;q12), leading to the definitive diagnosis of DSRCT. Of note, a positive cytokeratin staining is usually expected in DSRCT [13]. However, the tumor in our patient presented a negative staining. Although the absence of pan cytokeratin expression is unusual, previous reports have been published in the literature of pan cytokeratin-negative DSRCT [27–29]. The detection of the EWSR1::WT1 fusion is therefore desirable for DSRCT diagnosis [28].

It is well known that DSRCT are highly aggressive pediatric tumors [10,30]. The severe clinical course of the DSRCT observed in our patient, including multiple metastasis at diagnosis and poor treatment response, is similar to what has been reported in other individuals with this tumor type [30,31]. Although the patient was treated with multimodal therapy, as suggested for DSRCTs [10], the response was poor, and the patient has now been put on palliative care. This is also in line with observations on other individuals with DSRCT, in which the prognosis is poor, with a 5-year survival rate below 25% [32,33]. Finally, as in most DSRCT diagnoses [11,30,31], our patient was young and male. All in all, the clinical course of the DSRCT in our patient resembled that of other individuals with DSRCT not diagnosed with LGMD.

Previous studies on MD mice models report an increased risk of mixed sarcomas in aged animals, including in LGMD [14–19]. CAPN3 knockout mice models develop mixed sarcomas with characteristics of rhabdomyo-, fibro-, and liposarcomas [14], whereas the patient presented in this report was diagnosed with a sarcoma of uncertain differentiation, according to the classification of tumors from the World Health Organization [13].

In addition, 7 patients with concomitant LGMD and cancer have been reported in the literature [20–22]. Apart from 1 single individual with liposarcoma [21], 3 patients with melanoma [20], and 3 with myeloma [22] and LGMD have been described. Sarcoma presentations in patients with Duchenne MD have also been described [34–40]. Finally, a recent Swedish population-based epidemiological study conducted by our group, in which 2355 patients with MD were included, showed an increased risk of pediatric central nervous system tumors and gliomas and adult pancreatic and non-thyroid endocrine tumors in individuals with MD. An increased risk of sarcoma was not observed. However, it was not possible to evaluate the cancer risk and risk spectrum in individuals with LGMD exclusively [41]. Thus, the cancer incidence in this specific patient group remains unclear and needs further exploration.

Multiple theories exist about the events that could possibly lead to cancer development in dystrophic muscle. Schmidt et al showed that there is genetic instability, including DNA damage, aneuploidies, and increased double strand break repair in muscles from mice models and patients with MD [14]. This instability was observed as early as in myoblasts, in line with recent results suggesting that in MD mice, mixed sarcomas arise from muscle stem cells [19]. It has been proposed that conflicting differentiation signals given to myoblasts in MD are permissive for tumor formation [42]. Chronic inflammation and a tumor suppressor role in some MD disease-causing genes are additional proposed mechanisms for carcinogenesis in MD [42]. Understanding the risk of DSRCT and cancer in general in individuals with LGMD could have important implications for patients, including surveillance and improved genetic counseling [26]. However, before changes in clinical practice for patients with LGMD can be considered, further studies are needed to better understand their DSRCT risk and to delineate their cancer risk spectrum.

Conclusions

We describe the clinical, genetic, and molecular characteristics of a patient with LGMDR1 who developed a DSRCT. Very limited information is available about the incidence of cancer in this patient group. Therefore, increased awareness, further case reports, and epidemiological studies are warranted to better understand the link between pediatric cancer and LGMD.