Ursula Matulonis, MD, on Concordance of CA-125 and Radiographic Progression in Ovarian Cancer

By Mark Fuerst

About half of patients with platinum-sensitive relapsed ovarian cancer (PSROC) on maintenance poly (ADP-ribose) polymerase inhibitor (PARPi) therapy who showed RECIST progression did not have cancer antigen-125 (CA-125) progression, challenging current guidelines, according to a new study.

Angelina Tjokrowidjaja, MBBS, PhD, of the University of Sydney in Australia, and colleagues explained that although CA-125 is recommended by treatment guidelines and is widely used to diagnose ovarian cancer recurrence, the value of CA-125 as a surrogate for disease progression and its concordance with radiologic progression are unclear. The team’s study in the Journal of Clinical Oncology examined the concordance of these two types of progression for PSROC patients.

Co-author Ursula A. Matulonis, MD, chief of the Division of Gynecologic Oncology at Dana-Farber Cancer Institute and and Harvard Medical School in Boston, details the data in the following interview.

What does the study add to the literature?

Matulonis: In this pooled analysis that included over 1,200 patients with PSROC treated in four randomized trials of maintenance therapy with a PARPi (Study 19, SOLO2, ARIEL3, and NOVA), we found that approximately one in two patients with radiologic progression did not have CA-125 progression. Of these, the majority had normal CA-125 at the time of computed tomography (CT) progression.

This data saw that for those on PARPi in the recurrent setting, CA-125 progression by itself may not necessarily be a reliable marker for cancer growth. In this situation, CA-125 testing was not adequate in up to 50% of patients. These are eye-opening results. We should do more periodic testing to look at progression and cancer growth.

If we are going to pick up cancer growth, it’s best to do so in an earlier setting before significant symptoms appear. Patients may then be eligible for surgery or a clinical trial.

PARPi have side effects, including a drop in blood count, gastrointestinal toxicity, fatigue, and increased risk of secondary malignancy. There’s also financial toxicity. These drugs are expensive and some patients pay out of pocket. For all these reasons, detecting recurrence in its earliest form would be ideal.

What are the highlights of the study?

Matulonis: Of the 859 patients (68%) without CA-125 progressive disease, 382 also did not have RECIST [Response Evaluation Criteria in Solid Tumors] progressive disease, for a negative predictive value of 44%. Within the treatment arms, positive predictive value remained high (91%), but negative predictive value was lower on placebo (25%) than with PARPi (53%).

Of 477 patients with RECIST-only progressive disease, most (95%) had a normal CA-125 at the start of maintenance therapy, and the majority (64%) had CA-125 that remained within normal range.

The types of recurrence had different profiles. Solid organ recurrence without peritoneal disease was more common for those with RECIST-only progression versus those with both CA-125 and RECIST progression (36% vs 24%). Cancer cells around the intestines are more consistent with elevated CA-125 than with solitary liver or lung metastases.

What is the importance of the probability of disease when interpreting negative test results?

Matulonis: The presence of normal CA-125 potentially gives false reassurance to the oncologist and patient. It’s important to do periodic scans and do CA-125 testing, but also to listen to the patient and her symptoms. Subtle symptoms can be important for recurrence, such as difficulty with digestion, abdominal bloating, and constipation. These symptoms may indicate where the recurrence is, and are hard to see on CT scans.

How are the findings potentially generalizable to patients in first-line or later-line maintenance settings?

Matulonis: These trials are all in the recurrent maintenance setting with no cap on how many prior lines of therapy. It’s important for upfront trials, such as SOLO1 with olaparib, and studies with niraparib to do the same type of analysis. I do periodic scans in the upfront setting, where patients are on a PARPi for a defined period of time, because I want to pick up disease earlier.

What are the implications of the findings for clinical practice guidelines and future research?

Matulonis: These results can be put into guidelines, but the results are important enough to use right now. Oncologists should implement periodic radiographic imaging for patients on PARPi, and also consider doing this in the newly diagnosed setting.

We could do longer-term follow-up and similar analysis in newly diagnosed patients. No trial has examined the interval of surveillance scans. We don’t know the ideal interval. Is it every 3, 4, or 6 months? What’s the ideal scan — Is it CT, PET or PET/CT? A cost-benefit analysis would be another way to look at these randomized trials.

What’s the main take-home message for practicing oncologists?

Matulonis: Based on this pooled analysis, patients with PSROC on a PARPi should continue CA-125 testing, but clinicians should also incorporate radiographic imaging as well. Half the patients with recurrent disease and normal CA-125 will have progression on CT that would not have been detected on CA-125 alone.

These are practicing-changing results. By detecting recurrence early, we can take a patient off a PARPi that has the potential for significant side effects, reduce the cost of treatment, and think about sending the patient to surgery or into a clinical trial.

Read the study here and expert commentary about it here.

This article was published by: MedPageToday