It’s been a long time since I’ve written much about the one area where even quacks, cranks, and antivaxxers can’t argue that I don’t have expertise: Breast cancer. Oddly enough, it’s been a while since I’ve written about it, other than early after the introduction of COVID-19 vaccines, when radiologists were noting reactive enlarged lymph nodes under the arm after vaccination that were sometimes leading to unnecessary biopsies Oddly enough, even though I was aware of this study last month and had been meaning to write about it, what tweaked me to actually do so was an article that I encountered from an antivaxxer whom I’ve written about periodically, Maryanne Demasi, DCIS: should we stop calling it “breast cancer”?, with a blurb of, Renewed calls to minimise the overdiagnosis and unnecessary treatment of breast abnormalities.
Well, yes and no.
Demasi’s post is a paid Substack post; so I can’t read the whole thing. However, I distinctly remember it showing up at first as the full post. What I can say is that she is exaggerating when she extrapolates from the results of this study an implication that mammographic screening is useless. I also find it telling that, in her post on X, the hellsite formerly known as Twitter, she tags certain contrarians, specifically Drs. Vinay Prasad and Peter Gøtzsche, the latter of whom was her co-author on a study that used deceptive methodology to exaggerate the risk of COVID-19 vaccination. Also, keep in mind that Ms. Demasi is not exactly the greatest source for a take on this study. That being said, the introduction sounds deceptively reasonable:
The debate over how to manage ductal carcinoma in situ (DCIS) has gained momentum following a recent study suggesting that surgery might not be necessary for women diagnosed with the condition.
Often described as “stage zero breast cancer” or “a very early form of breast cancer,” DCIS is not actually cancer but rather a cluster of abnormal cells contained within the milk ducts.
Despite its non-cancerous status, the use of terms like “carcinoma” or “cancer” can cause considerable anxiety, prompting many women to choose aggressive treatments such as surgery and radiation—interventions that may not offer any clear benefit to their outcomes.
Last month, a new randomised trial, published in JAMA once again, challenged the necessity of surgery for DCIS.
She’s referring to the COMET trial, which stands for Comparing an Operation to Monitoring, with or without Endocrine Therapy. I’ll admit that it’s a rather tortured acronym, but the name basically describes the study: It’s a multicenter randomized controlled clinical trial that compares standard-of-care therapy, which includes surgery, radiation, and endocrine therapy (estrogen-blocking drugs like tamoxifen or aromatase inhibitors) with close monitoring and intervention only if there are signs of progression. I also note something that Ms. Demasi neglects to mention, namely that the trial is only enrolling women with low-risk DCIS. What that means specifically, I’ll get to in a minute, but we’re not talking extensive DCIS or DCIS with features that suggest it has a high propensity to progress to invasive cancer (or that it is likely to have already progressed, with the core needle biopsies having missed the invasive component. I also can’t help but mention that “de-escalation,” sometimes referred to colloquially as “less is more,” is not a new or radical concept when it comes to breast cancer, and particularly its precursor DCIS. Debates over the best treatment for DCIS (or whether it even needs to be treated at all) go way back. I’ve even written about some of them, such as in my 2015 post How should we treat “stage 0” breast cancer? It’s therefore not entirely surprising that the study by Hwang et al, which I’ll discuss in more detail shortly, found no statistically significant difference in the primary outcome examined, the 2-year cumulative risk of ipsilateral (same side) invasive cancer diagnosis, according to planned intention-to-treat and per-protocol analyses. The authors concluded:
Women with low-risk DCIS randomized to active monitoring did not have a higher rate of invasive cancer in the same breast at 2 years compared with those randomized to guideline-concordant care.
First, though, a refresher course on what DCIS is.
DCIS: Cancer, premalignant lesion, or something else?
DCIS is, as Ms. Demasi states, commonly referred to as “stage 0” breast cancer. Personally, I’ve always had a bit of a problem with that designation and often say so to patients, preferring to categorize it as a premalignant lesion that, if left alone, can go on to turn into invasive breast cancer. The lesion is characterized by milk duct cells that appear malignant but remain confined to inside of the milk ducts. In other words, they haven’t invaded the tissue surrounding the ducts yet. It is also true that, even now in 2025, general, DCIS is treated similarly to breast cancer, with surgical excision, either by mastectomy or breast-conserving surgery, followed by radiation therapy if breast conserving surgery is used. Then, depending on its hormone receptor status, adjuvant treatment consists of blocking estrogen for five years. The only definite difference in treatment that I tell patients is that we never use chemotherapy to treat DCIS. (It is a common misconception that DCIS can require chemotherapy.)
The rationale for this treatment strategy is the view of DCIS as being a precursor to fully invasive breast cancer and that treating the DCIS will prevent the development of breast cancer. Indeed, in the accompanying editorial (which I’ll get to later), Drs. Monica Morrow and Andrea Barrio use a term for DCIS that I very much like, nonobligate precursor of invasive breast cancer. In other words, DCIS has a propensity to progress to invasive breast cancer, but it doesn’t always, and we’ve been arguing about what percentage of DCIS lesions progress and how fast for the last 100 years. Personally, I like to refer to DCIS to patients as the last stop on the train to breast cancer, which communicates the idea that DCIS is late in the process of cancer development but also that it’s not guaranteed to progress to become cancer.
Over the last three of decades, however, it has become clear that not all DCIS is created equal. Much of it will never progress to breast cancer in the lifetime of the woman (particularly if the woman is older, which means less time for fully malignant transformation to occur). Evidence suggesting this includes studies showing an increase in DCIS incidence by 16-fold since the 1970s, when mammography started to be introduced on a large scale, with little change in the incidence of invasive cancer. Today, 20-25% of mammography-detected breast cancer diagnoses are DCIS; fifty years ago, DCIS was an uncommon diagnosis, except associated with an invasive cancer. Now it is common, all thanks to mammographic screening. One model of breast cancer progression is illustrated below:
Screening for breast cancer tends to be based on the concept that progression from normal duct cells to invasive cancer follows more or less a linear progression, and then invasive cancer grows in the primary organ and eventually spreads, either through lymph vessels or the blood, to distant sites. By this paradigm, effective screening should result in finding disease earlier and thus preventing patients destined to develop metastatic disease from progressing to that point because the disease is removed before it can metastasize. This should result in what is referred to as a “stage shift,” in which the incidence of metastatic disease at diagnosis and of locally advanced disease declines, “shifting” the stage lower, to smaller, more localized disease. Unfortunately, although there is evidence of a small stage shift for breast cancer, the decline in incidence of advanced disease is far lower than the amount of DCIS diagnosed, meaning that most DCIS is almost certainly overdiagnosed, with large numbers of DCIS that never would have progressed to invasive cancer being treated as though they would.
There is also additional complexity in that the above linear model might be only one of multiple pathways by which breast cancer develops, and DCIS exhibits considerable heterogeneity, hence the designations of low, intermediate, and high risk DCIS. A lot of work has been done trying to identify the molecular mechanisms through which DCIS progresses (or doesn’t progress) to invasive ductal carcinoma (IDC)—I even co-authored a review article on this topic over a decade ago—but, suffice to say, there remains controversy. A recent review article provided this excellent timeline of concepts and research regarding DCIS:
For those who are interested, there are a number of models being tested other than just the linear model above:
Another way of looking at some of these models is here:
We don’t need to go into all of these models. Again, suffice to say that a significant percentage of DCIS lesions never progress (and a significant percentage do). We just don’t know for sure the percentages and how rapidly. Moreover, questions have been raised regarding the value of surgery in decreasing the risk of dying of breast cancer for a woman diagnosed with DCIS. Indeed, the post that I referenced from 2013 discussed a study concluding that aggressive treatment of nearly all DCIS does not appear to lead to a reduction in breast cancer mortality, which was consistent with an analysis of the NSABP B-17 and B-24 trials published four years prior. Basically, adding radiation therapy after lumpectomy also appears to have no effect on the risk of breast cancer mortality and might even be slightly higher. What this result suggests is that radiation therapy might not be necessary for the majority of DCIS cases. These are the sorts of studies and data that inspired the COMET trial.
COMET trial
The COMET trial is a pragmatic randomized controlled clinical trial of active monitoring of DCIS versus surgery and standard of care. The COMET schema was published in 2019:
As reported by Hwang et al, the flow was:
So what did Hwang et al report in the first report on the results of the COMET trial? In brief:
Trial participants were:
The study population included women with newly diagnosed DCIS who were aged 40 years or older and had screen-detected, nuclear grade 1 or 2 estrogen and/or progesterone receptor–positive (≥10% staining for Allred Score ≥40), ERBB2-receptor negative (immunohistochemistry scores of 0, 1+, or 2+ if tested) disease, without evidence of invasive cancer. Concurrence of 2 pathologists’ reviews of diagnostic specimens was required to confirm eligibility. Patients with breast symptoms or mass on baseline breast imaging were excluded.
While guideline-concordant care consisted of:
Patients randomized to guideline-concordant care had usual-care treatment for their diagnosis, including surgery. The choice of mastectomy or breast-conserving surgery was made by patients together with their surgeons. Patients undergoing breast-conserving surgery were offered adjuvant radiation treatment in accordance with standard practice. Diagnostic mammograms were required every 12 months for both the affected (if not treated with mastectomy) and unaffected breast.
And active monitoring consisted of:
Patients in the active monitoring group were regularly monitored with imaging and physical examination. Diagnostic mammograms were required every 6 months for the affected breast as part of active monitoring and every 12 months for the unaffected breast. Patients with a new breast mass, nipple/skin changes on physical examination, or imaging findings concerning for disease progression (eg, a new mass, new architectural distortion, and/or increase in extent of calcifications ≥5 mm in ≥1 dimension) were recommended to undergo core needle biopsy. Surgical intervention was required if the needle biopsy identified invasive cancer. For benign breast changes, atypia, or DCIS, continued active monitoring was recommended. Patients who wished to have surgery at any time, for any reason, proceeded to surgery in consultation with their treating surgeon, with the reason for surgery and pathology diagnosis at surgical excision recorded.
That last part is important. Without the ability of the patient to choose to switch to standard-of-care at any time she wished, the trial could not be ethical. Come to think of it, I would marvel that the investigators got this study through an institutional review board (IRB), but the safeguards to avoid harm to study subjects appear more than adequate.
Here is a chart illustrating the active monitoring protocol:
As for the results, as of two years, they are quite striking in their lack of difference between groups:
The median age of the 957 participants analyzed was 63.6 (95% CI, 55.5-70.5) years in the guideline-concordant care group and 63.7 (95% CI, 60.0-71.6) years in the active monitoring group. Overall, 15.7% of participants were Black and 75.0% were White. In this prespecified primary analysis, median follow-up was 36.9 months; 346 patients had surgery for DCIS, 264 in the guideline-concordant care group and 82 in the active monitoring group. Forty-six women were diagnosed with invasive cancer, 19 in the active monitoring group and 27 in the guideline-concordant care group. The 2-year Kaplan-Meier cumulative rate of ipsilateral invasive cancer was 4.2% in the active monitoring group vs 5.9% in the guideline-concordant care group, a difference of −1.7% (upper limit of the 95% CI, 0.95%), indicating that active monitoring is not inferior to guideline-concordant care. Invasive tumor characteristics did not differ significantly between groups.
Here are the recurrence curves for IDC:
Also, at the time of analysis, six patients had died, two in the guideline-concordant care group and four in the active monitoring group, but none of them died of breast cancer.
The authors also noted the possibility of selection bias. You might have noticed how many patients in each group switched groups: 57/484 in the active monitoring group switched over to guideline-concurrent care, and 227/473 in the guideline-concurrent group crossed over to the active monitoring group. The design of the study was such that a 30% nonacceptance of allocation was anticipated and accommodated, but the nonacceptance rate in the guideline-concordant care group was 44%, about which the authors commented that “we did not anticipate the strong preference for monitoring over surgery among study participants.” The authors go on to note:
Recognizing the unexpectedly high rate of nonadherence to allocated intervention with potential for bias due to self-selection of treatment, we evaluated patient characteristics in 4 subgroups: those who were randomized to guideline-concordant care and received guideline-concordant care; those who were randomized to guideline-concordant care but received active monitoring; those who were randomized to active monitoring and received active monitoring; and those who were randomized to active monitoring but received guideline-concordant care. We found no obvious imbalance between groups (eTable 1C in Supplement 1).
This is somewhat reassuring, although the authors also note that they “cannot exclude the introduction of important unmeasured differences between groups, leading to selection and participation biases despite the prospective randomized design.”
A couple of things need to be emphasized here. First, even though the rate of ipsilateral invasive cancer was lower than I would have guessed, two years is not a very long time in the natural history of breast cancer. In other words, all this study shows is that, for short term outcomes, active monitoring is safe. The survival curves could, however, diverge with more time. Breast cancer, particularly estrogen receptor-positive breast cancer, can be an indolent disease, with ten or more years required to detect differences in outcomes. The authors recognize that and note:
Longer follow-up will help determine whether active monitoring offers durable safety and acceptability for patients in the management of this low-risk disease.
In other words, these results are preliminary, but encouraging.
What does this all mean?
Overall, from my perspective, the findings of the COMET trial thus far are modest. The results thus far suggest that “low risk” DCIS can be safely treated in the short term with active monitoring but have not yet demonstrated long term safety and efficacy. You wouldn’t know that looking at Demasi’s posts:
Let’s just say that nothing in the JAMA Open Network article reporting initial results of the COMET study questions whether we should be calling DCIS cancer or not. That’s really rather a separate debate that should be based on biology and clinical behavior rather than just one trial.
The accompanying editorial by Drs. Monica Morrow and Andrea Barrio, Is It Time to Abandon Surgery for Low-Risk DCIS?, is worth reading to put the results of this study in a much more reasonable and proper perspective. (Spoiler alert: Betteridge’s law of headlines applies, and the answer to the question is no, at least not yet.) First, Morrow and Barrio note that existing evidence from trials of surgical excision versus observation suggest that, as I mentioned above, for women diagnosed with DCIS the curve plotting incidence of invasive cancer versus time never plateaus. We also already have evidence from prior trials that IDCs detected on active monitoring of DCIS tend to be larger than those detected as recurrences after surgery for DCIS. Moreover, “low grade” DCIS does not necessarily mean that cancers that develop from the DCIS will be low grade. This leads them to come to a much less concrete conclusion that Demasi:
The COMET trial provides reassurance that in patients with DCIS and a limited life expectancy, surgical therapy, and possibly even endocrine therapy, can be omitted, although the more clinically relevant question is why such patients would be having mammographic screening to begin with. For the remainder, who constitute the large majority of women with DCIS, the safety of active monitoring remains in question. The long-term risk of invasive cancer development with active monitoring is unknown at this time but, based on ECOG-ACRIN E5194 and RTOG 9804, will be at least 0.6% per year, and possibly higher given the absence of surgical excision. Women enrolling in COMET demonstrated a strong preference for active monitoring; almost half of those randomized to guideline-concordant care did not have the prescribed surgery. However, COMET participants may not be representative of the population of women with DCIS at large. Population-based studies of women with DCIS have shown that concern about recurrence, invasive or DCIS, is the most important factor in patient selection of type of surgical therapy,10 so the proportion of women for whom any potentially avoidable risk of recurrence is unacceptable is unclear. Longer follow-up will also clarify the benefit of endocrine therapy in a successful active monitoring strategy. Acceptance of endocrine therapy following surgical excision of DCIS has historically been low, and adherence to 5 years of endocrine therapy is poor even in women with invasive cancer.11 Whether 5 years or longer of endocrine therapy is preferable to a brief outpatient surgical procedure remains to be seen. Meanwhile, adoption of active monitoring as a standard management strategy for low-risk DCIS requires longer follow-up for the majority of patients. The initial report of COMET does not provide the safety data necessary to omit surgery as part of DCIS management.
Precisely.
COMET suggests that there is a group of women with DCIS for whom surgery is unnecessary, although at the cost of every six month monitoring and likely repeat biopsies for the rest of their lives. That’s all we can say so far. It will take several more years before the data from COMET mature enough to let us say anything more definitive.
