Patient-Reported Outcomes Predict Survival in Cancer Trials

By Will Boggs MD

December 09, 2019

NEW YORK (Reuters Health) - Patient-reported outcomes (PROs) independently predict overall survival in randomized clinical trials on cancer, according to a systematic review.

Incorporation of PROs in cancer care allows for increased focus and more accurate information on issues that matter to patients. An increasing body of evidence suggests that baseline PROs are significant predictors of overall survival in patients with cancer.

Dr. Andrew Bottomley from European Organization for Research and Treatment of Cancer (EORTC), in Brussels, and colleagues examined the extent to which PROs have prognostic value across different types of cancer in their systematic review of 44 studies, including more than 28,000 patients with 13 different cancer types.

The main PRO tools used to assess these patients were the EORTC Quality of Life Core Questionnaire (22 studies, 50%) and the Functional Assessment of Cancer Treatment questionnaire (13 studies, 30%).

In 41 (93%) of the studies, at least one PRO domain was found to be significantly associated with overall survival after adjusting for other clinical variables, the team reports in The Lancet Oncology.

The most commonly reported independent prognostic PROs were physical functioning (17 studies, 39%) in nine types of cancer, baseline global health and quality of life (GHQ) scores (15 studies, 34%) in eight types of cancer, and pain (seven studies, 16%).

The addition of these PROs to other clinical factors, however, conferred only a small improvement in prognostic power for predicting overall survival in patients.

Similar prognostic PROs were identified for specific cancer types, but the same PROs were not always found to be prognostic in all studies of the same cancer type.

Moreover, some studies showed unexpected results, such as significantly longer overall survival of patients with brain cancer who had reduced social functioning, lower emotional functioning, and more communication deficits.

The three studies not identifying an association between PROs and overall survival included two studies of patients with advanced head and neck cancer and one of patients with stage-I-IV esophageal cancer.

The researchers acknowledge that "the statistical evidence for the increased predictive accuracy of PROs is scarce, and assessing the magnitude of effects remains complex. Therefore, more quantitative work is required to better understand how and in which clinical settings PROs should be used as stratification factors."

"Future prognostic studies should also report both the statistical and clinical significance of PROs to better capture the magnitude of effects, which would allow for a more precise estimate of the prognostic value of PROs than reporting either the statistical or clinical significance of PROs alone," they conclude.

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

SOURCE: The Lancet Oncology, online December 2, 2019.