This article is more than 5 years old. Information may no longer be current.
Conditional survival estimates valuable for intrahepatic cholangiocarcinoma
Click Here to Manage Email Alerts
Conditional survival estimates may provide more accurate quantitative information than OS about the changing survival probability for patients undergoing liver resection for intrahepatic cholangiocarcinoma, according to study results.
Timothy M. Pawlik, MD, MPH, PhD, chief of the division of surgical oncology and director of The Johns Hopkins Medicine Liver Tumor Center, and colleagues conducted a retrospective, international study to compare OS and conditional survival for patients with intrahepatic cholangiocarcinoma (ICC) to determine the impact of several clinicopathologic prognostic factors on each survival measure. Conventional OS measures are determined by static factors at or near the time of surgery, whereas conditional survival — defined as the probability of surviving for an additional 3 years — is estimated by including the years a patient has already survived, according to study background.
Timothy M. Pawlik
The researchers identified 535 patients who underwent resection of ICC between 1990 and 2013.
The study data indicated that OS decreased over time, from 39% at 3 years to 16% at 8 years (P = .002), whereas the 3-year conditional survival increased among those patients who survived.
Therefore, the 3-year conditional survival for patients who had already survived 5 years was 65% and the 8-year OS rate was 16% (P = .002).
Poor OS was associated with large tumor size (HR = 1.02; 95% CI, 1-1.05), multifocal disease (HR = 1.49; 95% CI, 1.19-1.86), lymph node metastasis (HR = 2.21; 95% CI, 1.67-2.93) and vascular invasion (HR = 1.39, 95% CI, 1.1-1.75).
In each of those subgroups, the conditional survival exceeded the actuarial OS. For example, patients with vascular invasion had an OS rate of 15% at 6 years compared with a conditional survival of 50% at 3 years. Likewise, patients with lymph node metastasis had an OS of 11% at 6 years and a conditional survival of 49% at 3 years.
The investigators admitted there were several limitations with their study, including selection bias as well as the fact that the multicenter study led to heterogeneity in the therapeutic approaches. Also, the researchers did not analyze the impact of adjuvant therapy and its potential effect on OS and conditional survival.
“The importance of the study is that it provides data for patients with ICC about survival following surgery in a more dynamic matter,” Pawlik told HemOnc Today. “Most patients who return to a clinic years after an operation want to know their survival probability based on the fact that they have survived to the current time — not the odds of survival based on factors from the time of surgery, which in some instances can be 2, 3 or even 5 or more years ago.
“Conditional survival provides patients information about their prognosis in a more dynamic way that takes accrued survival time into account,” Pawlik said. “This information can be used in counseling patients who are in surveillance or survival programs after being treated for their surgery.”
The researchers concluded that conditional survival estimates would be more accurate when determining prognosis over time than OS for patients with ICC.
Conditional survival quantifies the knowledge surgeons have had regarding improved survival outcomes in patients who have lived longer without recurrence, Jake E. Krige, MSc, FRCS, FCS(SA), and Delawir Kahn, ChM, FCS(SA), both of the University of Cape Town, South Africa, wrote in an accompanying editorial.
“Some would argue that conditional survival uses a biased advantage by deferring a decision and later identifying and selecting survivors who will invariably do better than the initial cohort in whom survivorship is diminished by including overall mortality,” Krige and Kahn wrote. “The drawbacks of conditional survival data are that increasing accurate information is delayed and only becomes available with the passage of time, during which precision is augmented by using progressive chronological data maturity. In contrast, traditional prognostic scoring systems are derived from information obtained solely at the time of surgery and therefore become increasingly inaccurate as time passes from the inception of surgical intervention.” – by Anthony SanFilippo
For more information:
Timothy M. Pawlik, MD, MPH, PhD, can be reached at 600 N. Wolf St., Blalock 688, Baltimore, MD, 21287; email: tpawlik1@jhmi.edu.
Disclosure: The researchers report no relevant financial disclosures.
Click Here to Manage Email Alerts