Therapy susceptible germline-related BRCA 1-mutation in a case of metastasized mixed adeno-neuroendocrine carcinoma (MANEC) of the small bowel

Small bowel carcinomas are rare and account only for 0.5% of all cancers. The 5 year overall survival rates for patients with small-bowel adenocarcinoma is 34.9%. To best of our knowledge no valid prognostic or survival data exist to the mixed adeno-neuroendocrine carcinoma (MANEC)-subtype of the small bowel. There is a strong need for a more effective and personalized systemic treatment option in carcinoma of the small bowel. Most recently Schrock, A et al. [1] described molecular alterations in a very large cohort of 317 small bowel carcinomas. According to this study microsatellite-instability (MSI) was present in 7,6% and are characterized by the following mutations like KRAS (53,6%), APC (26,8%), CDKN2A (14,5%), SMAD4 (17,4%) and BRAF (9,1% but only 10% of these have the typical V600E mutation) as well as ERBB2 mutation in 8,2% of the analysable tumors.

The most important pathway in DNA double strand break (DNA-dbs.) repair is homologous recombination (HR). The HR pathway is dependent on different proteins including BRCA1 and BRCA2 and others (e.g. MNR complex, CtIP, MRE11, RAD51, ATM, H2AX, PALB2, RPA, RAD52) [2, 3].

Effective DNA repair is essential to maintain DNA integrity – so BRCA1/2 deficient cells show a high degree of chromosomal instability which increases the risk of malignant transformation [4, 5]. Germline mutations of BRCA1/2 (gBRCA-mut) were described in carcinoma of the ovary in 5–18% and in some other solid malignant tumors including breast carcinoma, pancreatic carcinoma and less frequent in prostate carcinoma. It became apparent that BRCA mutations also occur in sporadic carcinomas and these tumors respond to Poly (ADP-ribose) polymerase- (PARP) inhibition. The full significance of somatic BRCA mutations as a biomarker for the prediction of therapeutic response to PARP inhibition is not entirely clear but ovarian cancer patients with a somatic BRCA mutation are equally likely to benefit from treatment with PARP inhibitors as those with a germline mutation of BRCA. Some authors suggest to perform somatic BRCA testing in routine clinical diagnostics of all advanced tumors, which are potentially amenable to PARP inhibitors [3, 6‐11].

Each BRCA variant was classified according to the established IARC classification. This classification ranges from class 1 variants (not pathogenic or of no clinical significance) via class 2 (likely not pathogenic or of no clinical significance), class 3 (uncertain), class 4 (likely pathogenic) to class 5 variants (definitively pathogenic) according to different databases (compare: BRCA mutation database: http://​arup.​utah.​edu/​database/​BRCA/​orClinVardatabase:http://​www.​ncbi.​nlm.​nih.​gov/​clinvar/​ or Universal mutation database BRCA share: http://​www.​umd.​be/​BRCA1/​).

We describe the case of a 74-year old man presented to our hospital due to a liver tumor in the segments II and III as well as satellite metastatic lesions in the right liver lobe. After taken a liver biopsy from the tumor the diagnosis of an adenocarcinoma was established and interpreted as an unresectable cholangiocarcinoma, initially lacking the information of the primary tumor in the small intestine. The patient was treated with a mono-chemotherapy of Gemcitabine ((1000 mg/qm d1–8) 12/2015–03/2016) followed by selective internal radiotherapy (SIRT) analogous to the SIRCCA study protocol (ClinicalTrials.​gov Identifier: NCT02807181). Because of anemia and poor general condition the administration of cisplatin was withheld. In the further course of treatment an additional tumor site in the ileum was detected and progression with cerebral metastasis occurred. The ileum tumor and one brain metastasis were removed. The ileum tumor was classified as a mixed adeno-neuroendocrine carcinoma (MANEC). Tumor manifestations in the liver and in the brain could be characterized as metastases from the adenocarcinoma component (Fig. 1).

In the further course we patient was treated with capecitabine (500 mg 2–0-2, 5d/week) 05–09/2016). This treatment finally failed 4 months later and all liver metastases showed progression. The treatment failure was the reason to analyze the tumor-DNA for molecular alterations using two panels of 14 genes including BRCA 1 and BRCA 2 (452 amplicons in total; compare versions below). We verified the same BRCA1 mutation (c.3700_3704delGTAAA p.V1234Qfs*8) with a very high allele frequency up to 96% in all locations using ultra deep sequencing in the formalin-fixed paraffin embedded material (Fig. 2). According to different databases (UMD, ARUP, ClinVar) this BRCA1 mutation was classified as pathogenic (class 5). The MANEC was microsatellite stable. The mutational hot spots of all other analyzed genes (KRAS, NRAS, HRAS, BRAF, DDR2, ERBB2, KEAP1, NFE2L2, PIK3CA, PTEN, RHOA and BRCA2) were wild type.

According to the encouraging experiences using combined platin-based chemotherapy and PARP-inhibitors in BRCA-mutated ovarian carcinomas the molecular results have been the trigger for the personalized therapeutic approach using combined chemotherapy of carboplatin, paclitaxel and the PARP1-inhibitor olaparib analogous to the treatment protocol of Oza et al. [14] following a bridging therapy with capecitabine (olaparib 200 mg capsules twice daily, administered orally on days 1–10 of each 21-day cycle) plus paclitaxel (175 mg/m², administered intravenously on day 1) and carboplatin (4 mg/mL per min) in total five cycles for the following 4 months.

All tumor manifestations demonstrated with a significant tumor regression in the MINT analysis and Ga-68-Dotatate PET-CT after 3 months of treatment (mind analysis Table 1 and Ga-68-Dotatate PET-CT Fig. 3) – and the persistent highly necrotic liver metastasis of 6 cm could be removed. In his most current follow up from December 2017 (25 months after his primary diagnosis) the patient is in a very good general condition without evidence for further metastases (compare time table, Table 2).

To the best of our knowledge this is the first description of pathogenic and treatable germline-related BRCA1 mutation in mixed adeno-neuroendocrine carcinoma (MANEC) of the small bowel. There was no known personal history of a previous malignant tumor. Nevertheless, the high allele frequency of the BRCA1 mutation in all tumor manifestations suggested a germline background. Consequently, we tested histologically tumor-free ileum mucosa and were able to detect the same BRCA1 mutation with an allele frequency of 48%. This finding clearly identifies the mutation as germline and a deletion of the wild type BRCA1 allele in the tumor. The homogeneous distribution of BRCA1 mutated tumor cells in the primary tumor and in the different metastasis analyzed here suggests a clonal origin and a causative role of this second hit. Even though a gBRCA1 mutation could be a marker for prostate cancer we did not find any indications for prostate cancer [7].

According to treatment protocols established in BRCA mutated ovarian carcinomas we successfully applied combined chemotherapy of carboplatin, paclitaxel and the PARP1-inhibitor olaparib. Initially, we assumed a somatic background of the BRCA1 mutation in our treated patient, because there was no known personal history of a prior malignant tumor in the 74 year-old man. The germline background of the BRCA1 mutated small bowel carcinoma (like in ovarian adenocarcinoma) is especially promising for a successful combination therapy using platinum-based chemotherapy and a PARP inhibitor (e.g. Olaparib). It has to be questioned whether the somatic pathogenic BRCA mutation itself is the main predictor of an effective PARP inhibition [3].

Beyond BRCA1/2 somatic mutations or the promoter methylation of BRCA some other alterations are potential markers of an effective PARP inhibition (like microsatellite-instable cancers with MRE11-dominant negative mutations, PTEN deficiency, ATM mutation and FANCF promoter methylation) [15].

Additionally, first evidence came up that gBRCA mutated ovarian carcinoma are sensitive to therapies using immune-checkpoint inhibition most likely due to its higher mutational load compared to wild type ovarian carcinoma [16]. It remains to be shown whether our BRCA mutated MANEC also benefit from immune-checkpoint inhibition as an additional future treatment option.

The germline background of our patient was unexpected because he is a 74-year old man with no previously known tumor burden. Currently, there is increasing evidence that the first germline associated tumor manifestation (e.g. endometrial carcinoma in Lynch syndrome) may occur beyond the sixth decade of living which means that we should be aware of the familiar background in tumors of older patients as well [17].