New Tool Predicts Survival in Patients with Head and Neck Tumors

Zosia Chustecka

March 19, 2007

March 19, 2007 — Cancer-survival statistics are usually calculated from the time of diagnosis, but for patients who have survived for some time already, the less commonly used "conditional-survival" (CS) calculation paints a clearer picture of their prognosis. Such a calculation has been made for the first time for patients with head and neck cancers, and the data often provide good news for these patients, researchers report in the April issue of Cancer.
"We found that most patients' prognosis improved after they survived longer periods of time from when they were first diagnosed and treated," commented one of the authors, Samuel Wang, MD, PhD, from the department of radiation medicine at the Oregon Health and Science University Cancer Institute, in Portland.

CS statistics have already been calculated for a number of other cancers, including breast and prostate, but never before for head and neck squamous cell carcinomas (HNSCC), as far as the authors are aware. From a practical point of view, these statistics are a "boon to both patients and clinicians," they comment. The data can be used as a "lookup table" to stratify patients over time, allowing patients to realize that, as time progresses, their risk for mortality changes. In addition, clinicians can use the data "in an effort to institute data-driven optimization of posttherapy surveillance."

Substantial Increase in Survival Over Time in Many Cases

The researchers used data collected by the Surveillance, Epidemiology and End Results (SEER) Program of the National Cancer Institute (NCI). They identified 76,181 patients with HNSCC between 1978 and 1998 for whom there were at least 5 years of follow-up data.

At this 5-year end-point, the researchers calculated observed conditional survival (OCS), which is defined as the probability of surviving all-cause mortality, as well as relative conditional survival (RCS), which denotes the calculated probability of not dying of cancer. The RCS is useful to define the risk of dying from cancer, they explain, while the OCS is more useful for patient prognostication of raw risk for death, bearing in mind that patients with HNSCC may have significant mortality risk from tobacco and alcohol use.

They found that for the whole HNSCC population, the 5-year OCS increased from 47.8% at diagnosis to a plateau of 64.4% within 3 years.

The RCS showed that, broadly speaking, after 6 years there is an increasing risk of HNSCC patients dying from other disease processes (such as heart disease or stroke) rather than their primary cancer. "Thus, a 'generalized' HNSCC patient might be expected to be 'cured' after 6-year disease-free survival insofar as their risk of mortality due to cancer is equivalent to competing-cause mortality risk of a 'generalized' noncancer patient," the authors write. However, the crude overall survival is never quite equivalent to a noncancer population, as some risk of dying from the index cancer or associated sequelae is "overlaid" onto the risks that everyone faces.

However, the team also found some differences in the patterns of prognosis for the different subtypes of HNSCC. Most of the subsites showed substantial incremental increases in OCS, plateauing rather rapidly at 3 years after diagnosis, and this reflects the trend for HNSCC cumulatively, they comment. The greatest increase in CS was seen in oropharyngeal cancers, where the patient 5-year survival more than doubles over the first decade of surveillance postdiagnosis (from 26.5% to 60%).

However, there were some exceptions to this pattern. HNSCC of the lip or gum showed a negative OCS progression over time, with poorer survival at 5 years after diagnosis. This would suggest that such patients may need increased comparative vigilance, the researchers comment in their paper. Elaborating in comments to Medscape, however, first author Clifford David Fuller, MD, from the University of Texas Health Science Center at San Antonio, said: "At the present time, we do not advocate changing surveillance patterns based solely on our findings alone. The data we present should be utilized in concert with a thorough evaluation of the individual patient's specific disease status, treatment profile, current health, and known comorbidities, in addition to epidemiological data such as 5-year survival and conditional survival."

Conditional Survival Part of Spectrum of Patient Care

"Conditional survival should be considered within the context of the full spectrum of patient care and should not be used as a single reference for risk assessment," the authors write. They also note that risk stratification based on large-scale population data such as SEER must always be interpreted clinically for individual patients, whose specific risk factors must be carefully weighed.

The next step will be to confirm these findings with additional large databases, and the researchers plan to create an online tool that will let physicians determine a patient's estimated survival in real time. A preliminary version of an online conditional survival model has been submitted as an abstract to the American Society of Clinical Oncology annual meeting, but "we want to carefully verify that the observations we noted in this manuscript hold true in other clinical databases before releasing a final version for clinical use," Dr. Fuller commented.

"Our goal is to demonstrate that there is great utility in dynamically modeling changes in survival probability over time, as well as specifically illustrating the need for risk modeling tools in head and neck cancer," Dr. Fuller told Medscape.

"Our main message to oncologists (and their patients) is one of optimism and caution in the face of a still deadly disease," he continued. "Optimism because, even in those patients with initially dismal prognoses, the 5-year survival probability increases for a period of time after diagnosis. Caution is warranted, however, since our data suggest that this survival improvement is limited to a minority of patients and stabilizes after 2 to 3 years in most head and neck cancers."

Cancer. 2007;109:1331-1343.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: