Can MRI Be Used to Identify 'Biologically Inert' DCIS?

Pam Harrison

March 13, 2015

VIENNA — MRI of the breast can detect more biologically relevant ductal carcinomas in situ (DCIS) than mammography and could be used to stratify risk for a more targeted therapeutic approach, new findings suggest.

MRI should be used to arrive at treatment decisions that are better tailored to the prognostic profile of the tumor, says Christiane Kuhl, MD, of RWTH Aachen University, in Germany, who is a leading MRI expert.

"People have now shown that there are two different, distinct pathways along which DCIS progresses — the so-called low-grade molecular pathway, where low-grade DCIS behaves more or less indolently and may not progress through a lifetime or progresses very slowly, vs high-grade DCIS, which is more likely to progress to high-grade invasive cancer, and progress more rapidly," Dr Kuhl told Medscape Medical News.

"The problem is to predict which DCIS will progress and which DCIS will stay within the milk duct and not become an invasive cancer," she added. "With MRI, we have a biomarker for DCIS that helps us make this prediction."

The key feature separating low-grade, nonprogressive DCIS from tumors that are more likely to progress is the degree to which the lesions show enhancement on MRI.

As Dr Kuhl explained, enhancement in DCIS requires a chelate such as gadolinium to accumulate within the milk ducts.

"The chelate is a large molecule that needs to penetrate the ductal basal membrane, which is surprising, as it is usually not possible to do this," Dr Kuhl said.

"And that tells you that we have pathologically increased ductal wall permeability, which is a beautiful biomarker for protease activity."

Thus, MRI can be seen as a biomarker for a process that has all the proteomic tools to help the tumor achieve invasive growth.

These properties include angiogenic activity required to feed the tumor and any invasive disease that arises from it and protease enzymes needed to digest the ductal basal membrane.

"We can exploit the fact that MRI is a biomarker and that depiction of DCIS depends on DCIS angiogenic and protease activity," Dr Kuhl confirmed.

Dr Kuhl and colleagues have repeatedly shown that DCIS tumors that do not enhance are more likely to be biologically inert and not possess the characteristics needed for invasive growth.

Conversely, those that exhibit ductal enhancement are most likely to be high-grade tumors and are made visible because of their proteomic characteristics.

In fact, the sensitivity of MRI increases with increasing biologic aggressiveness of DCIS, Dr Kuhl noted.

This contrasts to the sensitivity of mammography, which decreases with increasing biologic aggressive of DCIS, she added.

Important Clinical Implications

Being able to distinguish MRI-"occult" DCIS from those rendered visible by the chelating process has important clinical implications.

Currently, the American College of Radiology mandates biopsy for all lesions that are classified as "suspicious" on mammographic screening ― this category (BI-RADS 4) includes any lesons that cannot be classified as "definitely benign" or "definitely malignant."

Because of the high incidence of BI-RADS 4 findings, there are a predictably high number of unnecessary biopsies performed because of false positive results, Dr Kuhl pointed out.

In a recent study designed to assess the ability of MRI to more clearly distinguish BI-RADS 4 lesions as classified on mammography or ultrasound (Radiology. 2015;274:343-351), "we found that if you added MRI, the positive predictive value increased to over 75% for all mammographic findings, while the negative predictive value [NPV] increased to almost 100%," Dr Kuhl reported.

Specifically, in women without pure calcification, there were no false negative diagnoses with MRI, whereas in women with pure calcification, "we had three false negatives in 78 women, and all three of these were small, low-grade DCIS," Dr Kuhl said, adding that they had an additional three true positives, and all three were in women whose index lesion was BI-RADS 4 on original assessment.

These findings indicate that with an NPV of 100%, MRI helps avoid unnecessary biopsies in more than 90% of women, depending on the specific mammographic findings.

"We can also say that MRI failed to diagnose low-grade DCIS in three women, and this is exactly what's happening in prostate cancer right now, where we use MRI to find high-grade prostate cancer, and if the MRI is negative, we know that we are not dealing with high-grade disease, and urologists are happy to leave the tumor alone," she said.

This is the same direction that Dr Kuhl would like to see breast MRI take in the diagnosis and management of DCIS.

Currently, women who are diagnosed with DCIS — mammographically detected disease based on morphology alone, as Dr Kuhl pointed out ― are subjected to the full range of surgical and radiation therapy that women with more invasive cancers are offered.

Using MRI to depict biologically important DCIS would allow clinicians to stratify women already diagnosed with DCIS to those who do not require radiation therapy or who might be able to forego treatment altogether if their DCIS has no enhancing features on MRI.

"I'm not saying that all low-grade DCIS is unimportant, because there are women who die because of low-grade, invasive cancers," Dr Kuhl said.

"Rather, I would make a claim that nonenhancement DCIS is unimportant, and we will hopefully see a time when biomarker or "in vivo" imaging of DCIS will be considered as important as just having the morphological image of the breast provided by mammography," she said.

Potentially Important Observation

Asked to comment on Dr Kuhl's research, Daniel Kopans, MD, professor of radiology, Harvard Medical School, Boston, Massachusetts, told Medscape Medical News that her observations are potentially very important but that it may be premature to suggest that findings on MRI may be safely used to stratify women and DCIS lesions.

"Furthermore, this study does not prove that DCIS that does not enhance can be left untreated, since there was no treatment involved and no long-term follow-up," Dr Kopans said.

He does agree that Dr Kuhl's observations are "highly suggestive" that nonenhancing lesions are truly indolent lesions, but he emphasizes that even Dr Kuhl herself acknowledges that low-grade DCIS, given enough time, can progress to invasive breast cancer and lethal disease.

"I'm sure Dr Kuhl would agree that she does not have proof that nonenhancing DCIS lesions will not progress to invasive disease if given enough time," Dr Kopans observed.

"Additional prospective studies need to be done to see if this is a safe approach to DCIS."

Dr Kuhl and Dr Kopans have disclosed no relevant financial relationships.

European Congress of Radiology (ECR) 2015. Abstract A-153. Presented March 5, 2015.


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: