Read Original Release

A “practice-changing” test for DCIS — but does it impact results that matter to patients?

Journal of Surgical Oncology Publishes Study Showing Oncotype DX® DCIS Score™ Result Changes Treatment Recommendations in 30 Percent of DCIS Breast Cancer Patients Physicians Ranked Oncotype DX DCIS Score Most Important Factor in Treatment Planning

Our Review Summary

DCIS is usually diagnosed from a screening mammogram.

DCIS is usually diagnosed from a screening mammogram.

This company news release describes a study which found that the Oncotype DX DCIS Score test changed recommendations for radiation treatment among women with Ductal Carcinoma In Situ (DCIS) — a condition marked by abnormal, pre-cancerous cells inside the breast. The terminology employed by this news release is likely to be confusing to many readers. The body text includes a description of DCIS as a “non-invasive, indolent form of breast disease ” that often doesn’t progress to breast cancer. But the headline refers unambiguously to “DCIS Breast Cancer.” We think the release could have done a better job of capturing the controversy over the very name for this condition, what it means, and the confusion it causes women. There are also some important gaps in the information provided about the study and the test. The news release doesn’t mention the costs of this new test or acknowledge that the study did not assess clinical outcomes among patients (only changes in treatment recommendations that may or may not be appropriate). The release is also silent on the fact that the study was funded by Genomic Health, Inc, which makes the test, and that the study authors included company employees and researchers who are paid to speak and consult for the company.


Why This Matters

New tests come on the market all the time, but that doesn’t mean that insurance will cover them or that they will provide a real clinical benefit to patients. Currently, the Oncotype DCIS score is not included in guidelines from the National Comprehensive Cancer Network (NCCN) and is not uniformly accepted by oncologists as an important test to determine treatment recommendations. There is not enough good-quality evidence to determine if the test provides a benefit, and these new study results are unlikely to change that situation. The NCCN guidelines for DCIS state: “whole breast radiation therapy following lumpectomy reduces recurrence rates in DCIS by about 50%. Approximately half of the recurrences are invasive and half are DCIS. A number of factors determine local recurrence risk: palpable mass, larger size, higher grade, close or involved margins, and age <50 years. If the patient and the physician view the individual risk as “low”, some patients may be treated by excision alone. All data evaluating the three local treatments [lumpectomy with radiation, mastectomy, lumpectomy without radiation] show no differences in patient survival.”


Does the news release adequately discuss the costs of the intervention?

Not Satisfactory

The release never mentions the cost of the test and there is some concern that insurance carriers may not cover costs for tests of DCIS. Insurance coverage is variable and appeals do not always result in reversal of the denial.

Does the news release adequately quantify the benefits of the treatment/test/product/procedure?

Not Satisfactory

The news release highlights the fact that information from the test changed physician recommendations for treatment. Some physicians changed to recommending whole breast irradiation while others recommended against. A company official says the study is evidence that the test “can play a key role in enhancing cancer care for patients with DCIS, and we believe that these results will support expanded adoption of and access to the DCIS Score.”

But while changing practice recommendations may be an indication that the information provided by the test could be useful and beneficial, it’s not evidence that the test actually improves results for patients. Without clinical outcomes, it’s difficult to know the true extent of benefits — if any. The release should have qualified the findings with some mention of the fact that the study was not designed to assess outcomes that are meaningful to patients. Without such a warning it’s too easy for readers/patients to be misled about the real impact of the test. Any statement along those lines would have earned the release a Satisfactory rating here.

Does the news release adequately explain/quantify the harms of the intervention?

Not Satisfactory

The news release emphasizes the potential beneficial effect of test results on refinement of treatment recommendations. It doesn’t contemplate or even mention the possibility that an inaccurate test result could cause harm — for instance by recommending a conservative course of therapy when a more aggressive approach is warranted, or vice-versa. While the study itself didn’t assess outcomes or potential for harm, as with Benefits we’d expect some nod to the fact that this is an area of uncertainty that requires more research.

Does the news release seem to grasp the quality of the evidence?

Not Satisfactory

The study that’s the basis for this news release discussed a number of limitations that aren’t addressed in the release — for example that this was a utility study, not designed to assess outcomes, and that actual treatments were not assessed, just the change in treatment recommendations. As the study authors put it: “The primary limitation of this study is that, by design, this study is an analysis of how decisions are made and the impact of the DCIS Score result on the treatment recommendation, but not an assessment of whether those decisions were made appropriately.”

There was also selection bias in that they included patients only with perceived low recurrence risk. The study did not randomly select patients for inclusion and instead approached women with a low risk of local recurrence. The news release needed some acknowledgment of such limitations.

Does the news release commit disease-mongering?


The story did not commit disease mongering. Although it didn’t mention or emphasize the fact that some physicians don’t consider DCIS to be cancer at all, it does call DCIS “indolent” and says that it may be treated too aggressively in some cases.

Does the news release identify funding sources & disclose conflicts of interest?

Not Satisfactory

The original study listed financial conflict of interests among the study authors — including the fact that several study authors are paid speakers and consultants for Genomic Health, which makes the test. One of the authors is an employee of the company. But the new release did not provide any of this context.

Does the news release compare the new approach with existing alternatives?

Not Applicable

The release does not discuss alternatives to the Oncotype DX DCIS Score, except for the absence of such testing. But since there are no other commercially available tumor genomic tests for DCIS, we’ll rate the story Not Applicable here.

Does the news release establish the availability of the treatment/test/product/procedure?

Not Satisfactory

There was no mention on how available the Oncotype DX DCIS Score test is in clinical offices and no mention on whether it is covered by insurance. The test is, in fact, widely available — the physician sends a requisition form either to the pathology department, which will send the tissue to Genomic Health, or to Genomic Health, who will then request the tissue samples from pathology.

Does the news release establish the true novelty of the approach?


The news release states that there is no other commercially available tumor genomic test for DCIS.

Does the news release include unjustifiable, sensational language, including in the quotes of researchers?


The press release did not use unjustifiable language. But as noted above, we think the terminology used is confusing.

Total Score: 3 of 9 Satisfactory


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.