First reported case of a collision tumor composed of pancreatic adenocarcinoma and retroperitoneal liposarcoma: a case report

Collision tumors are rare cases with two tumor entities raised from two different cell types occurring synchronously in the same organ or in close anatomic location. While adenocarcinoma of the pancreas is the most common pancreatic tumor with an incidence of 10 per 100.000, retroperitoneal liposarcoma remains very rare. This is the first report of a collision tumor composed of these two tumor entities.

This is the first description of a collision tumor composed of pancreatic adenocarcinoma and a retroperitoneal liposarcoma. A PubMed search of “collision tumor” showed by the time 1054 publications. Most of the tumors are described in neurosurgery, dermatology and urology. Looking at visceral surgery most collision tumors are described at the stomach (e.g. GIST and adenocarcinoma). There are only 11 publications describing collision tumors with pancreatic involvement with a collision between ductal adenocarcinoma and neuroendocrine tumor both of the pancreas being the most frequently described [1–11]. There are only 8 publications describing a collision tumor with any sarcoma involved [10–17]. In those cases liposarcoma and leiomyosarcoma are the two most frequently mentioned entities.

In our case, pre-interventional CT showed a solid hypodense mass indicating a pancreatic neoplasm (Fig. 1a). We first diagnosed the liposarcoma by core needle biopsy. The liposarcoma belongs to soft tissue sarcomas (STS) and is the most common type of STS in the retroperitoneum (63%) [18]. While well differentiated liposarcoma (DL) is a locally aggressive neoplasia almost incapable of systemic spread, De-differentiated liposarcoma (DDL) metastasizes in 10–15% of cases. All types of WDL share the same genetic aberration. Amplification of the chromosomal region 12q13–15 containing the proto-oncogenes MDM2, CDK4 and HMGA2 is a pathognomonic feature. The circumstance that no pancreatic tissue could be found in the biopsy-specimen should therefore raise attention. The patient received a systematic left sided retroperitoneal compartment resection (Fig. 1b). The medial dissection plane was on the abdominal aorta, therefore sufficient for both, the WDL and the pancreatic ductal carcinoma. Yet, a systematic lymph node dissection as needed for the pancreatic carcinoma was not performed. In such cases the tumor should be treated as a pancreatic carcinoma with respect to the systematic lymph node dissection and the resection should be extended to the retroperitoneum as needed for retroperitoneal sarcomas. Otherwise, if the pancreatic adenocarcinoma was already diagnosed by biopsy, it would be conceivable that the resection had not been carried out to this extent. In this case, a liposarcoma would have been more difficult to detect.

In this special clinical context with a well differentiated liposarcoma and a nodal metastasized pancreatic adenocarcinoma the pancreatic cancer is crucial for the prognosis, so adjuvant treatment focusses on the adenocarcinoma.

Molecular-pathologic examination revealed a pathogenic BRCA 2 mutated pancreatic ductal adenocarcinoma (PDA). PDA remains one of the malignancies with the worst prognosis [19]. While most PDAs are thought to be sporadic, approximately 5 to 10% occur hereditarily. Multiple syndromes and diseases have been associated with an increased risk of developing pancreatic cancer, including hereditary breast and ovarian cancer with mutation of breast cancer 1 (BRCA1) and breast cancer 2 (BRCA2) [20], which lead to a deficiency in deoxyribonucleic acid (DNA) damage repair (DDR). Especially the homologous recombination repair (HRR) of DNA double strand is defective in BRCA deficiency, resulting in chromosomal instability and tumorigenesis [21]. Another family of DNA repair enzymes comprises Poly(ADP-ribose) polymerases (PARPs). PARPs are involved in crucial complementary repair processes. While HRR is ineffective in BRCA mutated cancer cells, PARP in turn plays a major role in repairing damaged DNA to maintain cell survival. A promising feature in the therapy of BRCA mutated cancers are PARP inhibitors and chemotherapeutic agents that induce DNA damage in the presence of impaired DNA repair [22].

According to german oncology guidelines, adjuvant treatment with Gemcitabine was recommended [23]. We plan an extension of the adjuvant chemotherapy with Capecitabine, according to the findings of the ESPAC-4-Study. Especially after negative margin resections, this study shows a survival benefit after treatment with Gemcitabine and Capecitabine compared to Gemcitabine monotherapy [24].

Because of the extended lymph node involvement of the pancreatic ductal adenocarcinoma the risk of recurrence was considered to be high [25, 26]. In case of tumor recurrence we plan a treatment with platin-based chemotherapy or specific PARP inhibitors. For BRCA2 mutated pancreatic cancer there is in vitro and in vivo data suggesting a higher sensitivity of tumor cells to DNA-crosslinking agents [27, 28]. Golan et al. showed in a retrospective study in patients with BRCA1/2 mutation a survival benefit for patients treated with platin based chemotherapy compared to non-platinum chemotherapy (22 vs. 9 months) [29].

PARP inhibitors have been proven to be a new treatment option for patient with germline-mutated ovarian and breast cancer [30, 31]. For ovarian cancer the European medicines Agency (EMA) has approved treatment of somatically mutated tumors with olaparib [32]. A phase II trial of olaparib monotherapy for patients with BRAC1/2 germline-mutated advanced pancreatic cancer showed 36,4% of patients to be progression-free at 6 months and 40,9% to be alive at 12 months [33]. At the moment there is one ongoing study in the USA which is investigating the benefit of a olaparib treatment for patients with a BRAC1/2 somatically-mutated pancreatic [34]. So far there is no evidence for the treatment with specific PARP inhibitors in the adjuvant setting [23].

There is no evidence for the use of adjuvant radiotherapy or chemotherapy after a R0-resection of a G1 liposarcoma, especially when as in our case the pancreatic ductal adenocarcinoma is the prognosis defining entity.

Collision tumors are rare and difficult to diagnose. The reported case demonstrates that inconsistent diagnostic results (such as imaging and pathology) should raise suspicion concerning the diagnosis. Awareness of these rare cases might protect us from underdiagnosing patients and therefore leading to better patient care.

There is evolving evidence that will lead to more personalized treatment options for somatic BRCA mutated pancreatic cancer.