Survival outcomes from the CALGB 40603 study in triple-negative breast cancer

Bookmark and Share
Published: 17 Dec 2015
Views: 5810
Dr William Sikov - Women & Infants Hospital of Rhode Island, Providence, USA

Dr Sikov talks to ecancertv at SABCS 2015 about survival outcomes from the CALGB 40603 study involving women with stage II-III triple-negative breast cancer (TNBC) who were treated with carboplatin, bevacizumab, or both in addition to standard neo-adjuvant chemotherapy.

The aim of the study was to assess the effects of adding carboplatin or bevacizumab on complete response (pCR) rates. Standard neo-adjuvant chemotherapy used in the trial was weekly paclitaxel then dose-dense doxorubicin plus cyclophosphamide.

The effect on pCR have been reported (J Clin Oncol 2015;33:13–21) and showed improved pCR with the addition of both the chemotherapy and the anti-angiogenic agent.

At SABCS 2015, the effects on the secondary endpoints of event-free (EFS) and overall survival (OS) were presented, with 3-year rates of 74.1% and 83.2%, respectively, although the study was not powered to assess the different treatment effects on either of these endpoints.

What it shows is that patients with TNBC who achieved pCR with study treatment had significantly better EFS and OS than patients who did not.

ecancer's filming at SABCS 2015 has been kindly supported by Novartis through the ECMS Foundation. ecancer is editorially independent and there is no influence over content.

You have been looking at pathological complete remission. It had a particular significance in the study you’ve quoted here but first of all could I get you to tell me what the issue was that you were trying to resolve with this study?

Certainly. This was a study looking at the addition of the chemotherapy drug carboplatin or the anti-angiogenic agent bevacizumab to a standard chemotherapy regimen in patients with clinical stage 2 and 3 triple negative breast cancer.

So what did you do?

Basically all patients received the same standard chemotherapy regimen, weekly paclitaxel for twelve weeks followed by four cycles of dose dense AC. To this the patient might have received in addition carboplatin, bevacizumab, neither or both. We presented the primary endpoint of this study two years ago which was to see whether the addition of either of these agents would increase the percentage of patients who achieve a pathologic complete response. The study was positive for those endpoints - the addition of either carboplatin or bevacizumab did significantly increase the percentage of patients in whom the cancer disappeared from the breast or the breast and the lymph nodes.

Now you’ve gone on, though, to look at survival. What did you find?

Now we’re an average of a little bit more than three years out from entry onto the study. The question is at that point, especially with triple negative breast cancer, which is such an aggressive form of breast cancer, at three plus years out many of the patients who will recur will have recurred and many of those who have recurred will have died. So the question is what was the impact of pathologic response on those endpoints and, although we did not have the statistical power to answer this question, we did look to see what effect carboplatin or bevacizumab would have on these endpoints.

First of all, then, what was the impact of PCR?

PCR has a very large impact on disease recurrence and death. There was a 70% improvement in disease recurrence and an 80% improvement in deaths in patients who achieved a pathologic complete response compared to those who did not. This translates into about a 20% absolute increase in the percentage of patients who are free of disease and are alive at that three plus year time point.

And did carboplatin or bevacizumab or both of them together influence that?

The answer is we don’t have the statistical power to really answer that question. While one would expect that by increasing the pathologic complete response rate we would improve event free and overall survival, the statistics work out in such a way that the difference is too small in a study of our size to prove that with statistical significance.

Saying that improving the pathological complete response rate influences survival, that’s not a big surprise is it? But what you really want to know is how to improve the PCR.

Yes, that’s true. We have seen this in the past, that pathologic complete response does improve outcomes down the road. A couple of take away points, though, from our study. One is that we significantly increased the overall pathologic complete response rate achieved compared to prior analysis, including the Cortazar meta-analysis. In our study 48% of patients achieved a pathologic complete response compared to about 34% in those prior combined analyses.

Now, is that because it was a good study or was it anything to do with the drugs you used?

We did have a higher pathologic complete response rate in the control arm of our study but by increasing the pathologic complete response rate with the addition of the carboplatin, the bevacizumab or both, we achieved an aggregate much higher pathologic complete response rate.

So what’s your feeling about this? You’re hinting that there might be an effect of using these agents, what’s your feeling?

I don’t want to overstate what I can really say based on these statistics. All I can say is that the addition of either of those agents does increase pathologic complete response rate and pathologic complete response rate is a very good thing.

What would you advise doctors to be doing now? The data are not all in, as you’ve quite clearly indicated, with a triple negative breast cancer patient.

I think it’s difficult what to advise. If a doctor is waiting for proof that the addition of, let’s say, carboplatin, since that’s the more easily available drug, to a standard chemotherapy regimen is going to improve long-term outcomes we don’t really have that data. On the other hand, if the doctor has a patient who has a larger tumour, palpable lymph nodes and where he thinks that response to neoadjuvant therapy may be very important in terms of what he can offer the patient down the road in terms of surgery or limiting axillary node dissection, the addition of carboplatin does seem to increase the likelihood that a patient will be eligible for breast conserving surgery or be able to avoid a full lymph node dissection.

These drugs have come with side effects so, whether it’s carboplatin or bevacizumab, what’s the deal there?

With carboplatin the major side effect that we saw was a higher rate of low blood counts, both a low white blood cell count, anaemia and thrombocytopenia, fortunately without a significant increase in the percentage of patients who had febrile neutropenia or required hospitalisation. Bevacizumab is associated with more different side effects including high blood pressure as well as higher incidence of low blood counts and complications of that. The way that the carboplatin is given may be important; in our study we gave the carboplatin at a high dose, an AUC of 6 every three weeks, the German study that was presented just before mine gave it as a weekly dose at an AUC of 2. I have to say that in my own practice when I choose to give carboplatin in the neoadjuvant setting off study I typically give it at that low weekly dose as opposed to a higher dose every three weeks. In my experience, and those of many of my colleagues, there seems to be less hematologic toxicity at that dosing schedule than the one used in our study.

So, in a nutshell, to sum up, what would you say is the best approach for neoadjuvant therapy for patients with triple negative breast cancer from the evidence you’ve got so far? Just in a few words to give the take home message.

The standard regimen still consists of weekly paclitaxel and dose dense AC given in either order although I personally prefer giving the weekly paclitaxel first. The addition of carboplatin, I think it has to be individualised, as the discussant for our abstract said, that there may be patients who seem to be at higher risk in whom the addition of carboplatin is reasonable because it may improve local responses and it may improve some of the options that you’re able to offer the patient. But for other patients who are already good surgical candidates there may be little benefit to doing so.

And bevacizumab?

Bevacizumab is more difficult because it has more toxicities, it’s also more expensive. I can’t really recommend adding bevacizumab to standard treatment at this time.

Is there anything you can say, though, about patients who still had disease in the breast after this treatment?

Yes, it’s important how much disease was left in the breast. Pathologic complete response is very black and white, no disease versus any amount. But what we built into the study was to look at patients who had minimal residual disease, something we call residual cancer burden class 1. A substantial fraction of patients, 13% of the study patients, achieved that near pathCR outcome. When we looked at those patients, they had the same good long-term prognosis as patients who had a pathologic complete response. So I would be very comfortable telling a patient of mine who was disappointed because she hadn’t achieved pathologic complete response, saying, ‘Yes, but in your case there was such minimal residual disease your outcome appears to be as good as the patients who achieved a pathologic complete response.’