Patients with pancreatic cancer who were treated at high-volume centers had a 16-month improvement in median overall survival (OS) versus those who received treatment at community medical centers.
In the retrospective study, researchers examined 139 patients with pancreatic cancer who were treated at high-volume centers and 106 patients treated in a community medical center. Patients at the high-volume center had a median OS of 43.6 months compared with 27.9 months for those treated in a community medical center (HR, 0.63;P<.01). The 5-year OS rates were 38.6% and 24.8%, between the high-volume and community centers, respectively (P<.01).
“I think that these findings are quite provocative. We saw both a superior median survival and 5-year OS in these patients,” said study presenter Margaret T. Mandelson, PhD, MPH, of the Virginia Mason Medical Center. “Our study lends some support to the concept of using high-volume centers for all therapy components for pancreatic cancer that is treated with curative intent.”
For the retrospective analysis, 245 patients with pancreatic cancer were identified in the Virginia Mason cancer registry, which is a multi-specialty group practice with over 400 physicians. Patients were treated with primary resection and were diagnosed between 2003 and 2014. Those treated with neoadjuvant therapy were excluded from the study, as were those who refused adjuvant therapy.
“We're in an unusual setting because digestive cancers comprise over 25% of all cancers that we treat. In fact, pancreas cancer is in the top 5 cancer treated at Virginia Mason,” said Mandelson. “We see about 300 newly diagnosed patients each year as well as at least 50 to 100 patients with recurrence. Approximately, 30% of patients will participate in clinical research.”
In the high volume centers, the median age of patients was 63.1 years whereas the median age was 68.2 years in the community setting. This difference in median age was statistically significant (P<.01). Other patient characteristics were well balanced between the arms for average distance traveled, gender, Medicare or Medicaid insurance status, and ECOG performance status.
A number of factors may have contributed to the improvement in survival, according to Mandelson. However, adjusting for distance traveled, age, and tumor characteristics did not alter OS findings, she added.
In terms of treatment type, some evidence suggests that patients treated in the community setting received single-agent treatment rather than a combination, although this was not thoroughly explored. Overall, 81% of patients received multi-agent chemotherapy and 53% had chemoradiation.
“We know that Virginia Mason has a strong setting for supportive care, and we try to maximize the patient’s tolerance to treatment,” said Mandelson. “We have a high rate of completion of treatment in this setting, and of course the impact of optimism and hope can’t be underestimated in this patient population.”
The study had a number of limitations, specifically a small sample size of patients treated over a wide timeframe. Moreover, the researchers did not assess the toxicity of adjuvant therapy and did not confirm that therapy was actually received. Patients are commonly undertreated in the community setting, Mandelson noted.
Strong and consistent evidence supports a volume-outcome relationship in resected pancreatic cancer, which was supported by findings from the study. However, further investigation of the differences in outcomes between high-volume centers and community settings is still warranted, to further uncover the factors that are impacting outcomes. Future studies should assess patterns of care and their impact on OS, noted Mandelson.
“I think the reason for the positive outcome probably has to do more with the processes and extensive number of physicians with different specialties that can be brought to these centers for patients with pancreatic cancer. This is compared to just volume alone for the positive outcome,” said James L. Abbruzzese, MD, professor of Medicine, associate director of Clinical Research and Training, Duke University. “If we accept the results of the respective study, then there are two possible explanations. Volume is obviously one, but I am going to try to suggest that there are other factors that are at least as important or even more important.”