DISCUSSION

The decision to pursue palliative therapy in patients with pancreatic adenocarcinoma is likely multifactorial, and is guided by factors such as location and character of symptoms, patient comorbidities, and patient and treating physician preferences. Focusing on a cohort of patients presenting with metastatic disease, for whom palliative therapy would likely comprise an important component of the treatment strategy, we found that that only a minority of patients (16.8%) had evidence of receiving a palliative intervention. The interventions designated as “palliative” did significantly increase during the time period, from 12.9% in 2003 to 19.2% in 2011. However, this subset still represents only a small proportion of the patients who would likely benefit from referral for palliative services. A similar trend was observed in a recent study among patients in the Netherlands, which showed an increasing proportion of patients (10% to 27%) with metastatic disease undergoing palliative chemotherapy (1). Although these observed increases are encouraging, they suggest that further efforts are necessary to promote and provide palliative therapy in this population with incurable disease.

The most common palliative modality identified in this study was systemic therapy. Pancreatic adenocarcinoma is often characterized as a systemic disease owing to its rapid infiltration and spread, limiting detection of localized disease to approximately 15–20% of incident cases (1,8). As such, treatment with systemic chemotherapy comprises the mainstay of palliative therapy for advanced pancreatic cancer (8). Multiple reports in the literature suggest a survival benefit with a variety of chemotherapy regimens, including single-agent gemcitabine or combination regiments such as FOLFIRINOX (3,8-10). Among our cohort, receipt of palliative chemotherapy was associated with prolonged survival. Aside from its modest survival benefit, a recent systematic review by Kristensen et al. highlighted its role in improving quality of life and pain scores in patients with metastatic disease (11). The goal of minimizing symptoms, either due to the disease or to the agent itself, should guide the choice of chemotherapy regimen. FOLFIRINOX has been shown to provide an incremental survival benefit over older agents, such as gemcitabine, but its higher incidence of side effects detracts from its appeal as a palliative regimen (3,10,12).

For patients with more localized symptoms, such as gastric outlet or biliary obstruction, palliation with surgical or endoscopic approaches may provide symptomatic improvement (2). Symptomatic biliary obstruction develops in approximately 80% of patients who have disease in the pancreatic head, and may cause pruritus, cholangitis, anorexia, malnutrition, or even hepatic failure (2). Surgical bypass (such as hepaticojejunostomy or choledochojejunostomy) traditionally comprised the mainstay of therapy for these symptoms. Recently, however, in accordance with the goal of minimizing therapeutic morbidity in the face of incurable disease, endoscopic therapies with stent placement are being increasingly utilized for both duodenal and biliary obstructions (2,3,13).

For gastric outlet obstruction, both palliative resection and bypass have been described to alleviate obstructive symptoms (2,3,8). Currently, palliative bypass is preferred, as a survival benefit has not been consistently shown for either approach, and bypass is associated with a lower morbidity. The risk of developing gastric outlet obstruction is so high among patients with biliary obstruction that prophylactic gastrojejunostomy is now recommended in patients undergoing palliative hepaticojejunostomy (2,3). Within our study, surgical palliation offered a small benefit and showed decreasing utilization over time, although distinction between approach (surgical versus endoscopic) and type (gastric, biliary, or both) could not be made.

Pain management forms an essential aspect of palliative care. Among patients receiving palliative interventions in our cohort, those receiving pain managements as the primary palliative therapy increased significantly from 2003 to 2011. The abdominal and back pain frequently experienced by patients with pancreatic cancer can be debilitating, and severely affect a patient’s quality of life (3). As with surgical therapy, radiation therapy can also provide palliation of localized symptoms. It is less commonly used than either systemic or surgical intervention, but can be an important adjunct to target specific areas of pain (14,15). Furthermore, although brain and bone involvement is uncommon among patients with pancreatic cancer, radiation therapy offers an important option for patients with pathologic fractures or neurologic impairment (14).

Unfortunately, our study indicates that there may be racial and ethnic disparities in providing palliative interventions to patients with unresectable pancreatic cancer. Both black race and Hispanic ethnicity were associated with lower odds of receiving designated forms of palliative care, despite controlling for demographic factors such as insurance status, income, facility type, and regional variation. This observation is not unique to our study; a recent review highlighted disparities in hospice utilization and the assessment and treatment of pain among black and Hispanic patients across a variety of advanced diseases (16). Another study utilized the Surveillance Epidemiology and End Results (SEER) database to identify cases of pancreatic adenocarcinoma and evaluate overall trends in treatment. Of the patients undergoing therapy that was determined to be palliative, patients of black race had lower odds of receiving chemotherapy, and those of Hispanic ethnicity were less likely to receive radiation (17). Although another study suggested that some of these disparities might be attenuated in the inpatient setting, it is evident that further efforts to eliminate these differences are necessary (18).

Other differences in provision of palliative therapy were noted among our study cohort. While gender does not appear to have a significant role, several other patient factors may be associated with the decision to offer palliative services. Interestingly, while patients with more comorbidities are more frequently offered palliative therapy, age had the opposite association: with increasing age, the odds of undergoing a palliative intervention decreased (Table 2). This finding was also reflected in the study by Bernards et al., which noted a lower use of palliative chemotherapy among older patients in the Netherlands (1). Socioeconomic factors also appeared to affect provision of palliative therapy among our cohort; those with higher income and private health insurance were less likely to receive palliative interventions. This finding is in distinction to that observed in the Netherlands, where those of higher socioeconomic status were more like to undergo palliative chemotherapy. Further investigation into causes of these differences is important.