Decision-analytic model guides cervical cancer treatment

March 25, 2009

MRI and/or PET/CT can help physicians determine appropriate treatment for cervical cancer patients as well as prevent unnecessary therapy, according to a study published in the American Journal of Roentgenology.

MRI and/or PET/CT can help physicians determine appropriate treatment for cervical cancer patients as well as prevent unnecessary therapy, according to a study published in the American Journal of Roentgenology.

Treatment of patients with cervical cancer involves two options. The first is surgery with or without postoperative adjuvant chemoradiation (chemotherapy and radiation therapy). The second is primary chemoradiation.

One problem, however, is overtreatment. Overtreatment has not been studied well in the literature and should become a factor in deciding how to stage patients up front, said lead author Dr. Pari Pandharipande from the Institute for Technology Assessment and Massachusetts General Hospital.

In their decision-analytic model, the researchers found that PET/CT performed better than MRI across all outcomes as a stand-alone strategy. It was preferred to other strategies for triaging patients to the correct primary therapy (AJR 2009;192:802-814).

The researchers developed the model to predict outcomes for a hypothetical patient cohort using International Federation of Gynecology and Obstetrics standards for IB cervical cancer patients who underwent pretreatment MRI, PET/CT, MRI and PET/CT, or no imaging. Pandharipande and her team compared five-year overall survival, percentage of patients receiving correct primary therapy, and percentage of patients spared trimodality therapy (surgery followed by chemoradiation).

Five-year survival was highest with no imaging at 92.37%, followed by PET/CT at 92.36%, and combined MRI and PET/CT at 92.30%. PET/CT had the highest percentage of triage to correct primary therapy at 89.27%, and MRI (alone) had the lowest with 68.21%. MRI combined with PET/CT resulted in the highest rate of avoidance of trimodality therapy (95%), while the lowest rate occurred when patients weren't imaged at all (82.32%).

While a combined MRI and PET/CT strategy is the best to use to avoid trimodality therapy, that may not be the primary goal for some patients.

"If the study had shown there was one particular staging strategy that was best no matter what your outcome of interest was -- survival, triage to optimal therapy, or reduction of overtreatment -- then it would be in a way simpler. What we found at this point is that these staging decisions still need to be made at the individual patient level," Pandharipande said.

Essentially, the study highlights what information to consider in each scenario, according to Pandharipande.

Many radiology studies focus on test performance -- what is a modality's sensitivity and specificity. This is important information during the development, implementation, and evaluation of the widespread use of an imaging technique, she said.

"The next step is how does it affect patients down the line? And that's where my body of research is centered," Pandharipande said.

Since the researchers used a decision-analytic model, the reader has to accept certain assumptions that aren't necessarily the case.

"There are some assumptions they make that are either not clearly spelled out, like stage distribution, or the assumptions that are spelled out I don't necessarily agree with," said Dr. Michael Gold, chair of the ongoing Gynecologic Oncology Group study on the utility of cervical cancer treatment planning with PET/CT and MRI (GOG 233/ACRIN 6671).

Some of the assumptions Gold, an associate professor at Vanderbilt University Medical Center, does not agree with are the percentage of stage IB cervical cancer patients with positive lymph nodes as well as the standard of care at MGH, where anyone with pelvic lymph node metastases gets extended field radiation.

As far as the usefulness of the study, it won't change things much, according to Gold. It's still going to be the gyn/oncologists or the radiation oncologists who send patients in for study -- one will want to operate, the other will want to irradiate, and neither will look for a reason to abstain.

For more information from the Diagnostic Imaging and SearchMedica archives:

Functional MRI boosts early staging of cervical cancerPET emerges as staging exam for cervical cancerPET/CT proves better than MRI for cervical cancer therapy decision making