"In my last video I told you I am frustrated with the way DCIS is managed. Not because surgeons are doing the wrong thing — most of us are following evidence-based guidelines. But because the evidence base itself has significant gaps. And some of what we are recommending may not be delivering the benefit we think it is."
This is the video for people who want the full picture — not reassurance, not a simplified protocol, but the honest evidence on the decisions that follow a DCIS diagnosis. We are going deeper on size, the Van Nuys Prognostic Index, radiation, and the tamoxifen question that I think deserves a far more honest conversation than it typically gets.
Size — What It Actually Tells Us (and What It Doesn't)
Size is often one of the first things your surgeon mentions after a DCIS diagnosis, and it feels intuitive: bigger means more dangerous, smaller means less dangerous. But with DCIS, size is more complicated than that.
What size tells us: the extent of the DCIS — how much of the duct system is involved. It influences whether lumpectomy is technically feasible, and it is one of four variables in the Van Nuys Prognostic Index.
What size does not reliably tell us: biological behaviour. A 2-centimetre area of low-grade DCIS may carry a lower risk of clinically meaningful progression than a 5-millimetre area of high-grade disease. Size and grade together tell a more complete story than either alone.
There is also the measurement problem. DCIS can be discontinuous — appearing across different ducts — which makes accurate measurement on imaging genuinely difficult. Your radiologist and your pathologist may report different sizes for the same lesion, and both may be correct within the limits of their methods.
When your surgeon mentions size, ask: "Size in the context of what grade, what margins, and what age?" Size alone is not sufficient to drive the treatment recommendation.
The Van Nuys Prognostic Index — Useful Tool or Overused Formula?
The VNPI was developed by Melvin Silverstein at the University of Southern California, first published in 1996 and updated in 2003 to add patient age. It combines four variables — tumour size, margin width, nuclear grade with or without comedo necrosis, and age — each scored 1 to 3, giving a total range of 4 to 12.
| Variable | Score 1 ✅ | Score 2 ⚠️ | Score 3 ❌ |
|---|---|---|---|
| Tumour size | ≤15mm | 16–40mm | >40mm |
| Margin width | ≥10mm | 1–9mm | <1mm |
| Grade + necrosis | Low, no necrosis | Intermediate | High grade |
| Patient age | ≥61 years | 40–60 years | ≤39 years |
In New Zealand and Australia, over 70% of DCIS cases are managed with reference to the VNPI — compared to less than 16% in the UK. Asking your surgeon for your score is entirely appropriate.
But Here Is the Honest Analysis
The VNPI was derived from a single institution's retrospective data. It has never been externally validated in a large, independent, prospective cohort. The scoring criteria have changed across versions — meaning a 1996 score is not comparable to a 2003 one. And it does not incorporate biological markers such as hormone receptor status or HER2, which we now know carry prognostic significance.
More fundamentally: the VNPI predicts local recurrence risk — the chance of DCIS or invasive cancer returning in the same breast. What it does not predict is mortality. And for DCIS, the relationship between local recurrence and survival is not straightforward. A local recurrence is not the same thing as a life lost.
The VNPI is a useful framework for discussion, not a formula that produces the right answer. It should inform the conversation — not replace it.
🇳🇿🇦🇺 The high VNPI adoption in NZ and AU means your surgeon is likely using it or something equivalent. Ask: "What is my VNPI score, and how is it influencing the recommendation?"
Grade and Timing — The Window of Treatment
Grade and timing are connected in a way that matters clinically. For high-grade DCIS, roughly 50% of cases that will progress do so within five years — which is the biological argument for prompt treatment. Not because delay of a few weeks is catastrophic, but because the window of opportunity to intervene before progression may be meaningfully shorter than for low-grade disease.
For low-grade DCIS, the progression timeline is so extended — potentially decades — that it introduces the legitimate question of whether progression would occur within a woman's expected lifespan at all, particularly in older patients. This is the biological underpinning of the active surveillance argument.
Radiation — When Does It Actually Help?
Radiation after lumpectomy for DCIS reduces the relative risk of local recurrence by approximately 50%. That sounds substantial. But relative risk reduction can be misleading when you don't know what it is relative to.
Here is the key distinction: relative vs absolute risk reduction.
- If your baseline recurrence risk without radiation is 30%, a 50% relative reduction brings it to 15% — an absolute benefit of 15 percentage points. Meaningful.
- If your baseline risk is 10%, the same 50% relative reduction brings it to 5% — an absolute benefit of only 5 percentage points. Same relative number. Very different clinical significance.
The question to ask when radiation is discussed is not just whether it reduces recurrence — it does. The question is: by how much, for my specific grade, size, and margin width? What is my absolute risk reduction, not just my relative risk reduction?
Endocrine Therapy — The Honest Evidence
This is the section I find most important to talk about honestly. In countries where ER testing of DCIS is routine — primarily the US and UK — women with ER-positive DCIS may be offered tamoxifen or an aromatase inhibitor after surgery, to reduce the risk of developing invasive breast cancer in either breast. In New Zealand and Australia this is not standard practice — and the reason for that is worth understanding.
What the Evidence Shows
The NSABP B-24 trial showed tamoxifen after lumpectomy and radiation for DCIS reduced breast cancer events at five years from 13.4% to 8.2% — a relative reduction of around 37%, and an absolute reduction of approximately 5 percentage points. That was the evidence that established tamoxifen as an option for ER-positive DCIS in the US.
What Those Numbers Don't Tell You
They do not tell you what happens to women who take tamoxifen for five years in terms of quality of life. Tamoxifen in premenopausal women causes menopausal symptoms — hot flushes, joint pain, mood changes, vaginal dryness, and sexual dysfunction — that are real, persistent, and in some women, severe. Aromatase inhibitors in postmenopausal women carry risks of bone density loss, fracture, and cardiovascular effects. These are not trivial side effects accumulating over five years.
The absolute reduction in all-cause mortality from tamoxifen for DCIS is not clearly established. We know it reduces breast cancer events. We do not have strong evidence that it reduces breast cancer death specifically from DCIS — in part because mortality from DCIS is already very low, and in part because the trials were not powered for this endpoint.
What this means clinically: we may be adding meaningful morbidity — five years of significant side effects — for a treatment whose benefit is primarily measured in event reduction rather than mortality reduction. In a disease where 10-year survival exceeds 98%, that trade-off deserves to be made explicitly, with full information, by each individual patient.
There is a further concern emerging in the literature. If a proportion of DCIS would never have progressed regardless of treatment, endocrine therapy in those women does not reduce their risk of future invasive cancer — it simply exposes them to five years of side effects for no benefit. We still cannot reliably identify those women, which means some of the women currently taking tamoxifen for DCIS may be doing so unnecessarily.
Endocrine therapy is not wrong for DCIS. For some women — particularly younger women with high-grade, ER-positive disease and significant family history — the benefit-to-burden calculation may genuinely favour treatment. But this decision should never be automatic. It should be an informed, individualised conversation that includes the absolute numbers, the side effect profile, and an honest acknowledgement of what the evidence does and does not show.
🇳🇿🇦🇺 Endocrine therapy not being routine in NZ and AU does not mean local surgeons are behind the evidence — it reflects a considered position that the benefit-to-burden ratio for this specific indication is not clearly established. If you are in NZ or AU and a clinician is recommending tamoxifen for DCIS, it is entirely appropriate to ask: what is the absolute benefit estimated to be for my specific case?
What to Take From This
DCIS is not a simple diagnosis. And the treatment decisions that follow are not simple either. Size matters — but in the context of grade. Grade matters — but in the context of timing and age. The VNPI is useful — but it is a framework, not a formula. Radiation reduces recurrence — but the absolute benefit varies enormously. And endocrine therapy may help some women — but the evidence for mortality benefit from DCIS-specific treatment is not as robust as it is sometimes presented.
What I want you to take from this post is not doubt — but precision. Precision about what questions to ask, what numbers actually matter, and what the evidence does and does not support.
Continue Reading: The DCIS Series
Book a consultation: mybreastcarecentre.co.nz