Pancreatic acinar cell carcinoma is a rare entity with heterogenous clinicopathological spectrum of disease which impacts therapeutic management and patient outcomes [17]. Imaging diagnosis of pACC is challenging due to limited literature and in this study we investigated the role of imaging as a biomarker for predicting outcome and association with survival.

We found that a majority of pACC were large solid masses in the pancreatic head with ill-defined margins. pACC tends to present at a larger size at the time of diagnosis compared to PDAC, with a reported average size ranging from 4 to 10 cm [6, 8, 18]. In our study, the average maximal diameter was 5.0 cm while the median maximal diameter was 4.1 cm. Most (53.1%) presented in the pancreatic head and 71.4% were metastatic at presentation, in keeping with prior studies [6, 19, 20]. Biliary ductal dilatation was uncommon, with only 22.4% demonstrating pancreatic ductal dilatation, helping to distinguish from other entities such as PDAC in which obstructive biliary dilatation is more common and present in up to 90% of cases [21]. Bhosale et al. reports calcifications in 6% of tumors, in keeping with our study results in which 3/49 (6.1%) demonstrated calcifications. The relative absence of calcifications in pACC can help distinguish it from tumors where calcifications are more frequently associated with the tumor such as pancreatic neuroendocrine tumors, and cystic neoplasms including solid pseudopapillary neoplasm, serous cystic neoplasm, and mucinous cystic neoplasm [22].

All lesions demonstrated contrast enhancement on CT and MRI. On portal venous phase of CT, the majority were hypo- or isodense (82.6%) to the normal pancreatic parenchyma, thought to reflect the lesion’s hypovascularity, and tended to have a heterogenous enhancement pattern (79.6%). These findings affirm those of previously reported limited patient series [6, 8]. Hence, pACC should be considered in the differential diagnosis for large heterogeneously enhancing hypodense pancreatic masses in the absence of biliary or pancreatic ductal dilatation [7, 8, 19, 23]. While pACC and PDAC share imaging features, key distinctions exist. pACC typically appears as a large heterogenous mass without significant biliary dilatation, whereas PDAC often presents with biliary obstruction. Both demonstrate diffusion restriction and unlike previous studies, the present study found that pACC often to have ill-defined margins resembling PDAC.

Pancreatic neuroendocrine tumor may be another differential consideration on pathology for these pancreatic neoplasms, and imaging may be able to help in recognizing certain distinguishing features. Well-differentiated pancreatic neuroendocrine tumors (grade 1 and 2) are typically small well-defined lesions with intense homogenous enhancement [24]. This appearance contrasts with the large ill-defined heterogeneously enhancing imaging presentation of pACC, although high grade pancreatic neuroendocrine tumors may potentially demonstrate greater degree of inhomogeneity and heterogenous enhancement pattern. On MRI, case studies and limited case series have suggested that pACC shows diffusion restriction and heterogenous enhancement pattern [25, 26]. These findings are affirmed in our study in which 89.4% demonstrated diffusion restriction, and 59.1% with heterogenous enhancement.

Approximately 50% of the cohort died after a median follow up time of 853 days. According to small-cohort clinicopathologic analysis performed in 1992, age and pathologic stage were associated with survival in which patients who presented before the age of 60 survived twice as long as older patients [27]. Indeed, our study affirmed that location within the pancreas did not demonstrate significant correlation with survival, and higher pathologic stage and older age were associated with worse survival, with older age being significant on multivariate analysis. Treatment modality was also associated with survival, in which surgical resection of the primary tumor was associated with improved survival compared to non-surgical management.

We found that a few imaging features were associated with survival. T1 hyperintensity within the tumor on MRI was associated with worse survival; given our knowledge of which materials have intrinsic T1-shortening properties, the presence of T1 hyperintense signal in the pancreatic mass may relate to intralesional hemorrhage, necrosis with proteinaceous debris, or the presence of intravascular thrombi. Multivariate analysis demonstrated that the imaging presence of hypoattenuating components corresponding to pathology-proven necrotic contents were associated with worse survival. These results align with recent studies indicating tumor necrosis on imaging may be predictive of tumor aggressiveness in other pancreatic tumors including PDAC; Anderson et al. (2023) notes association of MRI-evident necrosis with larger size PDAC tumors as well as higher likelihood of regional lymphadenopathy and metastases [28, 29]. Previous literature supports the finding in which hypoattenuating pancreatic cancers tended to correlate with necrosis and dense fibrous stroma from desmoplastic reaction, which portend poorer prognosis [10, 30]. Notably, our study found that encasement of the splenic vein was associated with worse survival compared to no involvement. These findings represent an extension of previous literature illustrating poor prognosis of pancreatic cancers with splenic vessel involvement. Crippa et al. (2018) had demonstrated in a meta-analysis of patients with PDAC involving the pancreatic body and/or tail that splenic vessel involvement was associated with worse survival, reflecting stigmata of more aggressive disease [31]. Recognizing the clinical implications of imaging features such as T1 hyperintensity, hypoattenuating necrosis, and vascular invasion is crucial, as their correlation with tumor aggressiveness may influence treatment decision-making. These features have the potential to serve as imaging biomarkers for risk stratification, facilitating therapeutic optimization, prognostic assessment, and treatment selection.

Compared to previous studies on pACC, our study contributes to the growing knowledge of imaging characterization on both CT and MRI while also providing novel insights by correlating imaging features with survival outcomes, an aspect not previously investigated in the literature. Prior studies with limited sample sizes have included a maximum of 30 patients, with most cohorts comprising fewer than 20 individuals [6,7,8]. These studies primarily investigated the CT characteristics of pancreatic acinar cell carcinoma (pACC), with Tatli et al. (2005) being the only study to evaluate MRI features, albeit in just two patients [6]. The variability in study methodologies is likely attributable to the small sample sizes inherent to this rare malignancy. Patient demographics were largely consistent across studies, with a mean age in the 60s and a predominance of male patients. The overall CT features of pACC seen in the present study are consistent with previous studies, confirming that pACC typically presents as a large, heterogeneous mass, with biliary ductal dilatation and calcifications being uncommon features.

This was a retrospective single-institution study with its inherent limitations related to such a study design. Further validation of these findings in a multi-center or larger cohort would be beneficial. Assessment of prognostic features was limited for patients whose care was transferred to outside institutions, in which subsequent records were unavailable for review. The number of lesions demonstrating T1 hyperintensity and splenic vein involvement represented the relatively smaller but significant proportion of patients who may have worse prognosis. Due to the long inclusion period required for the identification of this exceptionally rare pancreatic neoplasm, the total number of available MRI studies for analysis was limited. Furthermore, some MR examinations lacked complete data for the assessment of parameters such as diffusion-weighted imaging and dynamic contrast enhancement patterns. Variability in the imaging phases of contrast-enhanced CT limited the ability to perform a comprehensive analysis of temporal enhancement patterns. For patients who underwent biopsy only without surgical resection, radiologic-pathologic correlation was not feasible. Nonetheless, the present study offers valuable new information to bridge our understanding of this rare disease entity.

In conclusion, pancreatic acinar cell carcinoma often presents as a large ill-defined mass without biliary ductal dilation. T1 hyperintensity, presence of hypoattenuating necrotic components, and splenic vein invasion may serve as independent predictors of survival, and our results contribute to the growing body of knowledge on the diagnostic and prognostic imaging features of this rare pancreatic malignancy.