Synchronous recurrence of concurrent colon adenocarcinoma and dedifferentiated liposarcoma

Abstract

A 62-year-old man presented with concurrent sigmoid colon adenocarcinoma and small bowel mesenteric dedifferentiated liposarcoma. Following surgical resection of the colon cancer, complete excision of the mesenteric sarcoma and adjuvant folinic acid, fluorouracil and oxaliplatin (FOLFOX) chemotherapy, the patient demonstrated no radiological evidence of disease for more than 2 years. The patient then developed synchronous recurrence of both cancers: the colon cancer metastasised to the liver and a pelvic lymph node, and the liposarcoma recurred in the original location. The patient underwent additional chemotherapy with complete response of the metastatic colon cancer and stable disease for the liposarcoma. The recurrent mesenteric tumour was subsequently resected. Although concurrent cancers have been reported, this unique case of synchronous recurrence raises interesting hypotheses regarding host–tumour interaction and immune surveillance.

Background

Concurrent cancers have been reported, including our previous work in patients with well-differentiated (WD)/dedifferentiated (DD) liposarcoma.¹ WD/DD liposarcoma is a rare cancer of putative adipocyte origin that features characteristic amplification of 12q13-15, which includes genes such as MDM2 and CDK4. WD tumours are low grade and generally indolent, without the potential for metastasis; in contrast, tumours that have a DD component are high grade, with an approximately 20%–30% distant metastatic rate, most commonly to the lungs.² While there are hereditary syndromes associated with colon adenocarcinoma, except for isolated case reports (eg, neurofibromatosis),^3–5 there are no defined hereditary syndromes associated with WD/DD liposarcoma. In this case report, we describe a patient who had concurrent colon adenocarcinoma and DD liposarcoma, treated to no evidence of disease, followed by synchronous recurrence of both cancers. This pattern of synchronous recurrence for two different cancers has not been previously reported and raises interesting hypotheses for further study.

Case presentation

A 62-year-old man underwent screening colonoscopy, and a mass was identified in his sigmoid colon. A biopsy of the sigmoid mass demonstrated adenocarcinoma. A contrast-enhanced, staging CT of the chest, abdomen and pelvis was negative for any other sites of disease. The patient then underwent a laparoscopic converted to open sigmoid colon resection. During initial laparoscopy, a small (<2 cm) firm nodule was found in the mid–small bowel mesentery, at which point the operation was converted to open to better assess the entire abdomen for other possible sites of disease. Incisional biopsy of the mesenteric mass was interpreted on frozen section as suggestive of sclerosing mesenteritis. The remainder of the mass was excised completely, and metal clips were left in place to mark this area. The patient recovered from this operation without complications. Final pathology for the colon resection demonstrated a T3 N2a poorly differentiated colon cancer; margins were negative for disease. The mesenteric tumour was found to have amplification of carboxypeptidase M, a gene just adjacent to MDM2, leading to the final diagnosis of high-grade DD liposarcoma. Histology images for the primary malignancies are shown in figure 1A,B.

Figure 1.

Histology images of original malignancies and synchronous recurrence. H&E staining. (A) Primary sigmoid colon adenocarcinoma (original magnification, ×10). (B) Primary DD liposarcoma (original magnification, ×40). (C) Recurrent colon cancer (original magnification, ×10). (D) Recurrent DD liposarcoma (original magnification, ×40). DD, dedifferentiated.

Investigations

A whole-body positron emission tomography (PET)/CT scan performed 6 weeks after the surgery did not demonstrate any sites of disease or areas of 18F-fluorodeoxyglucose (FDG)-avidity, confirming complete macroscopic resection of both primary cancers.

The patient was also referred to a genetic counsellor. The patient’s family history was significant for a mother with lung cancer, a maternal grandmother with liver cancer and a sister with breast cancer. Based on this family history, the Amsterdam criteria for assessing the likelihood of Lynch syndrome was not met, with the probability of having an MLH2, MSH2, or MSH6 mutation projected to be 0.01%. In support of this, immunohistochemistry performed on the colon cancer specimen confirmed normal expression of MLH1, MSH2, MSH6 or PMS2 protein. The risk of other hereditary cancer syndromes, including Li Fraumeni and others, was discussed; however, no further genetic testing was performed.

Treatment

Given the nodal positivity of his colon cancer, the patient was referred to a medical oncologist and started on adjuvant folinic acid, fluorouracil and oxaliplatin chemotherapy. After multidisciplinary tumour board discussion, the consensus was that no further treatment was needed for the liposarcoma, given its small size and incidental nature. After finishing chemotherapy, the patient continued to be monitored closely in surveillance for both cancers. A total of five CT scans were performed during the course of over 2 years, all of which were negative for any evidence of disease.

Outcome and follow-up

Almost 2.5 years following his surgery, the patient developed synchronous recurrence of colon cancer and mesenteric liposarcoma. The CT scan showed metastatic disease in the liver (one solitary liver metastasis at the dome of the left liver measuring 2 cm) and an enlarged pelvic lymph node that was subsequently needle biopsied percutaneously under image (CT) guidance. Final pathology was consistent with colon adenocarcinoma (figure 1C). A new 2 cm mass was also noted in the mesentery in the area of the metal clips. Needle biopsy was attempted but aborted due to the lack of a safe window for access. On a repeat PET scan, FDG-avidity was noted with all three sites of disease with no evidence of disease in the lungs or other visceral organs (figure 2).

Figure 2.

Initial positron emission tomography scan at the time of synchronous recurrence. 18F-fluorodeoxyglucose-avidity was noted at all three sites: left liver, mesentery (liposarcoma) and pelvic lymph node (arrows). The 2 cm mesenteric liposarcoma had a maximum standardised uptake value of 5 (circle, adjacent to the previously placed surgical clips).

The patient was started on folinic acid, fluorouracil and irinotecan (FOLFIRI) chemotherapy for the recurrent colon cancer, which was given priority given its metastatic nature, with plans for subsequent treatment of the liposarcoma. An interval PET scan showed complete response in the liver and lymph node, with no further sites of disease. The mesenteric tumour remained stable in size but showed metabolic disease progression with increased FDG-avidity (with a maximum standard uptake value (SUV) from 5.0 to 8.0).

Discussion was carried out with the patient and he was given the option of chemotherapy for the recurrent liposarcoma versus surgical resection. The tumour was not causing any symptoms and there was no concern for impending bowel obstruction. The patient also had a premalignant polyp in the right colon identified by surveillance colonoscopy. The decision was made to pursue surgery, and the patient underwent small bowel resection for the mesenteric tumour and right hemicolectomy for the polyp. As the right colon polyp had a flat base, it was thought that an endoscopic polypectomy would not be complete. Since surgery was already planned for the recurrent mesenteric liposarcoma, it was decided that the polyp would be addressed at the same time. At the time of surgery, no other sites of disease were found. Additional intraperitoneal fat and retroperitoneal fat were sampled. Final pathology demonstrated no evidence of adenocarcinoma at the site of the colon polyp, DD liposarcoma (FNCLCC grade 2/3) in the small bowel mesentery (figure 1D) and normal adipose tissue in the fat samples. The diagnosis of liposarcoma was confirmed, with molecular studies showing MDM2 amplification (figure 3).

Figure 3.

Recurrent DD liposarcoma in the mid–small bowel mesentery. (A) Intraoperative image of the tumour (arrowhead) just prior to en bloc resection of the small bowel. (B) Immunohistochemical staining showing strong nuclear reaction for MDM2, confirming diagnosis of DD liposarcoma. Fluorescence in situ hybridisation was also done, which showed amplification (not shown). DD, dedifferentiated.

At the time of this publication, the patient is 4.2 years from his initial surgery, 1.7 years from synchronous recurrence and 10 months from second surgery, without evidence of disease based on follow-up CT.

Discussion

Based on a case series and analysis of a national cancer database, our previous work reported that some patients may be at risk of developing concurrent liposarcoma and another primary malignancy.¹ There are, in fact, case reports in the literature of concurrent liposarcoma and colorectal adenocarcinoma occurring in the same patient.^{6 7} Our current case, however, is unique with synchronous recurrence of two cancers.

Given this patient’s family history of cancer, an unidentified hereditary syndrome or at least a strong genetic predisposition likely exists. A recent study of 1162 patients with various types of sarcoma found that more than half of these patients had an excess of pathogenic germline monogenic and polygenic variants.⁸ Still, we are confident that even in the context of a potential germline mutation, the clinical pattern of recurrence in this patient is indeed recurrence as opposed to new primary cancers. For the colon cancer, with node positive disease, the patient had a higher likelihood of distant recurrence (liver) and, with negative margins, a lower likelihood of local recurrence.⁹ For the liposarcoma, as the discovery was incidental, formal resection was not done initially, and the excisional biopsy likely left behind microscopic disease that subsequently developed into local recurrence. Importantly, the patient had no macroscopic evidence of disease by serial imaging for an interval time of more than 2 years. We propose that this phenomenon of synchronous recurrence of two cancers is a result of systemic cancer evasion of immune pressure. Such escape of immune pressure may occur in either an antigen-specific or a non-specific manner.

Recently, mutation-derived ‘neoantigens’ have been described as a primary source of antigen-specific recognition of cancer cells by the immune system, especially T cells.¹⁰ Shared mutations in synchronous tumours have been described.¹¹ Additionally, various driver mutations in genes such as PTEN, KRAS, PK3CA and TP53 are commonly found across several malignancies.¹² We suggest that the occurrence of shared neoantigens between synchronous tumours could modulate immune responses to concurrent tumours. Chronic tumour-derived neoantigen in one primary tumour site could also ultimately elicit T cell exhaustion in tumour specific T cells.¹³ Development of such T cell exhaustion may therefore also impair T cell responses to another synchronous tumour site.¹⁴

Synchronous tumours may also elicit immune suppression with non-antigen-specific pathways. Increases in circulatory levels of numerous immune inhibitory cytokines, such as transforming growth factor beta (TGF-β), interleukin (IL)-10, IL-6 and IL-4, are also found to occur in tumour-bearing patients and could lead to a systemic dampening of immune responses.¹⁵ Alternatively, tumour induction of increased levels of tumour-associated macrophages and myeloid-derived suppressor cells in circulation could similarly allow for systemic suppression of immune responses.¹⁶ Finally, tumour-driven metabolic changes within and beyond the tumour microenvironment may lead to conditions unfavourable for immune-mediated regulation of cancer growth as immune cell function is inhibited by competition for glucose and amino acids with cancer cells.^{17 18} It is therefore possible that immunosuppressive effects of one primary tumour contributed to immune suppression at another synchronous tumour site.

Interestingly, for this case, organised aggregates of immune cells known as tertiary lymphoid structures (TLS) were noted in both the primary colon cancer (figure 4A) and the liposarcoma (figure 4B). TLS have been identified in a variety of other cancers and have been shown to represent developing antigen-driven adaptive immune responses.¹⁹ In this patient, the synchronous presence of TLS may be coincidental or, alternatively, may suggest a broad, universal immune response directed at two different cancers. Failure to mount an effective antitumour response within either cancer over time could then conceivably result in systemic failure leading to synchronous recurrence.

Figure 4.

Histology images showing TLS. (A) TLS present in the primary colon adenocarcinoma, indicated by the blue arrowhead. (B) TLS were also identified in the primary liposarcoma, indicated by blue arrowheads. TLS, tertiary lymphoid structures.

Further research investigation with this case is ongoing. Genomic sequencing of both cancers is being done to determine whether a common germline mutation exists between the two cancers. In addition, further immunological studies are being performed to assess for differences in the immune response within the microenvironment of each cancer.

Learning points.

• Synchronous recurrence can occur in patients with concurrent primary malignancies.

• Management and surveillance of patients with concurrent malignancies must take into consideration both cancers individually and in the context of the other.

• Synchronous recurrence may provide a unique insight into host-tumour interaction and immune surveillance.