The antibodyâ€“drug conjugate sacituzumab govitecan (IMMU-132) received a breakthrough therapy designation based on phase II trial data showing a response rate of 31% in heavily pretreated patients with metastatic TNBC. However, treatment advances are lagging behind those seen in other breast cancer phenotypes.
Adam M. Brufsky, MD, PhD
Compared with estrogen receptor (ER)—positive and HER2-positive breast cancer, triple-negative breast cancer (TNBC) is typically more aggressive and carries a higher mortality rate. Several ongoing clinical trials are attempting to gain ground. In February, the antibody–drug conjugate sacituzumab govitecan (IMMU-132) received a breakthrough therapy designation based on phase II trial data showing a response rate of 31% in heavily pretreated patients with metastatic TNBC. However, treatment advances are lagging behind those seen in other breast cancer phenotypes.
Panelists taking part in a recent OncLive Peer Exchange® program provided context to the emerging evidence in TNBC, highlighting some of the key clinical trial data presented at the 2015 San Antonio Breast Cancer Symposium (SABCS). Moderated by Adam M. Brufsky, MD, PhD, the panel shed light on the impact this recent evidence may have on clinical practice.One area open for debate is whether platinum agents should be added on to standard chemotherapy in the neoadjuvant setting of TNBC. Data from two clinical trials presented at 2015 SABCS pulled back one layer of the onion regarding neoadjuvant use of carboplatin, but are the data robust enough to change what clinicians do in clinical practice?The phase II GeparSixto study evaluated the addition of carboplatin to weekly paclitaxel plus non-pegylated liposomal doxorubicin (Myocet) for 18 weeks in TNBC and HER2-positive breast cancer. Patients with TNBC (n = 315) received concurrent bevacizumab, while those with HER2-positive disease received trastuzumab and lapatinib. The carboplatin dose was reduced from area under the curve (AUC) 2.0 to AUC 1.5 to improve tolerability.1
In patients with TNBC, carboplatin improved the pathologic complete response (pCR) from 36.9% to 53.2% (P = .005). However, the pCR improvement was not statistically significant in the HER2-positive subgroup (36.8% vs 32.8%, respectively; P = .6).1
The study also showed a disease-free survival (DFS) benefit with carboplatin in the TNBC group after a median of 35 months (85.8% vs 76.1%; P = .0350).1
In addition, in patients with BRCA wild-type tumors, the odds ratio (OR) for pCR was 2.09 in favor of carboplatin treatment compared with controls (50.8% vs 33.1%; P = .005). In the BRCA-mutant patients, the OR for pCR was 1.6 in favor of carboplatin, but the difference was not statistically significant (61.5% vs 50.0%; P = .413).1
Hope S. Rugo, MD
Prior to this study, it was thought that platinum compounds are not active in patients without germline BRCA mutations, said Christy A. Russell, MD. She noted that before she saw the GeparSixto findings, she was not in favor of adding carboplatin to neoadjuvant chemotherapy. She commented that the advantage in BRCA wild-type TNBC was seen “not only in pathologic complete response rate, but also what appears to be translating into disease-free survival by the addition of carboplatin. And it has now made me think about it a little more. I have not yet added it.”Joyce A. O’Shaughnessy, MD, pointed out that updated data from the CALGB 40603 trial, also presented at the conference, showed higher pCR, but no DFS improvement when carboplatin is added to once weekly paclitaxel in patients with stage II/III TNBC.2
The standard chemotherapy regimen for this trial was weekly paclitaxel 80 mg/m2 for 12 courses plus dose-dense doxorubicin and cyclophosphamide. Patients were randomized in a 2 x 2 schema to receive the standard regimen alone or the standard plus one of these regimens: bevacizumab every 2 weeks for 9 cycles; carboplatin AUC 6 every 3 weeks for 4 cycles; or the combination of bevacizumab and carboplatin.2
The pCR rates in the breast, the primary endpoint of the initial study, were 60% for patients receiving carboplatin and 46% in patients who did not receive carboplatin, an increase of 76% (P = .0018).2 In follow-up data presented at 2015 SABCS, Sikov and colleagues reported that patients who achieved a pCR on any arm of the study achieved far superior event-free survival and overall survival outcomes compared with those who did not have a pCR.2
Panelists discussed the regimen and the dosing in the trials to help give the results more context. One noteworthy difference was the standard chemotherapy used. Myocet is not FDA approved, and thus the standard used in the GeparSixto trial is not representative of a US regimen. The drug is approved in Europe in combination with cyclophosphamide. The standard chemotherapy in the CALGB 40603 study was typical, noted the panelists.
Sara A. Hurvitz, MD
Brufsky theorized that the control arm in the GeparSixto study was likely an inferior regimen, potentially causing carboplatin to exert more of an effect.
An additional factor to consider in these trials is that the dose of carboplatin was lower in the GeparSixto study versus the CALGB 40603 (AUC 1.5 vs AUC 6, respectively). The panelists agreed that these factors, combined with the small sizes of the trials, made it challenging to draw any conclusions about the impact of adding a platinum agent in this setting.
“We really need larger trials powered for disease- free survival,” O’Shaughnessy said. However, she added that it may be appropriate to try adding on carboplatin in patients who are very high risk. “What I’ve done personally, and I think many of us may do, is when you see these poor responders, we tend to add,” agreed Hope S. Rugo, MD. “So, [for] poor responders with triple-negative disease, I’ll add platinum as opposed to starting with it since we already know we get a reasonable pCR rate in those patients.”Data from a separate study presented at 2015 SABCS examined the use of capecitabine in the adjuvant setting.
The phase III CREATE-X trial enrolled patients with pathologic residual invasive disease following surgery and neoadjuvant therapy with an anthracycline, taxane, or both. It included hormone receptor (HR)—positive and HR-negative subsets, the latter of which included some patients with triple-negative disease. Patients were randomized to standard treatment with or without 8 cycles of capecitabine. The primary endpoint was DFS.3 The 3-year overall DFS rate was 82.8% with capecitabine compared with 74.0% in the standard therapy arm. The benefit with capecitabine was less pronounced in the HR-positive group (n = 561; HR, 0.84). Those with HR-negative disease (n = 296) had a 42% reduction in the risk of recurrence with capecitabine (HR, 0.58).3
Joyce A. O’Shaughnessy, MD
The subset of patients with TNBC showed the most benefit, said O’Shaughnessy. “My personal opinion is that it is something worth discussing with patients,” she added, noting that more clinical studies are needed to confirm the benefit. Rugo commented on the dose of capecitabine, noting that the drug metabolizes differently in some ethnic populations. “I might have had two patients ever tolerate 2500 mg/m2 of capecitabine, and that’s the dose they gave, and their dose intensity was very good,” she said. “And we know that the metabolism of capecitabine is different in many, not all, Asian people based on differences in pharmacogenomics.”
O’Shaughnessy stated that previous trials have shown a systemic benefit with capecitabine in triple-negative disease. Brufsky asked the panelists whether the evidence to date convinces them to use adjuvant capecitabine. Rugo and Sara A. Hurvitz, MD, both said they would consider using capecitabine in patients with significant residual disease following good neoadjuvant therapy. “I think it warrants a discussion with patients who have had standard therapy and have a lot of residual disease if they are hormone receptor—negative,” Hurvitz said.The discussion turned to the use of eribulin mesylate in metastatic disease. The results from the phase III 301 study found that eribulin mesylate was not superior to capecitabine in locally advanced or metastatic breast cancer following treatment with anthracycline and taxane chemotherapy.4
However, subsequent subanalyses revealed certain subgroups did show a benefit.5 Those with triple-negative disease had a median overall survival of 14.4 months with eribulin mesylate compared with 9.4 months with capecitabine. (HR = 0.70; 95% CI, 0.55-0.91; P = .01).
“It truly is that new drug we needed after the anthracycline, taxane, capecitabine [therapies],” O’Shaughnessy said. “And now we have a non— cross-resistant drug both in the ER-positive and the triple-negative [patients].”
Christy A. Russell, MD
Rugo stated that there is a learning curve with using the drug because of the risk of mucositis. She uses a lower dose than the package insert dose of 1.4 mg/m2. She said she has started patients at .4 mg/m2, and gradually increased the dose by .1 mg/m2, and has “had it be a treatment which works when people’s livers are failing and gives them another 8 months of life,” Rugo remarked.
In liver failure, Rugo and O’Shaughnessy agreed that the bilirubin level should guide the starting dose, not the Child-Pugh score. “I’ve given it to people with bilirubins of 6 and 7, but then you should really be giving .4 [mg/m2],” said Rugo. “And you still are going to get mucositis, and it’s really pretty impressive, but it works.”Another study presented at the SABCS conference was the phase Ib KEYNOTE-028 trial, which showed that the PD-1 inhibitor pembrolizumab had an overall response rate of 12% in PD-L1 positive patients with ER-positive/HER2-negative advanced breast cancer. Among those with TNBC, the ORR was 18.5%.6
By contrast, the PD-L1 inhibitor avelumab was less effective against mixed phenotypes of breast cancer in the JAVELIN trial.7 The ORR for ER-positive disease was 2.8% (2 of 72 patients) compared with 8.6% in triple-negative disease (5 of 58 patients). The response rate was higher among patients with TNBC with >10% PD-L1 expression in immune cells within the tumor; the response rate in this category was 44.4% (4 of 9 patients). The PD1/PD-L1 immunotherapy response rates are not as impressive as what has been observed in other cancers, noted the panelists. “One of the theories right now is that the reason melanoma and renal cell cancer are so responsive is that they have a lot of neoantigen diversity, and breast cancer and prostate cancer don’t,” Brufsky said. The panelists agreed that more factors likely must be considered to making immunotherapy work in breast cancer than focusing on PD-L1 expression. One variable to consider is the role of tumor-infiltrating lymphocytes, and their impact on prognosis, Russell noted.
“The area of immune therapy is going to be very exciting to look at, and I think we are all hoping to crack the nut on triple-negative disease,” Hurvitz said. “We have to get better at distinguishing disease subtypes and developing targeted therapies.