Although “treatment advances” and “precision medicine” are today’s buzz words in oncology, they don’t apply equally to all malignancies. For instance, median overall survival for newly diagnosed patients with metastatic pancreatic cancer has not improved much over the past 20 years, according to analysis of more than 50,000 patients.1 However, more patients are achieving “long-term” survival, but it is defined as 12 months or more, and this fraction of patients is still limited to about 12%, investigators reported at the 2017 Gastrointestinal Cancers Symposium.
“We asked, what has happened to overall survival over 2 decades, and how have new and more aggressive treatments influenced this? Additionally, patient outliers including short- and long-term survivors were identified,” said Talia Golan, MD, of The Oncology Institute Sheba Medical Center in Tel HaShomer, Israel. “We felt that the improvements in medical care would make a difference.”
We thought the proportion of patients dying within 2 months would have decreased a lot more, perhaps down to 30%. I think this would come as a surprise to oncologists at academic centers….— Talia Golan, MD
Tweet this quote
Instead, the analysis indicated only minor changes, at best, and 50% of patients still survive no more than 2 months. “That’s not good enough,” Dr. Golan admitted. “Our most important conclusion was clinical studies, specifically incorporating tissue acquisition, are urgently needed in order to further investigate this substantial patient population who have yet to benefit from advances in modern medicine.”
Although the analysis was performed in Israel, the investigators used the Surveillance, Epidemiology, and End Results (SEER) database, which included 57,263 persons diagnosed in the United States between 1993 and 2013. The proportion of patients diagnosed with metastatic pancreatic cancer during this period remained stable: 54.9% in 1993 and 53.7% in 2013. Median age at diagnosis was 69 years, with a slight decrease in age at diagnosis over time. Of patients with documentation of tumor grade, 43.1% had tumors that were well to moderately well differentiated, and 56.9% had tumors that were poorly differentiated or undifferentiated.
Median Overall Survival Unchanged
Median overall survival of the entire metastatic population remained at 2 months, unchanged over time. Nonetheless, the hazard ratio (HR) for risk of death has shown a consistent and statistically significant improvement from 1993 until 2013 (HR ratio for death in 2013 was 0.71 compared to in 1993; confidence interval [CI] = 0.66–0.76, P = < .001).
“It appears that the prognosis has improved principally among younger, but not older, patients,” added Dr. Golan. Median age of short-term survivors was 72 years, although it was 63 years for long-term survivors (P < .001).
A more encouraging finding was that a greater proportion of patients are surviving at least 12 months, but again, the change was small, Dr. Golan revealed.
Short-term survivors were defined as patients who die within 2 months of diagnosis, whereas long-term survivors are those living at least 12 months. The percentage of short-term survivors decreased from 63.5% in 1993 to 50.6% in 2013 (P < .0001).
“We thought the proportion of patients dying within 2 months would have decreased a lot more, perhaps down to 30%,” she said. “I think this would come as a surprise to oncologists at academic centers, who tend to see patients with better performance status, higher functional capacity, and an ability to enroll in clinical trials. That’s a selected patient population. This study was an opportunity to see what’s happening in the real world,” she said.
The subgroup of patients achieving long-term survival increased from 4.9% to 12.7% (P < .0001). Long-term survivors were more likely to be younger, white or Asian-Pacific, married, and to have a lower-grade tumor located in the head of the pancreas.
“So, there are changes in long-term survival. We are slowly seeing them, but they are still very small,” she noted. ■
Disclosure: Dr. Golan reported no potential conflicts of interest.
Reference
