UCPOGC is a rare pancreatic tumor that may arise anywhere along the pancreas and typically presents in the fifth to seventh decades of life with symptoms such as abdominal or back pain, palpable mass, weight loss, and/or jaundice [1, 4,5,6,7, 10]. Upon presentation, our patient only experienced weight loss, but it was unclear whether this was purely intentional or secondary to the tumor. Asymptomatic cases have been reported, though UCPOGCs in asymptomatic cases tend to be smaller in comparison [11,12,13]. At their time of discovery, UCPOGCs are generally quite large, measuring an average of 8 cm, which may be due to their rapid growth [14, 15]. Our specimen was 11.0 × 8.1 × 7.0 cm which is relatively large [1, 6, 7], so it is surprising that our patient had no bulk or obstructive symptoms. In contrast, a 57-year-old male presenting with a nearly identical tumor size and location experienced epigastric pain and weight loss .
Involvement of the spleen in cases of UCPOGC is well-described [1, 4, 5]. Our specimen analysis revealed the primary site in the distal pancreas, with the splenic involvement because of local invasion by the primary pancreatic UCPOGC, a previously described phenomenon .
Our patient had an elevated CA 19-9 of 138 U/mL (normal < 35 U/mL) but normal serum CEA of 2.4 ng/mL (normal < 3.0 ng/mL) and fluid CEA of 2.6 ng/mL (normal 0.0–3.8 ng/mL). In previous cases of UCPOGCs, both normal and elevated levels of CA 19-9 and CEA have been described [1, 4, 5, 9, 11, 16, 17].
The sensitivity of CT, MRI, and EUS for the detection of UCPOGC is similar to that for the detection of traditional PDAC . The specificity of these imaging modalities remains in question, though there are some features that may help distinguish UCPOGC from PDAC. On CT, UCPOGCs are described as heterogenous masses with cystic or solid features [1, 4,5,6,7, 10, 16]. Contrast enhancement may or may not be present. In this case, contrast-enhanced CT revealed a heterogeneous and hypodense mass. This mass features thick solid peripheral contrast enhancement, with suggestion of central necrosis. CT features that may help differentiate UCPOGC from PDAC include the presence of contrast enhancement and regularity of the tumor border. PDACs are typically hypoattenuating tumors with ill-defined borders, whereas UCPOGCs are well-defined and contrast enhancement is not uncommon [16, 18]. The hyperemic nature of UCPOGCs is expected given the rapid growth of UCPOGCs .
In our case, abdominal MRI of the UCPOGC demonstrated heterogenous intermediate T2-signal, marked restricted diffusion, and heterogenous peripheral T1 fat saturated post-contrast enhancement. In the literature, UCPOGCs have a variable appearance on MRI, with no consistent signal intensity described for each MRI sequence. UCPOGCs have been described as hypo- or hyperintense on T1, T2, and diffusion-weighted imaging [4, 10, 14]. Isointense UCPOGCs on T2 and diffusion-weighted sequences have also been described . In contrast, PDACs exhibit low signal intensity on T1-weighted imaging, but also appear variable on T2 and diffusion-weighted sequences . In our study, EUS of the UCPOGC revealed a heterogenous, isoechoic to hyperechoic mass. The heterogenous echotexture of UCPOGCs on EUS is a well-described feature [19,20,21], and its presence may help distinguish UCPOGCs from PDACs, which are more uniformly hypoechoic .
UCPOGCs are characterized by pleomorphic mononuclear neoplastic cells with intermixed, benign osteoclast-like giant cells (OGC), which are the histopathologic hallmark of UCPOGCs. In our case, the OGCs were CD68 positive and cytokeratin AE1/AE3 negative. This immunohistochemical profile is consistent with previously reported cases of UCPOGC and reinforces the idea that OGCs are derived from a benign histiocytic cell population [1, 11, 14, 20, 22]. OGCs are thought to be secondarily recruited into the tumor via chemotaxis , and their presence is associated with an improved prognosis compared to PDACs and undifferentiated carcinomas of the pancreas without OGCs [1, 8, 9].
Muraki et al. reported a 5-year survival of 59.1% and median survival of 7.67 years for patients with UCPOGC, versus 15.7% 5-year survival and 1.59-year median survival for PDACs . As seen in our patient, UCPOGCs are associated with concomitant PDAC with composite tumors in up to 76% of cases , which results in a worse prognosis [8, 9]. Luchini et al. reported a median 15-month survival in a group of 9 patients with UCPOGC with associated PDAC, and our patient exhibited 11-month survival post-op . In our case, the PDAC component stained positive for CDX2, which is expressed in normal pancreatic ductal epithelium and appears to be downregulated during the transition to PDAC . Its expression in PDAC may indicate a worse prognosis , although it is unclear if a CDX2-positive PDAC component in UCPOGC is associated with worse outcomes. The UCPOGC in this case also stained positive for INI-1, which is consistent with Reid et al. study , where three out of three UCPOGCs retained INI-1 stain. INI-1 staining may be a helpful distinguishing feature, as Agaimy et al. demonstrated loss of INI-1 in all four monomorphic anaplastic subtypes of pancreatic undifferentiated rhabdoid carcinoma .
The optimal treatment for UCPOGCs remains unstandardized. Surgery is considered first-line treatment , and in our case, the UCPOGC was removed via en-bloc resection. The roles of chemotherapy and radiation are unclear. There are case reports of successful treatment with adjuvant chemotherapy, with one patient recurrence-free 10 years after en-bloc resection of UCPOGC with adjuvant gemcitabine . However, UCPOGCs unresponsive to adjuvant chemotherapy have also been described . In our case, the patient was treated with adjuvant mFFX based on the presence of concurrent PDAC but developed liver metastasis while on treatment. As a result, he was switched to gemcitabine plus nab-paclitaxel which has demonstrated efficacy in undifferentiated pancreatic carcinomas , however the patient unfortunately expired after two cycles. Adjuvant radiation has theoretical benefit given the radiosensitivity of giant cell tumors of bone, although its effectiveness has not been fully established . There is a paucity of data regarding therapy for metastatic UCPOGC, with some reports of benefit with pembrolizumab [28, 29]. Interestingly, expression of programmed death-ligand 1 is a negative prognostic factor for UCPOGCs , which may explain the efficacy of pembrolizumab in metastatic disease.
In summary, UCPOGCs are extremely rare pancreatic malignancies with a unique histopathology and clinical course. Certain radiologic findings may be helpful to distinguish UCPOGC from PDAC, but they are generally non-specific and UCPOGC remains a pathological diagnosis. If possible, surgery is generally accepted as the first-line treatment option, whereas the roles of chemotherapy and radiation are unclear, and their evidence is primarily anecdotal. Due to the exceeding rarity of this tumor, large-scale clinical studies are unlikely to be feasible, therefore contributing individual insights and experiences will be important in promoting better understanding and care for patients with this devastating disease.