Identification of dual STRN-NTRK2 rearrangements in a high grade sarcoma, with good clinical response to first-line larotrectinib therapy

Abstract

Background

Among the three NTRK genes, NTRK2 possesses a tremendous structural complexity and involves tumorigenesis of several types of tumors. To date, only STRN and RBPMS are identified in the fusion with NTRK2 in adult soft tissue tumors. More recently, the highly selective Trk tyrosine kinases inhibitors, including larotrectinib and entrectinib, have shown significant efficacy for treating tumors harboring NTRK fusions and were approved by FDA.

Case presentation

We report a case of sarcoma in a 35-year-old female harboring two STRN-NTRK2 gene fusions, with a good clinical response to first-line larotrectinib treatment. Core biopsy of the 16.5 cm gluteal mass showed a high-grade mesenchymal neoplasm with features reminiscent of a solitary fibrous tumor, but negative for STAT6. In-house next-generation sequencing gene fusion panel showed two in-frame STRN-NTRK2 fusions, which contain the same 5’ partner sequence (exon 1–3) of STRN, and the 3’ fusion partner starting from either the exon 15 or the exon 16 of NTRK2. Due to the large size and location of the tumor, first-line neoadjuvant therapy with larotrectinib was initiated. The patient has an excellent clinical response with an 83% tumor size reduction by imaging. The tumor was subsequently completely resected. After 130 days, larotrectinib was reinitiated for lung metastasis (up to 7 cm), and a complete resolution was achieved. When compared with NTRK1 and NTRK3, NTRK2 fusions are the least common. Of note, the only other report in the literature on NRTK2 fusion-positive sarcoma also showed solitary fibrous tumor (SFT)-like morphology, and the patient responded well to larotrectinib as the second line adjuvant therapy.

Conclusions

In conclusion, the identification of NTRK2 fusions in patients with soft tissue tumors could significantly improve the clinical outcome through selective NTRK inhibitor therapy, especially in the first-line setting. Prompt RNA-based NGS testing at initial diagnosis may benefit these patients. Our case is among the first few in the literature on NTRK2 fusion sarcoma with first-line larotrectinib therapy in the primary and metastatic setting, with good clinical response and minimal side effects.

Background

TrkA, TrkB, and TrkC are receptor tyrosine kinases encoded by NTRK1-3 genes. They are activated by binding with neurotrophins [1,2,3]. The Trk proteins share highly homologous sequence and structural features, including an extracellular region, a transmembrane region, and an intracellular tyrosine kinase domain, which upon activation, transduces the downstream signaling via MAPK, PI3K, and PKC pathways [4]. Intrachromosomal or interchromosomal NTRK gene fusion is the most noticeable underlying mechanism of oncogenesis. The fusion protein contains in-frame N-terminal amino acids from the fusion partner and the C-terminal amino acids containing the tyrosine kinase domain of the Trk receptor, resulting in a constitutively active chimeric kinase [1,2,3,4,5,6,7]. More than 80 genes have been identified as partners in the fusion to NTRK genes [8].

Among the three NTRK genes, NTRK2 possesses a tremendous structural complexity and involves oncogenesis of several types of tumors. To date, there are approximately 42 partner genes involving NTRK2 rearrangement [2, 3, 5, 8,9,10,11,12], and only STRN and RBPMS are identified in the fusion with NTRK2 in adult soft tissue tumors [3] (Table 1).

Table 1 Review of reported NTRK2 fusion tumors with different partner genes

More recently, the highly selective Trk tyrosine kinases inhibitors, including larotrectinib and entrectinib, have shown significant efficacy for treating tumors harboring NTRK fusions and were approved by FDA [2, 7, 13, 14].

Herein, we report the identification of NTRK2 gene rearrangement by initial RNA-based NGS, with the efficacy of larotrectinib treatment in an adult patient with a sarcoma harboring STRN-NTRK2 fusions. The results demonstrate the importance of the identification of NTRK alteration in soft tissue sarcoma, and the urgent need for mechanistic study for resistance to targeted therapy in NTRK2 fusion sarcoma.

Case presentation

A 35-year-old woman presented to our oncology clinic for a second opinion for a right gluteal sarcoma initially diagnosed outside the continental United States. She complained of nausea, fever, and worsening pain and numbness form the affected area. The mass had been increasing in size. MRI showed a large, heterogeneous, T2 hyperintense, vascular, and partially necrotic and enhancing mass, centered in the right gluteus medius and maximus muscles, with extension into the right paraspinal musculature, and invasion of the right posterior iliac bone. The initial MRI (50 days ago) showed the mass measuring 10.2 × 9.7 × 7.0 cm (Fig. 1B), and measured 16.5 × 12.9 × 10.4 cm (AP x TV x CC) (Fig. 1C) at first encounter at our institution. PET scan showed no metastasis.

Fig. 1

Clinical timeline (A) and imaging (B-E). Before treatment (B: - Day 50 & C: Day 0), a heterogenous mass (yellow arrows in B-E) centers in the right gluteus medius and maximus muscles with extension into the right paraspinal musculature, and invasion of the right posterior iliac bone. Imaging on post-treatment Day 60 (D) and Day 170 (E) show significant decrease in tumor size (MRI#4 of Day 80 is not shown here). 130 days after surgery, a 6.9 cm solid mass in the lower lobe of left lung is biopsied and proven to be a metastatic lesion (F, Day 320, blue arrow with circle), with near complete resolution after the resumption of Larotretinib treatment (G: Day 455, blue arrow with circle & H: Day 560)

A core biopsy was performed and showed a highly cellular mesenchymal neoplasm consisting of round/ovoid cells, with eosinophilic cytoplasm, round to ovoid nuclei with mostly inconspicuous nucleoli, and no significant nuclear pleomorphism (Fig. 2A and B). The tumor was rich in vasculature, with focal staghorn-type vessels (Fig. 2A). Brisk mitotic activity (21 per 10 high-power fields) and tumor necrosis were identified (Fig. 2C). Histologically, the tumor was reminiscent of a malignant solitary fibrous tumor (SFT). It was diffusely positive for CD34 (Fig. 2D), but negative for STAT6 (data not shown). The tumor was also positive for FLI1 (patchy), TLE1 (patchy), and CD99 (focal and weak), and was negative for AE1/AE3, S100, SOX10, ASMA, desmin, ERG, and CD31. LCA stain highlighted the intermixed inflammatory cells (data not shown).

Fig. 2

Pathology of the core biopsy. H & E sections show a highly cellular mesenchymal neoplasm of round/ovoid cells with eosinophilic cytoplasm, round to ovoid nuclei and mostly inconspicuous nucleoli without significant nuclear pleomorphism. The tumor is rich in the vasculature with staghorn-type vessels (A & B), and brisk mitotic activity (C, arrows). Tumor is positive for CD34 (D). Magnifications: A, 100X; B-D, 400X

In-house RNA-based NGS gene fusion panel analysis (Archer FusionPlex™ Comprehensive Thyroid and Lung Panel, 18 gene fusion panel) was performed, and two concurrent STRN-NTRK2 fusions were identified. The NGS data showed that both fusions were in-frame and located at the exon-exon boundary. Both fusion RNAs had the same 5’ partner sequence (exon 1–3) of the STRN gene (chr2, NM_003162.3, breakpoint: 37,143,221, Fig. 3A and B). One of the 3’ fusion sequence started from the exon 15 of NTRK2 (chr9, NM_006180.4, breakpoint: 87,475,955, Fig. 3A), and the other 3’ fusion sequence began from exon 16 of NTRK2 (chr9, NM_006180.4, breakpoint: 87,482,158, Fig. 3B). Of note, the fusion product containing exon 1–3 of STRN and exon 16–21 of NTRK2 was the same as the one which was previously identified as an undifferentiated sarcoma in a pediatric patient [14]. The second fusion pattern starting from exon 15 of the NTRK2 had been identified in a ganglioglioma harboring TLE4-NTRK2 fusion.

Fig. 3

Molecular characterization of tumor sample by NGS. Two in-frame STRN-NTRK2 fusions with same 5’ partner sequence (exons 1–3) of STRN gene identified by NGS.

Due to the location and extent of the tumor, it was deemed unresectable or would require a morbid upfront surgery, i.e., hemipelvectomy. Given the NGS findings, multidisciplinary tumor board was held, and the patient was started on the selective NTRK inhibitor, Larotrectinib, at 100 mg, BID, as the first-line therapy. The patient noticed a quick amelioration of tumor-related pain and was able to discontinue all pain medicines (previously on 20 mg oxycontin plus 10 mg oxycodone immediate release every 6 h as needed). The patient also reported a significant shrinkage of the tumor after the initial 7-day treatment, and it continued to improve in the following 45 days. On post-treatment day 50, MRI was repeated and showed that the tumor had significantly decreased in size to 6.6 × 7.7 × 7.4 cm (Fig. 1D), approximately 83% reduction compared with the tumor size before treatment. The previously noted focus of enhancement in the right paraspinal musculature was not present in this imaging.

The patient was followed up closely, and MRI was repeated on post-treatment day 90 and day 170. The size of tumor remained stable during this interval (Fig. 1E, day 90: 6.3 × 7.4 × 5.8 cm; day 170: 6.0 × 7.1 × 5.1 cm). During the later period of treatment, the patient started to feel a few nodules around the original tumor expanding with new significant pain. Metastatic workup showed no metastases, and on post-treatment day 190, the patient underwent radical resection of right gluteal soft tissue sarcoma and right ilium, which was uneventful.

The resection specimen revealed a 6.1 × 7.5 × 3.6 cm ill-defined, heterogeneous, fleshy, solid and necrotic mass with a tan-pink cut surface. Histologically, the tumor showed high-grade undifferentiated sarcoma with treatment effects (granulation tissue, myxoid change, fibrosis, foamy macrophages, and necrosis (50% of tumor volume)). It focally invaded into the ilium. Resection margins are free of tumor.

After wound healing was completed, the patient was started on adjuvant radiation therapy (a total of 5000 cgy) given this being a high grade tumor > 10 cm. Larotrectinib was held during adjuvant radiation due to a lack of data on concurrent radiation therapy. Within six weeks after completion of radiation, while still off larotrectenib, patient presented with chest pain and shortness of breath, and chest tightness. Imaging studies showed multiple new large lung nodules. A biopsy of one of the nodules showed metastatic sarcoma. The in-house NGS gene fusion panel (Archer FusionPlex™ Pan Solid Tumor v2 panel, ArcherDX/Invitae, 99 gene panel) showed persistent NTRK2 fusions. The patient was then restarted on palliative larotrectenib, and the symptoms improved within a few days. CAT scan repeated 60 days after initiation of palliative larotrectenib showed complete resolution of her lung metastases. The patient continued larotrectinib with response for 22 months before developing symptomatic right parietal lobe and leptomeningeal disease, and biopsy showed metastatic sarcoma with fusion panel again positive for NTRK2-STRN fusions. Imaging studies to evaluate status of disease showed persistent extracranial response to larotrectenib. Patient was started on compassionate use of second generation NTRK inhibitor PBI-200 with known blood brain barrier penetration. The brain metastasis biopsy was sent for outside laboratory NGS to look for mutations. However, none of the known on-target resistant mutations of TRKB, including solvent front mutations (e.g. TRKB^G639R), gate keeper residue mutation (TRKB^F633L), xDFG motif mutation (TRKB^G667C) [15], and none of the known off-target mutations (e.g., MET amplification, BRAF^V600E mutation, and KARS mutations) were identified in this case [15]. Patient had disease control for 3 months on the second generation NTRK inhibitor with eventual worsening of intracranial disease and succumbed to her disease.

Discussion and conclusions

NTRK fusions are identified in 0.31% of adult tumors and 0.34% of pediatric tumors. The most common gene was NTRK3 (0.16% of adult tumors), followed by NTRK1 (0.14% of pediatric tumors), with NTRK2 being the least common (0.06% of adult tumors) [5]. The most commonly seen tumor types driven by NTRK gene fusion are mammary analog secretory carcinoma, secretory breast carcinoma, and infantile fibrosarcoma (ETV6-NTRK3 fusions, 70–91%) [5]. Other tumor types with very low incidence (< 2%) include thyroid cancer, breast carcinoma, non-small cell lung cancer, colorectal cancer, melanoma, brain tumors, and sarcoma [5].

Among all the tumors harboring NTRK rearrangements, soft tissue tumors are attracting more diagnosis-oriented attention due to the equivocal histologic findings and available targeted therapy. Selective Trk inhibitors such as larotrectinib and entrectinib emerged as the treatment for tumors harboring NTRK fusions with FDA approval [2, 7, 13, 14]. However, the complexity of the NTRK2 gene structure, the diversity of fusion partners, the rarity of the clinical cases, and the limited availability of the detection modalities make the diagnosis and treatment more difficult.

For detection of NTRK fusions, several methods are available: immunohistochemistry (IHC, pan-TRK antibody), fluorescence in situ hybridization (FISH), DNA-based NGS, and RNA-based NGS sequencing [16]. In the largest study to date (33,997 cancer cases), DNA-based sequencing showed an overall sensitivity of 81.1% (60/74 cases) and specificity of 99.9% for detection of NTRK fusions, when compared to RNA-based sequencing [17]. More specifically, the sensitivity for DNA-based NGS to detect NTRK1 fusion is 96.8% (30/31 cases), 0% for NTRK2 fusion (0/4 cases), and 76.9% for NTRK3 fusions (30/39 cases). For IHC, an overall sensitivity is 87.9%, and specificity is only 81.1% [17]. In fact, for our patient, multiple specimens were also sent to outside commercial laboratories for comprehensive sequencing, and the NTRK2 fusions were missed twice by two different commercial laboratories using DNA-based NGS methods. Both of the labs repeated the testing using RNA-based NGS, and detected the NTRK2 fusion. Based on ours and others’ experiences, RNA-based sequencing appears to be the optimal way to identify NTRK fusions, especially NTRK2 fusions, because the splicing out of introns simplifies the technical requirements of adequate coverage, and detection of RNA-level fusions provides direct evidence of functional transcription, therefore, should be the test of choice when possible.

The results of STRN-NTRK2 fusion from our case show some common as well as unique features when compared with previously reported cases. First, most of these fusion proteins contain the C terminal fragments translated from either exon 15 or exon 16 of NTRK. The breakpoint of STRN is a recurrent site in various tumors. The product from STRN contributes a coiled-coil domain for potential oligomerization. These findings lead to the recognition of these breakpoints as hot spots as well as a common fusion pattern of these two genes in soft tissue tumors. Second, the incidence of STRN-NTRK2 fusion might be underestimated in soft tissue tumors if the RNA-based NGS panel is not used in the clinical setting. Third, the tumor in our case showed SFT-like histology, and NTRK tumors should enter the differential diagnosis of SFT-like tumors with negative STAT6. Finally, the initial and prolonged good clinical response to TRK inhibitor therapy (primary and extracranial metastases), and the ultimate resistance in cranial disease highlighted the urgency for mechanistic study as well as development of newer and better drugs.

In conclusion, the identification of NTRK2 fusions in patients with soft tissue tumors could significantly improve the clinical outcome through selective NTRK inhibitor therapy, especially in the first-line setting. Larotrectenib is very effective in treating either primary or metastatic tumors harboring NTRK rearrangement. Due to the lack of specific morphologic pattern and IHC profile, prompt RNA-based NGS testing at initial diagnosis may benefit these patients. However, with the use of TRK inhibitors for about 3 years now, we have started to see development of resistance in these patients to first-line TRK inhibitors, and research is urgently needed to elucidate mechanism of resistance, especially in the NTRK2 fusion tumors, to develop better 2nd and 3rd generation NTRK inhibitors and alternative treatment regimens.