A new journal, JAMA Oncology, has published a report on a controversial breast condition called DCIS. This stands for ductal carcinoma in situ. In an unknown fraction of cases, DCIS leads to invasive breast cancer. The Times placed an article about the analysis on its home page and front cover, which renewed discussion.
What’s controversial about DCIS is whether or not it should be considered “cancer.” Most oncologists and pathologists agree that it’s a Stage 0 tumor. That’s because in DCIS, the cells lining the milk ducts (“ductal”) in the breasts, which might deliver milk to a woman’s nipple, have abnormal, malignant-appearing features. DCIS is non-invasive: by definition, the abnormal cells haven’t broken through the ducts into other breast tissue.
Why the articles, in both the medical journal and news, hit a nerve is because DCIS is so common. Each year, some 60,000 U.S. women are informed they have this condition. Because DCIS is usually detected by biopsy after screening, and not by a woman feeling a lump (there is no lump), the known incidence has shot up in parallel with mammography, since the late 1980s.
Today, DCIS accounts for approximately 1 in 5 breast cancer-ish diagnoses in the United States. The practical problem with DCIS is that most oncologists, surgical and medical, will admit they’re not sure what to do about it.
We could label DCIS as cancer or not cancer, or “premalignant,” but that wouldn’t solve the dilemma of how to best to treat it. This new study suggests that DCIS is associated with greater risk of death from breast cancer in black women and in younger women. Instead of brushing off DCIS (“don’t worry”), or not telling women about it (as some have suggested), we should promote research to better understand the condition and figure out how to best manage it, for the future.
The condition tends to be more diffuse, i.e. multifocal – meaning that spots of abnormal cells aren’t connected – compared to invasive breast cancer. So lumpectomy with an aim to carve it out can be tricky; it’s hard, though not necessarily impossible, to get clean surgical margins. Radiation is often given for DCIS, after lumpectomy. But that carries toxicity and risk, which is one of the reasons that some women choose, and some doctors recommend, mastectomy.
The benefit of giving radiation – or any treatment, for DCIS – is uncertain. That’s because in the 1990s, when the condition started appearing after mammography, many (in my community, most) doctors recommended mastectomy for DCIS, because of its diffuse nature and the general view, back then, that DCIS is an early-stage cancer or, at least, premalignant. The fact is, we had no data to guide us then, and we still have very little, in 2015. Which is why the JAMA Oncology paper is helpful.
Although retrospective, and with too-short follow-up, the new article provides a lot of information on what happens to many women (over 108,000) who received a diagnosis of DCIS between 1988 and 2011. Of note, the average age of women in this analysis, culled from cancer registry (SEER) data, was 54 years. The study suggests future directions of research, and questions women might ask of their physicians before choosing treatment for DCIS.
The researchers, based at Toronto’s Women’s College Hospital, looked only at cases of women with DCIS up to age 70 years. Overall, what they found is good news: the death rate from breast cancer after DCIS, calculated by careful extrapolation to a 20-year mark, was only 3.3%, overall. In other words, this condition shouldn’t be terrifying, because it’s an unlikely killer. However, this apparently low death rate should be considered in the context of almost all of these under-70 women having received surgical and other therapy; it could be that many or perhaps most of the 108,000 women would have gone on, without intervention, to develop invasive or metastatic and incurable disease. No one can say, based on this analysis.
A major, and unfortunate – i.e. terrible, caveat is that death rates from breast cancer, within 20 years, for women who had DCIS diagnosis before age 35 was 7.8% (i.e. nearly 8% of those young women with DCIS died from breast cancer before reaching the age of 55 years). And that for black women in the registry, the death rate was 7% (vs. 3.0% for non-Hispanic whites). One might infer from that observation, if blacks are getting less treatment or delays due to a correlation of having lesser insurance, that therapy might make a very positive impact. Again, that is uncertain.
The paper has several limitations. Among those, the “real” follow-up averaged only 7.5 years. That’s short for any analysis of non-metastatic breast cancer, and especially so if the mean age for those studied was only 54 years at diagnosis. Another problem is that it’s not a randomized prospective study. As far as conclusions about occurrence of invasive breast cancer in the same or other breast after DCIS, or in other parts of the body, I think it’s muddled. Because women with DCIS are prone to developing new, invasive breast cancer, it could be that later metastases arise due to tumors that grew apart from the DCIS.
Complicating the picture – but in a way that will, ultimately help – is pathology information that has emerged since 1990. Doctors now appreciate that DCIS is not all the same. Just as polyps and moles and other premalignant or Stage 0 cancers vary, so does DCIS. Some cases have genetic and molecular features that make them more likely to become invasive and cause harm. A few months ago I wrote about Oncotype Dx, a pathology test (offered by Genomic Health, and covered by some insurers including Medicare) that can distinguish between DCIS cases that are likely to recur, and probably warrant treatment, and those which are unlikely to recur, and so might be safely monitored after biopsy or lumpectomy.
Given the huge number of DCIS cases, and the intolerably high death rate from breast cancer – a problem that can’t be ignored, this is an important area of research, so that women get neither under-treatment nor over-treatment. That’s my bottom line.
