Conditional survival provides better estimate of cholangiocarcinoma prognosis

Last Updated: 2015-04-07

By Will Boggs MD

NEW YORK (Reuters Health) - Conditional survival estimates, which take into consideration the years already survived, better inform patients with intrahepatic cholangiocarcinoma (ICC) of their likely prognosis after liver resection, according to a retrospective international study.

"Physicians need to modify their estimate of prognosis when counseling patients the longer out the patient is after the initial treatment," Dr. Delawir Kahn, from the University of Cape Town Health Sciences Faculty, Cape Town, South Africa, told Reuters Health by email. Dr. Kahn, along with Jake E. Krige, also of Cape Town, wrote an editorial related to this report,

Conventional overall actuarial survival estimates for ICC are based on static factors determined around the time of surgery, and some of these prognostic factors have been called into question.

In other settings, conditional survival estimates have been shown to offer a better guide to long-term prognosis and to provide information important for guiding surveillance.

Dr. Timothy M. Pawlik, from the Johns Hopkins University School of Medicine, Baltimore, Maryland, and colleagues used data from an international multi-institutional database to analyze conditional survival in 535 patients who underwent curative intent liver resection for ICC.

The median overall survival was 27.4 months, with most disease-specific deaths (65.6%) occurring within 24 months after surgery. Larger tumor size, multifocal disease, vascular invasion, lymph node metastasis, and advanced American Joint Committee on Cancer (AJCC) stage significantly worsened overall survival.

Actuarial survival decreased progressively from 75% at one year to 39% at three years, 25% at five years, and 16% at eight years, according to the April 1 JAMA Surgery online report.

In contrast, conditional survival increased over time among the patients who remained alive. For example, the probability of surviving to year 8 after already having survived to year 5 was 65%, compared with the actuarial eight-year survival of 16%.

The conditional survival probabilities of an additional three years were 38% for patients who had already survived for one year, 43% for those who had already survived for two years, and 44% for those who had already survived for five years.

The calculated three-year conditional survivals exceeded actuarial survival for all prognostic subgroups, and the differences were more substantial among patients initially predicted to have worse prognoses.

"Conditional survival estimates may provide critical quantitative information about the changing probability of survival over time among patients undergoing liver resection for ICC and therefore can be of significant value to patients and health care professionals," the researchers conclude.

"The benefit of using conditional survival is mainly for the patient," Dr. Kahn said. "The longer out you are after the treatment of certain cancers, the greater the chance of surviving longer term. This is important to know from a patient point of view; e.g., a patient who at the time of the operation only has a 40% chance of surviving five years; but if three years later he/she is still alive, the chances of getting to five years would be much greater. This is extremely valuable and reassuring for the patient."

"Does it influence management?" Dr. Kahn wondered. "Possibly, but in a more subtle manner."

Dr. Gregory J. Gores, from Mayo College of Medicine, Rochester, Minnesota, told Reuters Health by email that he was "not overly impressed with this paper," partly because it "only takes into account patients who have surgery, which is the minority of patients."

" may help counsel patients who are out several years in regards to survival, need for surveillance, and help focus them on other long-term health risks, such as cardiovascular risk factors, which are usually not relevant when faced with short-term survival risks," Dr. Gores said. "The real question is whether patients need adjuvant therapy post-operatively, and this paper does not help in this regard."

Dr. Boris Blechacz, from the University of Texas MD Anderson Cancer Center, Houston, agreed with the points made by Dr. Gores: "The study addresses a very small patient population as the majority of cholangiocarcinoma patients (in particular intrahepatic cholangiocarcinoma patients) are diagnosed at advanced-stage disease and up to 45% of the few deemed resectable are found to be unresectable during laparotomy."

"The effect of age on conditional survival has not been sufficiently addressed but would be useful in decision making for patients with advanced age and increased surgical risk," Dr. Blechacz said. "There are studies on other malignancies that have shown age-specific variations in conditional survival up to five years post-operatively."

"In summary," Dr. Blechacz concluded, "the study might help to put a number on relative survival in a very small patient population, thereby giving these patients a quantifiable prognosis and guide other decisions in regard to their preventive care. However, the study does not help to guide treatment does not answer some of the important unanswered questions (i.e., adjuvant treatment)."

Dr. Pawlik did not respond to a request for comments.

The authors reported no external funding or disclosures.

SOURCE: http://bit.ly/1GIkToY and http://bit.ly/1Ph7ubv

JAMA Surg 2015.