BARCELONA, Spain — Active monitoring may be a feasible alternative to immediate surgery for some women with low-risk ductal carcinoma in situ (DCIS), but results from two phase 3 studies presented at 15th European Breast Cancer Conference suggested that longer follow-up is needed before practice changes.

In an interim analysis of the European LORD trial, the ipsilateral invasive breast cancer (iiBC)-free rate was similar in both the active surveillance and surgery groups, at 94.3% vs 90.6% at 2 years and 86.6% vs 88.4% at 4 years.

Study investigator Jelle Wesseling, MD, PhD, of Netherlands Cancer Institute in Amsterdam and Leiden University Medical Center in Leiden, both in Netherlands, said that the incidence of iiBC was “really low” in the trial and that these early data suggested that active monitoring was “not less safe than standard treatment.”

Results from the UK LORIS trial were more mixed. At 5 years, 87.5% of women in the active-monitoring group and 94.2% of those in the surgery group were free of iiBC.

“We have insufficient data to conclude that active monitoring is noninferior to surgery for low-risk DCIS,” said LORIS study investigator Matthew Wallis, MB ChB, of Cambridge University Hospitals NHS Foundation Trust, Cambridge, England. The trial did not meet its prespecified 2.5% noninferiority margin. Although a later clinical survey suggested that a 5.0% margin might have been acceptable, the data were still too uncertain to support active monitoring as an acceptable alternative to surgery.

The invited discussant, Shelley Potter, MBChB, PhD, professor of surgical oncology at the University of Bristol, Bristol, England, said, “Active monitoring may be noninferior but LORIS is underpowered.”

As for LORD, interpretation depended in part on how events were defined. In the trial’s initial intention-to-treat analysis, iiBC detected at primary surgery in the standard-treatment group was not counted as an event, which made active surveillance appear inferior. When those cases were included in a cohort analysis, cumulative iiBC incidence was 9% in the standard-treatment group and 6% in the active-surveillance group.

Based on currently available data, Potter said, “Active monitoring for low-risk DCIS — we’re getting there, we’re just not there quite yet.”

Four Trials, One Question

Potter noted that LORIS and LORD were part of a larger group of trials evaluating whether active surveillance could be a safe alternative to surgery for women with low-grade DCIS. The others are the US COMET trial and Japanese LORETTA study.

“Over-diagnosis, and more specifically over-treatment, is a real issue for patients with DCIS,” Potter said. “DCIS is a disease of screening, and 25% of all the breast cancers detected are DCIS. Traditionally, all DCIS has been managed the same, but our understanding has evolved,” she said.

DCIS is highly heterogeneous, and not all cases will progress to invasive cancer, Potter said. She added that surgery might be appropriate for patients with high-risk DCIS who are likely to progress, but probably not for indolent lesions that are unlikely to do so. The challenge is to accurately identify which patients are at low vs high risk, she said.

Active monitoring also appears to have high levels of patient acceptance. In the COMET trial, crossover was built into the study design and included as an endpoint; more patients crossed from surgery to active monitoring than in the other direction, suggesting that many patients were comfortable choosing monitoring over immediate surgery.

LORETTA was stopped by its data safety monitoring board when the number of iiBC cases breached the prespecified unacceptable threshold that had been set.

LORIS Misses Original Noninferiority Target

Wallis said that LORIS was the first trial to examine whether active monitoring is safe for patients with low-grade (grade 1/2) DCIS. The trial began in 2014 with an initial accrual target of 932 participants, but poor recruitment led to a revised target of 188. A total of 181 women were randomized.

Because many patients did not remain in their assigned group, a per-protocol analysis was not feasible. The researchers instead analyzed patients according to the treatment they actually received.

In the as-treated analysis, iiBC occurred in 12 of 112 women (10.7%) in the active-monitoring group and 2 of 69 (2.9%) in the surgery group. There were no breast cancer deaths, and two patients in each group died of other causes.

Pathology results showed that the median size of the largest invasive focus was 7.0 mm in the active-monitoring group vs 3.5 mm in the surgery group.

There was “no detriment to psychological well-being” with active monitoring vs surgery, Wallis said. Patient-reported outcome measures, including those measuring anxiety or general health status, showed no difference between the groups.

LORD Shifts From Randomization to Patient Preference

LORD was originally set up to be a randomized controlled trial, but it soon became apparent that randomization would not be feasible because many women had strong treatment preferences, Wesseling said.

In a late-breaking oral presentation, he presented data on 1365 participants,1032 of whom opted for active surveillance over surgery.

An initial intention-to-treat analysis indicated that active surveillance was likely inferior to surgery in terms of iiBC-free survival at all the time points considered (6, 12, 24, 26, and 48 months), with an overall hazard ratio of 6.303 strongly favoring surgery.

The trial’s data safety monitoring board recommended that inclusion into the trial be stopped but to continue patient follow-up.

However, cases of iiBC found at primary surgery were not included in the initial analysis, Wesseling said. When they were, the results changed, with higher iiBC-free rates in the active surveillance group before 24 months but not after. This could be due to delayed detection in the active surveillance group, he said.

“Longer follow-up is needed of course to draw final conclusions,” said Wesseling, noting that LORD will continue as a long-term observational registration study.

In a press release issued by the conference organizers, Wesseling said, “For women with low-risk DCIS, these interim results are reassuring. There is currently no indication that active surveillance leads to worse early outcomes compared with immediate surgery.”

Active surveillance means that surgery “remains possible at any time,” added Wesseling. “This study contributes important evidence to the discussion about reducing overtreatment in breast cancer care.”

Isabel Rubio, MD, who was not involved in the study, said in the same press release, “The LORD trial is one of the largest prospective studies worldwide focusing specifically on patients with low-risk DCIS.”

Rubio, who is head of breast surgical oncology at Clínica Universidad de Navarra in Madrid, Spain, added, “With longer follow-up, these findings may help support a more de-escalated approach to care for selected patients, suggesting that carefully monitored active surveillance could offer outcomes similar to surgery, while avoiding overtreatment.”

The LORD trial was funded by KWF Dutch Cancer Society and the Rising Tide Foundation of Clinical Cancer Research. LORIS was funded by the National Institute of Health and Social Care Research. Wesseling, Jones, and Potter had no relevant financial relationships to report.

Sara Freeman is a medical journalist based in London, England.