Soft‐tissue sarcomas are a group of mesenchymal tumors with limited treatment options. Building on reports of varied rates of alterations in genes associated with the homologous recombination pathway, this article characterizes the frequency of homologous recombination DNA repair pathway alterations in soft‐tissue sarcoma subtypes and highlights the unique nature of leiomyosarcoma.
Keywords: Homologous recombination, DNA repair, Sarcoma, Poly (ADP‐ribose) polymerase inhibitors
Abstract
Background.
Soft‐tissue sarcomas (STS) describe a heterogeneous group of mesenchymal tumors with limited treatment options. Targeted therapies exist for BRCA1/2 gene alterations, but their prevalence and role have not been fully described in STS. Here, we present the largest effort to characterize the frequency of homologous recombination (HR) DNA repair pathway alterations in STS subtypes and highlight the unique nature of leiomyosarcoma (LMS).
Materials and Methods.
DNA sequencing data were analyzed for HR pathway alterations for 1,236 patients with STS. DNA sequencing data from an additional 1,312 patients were used to confirm the prevalence of HR pathway alterations in LMS. Four uterine LMS (uLMS) patients with functional BRCA2 loss were evaluated for response to poly (ADP‐ribose) polymerase (PARP) inhibition.
Results.
In an unselected STS study population, BRCA2 alterations were identified in 15 (1%) patients, and homozygous BRCA2 loss was detected in 9 (<1%). However, subset analysis revealed that these BRCA2 alterations were concentrated in uLMS as compared with any other STS subtype. Notably, 10% of uLMS tumors had a BRCA2 alteration. We further report that PARP inhibitors had demonstrated durable clinical benefit in four uLMS patients with BRCA2 loss.
Conclusion.
HR pathway alterations are rare in most STS. However, we identify uLMS to be enriched for BRCA2 loss and report the positive outcomes of a series of patients treated with PARP inhibitors. Our data suggest that patients with uLMS should be considered for somatic BRCA2 profiling. Prospective trials are necessary to confirm the efficacy of PARP inhibition in uLMS.
Implications for Practice.
Soft‐tissue sarcomas are a highly morbid, diverse set of tumors with limited treatment options. This study identifies an increased prevalence of functional BRCA1/2 loss in patients with uterine leiomyosarcoma (uLMS). It also presents four patients with uLMS and BRCA2 loss who achieved durable clinical benefit from poly (ADP‐ribose) polymerase inhibition. These data suggest that patients with uLMS in particular should be screened for BRCA1/2 alterations and may benefit from treatment targeted to these alterations.
Introduction
Soft‐tissue sarcomas (STS) are molecularly heterogeneous mesenchymal tumors that constitute 1% of all adult malignancies [1]. Patients with advanced STS have a dismal prognosis; thus, the identification of targetable pathway alterations is critical to altering the course of these morbid diseases. A dominant subtype of STS is leiomyosarcoma (LMS), which is itself heterogeneous with derivative tissues including uterine, cutaneous, and blood vessel origin [2], [3], [4], [5], [6], [7], [8].
Previous studies in STS have reported a varied rate of alterations in genes associated with the homologous recombination (HR) pathway including the breast cancer 1 and breast cancer 2 genes (BRCA1 and BRCA2; BRCA1/2; supplemental online Table 1) [9], [10], [11], [12]. In breast and ovarian cancers, deleterious alterations in BRCA1/2 result in reduced double‐stranded DNA‐break repair capacity via the HR pathway, which sensitizes them to treatment with poly (ADP‐ribose) polymerase (PARP) inhibitors [13], [14], [15]. To this end, we report on a series of patients in our practice with functional BRCA2 losses who received clinical benefit from PARP inhibition after failing multiple lines of therapy.
Table 1. HR pathway alterations in soft‐tissue sarcomas.
Data are presented as n (%).
Abbreviations: HR, homologous recombination; LMS, leiomyosarcoma; LPS, liposarcoma.
Based on these findings, we set out to determine the prevalence of HR pathway alterations in STS. To better characterize the prevalence of pathogenic alterations in the HR pathway in STS, we aggregated 1,236 patients diagnosed with STS from three cohorts: The Cancer Genome Atlas (TCGA), Genomics Evidence Neoplasia Information Exchange (GENIE), and our internal patient cohort (OSUCCC). Clinical data were used as available to identify trends in patients with genomic variation in the HR pathway. To confirm gene alterations in LMS, we obtained sequencing data from 1,312 LMS patients from the Foundation Medicine (FMI) databank. Finally, we report that BRCA1/2 alterations are particularly prevalent in uterine LMS (uLMS).
Materials and Methods
Genomic alteration data were aggregated as listed below. Data for genes associated with the HR pathway (supplemental online Table 2) were queried and clinical data were collected as available [14]. Gene alterations were defined as follows: genomic loss referred to homozygous deletion (−2 in GENIE; ≤0.3 in TCGA); genomic sequence alteration referred to any pathogenic sequence alteration in a gene including single nucleotide polymorphisms, frameshifts, and splice site variants.
Table 2. Demographics for clinical analysis.
Abbreviations: OSU, internal patient cohort; TCGA, The Cancer Genome Atlas.
Data Sets
TCGA data were downloaded from the Genomic Data Commons (https://gdc.cancer.gov/) portal using a Binary Alignment/Map (BAM) slicing application programming interface and custom Python download scripts, and was limited to samples in the STS cohort. Hg38 was used as the reference genome. The BAM files were converted to mpileup format using Samtools (version 1.3.1); with a minimum read mapping quality of 20. Matched normal and tumor mpileup pairs were run on VarScan2 (version 2.4.1). The VarScan2 genomic alteration data were annotated using ANNOVAR (version date 2016‐02‐01). Copy number data were assessed using the mean copy number over the selected genomic region.
GENIE data were downloaded from Sage Bionetworks (release 1.0.01; https://synapse.org/genie). Samples identified as sarcoma subtypes by their Oncotree code (http://oncotree.mskcc.org/) included 583 distinct patients with both copy number and genomic alteration data available. Losses were identified based on provided copy number levels, corresponding to −1 (shallow or heterozygous loss) and − 2 (deep or homozygous loss). Genomic alterations classified as missense, nonsense, nonstop, or frameshift genomic alteration were assessed.
OSUCCC data were obtained by chart review from 398 consecutive STS patients who had genomic sequencing and clinical data available (December 2012 to July 2017, Institutional Review Board Approval: 2016C0113). All patients were genomically profiled using the FoundationOne Heme platform [16]. Genomic results were taken directly from the patient reports.
The FMI LMS validation data set contained 1,312 unique LMS patients assayed as previously described [16] from their internal database. Limited pathologic subtyping were included with these samples, which precluded their use in specifically validating the uLMS subset.
Calling of Genomic Alteration Pathogenicity
Pathogenicity of the genomic alterations were arbitrated using ClinVar's measure of clinical significance [17]. Only genomic alterations considered pathogenic or likely pathogenic were selected for further study. A summary of all detected genomic sequence alterations and their pathogenicity as determined by ClinVar is reported in supplemental online Table 3.
Table 3. Clinical outcomes for leiomyosarcoma patients receiving poly (ADP‐ribose) polymerase inhibitors.
Olaparib was used in combination with letrozole. Trabectedin was added to this regimen after 5 months.
Patient with a past medical history of total hysterectomy initially diagnosed with small‐bowel leiomyosarcoma. Strong estrogen receptor/progesterone receptor staining suggest uterine in origin.
Abbreviations: PR, partial response; SD, stable disease; uLMS, uterine leiomyosarcoma.
Statistical Analysis
Data were analyzed using R (version 3.4.1) in RStudio (version 1.0.153). Comparison of the frequency of genomic alteration was evaluated using Pearson's chi‐squared test. Continuous data were evaluated using the Kruskal‐Wallis test.
Patient Case Series
Retrospective data from all available LMS patients who had a functional loss in BRCA2 and who received PARP inhibitors at the OSUCCC and Johns Hopkins institutions were obtained. Olaparib (AstraZeneca) was obtained off label.
Results
Overview of Analysis
We aggregated data from 1,236 patients from three data sets (All STS; Fig. 1). Here we define HR pathway genes as those to be previously defined per the ARIEL3 trial of the PARP inhibitor rucaparib [14]. After identifying an enriched population exhibiting an increased prevalence of HR pathway alterations in the LMS subtype of STS, we selected 170 patients from the TCGA and OSUCCC databases with LMS for whom detailed pathology was available for additional analysis (LMS Subset; Fig. 1). To confirm that this cohort reflected the HR pathway alterations of LMS as a whole, we obtained genomic data from 1,312 patients from FMI (LMS Validation Set; Fig. 1). A schematic detailing the data sets used in this analysis can be found in Figure 1.
Figure 1.
Our initial analysis of the frequency of BRCA variants in STS included three data sets: GENIE, TCGA, and our internal data set. To confirm the prevalence of HR pathway alterations in LMS, we obtained sequencing data from FMI for all patients in their database who were diagnosed with LMS. After assessing the frequency of BRCA variants in STS, we found that BRCA2 loss was most common in the LMS subtype of STS. To assess the clinical phenotype of BRCA variation, we combined the clinical data from the TCGA as well as our internal data set. Schematic of data sets used in our analysis.
Abbreviations: FMI, Foundation Medicine; GENIE, Genomics Evidence Neoplasia Information Exchange; HR, homologous recombination; LMS, leiomyosarcoma; OSUCCC, internal patient cohort; STS, soft‐tissue sarcomas; TCGA, The Cancer Genome Atlas.
HR Pathway Alterations Are Rare in Unselected STS
In 1,236 STS patients in the All STS Set (Fig. 1), the overall frequency of gene alterations in the HR pathway was 11%. This percentage was not different among individual data sets (GENIE 11%, OSUCCC 13%, TCGA 11%; p = .6). However, loss of non‐BRCA genes in the HR pathway data set were different as outlined in supplemental online Figure 1 (GENIE 3%, OSUCCC 1%, TCGA 5%; p = .02). BRCA1/2 alteration incidence was not significant across the data sets (GENIE 1%, OSUCC 2%, TCGA 2%; p = .76).
LMS Had Nearly Twice the Number of HR Alterations as Other STS
To determine whether HR alterations were driven by particular STS subtypes, we subclassified our patients into the two common subtypes of STS, LMS and liposarcoma (LPS), and included any other subtype of STS in an “All Other Subtypes” classification. Notably, LMS had twice the number of HR alterations compared with our other classifications (LMS 19%, LPS 10%, All other 9%; p = .19; Table 1).
One hundred twenty‐five patients were identified to have an alteration in a non‐BRCA gene associated with the HR pathway. We divided this non‐BRCA HR group into gene losses (32 patients) and sequence alterations (93 patients). Although the prevalence of loss of non‐BRCA genes in the HR pathway did not vary significantly between specific subtype of STS (LMS 3%, LPS 4%, other 2%; p = .74), pathogenic sequence alterations were statistically more common in the LMS subtype (LMS 13%, LPS 6%, other 6%, p < .001). In this data set, five patients were found to have alterations in BRCA1 (two losses, three sequence alterations), four of whom were in the LMS subset. BRCA2 alterations were detected in 13 patients. Loss of BRCA2 was most common within the LMS subtype (LMS 3%, LPS 0%, other <1%; p < .001); however, BRCA2 sequence alteration rates did not differ significantly among subtypes (LMS <1%, LPS <1%, other <1%; p = .92).
FMI LMS Validation Data Set Confirms LMS HR Alterations Rates
To confirm that our LMS patients were representative of LMS as a whole, we compared them with genomic data from 1,312 LMS patients from the FMI database (Fig. 1; supplemental online Table 4). Of note, the diagnosis of LMS was based on data provided to FMI and not independently validated. Sixty‐five patients (5%) from the OSUCCC cohort were included in the FMI cohort. Patients in this larger confirmatory FMI cohort exhibited similar rates of BRCA1/2 alterations (FMI 5% vs. 5%; p = .99). Of the 30 genes that were profiled and assessed individually, only ATRX differed significantly. The patients in the FMI cohort exhibited higher rates of both ATRX loss (FMI 10%, vs. 1%; p < .001) and sequence alteration (FMI 26% vs. 8%; p < .001).
BRCA1/2 Alterations Are Most Commonly Found in uLMS
Based on the prevalence of HR pathway variants in LMS, we further analyzed the 170 LMS patients from the TCGA and OSUCCC cohorts (Fig. 1). Data from GENIE and FMI were excluded as they did not include detailed clinicopathologic data to allow reliable subtyping.
These 170 LMS patients (65 from OSUCCC; 105 for the TCGA) contained 61 uLMS and 109 nonuterine LMS. Demographic data for these patients are provided in Table 2. In this data set, 21 non‐BRCA HR genomic alterations were identified (3 loss, 18 sequence alterations). No patients in this cohort exhibited sequence alterations in BRCA1/2. Importantly, the uLMS constituted a statistically significant majority of the BRCA1/2 loss (six of seven patients; supplemental online Table 5). A single annotated patient had the only BRCA1 loss identified in this data set, as well as a BRCA2 loss. Overall, 3% of uLMS patients had a loss of a non‐BRCA HR gene as compared with 1% in the nonuterine LMS cohort, and 10% of uLMS patients exhibited a pathogenic genomic sequence alteration in a non‐BRCA HR gene as compared with 13% in the nonuterine LMS cohort (Fig. 2).
Figure 2.
Uterine pathology in LMS by HR pathway status. BRCA1/2 loss was more common in uterine LMS compared with nonuterine LMS.
Abbreviations: HR, homologous recombination; LMS, leiomyosarcoma; uLMS, uterine leiomyosarcoma.
BRCA1/2 Alterations Trend to a Poorly Differentiated Phenotype
We also assessed histologic characteristics as well as mitotic counts in LMS patients. No HR pathway alteration correlated with a statistically significant difference in histologic subtype or mitotic count. However, LMS patients with BRCA1/2 loss trended toward a more poorly differentiated histologic subtype (BRCA1/2 loss 57%, HR intact 36%, p = .56) as well as an increased mitotic count (BRCA1/2 loss 40.2 ± 34.8, HR intact 16.8 ± 17.7, p = .09).
Effect of PARP Inhibitor Therapy in LMS Patients Harboring a BRCA2 Loss
In patients with BRCA1/2 loss, PARP‐directed therapy may be clinically beneficial. Here, we report the clinical course of available LMS patients at two institutions harboring BRCA2 genomic alterations who were treated with PARP inhibitors (Table 3).
Patient 1.
A 58‐year‐old female diagnosed with uLMS was initially treated surgically for a 1‐year interval before her first progression. She was treated with nine cycles of gemcitabine and docetaxel before progression. Further treatment included pelvic mass resection, two cycles of radiation, and single‐agent cisplatin. An additional resection included portions of the small bowel followed by AIM [doxorubicin, ifosfamide, mesna] for three cycles, which was then discontinued because of poor tolerance. The patient then developed extensive bone metastasis. Based on a loss of BRCA2 in the patient's tumor, olaparib treatment was initiated. After 3 months of therapy with olaparib, these lesions became markedly less avid on positron emission tomography/computed tomography imaging. The patient has exhibited ongoing stable disease (SD) on olaparib for 17 months.
Patient 2.
A 65‐year‐old female diagnosed with poorly differentiated, high‐grade uLMS with metastases to the bone and lungs upon presentation was initially treated with gemcitabine and docetaxel for three cycles. She then progressed through two cycles of doxorubicin and 4 months of pazopanib with a new metastasis to the brain, followed by nine cycles of trabectedin before further tumor progression. Based on loss of BRCA2 in the patient's tumor, olaparib was given for 5 months before a subclinical progression, an increase in lesion size less than progression as defined by RECIST v1.1, prompted a change in this patient's regimen. Based on this patient's history of good response to trabectedin, promising preliminary data for the combination of olaparib and trabectedin in STS [15], [18], and a lack of therapeutic options available for this patient, trabectedin was added to olaparib therapy. This regimen stabilized this patient's disease for 15 months before progression.
Patient 3.
A 72‐year‐old female with a past medical history significant for total hysterectomy and bowel resection was diagnosed with high‐grade LMS of the small intestine. The tumor was surgically resected and recurred 3 years later. Later staining of the tumor exhibited highly positive estrogen and progesterone receptor status (>75%) suggesting a uterine origin [19]. Because of comorbidities, intensive chemotherapy was not an option. Systemic treatment included pazopanib for 1 year with the addition of everolimus for an additional year because of subclinical tumor progression. After continued progression, the patient was initiated on olaparib based on a BRCA2 loss in the patient's tumor. She is now 16 months with SD and continues on therapy.
Patient 4.
A 54‐year‐old female diagnosed with high‐grade uLMS initially underwent a total abdominal hysterectomy and bilateral salpingo‐oophrectomy with identification of bladder and peritoneal implants at the time of resection. She received adjuvant gemcitabine and docetaxel in the context of recurrent disease for one cycle. This was followed by three cycles of ifosfamide and etoposide (IE) with concurrent radiation to the pelvis, given the appearance of further metastatic tumors. She underwent an additional resection followed by three additional cycles of IE. Sequential local therapies included video‐assisted thoracoscopic surgery to right and left lung, radiation to bilateral lung and adrenal lesions, multiple rounds of cryoablation, and radiation to pancreatic metastases. Unfortunately, the fiducial placement for the pancreatic radiation likely caused local spread of the uLMS to stomach lumen. Given the presence of a pathogenic BRCA2 sequence and patient preference, olaparib therapy was initiated. Ten weeks after initiation of olaparib, the gastric mass has decreased by 62%. No new tumors have appeared over this period, and the patient remains stable 16 weeks after initiating therapy.
Discussion
STS encompass a variety of pathologic tissue types with limited treatment options available [20]. Indeed, alterations in the HR pathway have previously been noted to be more frequent in uLMS [9], [10], [11], [12], and this lends credence that uLMS is a distinct subgroup from nonuterine LMS [6], [12], [21]. The frequency of these alterations, as well as their potential targetability with PARP inhibitors, has yet to be confirmed in a large patient cohort. To our knowledge, this study represents the largest characterization of HR pathway alterations in STS and LMS to date.
In an unselected STS population, HR alteration rates are low. Previous studies have reported variable frequencies of HR pathway alteration rates in STS [9], [10], [11], [12]. One potential cause of variability may be due to methods used to call pathogenicity. In this study, we focused on genomic sequence alterations judged to be clinically pathogenic by ClinVar that considers sequence alterations that have been vetted by a panel of experts [17]. By taking this conservative approach, we hoped to reduce the noise of benign genomic sequence alterations, and this may explain the lower overall prevalence of HR alterations in this analysis. Subsequent analysis using COSMIC as a reference for pathogenic mutations showed an increase in overall frequency of mutations but no significant change in their distribution (supplemental online Figure 2; supplemental online Table 6) [22].
LMS exhibits a significantly higher rate of HR pathway alterations compared with other subtypes of STS. HR pathway alteration rates in this study were confirmed by the FMI patient database. The FMI data set exhibited similar prevalence of HR pathway alterations with the exception of higher rates of ATRX alterations. This may have been due to the conservative nature of our sequence calling method using ClinVar, as well as our conservative cutoff to call homozygous loss. It is likely that our methods underestimate the true prevalence of ATRX alterations but thus may provide a lower bound of the incidence rate.
We report that approximately 10% of the patients with uLMS will have a functional BRCA2 loss—comparable to that in ovarian and breast cancers [23], [24]. Although previous studies suggest a correlation between HR pathway alterations and cancers of estrogen‐driven tissues [25], these data suggest that this correlation exists in STS. Additionally, a previous case‐control study by Ballinger et al. described an increased prevalence of BRCA1/2 alterations in patients who developed STS compared with healthy controls [26]. It is therefore unsurprising that there may be a subpopulation of STS driven by BRCA1/2 alterations, as proposed by this analysis. Moreover, LMS tumors with loss of BRCA1/2 trended toward being more pleomorphic and containing a higher mitotic count, reflecting previous studies suggesting that tumors with BRCA1/2 loss tend to be more neoplastic, rapidly dividing tumors [27].
Finally, as evidenced by our case series, PARP inhibition may represent a viable option for STS patients with functional BRCA1/2 loss. Prospective clinical trials are needed to better define and confirm the utility of PARP inhibition in this population. Currently, there are two clinical trials recruiting adult patients with any advanced solid malignancy harboring an HR pathway alteration (supplemental online Table 7). Expansion of studies of this type, or opening of trials specific to STS, would improve our understanding of the clinical actionability of these genomic alterations.
Conclusion
This is the largest study to specifically assess the frequency of HR pathway alterations in STS. We find that uLMS is associated with a high frequency of loss of genes in the HR pathway, specifically with BRCA1/2 alterations occurring in 10% of patients. Based on these preliminary data, we suggest that testing for BRCA2 loss in uLMS patients may aid in identifying targeted therapies. Prospective clinical trials will be needed to ascertain the response to this intervention.
See http://www.TheOncologist.com for supplemental material available online.
Footnotes
For Further Reading: Rebecca C. Arend, Michael D. Toboni, Allison M. Montgomery et al. Systemic Treatment of Metastatic/Recurrent Uterine Leiomyosarcoma: A Changing Paradigm. The Oncologist 2018;23:1533–1545.
Implications for Practice: Recent regulatory approval of novel treatment options has significantly expanded the therapeutic armamentarium, and the addition of these therapies has challenged clinicians to select and optimally sequence these compounds. This review will focus on examination of current guidelines and consensus building for optimal sequencing of systemic therapies for advanced or recurrent uterine leoimyosarcoma.
Author Contributions
Conception/design: Nathan D. Seligson, Esko A. Kautto, John L. Hays, James L. Chen
Provision of study material or patients: Nathan D. Seligson, Esko A. Kautto, Edward N. Passen, Colin Stets, Amanda E. Toland, Sherri Z. Millis, Christian F. Meyer, James L. Chen
Collection and/or assembly of data: Nathan D. Seligson, Esko A. Kautto, Sherri Z. Millis, Christian F. Meyer, James L. Chen
Data analysis and interpretation: Nathan D. Seligson, Esko A. Kautto, Amanda E. Toland, Sherri Z. Millis, John L. Hays, James L. Chen
Manuscript writing: Nathan D. Seligson, Esko A. Kautto, Edward N. Passen, John L. Hays, James L. Chen
Final approval of manuscript: Nathan D. Seligson, Esko A. Kautto, Edward N. Passen, Colin Stets, Amanda E. Toland, Sherri Z. Millis, Christian F. Meyer, John L. Hays, James L. Chen
Disclosures
Sherri Z. Millis: Foundation Medicine Inc (E); John L. Hays: Clovis Oncology, Merck (C/A). The other authors indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board
References
- 1.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin 2017;67:7–30. [DOI] [PubMed] [Google Scholar]
- 2.Guo X, Jo VY, Mills AM et al. Clinically relevant molecular subtypes in leiomyosarcoma. Clin Cancer Res 2015;21:3501–3511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Barlin JN, Zhou QC, Leitao MM et al. Molecular subtypes of uterine leiomyosarcoma and correlation with clinical outcome. Neoplasia 2015;17:183–189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Barretina J, Taylor BS, Banerji S et al. Subtype‐specific genomic alterations define new targets for soft‐tissue sarcoma therapy. Nat Genet 2010;42:715–721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Beck AH, Lee CH, Witten DM et al. Discovery of molecular subtypes in leiomyosarcoma through integrative molecular profiling. Oncogene 2010;29:845–854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cancer Genome Atlas Research Network . Comprehensive and integrated genomic characterization of adult soft tissue sarcomas. Cell 2017;171:950–965.e928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Italiano A, Laurand A, Laroche A et al. ERCC5/XPG, ERCC1, and BRCA1 gene status and clinical benefit of trabectedin in patients with soft tissue sarcoma. Cancer 2011;117:3445–3456. [DOI] [PubMed] [Google Scholar]
- 8.Xing D, Scangas G, Nitta M et al. A role for BRCA1 in uterine leiomyosarcoma. Cancer Res 2009;69:8231–8235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chudasama P, Mughal SS, Sanders MA et al. Integrative genomic and transcriptomic analysis of leiomyosarcoma. Nat Commun 2018;9:144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lucchesi C, Khalifa E, Laizet Y et al. Targetable alterations in adult patients with soft‐tissue sarcomas: Insights for personalized therapy. JAMA Oncol 2018;4:1398–1404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Movva S, Wen W, Chen W et al. Multi‐platform profiling of over 2000 sarcomas: Identification of biomarkers and novel therapeutic targets. Oncotarget 2015;6:12234–12247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cote GM, He J, Choy E. Next‐generation sequencing for patients with sarcoma: A single center experience. The Oncologist 2018;23:234–242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gelmon KA, Tischkowitz M, Mackay H et al. Olaparib in patients with recurrent high‐grade serous or poorly differentiated ovarian carcinoma or triple‐negative breast cancer: A phase 2, multicentre, open‐label, non‐randomised study. Lancet Oncol 2011;12:852–861. [DOI] [PubMed] [Google Scholar]
- 14.Coleman RL, Oza AM, Lorusso D et al. Rucaparib maintenance treatment for recurrent ovarian carcinoma after response to platinum therapy (ARIEL3): A randomised, double‐blind, placebo‐controlled, phase 3 trial. Lancet 2017;390:1949–1961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Laroche A, Chaire V, Le Loarer F et al. Activity of trabectedin and the PARP inhibitor rucaparib in soft‐tissue sarcomas. J Hematol Oncol 2017;10:84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Frampton GM, Fichtenholtz A, Otto GA et al. Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol 2013;31:1023–1031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Landrum MJ, Lee JM, Benson M et al. ClinVar: Public archive of interpretations of clinically relevant variants. Nucleic Acids Res 2016;44:D862–D868. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ordonez JL, Amaral AT, Carcaboso AM et al. The PARP inhibitor olaparib enhances the sensitivity of Ewing sarcoma to trabectedin. Oncotarget 2015;6:18875–18890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kelley TW, Borden EC, Goldblum JR. Estrogen and progesterone receptor expression in uterine and extrauterine leiomyosarcomas: An immunohistochemical study. Appl Immunohistochem Mol Morphol 2004;12:338–341. [DOI] [PubMed] [Google Scholar]
- 20.Bathan AJ, Constantinidou A, Pollack SM et al. Diagnosis, prognosis, and management of leiomyosarcoma: Recognition of anatomic variants. Curr Opin Oncol 2013;25:384–389. [DOI] [PubMed] [Google Scholar]
- 21.An Y, Wang S, Li S et al. Distinct molecular subtypes of uterine leiomyosarcoma respond differently to chemotherapy treatment. BMC Cancer 2017;17:639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Shihab HA, Gough J, Mort M et al. Ranking non‐synonymous single nucleotide polymorphisms based on disease concepts. Hum Genomics 2014;8:11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cancer Genome Atlas Network . Comprehensive molecular portraits of human breast tumours. Nature 2012;490:61–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cancer Genome Atlas Research Network . Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609–615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Paul A, Paul S. The breast cancer susceptibility genes (BRCA) in breast and ovarian cancers. Front Biosci (Landmark Ed) 2014;19:605–618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ballinger ML, Goode DL, Ray‐Coquard I et al. Monogenic and polygenic determinants of sarcoma risk: An international genetic study. Lancet Oncol 2016;17:1261–1271. [DOI] [PubMed] [Google Scholar]
- 27.Hanby AM, Kelsell DP, Potts HW et al. Association between loss of heterozygosity of BRCA1 and BRCA2 and morphological attributes of sporadic breast cancer. Int J Cancer 2000;88:204–208. [PubMed] [Google Scholar]