CME
Physicians: Maximum of 1.50 AMA PRA Category 1 Credits™
Released: August 17, 2022
Expiration: August 16, 2023
Joyce O’Shaughnessy, MD:
I will start with a discussion of, in my opinion, the most important data presented at ASCO 2022, the DESTINY‑Breast04 trial. This was a plenary presentation that garnered, to my knowledge, only the third standing ovation that has ever occurred at ASCO. DESTINY‑Breast04 is a phase III trial in patients with HER2‑low breast cancer, defined as a score of 1+ by immunohistochemistry or 2+ with negative fluorescence in situ hybridization.1,2 Approximately 60% of metastatic breast cancers (MBC) that are deemed HER2 negative express low levels of HER2, and HER2-targeted therapies do not show clinical activity against these cancers. This means that patients have fewer treatment options for their metastatic disease. DESTINY-Breast04 evaluated the anti-HER2 antibody–drug conjugate trastuzumab deruxtecan, which is approved for HER2-positive MBC, in patients with pretreated HER2-low metastatic disease. Trastuzumab deruxtecan (T‑DXd) consists of trastuzumab conjugated to the very potent topoisomerase‑1 inhibitor deruxtecan, which also has a bystander effect and can kill neighboring cells that do not overexpress HER2.
Joyce O’Shaughnessy, MD:
Patients in the DESTINY-Breast04 trial were endocrine therapy pretreated and had received 1-2 lines of chemotherapy for metastatic disease or had recurrence within 6 months of adjuvant chemotherapy. They were randomized 2:1, regardless of receptor status, to receive either T‑DXd or a standard chemotherapy of the physician’s choice, including capecitabine, eribulin, gemcitabine, paclitaxel, or nab-paclitaxel. In total, 557 patients were enrolled. Stratification factors were HER2-low status (IHC1+ vs IHC2+ and ISH-), number of prior lines of chemotherapy, and HR status. The primary endpoint was PFS in HR‑positive patients by blinded independent central review, and key secondary endpoints included PFS in all patients, OS in HR‑positive and in all patients, ORR, duration of response (DoR), and efficacy in HR-negative patients.
Joyce O’Shaughnessy, MD:
The treatment arms were well balanced. Slightly more than one half of patients, 58% in each arm, were HER2 1+, with the remainder HER2 2+. Approximately 70% of patients had liver metastases.
Joyce O’Shaughnessy, MD:
The patients had a median of 3 prior systemic therapies in the metastatic setting, with a median of 1 prior chemotherapy and a median of 2 prior endocrine therapies. Approximately 65% to 70% of patients had received a prior CDK4/6 inhibitor.
Joyce O’Shaughnessy, MD:
Regarding the primary endpoint of PFS in HR‑positive patients, it was a strikingly positive result with nearly a doubling of median PFS from 5.4 months with chemotherapy to 10.1 months with T‑DXd, with a hazard ratio of 0.51 (95% CI: 0.40-0.64) that was highly statistically significant (P <.001). The full patient population, which included HR-negative, HER2‑low patients, showed very similar data. Again, we see a significant benefit with T-DXd compared with chemotherapy, nearly a doubling of median PFS to 9.9 months compared with 5.1 months, and a hazard ratio of 0.50 (95% CI: 0.40-0.63; P <.001).
Joyce O’Shaughnessy, MD:
Of most importance, the OS data were also strikingly positive. There was a 6.4‑month absolute improvement in median survival in the HR‑positive patients, from 17.5 months with chemotherapy to 23.9 months with T‑DXd, with a hazard ratio of 0.64 (95% CI: 0.48-0.86) and P = .003. The results were very similar in the full patient population, including HR-negative patients.
Joyce O’Shaughnessy, MD:
The PFS effect of T-DXd in HR‑positive patients was very positive across all analyzed subgroups. There was no difference between patients with or without prior CDK4/6 inhibitor, or between those who were HER2 1+ vs 2+, showing that the T-DXd benefit was not a dose effect based on the expressed HER2 level. T-DXd was favored regardless of patients having 1 prior chemotherapy vs heavier pretreatment, age, or the presence of visceral disease. The PFS benefit of T-DXd was very strikingly positive in all of these subgroups.
Joyce O’Shaughnessy, MD:
The investigators conducted an exploratory analysis of PFS and OS in the small number of hormone receptor‑negative patients in the trial, and the results are very intriguing. There were only 58 patients in this group, 18 patients treated with chemotherapy and 40 treated with T‑DXd. Even with these small numbers, we see a strikingly positive survival effect, with an approximate tripling of median PFS from 2.9 months with chemotherapy up to 8.5 months with T‑DXd, with a hazard ratio of 0.46 (95% CI: 0.24-0.89). For OS, there was an approximately 10‑month improvement from 8.3 months median survival with chemotherapy up to 18.2 months with T‑DXd, with a hazard ratio of 0.48 (95% CI: 0.24-0.95). There are no P values here because these were exploratory analyses, but the confidence intervals did not cross 1.
Joyce O’Shaughnessy, MD:
Response rates also are much higher with T-DXd, approximately triple those with chemotherapy, in both HR‑positive and HR‑negative patients. Among all patients, 53% had a confirmed response with T‑DXd compared with 16% with chemotherapy. The median DoR also was longer with T-DXd, 10.7 months vs 6.8 months with chemotherapy.
Joyce O’Shaughnessy, MD:
Regarding safety, 4.6% of patients receiving T-DXd experienced some left ventricular dysfunction, primarily grades 1-2. T‑DXd had an excellent safety profile, with few grade ≥3 events; 13.7% of patients receiving T-DXd had grade 3/4 neutropenia vs 40.7% of patients receiving chemotherapy. The main toxicity seen with T‑DXd, as previously described, is nausea, which occurred at any grade in 73% of patients vs 23.8% with chemotherapy. Any grade fatigue was seen in 47.7% of patients in the T-DXd arm and alopecia in 38%.
Interstitial lung disease (ILD) or pneumonitis occurred in 12.1% of patients with T‑DXd and 3 of them, or 0.8%, died of ILD. Most cases of ILD were grade 1/2 and 1.3% were grade 3.
Joyce O’Shaughnessy, MD:
Based on the significant response and survival benefit with T‑DXd across all patient subgroups and the favorable safety profile consistent with what we have previously seen with T-DXd, Modi and colleagues concluded that T‑DXd now should be a new standard of care for patients with HER2‑low MBC. I certainly agree with that conclusion, including for patients with HR-negative MBC who would have otherwise been considered triple-negative, with HER2 1+ or 2+ cancers. For patients with HER2-low breast cancer who become endocrine therapy refractory, I think we would transition to T‑DXd rather than a single‑agent chemotherapy.
On August 5, 2022, the FDA approved a new indication for T-DXd based on the positive results reported for DESTINY-Breast04.47 T-DXd is now indicated for treatment of adults with unresectable or metastatic HER2-low breast cancer (defined as IHC 1+ or IHC 2+/ISH-) who had received prior chemotherapy for metastatic disease or who experienced disease recurrence during or within 6 months of completing adjuvant chemotherapy. This represents the first FDA approval of a targeted therapy for the HER2-low subtype of breast cancer.
Given these data and the FDA approval, it is becoming increasingly important to define low-level expression of HER2. We know there is a great deal of discordance among expert pathologists in delineating HER2 0 vs HER2 1+. There is more concordance regarding HER2 2+, but this will be an area where we need to talk with pathologists about reporting low levels of HER2 expression. For example, if a patient meets the ASCO/College of American Pathologists guideline of HER2 1+ as faint expression in >10% of tumor cells, oncologists need that result reported out as a 1+ so they can consider offering patients this new therapy which improves survival.
Joyce O’Shaughnessy, MD:
The phase III TROPiCS‑02 trial evaluated another antibody–drug conjugate, sacituzumab govitecan, in patients with HR‑positive, HER2‑negative MBC previously treated with endocrine therapy, CDK4/6 inhibitors, and chemotherapy.3 This agent is a first-in-class TROP-2–directed antibody conjugated to SN38 with a high drug:antibody ratio. It also has a bystander effect and can kill neighboring cells that express no or low levels of TROP-2. It is approved by the FDA for patients with triple‑negative breast cancer who have had 2 or more prior therapies for breast cancer and 1 or more for metastatic disease.
Joyce O’Shaughnessy, MD:
In this trial, 543 patients with HR‑positive, HER2‑negative metastatic disease were randomized 1:1 to receive sacituzumab govitecan or chemotherapy of the physician’s choice. The patients had to have prior endocrine therapy, a taxane, and a CDK4/6 inhibitor in any setting, and they had 2‑4 prior lines of chemotherapy for metastatic disease.
The primary endpoint was PFS by blinded independent central review, and secondary endpoints included OS, response, and safety.
Joyce O’Shaughnessy, MD:
Compared with DESTINY‑Breast04, TROPiCS-02 enrolled a more heavily pretreated population with greater tumor burden. Here, approximately 85% of patients had liver metastases. They had a median of 3 prior chemotherapy regimens for metastatic disease, compared with a median of 1 in DESTINY‑Breast04. And these patients had a median time from MBC diagnosis to randomization of 4 years, putting them well past the median survival of HR‑positive, HER2‑negative breast cancer.
Joyce O’Shaughnessy, MD:
The trial met its primary endpoint, with a median PFS by blinded independent central review of 5.5 months with sacituzumab govitecan vs 4 months with physician’s choice of chemotherapy, a statistically significant improvement with a stratified hazard ratio of 0.66 (95% CI: 0.53-0.83) and P = .0003. The landmark analyses at 6, 9, and 12 months all showed an improvement in the percentage of patients who were progression free. For example, at 12 months, just 7.1% of patients receiving chemotherapy were progression free vs 21.3% of patients receiving sacituzumab govitecan. Although the median PFS was not strikingly different between the 2 treatment groups, I think the durability of response is where we can see the real clinical benefit of this regimen to patients—the flattening tail of the PFS curve and a tripling of the number of patients reaching 1 year without progression.
The investigators reported a PFS benefit across subgroups as well, including older patients and those with visceral disease or very heavily pretreated, with 3 or more prior chemotherapy regimens for MBC. These results show that sacituzumab govitecan is an active agent in HR‑positive, HER2‑negative, heavily pretreated metastatic disease.
Joyce O’Shaughnessy, MD:
The OS data are still very immature. This is the first of 3 planned interim analyses, and at this point there is no difference in OS between the treatment groups.
Joyce O’Shaughnessy, MD:
The ORR was 14% with chemotherapy and 21% with sacituzumab govitecan, with an odds ratio of 1.63 and nominal P value of .03. The clinical benefit rate was 22% with chemotherapy and 34% with the sacituzumab govitecan, with an odds ratio of 1.84 and nominal P of .002. Here, we can see the longer DoR with sacituzumab govitecan, a median of 7.4 months vs 5.6 months with chemotherapy.
Joyce O’Shaughnessy, MD:
The trial did not reveal any surprises about safety. There were 6 deaths on the sacituzumab govitecan arm and one, a case of septic shock due to neutropenic colitis, was considered related to sacituzumab govitecan. The others were unrelated to the sacituzumab govitecan.
Joyce O’Shaughnessy, MD:
Overall, there was a low rate of serious toxicities. Only 4% to 5% of patients in each arm had febrile neutropenia. In the sacituzumab govitecan arm, 57% of patients had any-grade diarrhea and 9% had grade 3/4 diarrhea. Health‑related quality of life was higher with sacituzumab govitecan (P = .005), with delayed worsening of fatigue and global health status. There was no interstitial lung disease or treatment-related cardiac failure or left ventricular dysfunction with sacituzumab govitecan.
Joyce O’Shaughnessy, MD:
The TROPiCS‑02 trial met its primary endpoint, showing that sacituzumab govitecan significantly improved PFS by blinded independent central review compared with chemotherapy. No new toxicity signals emerged with sacituzumab govitecan. Although we are still awaiting mature OS data, based on these results, I think sacituzumab govitecan is an excellent option for heavily pretreated patients.
I think we will prioritize the earlier use of T‑DXd for HER2‑low patients, those who are 1+ or 2+ by IHC. For patients with no HER2 expression, I think sacituzumab govitecan would be a better choice compared with single‑agent chemotherapy, given the longer PFS and particularly the longer tail on the PFS curve.
What we don’t know yet is whether there will be cross resistance to sacituzumab govitecan following progression on T‑DXd, given that the payloads of these 2 antibody–drug conjugates target the same enzyme, topoisomerase‑1. Will there need to be time in between the 2 regimens? We simply don’t know. But I expect we may see future clinical trials prospectively evaluating that question in individual patients who have a long natural history.
Joyce O’Shaughnessy, MD:
The phase II MAINTAIN trial addressed the important question of whether there is any clinical benefit to continuing CDK4/6 inhibitor therapy after progression on a CDK4/6 inhibitor. CDK4/6 inhibition plus endocrine therapy is the preferred treatment for HR-positive, HER2-negative recurrent MBC in the first-line setting, and some observational data have suggested that some patients can get clinical benefit on a CDK4/6 inhibitor following progression on a prior CDK4/6 inhibitor.4 MAINTAIN assessed the efficacy of continuing ribociclib and switching endocrine therapies in patients with unresectable or metastatic HR-positive, HER2-negative breast cancer following progression on endocrine therapy and a CDK4/6 inhibitor.5
Joyce O’Shaughnessy, MD:
The trial enrolled 120 patients with estrogen receptor (ER) and/or progesterone receptor (PgR) expression ≥1%, HER2-negative MBC whose disease was progressing on an endocrine therapy, typically an aromatase inhibitor (AI), plus a CDK4/6 inhibitor. Almost all patients were progressing on palbociclib. A small fraction were progressing on ribociclib. These patients were randomized to receive ribociclib or placebo. In addition, all patients switched to a different endocrine therapy. If they were receiving an AI, they switched to fulvestrant; if they had been on fulvestrant, they switched to exemestane. Premenopausal patients were eligible for the trial if they were receiving a gonadotropin-releasing hormone agonist.
The primary endpoint of the study was locally assessed PFS. Secondary endpoints included tumor response and safety.
Joyce O’Shaughnessy, MD:
In total, 119 patients were randomized, of which 18% had had 2 or more prior endocrine therapies for metastatic disease. A very small minority, approximately 9%, had had chemotherapy for metastatic disease. The majority (87%) had received prior palbociclib, approximately 10% received prior ribociclib, and just 2 patients (3%) in the ribociclib arm and none in the placebo arm had had prior abemaciclib. The median duration of the prior CDK4/6 inhibitor was approximately 16 months in each treatment arm.
Joyce O’Shaughnessy, MD:
The primary endpoint of PFS was met. The median PFS was 2.8 months with placebo plus endocrine therapy vs 5.3 months with ribociclib plus endocrine therapy, with a hazard ratio of 0.57 (95% CI: 0.39-0.95) and P = .006. Endocrine therapy alone following progression on a CDK4/6 inhibitor was associated with a very low PFS, and very few patients were still receiving this therapy at 12 months. As in the TROPiCS-02 trial, here again we see a tail on the curve for patients receiving ribociclib; at 12 months, approximately 25% of patients were still receiving ribociclib plus endocrine therapy. Again, the duration of the response is intriguing.
Joyce O’Shaughnessy, MD:
The PFS benefit with ribociclib vs placebo was seen across patient subgroups, including age greater than or less than 65, receipt of prior palbociclib or ribociclib, prior duration of CDK4/6 inhibitor therapy, and presence of visceral disease. All of the subsets favored ribociclib plus endocrine therapy to some degree.
Joyce O’Shaughnessy, MD:
ORR was 11% with endocrine therapy vs 20% with the addition of ribociclib (P = .51). Clinical benefit rate was 25% with endocrine therapy vs 43% in the ribociclib arm (P = .06). We can also see the longer DoR with ribociclib, a median of 18.8 months vs 14.8 months with endocrine therapy alone.
Joyce O’Shaughnessy, MD:
The toxicity profile of ribociclib is as expected, with neutropenia being the main toxicity, occurring at any grade in 72% of patients and grade 3/4 events in 40% of patients. There were 2 treatment‑related deaths on the ribociclib arm, 1 from sepsis, neutropenia, and disease progression and 1 with pneumonia without fever or neutropenia. In the placebo arm, there was 1 treatment-related death due to pneumonia without fever or neutropenia.
Joyce O’Shaughnessy, MD:
The MAINTAIN investigators concluded that switching to ribociclib and a different endocrine therapy improved both median PFS and clinical benefit rate after progression on a CDK4/6 inhibitor, and that this approach had a manageable toxicity profile.
However, as Claudine Isaacs, MD, pointed out during her discussion of this presentation at ASCO, patients in this trial changed 2 things in their treatment: They changed their endocrine therapy, and they either continued CDK4/6 inhibition with a change to ribociclib in most patients, or received placebo. With 2 variables being changed, it is difficult to know precisely the contribution of receiving ribociclib as a second CDK4/6 inhibitor. Furthermore, MAINTAIN is a small, randomized phase II trial. Dr. Isaacs expressed the opinion that these results are not practice changing, and I agree. We generally do not change practice based on randomized phase II data. Phase III trials are ongoing, and we await those data.
For some patients with very long natural histories for whom we are looking for additional, well-tolerated options, I think it is not unreasonable to treat patients with a second CDK4/6 inhibitor at some point in their natural history. In my practice, however, I would generally not follow progression on a first‑line CDK4/6 inhibitor with another CDK4/6 inhibitor, pending additional data.
Joyce O’Shaughnessy, MD:
Finn and colleagues presented the final survival data from the phase III PALOMA‑2 trial of first‑line palbociclib vs placebo combined with letrozole in HR‑positive, HER2‑negative MBC.6 The primary analysis of this trial reported a 10‑month improvement in median PFS, from 14.5 to 24.8 months, with the addition of palbociclib to letrozole, with a hazard ratio of 0.58 (P <.001).7 This year, the investigators presented the final OS analysis with a very long median follow‑up of 90 months.
Joyce O’Shaughnessy, MD:
Reviewing the details of this trial, 666 patients with ER-positive, HER2-negative advanced breast cancer who had no prior treatment for advanced disease were randomized to receive letrozole plus either standard dose palbociclib (125 mg daily, 3 weeks on, 1 week off) or placebo on the same schedule. The primary endpoint was PFS by investigator, and one of the secondary endpoints was OS.
Joyce O’Shaughnessy, MD:
The patients in PALOMA-2 show a typical profile for patients receiving first‑line therapy. However, a caveat is that 22% of patients on each treatment arm had experienced disease recurrence within 12 months of the end of adjuvant endocrine therapy. Since these patients are receiving letrozole first-line, it’s very likely that almost all this adjuvant endocrine therapy was tamoxifen. That suggests that these 22% of patients had endocrine therapy‑resistant disease.
Joyce O’Shaughnessy, MD:
In the final OS analysis in the intention-to-treat (ITT) population, median OS was not different between the 2 arms—51.2 months with letrozole/placebo and 53.9 months with the addition of palbociclib, for a 2.7‑month absolute improvement in OS, with a hazard ratio of 0.96 that was not statistically significant.
Because the follow-up on this trial was so long, a median of 90 months, 13% of patients in the control arm and 21% in the palbociclib arm were lost to follow-up and are missing survival data. This is not surprising with such a long follow-up. A post hoc sensitivity analysis excluding those patients found an approximately 7‑month absolute improvement, but this was not statistically significant. Addition of palbociclib to letrozole increased the median time to chemotherapy by approximately 8 months.
The analysis also included a combined analysis of the survival data from PALOMA‑1 and PALOMA‑2. The preplanned combined survival analysis in the ITT population showed a 5‑month absolute improvement in survival with palbociclib, from 46.8 to 51.8 months, but that was not significant with a hazard ratio of 0.93.
An unplanned, exploratory subgroup analysis excluded those patients previously mentioned who had a treatment‑free interval less than 12 months since adjuvant endocrine therapy. In this subgroup, we do see an improvement in survival from 44.6 months with letrozole/placebo to 64 months with the addition of the palbociclib with a hazard ratio of 0.73 in this exploratory endpoint.
Joyce O’Shaughnessy, MD:
Looking at the subgroup analyses of survival, the disease‑free interval, which is really a treatment‑free interval, analysis is quite interesting. In the 22% of patients with the short, ≤12-month treatment‑free interval between the end of adjuvant endocrine therapy to recurrence, we see a hazard ratio of 1.0 with the addition of palbociclib. These patients do not derive benefit from palbociclib regarding OS. But the patients with more than 1 year between adjuvant endocrine therapy to recurrence have a hazard ratio for survival of 0.73. This finding suggests that palbociclib is improving survival in patients with endocrine therapy‑sensitive disease, which is consistent with what we have seen in other trials. For example, the PALOMA‑3 trial of fulvestrant with or without palbociclib showed a 10‑month improvement in survival in those with endocrine therapy–sensitive disease per European Society of Medical Oncology definition and no positive impact on survival in those with primary endocrine therapy‑resistant disease.8 The PALOMA-2 findings are consistent with those from PALOMA‑3.
Patients with bone‑only metastatic disease, who tend to be a very endocrine therapy–sensitive group, show a survival benefit with the addition of palbociclib, with a hazard ratio of 0.71. We also see a trend toward longer survival in patients older than 65 years of age, where again we tend to see more endocrine therapy–sensitive disease, with a hazard ratio of 0.87.
The MONALEESA‑2 trial showed a statistically significant improvement in OS with ribociclib plus letrozole versus letrozole alone.9 Cross‑trial comparisons are very challenging as there are always differences in study populations, trial design and in follow-up, but it is important to note that both PALOMA-2 and MONALEESA-2 had approximately the same proportion of patients with a treatment-free interval less than 12 months.
Joyce O’Shaughnessy, MD:
It is always important to look at post‑study systemic therapies patients received upon progression on their first‑line palbociclib vs placebo. These were very well balanced between the treatment arms, with the expected exception that more patients in the placebo arm, 27% vs 12% in the palbociclib arm, received a subsequent CDK4/6 inhibitor upon progression on trial. There are no obvious imbalances in postprogression therapies that could have confounded the survival results.
Joyce O’Shaughnessy, MD:
The final analysis did not reveal any new insights on toxicity. Neutropenia is the main toxicity, with grade 3 events occurring in 56% of patients receiving palbociclib and grade 4 events in 13%. But febrile neutropenia rates are generally low with palbociclib, in the range of approximately 1% to 2%.
Joyce O’Shaughnessy, MD:
In this final analysis, investigators concluded that the addition of palbociclib to letrozole as first-line treatment for ER-positive, HER2-negative advanced breast cancer did not significantly improve OS. However, patients in both treatment arms overall did well, with median survival >50 months. As the investigators note, the high number of patients with missing survival data, particularly in the placebo arm, somewhat confounds interpretation of the OS data.
Joyce O’Shaughnessy, MD:
For patients whose disease progresses on a CDK4/6 inhibitor, we want to consider treatment with agents that may address mechanisms of CDK4/6 inhibitor and endocrine therapy resistance. Everolimus has shown modest activity post progression on CDK4/6 inhibitor. However, the results from the phase II BYLieve trial suggest that alpelisib is active after progression on a CDK4/6 inhibitor for the 40% of patients whose breast cancers have PIK3CA mutations. Among patients in BYLieve most recently treated with a CDK4/6 inhibitor and an AI, clinical benefit rate with alpelisib and fulvestrant was nearly 50% and median PFS was approximately 7 months.10
The AKT inhibitor capivasertib, which inhibits AKT1, AKT2 and AKT3, is an exciting agent in this context. The phase II FAKTION trial evaluated the addition of capivasertib to fulvestrant in patients with HR‑positive, HER2‑negative locally advanced unresectable or MBC that had either recurred or progressed on AI.
The primary analysis reported an improvement in median PFS from 4.8 months with fulvestrant/placebo to 10.3 months with fulvestrant plus capivasertib, which was highly statistically significant with a hazard ratio of 0.58 (95% CI: 0.39-0.84; P = .0044).11 At ASCO 2022, investigators presented an updated analysis, including updated PFS and safety data, a mature survival analysis, and a secondary analysis of the results according to molecular alterations in the PI3K pathway.12,13
Joyce O’Shaughnessy, MD:
FAKTION is a randomized phase II trial of 140 patients who could have had up to 1 prior line of chemotherapy and 3 prior lines of endocrine therapy for metastatic disease, no prior fulvestrant or mTOR/PI3K/AKT inhibitor. Patients were randomized, without being selected upfront for a molecular alteration in the PI3K or AKT pathway, to receive fulvestrant plus placebo or fulvestrant plus capivasertib, which is given 400 mg BID orally, 4 days on, 3 days off each week.
The primary endpoint was investigator‑assessed PFS in the ITT population. Secondary endpoints included survival in the ITT population, and in patients with PI3K pathway alterations. This secondary analysis looked at all known PIK3CA activating mutations, AKT1 E17K mutations, and PTEN alterations by next-generation sequencing.
Joyce O’Shaughnessy, MD:
Patient baseline characteristics were well balanced across the 2 arms of the trial. Most patients, 82% to 83%, had ductal breast cancer and 70% of patients had measurable disease. Patients had a median of 1 prior line of endocrine therapy, and approximately one half of the patients had had prior adjuvant chemotherapy.
Joyce O’Shaughnessy, MD:
The updated PFS and OS analyses both showed significant improvements with the addition of capivasertib. The median PFS was 4.8 months with fulvestrant and placebo post progression on an AI vs 10.3 months with fulvestrant plus capivasertib, with a hazard ratio of 0.56 (95% CI: 0.38-0.81) with P = .0023. Likewise, OS was 23.4 months in the placebo arm compared with 29.3 months with capivasertib, with a hazard ratio of 0.66 (95% CI: 0.45-0.97) and P = .035, a significant increase. These results show us that adding capivasertib to fulvestrant has a very favorable impact on the natural history of the disease in unselected patients progressing on AI.
Joyce O’Shaughnessy, MD:
In the updated safety analysis, the main toxicity of capivasertib is diarrhea, with 83% of patients in this arm experiencing any-grade diarrhea, 14% of which were grade 3 events. Rash occurred in 54% of patients, 20% of which was grade 3. Patients had some hyperglycemia, but the incidence is lower than we typically see with alpelisib. With capivasertib, 38% of patients had grade 1/2 vomiting and 16% had mucositis.
Joyce O’Shaughnessy, MD:
Next, we will look at the prespecified secondary analysis done using an expanded NGS analysis to group patients based on whether they had a pathway alteration, defined as PIK3CA activating mutation, AKT1 E17K mutation, or PTEN alteration. Genomic testing of tissue and/or plasma was performed for 80% of the patients. There was good concordance between NGS and ctDNA regarding detection of these alterations. This expanded testing identified pathway alterations in 25% of the tumors originally classified as pathway nonaltered.
Joyce O’Shaughnessy, MD:
Approximately 60% of patients’ cancers had a pathway alteration in PI3K, AKT1, or PTEN, 39 patients in the capivasertib arm and 37 patients in the placebo arm. The baseline characteristics of the expanded pathway‑altered vs the pathway-nonaltered subgroups were well balanced regarding key prognostic factors.
Joyce O’Shaughnessy, MD:
In patients with a pathway alteration in PI3K, AKT or PTEN, addition of capivasertib was associated with a substantial improvement in median PFS, 12.8 months vs 4.6 months with fulvestrant and placebo, with a hazard ratio of 0.44 (95% CI: 0.26-0.72) and P = .0014, a very significant result.
Conversely, capivasertib was associated with a much more modest impact on PFS in the patients without a pathway alteration. Median PFS was 7.7 months in patients who received capivasertib vs 4.9 months who received fulvestrant and placebo, with a hazard ratio of 0.7 (95% CI: 0.40-1.25) that was not significant (P = .23). It’s worth noting that this group was small, with only 64 patients across both treatment arms.
Joyce O’Shaughnessy, MD:
For OS, the impact of pathway alterations was even more striking; in fact, only the pathway‑altered group showed a survival benefit with capivasertib. Among patients with a pathway alteration, median survival was 20 months with fulvestrant and 38.9 months with the addition of capivasertib, with a hazard ratio of 0.46 (95% CI: 0.27-0.79) and P = .0047, a very impressive outcome. But there was no impact at all on survival in the pathway non‑altered group, with median OS approximately 25-26 months in both treatment arms (hazard ratio: 0.86; 95% CI: 0.49-1.52; P = .60).
Joyce O’Shaughnessy, MD:
These data are very provocative and focus us on enriching the population for patients whose cancers have an alteration in the AKT/PI3K pathway. Patients in FAKTION had not had a prior CDK4/6 inhibitor, which is our first‑line treatment now. The ongoing CAPitello‑291 trial has finished enrollment in patients with HR‑positive, HER2‑negative MBC following recurrence or progression on an AI plus, for most patients, a CDK4/6 inhibitor (NCT04305496). This trial is in all‑comers, regardless of molecular alterations, and will also have a subset analysis of patients with PI3K pathway alterations. We eagerly await the results of CAPitello‑291, particularly in those patients who have had a prior CDK4/6 inhibitor.
Joyce O’Shaughnessy, MD:
We also saw an interesting analysis of the KEYNOTE-522 trial in early-stage triple‑negative breast cancer patients. This phase III trial evaluated pembrolizumab vs placebo in combination with neoadjuvant chemotherapy and as single-agent adjuvant therapy. The primary analysis reported a significant improvement in the pathologic complete response (CR) rate of 13.6% and, of importance, an increase in 3‑year event‑free survival (EFS) from 76.8% with preoperative chemotherapy alone to 84.5% with preoperative pembrolizumab with chemotherapy followed by adjuvant pembrolizumab in patients with stage II/III triple-negative breast cancer (TNBC). These results led to FDA approval of pembrolizumab as neoadjuvant therapy with chemotherapy and as adjuvant therapy in high-risk, early-stage triple-negative breast cancer.14-16 This current analysis of KEYNOTE‑522 is an exploratory analysis examining residual cancer burden (RCB) at the time of definitive surgery and its impact on EFS.17
Joyce O’Shaughnessy, MD:
Patients in KEYNOTE‑522 received preoperative paclitaxel and carboplatin followed by AC with either pembrolizumab or placebo. Then they had surgery, where definitive RCB was assessed. In the adjuvant phase, patients again received pembrolizumab or placebo, according to their initial randomization. Capecitabine was not given in the adjuvant setting.
As previously mentioned, the primary endpoints were pathologic CR (pCR) rate and EFS by local review. RCB was an exploratory endpoint.
Joyce O’Shaughnessy, MD:
The baseline characteristics of patients in this trial are well balanced between arms. Approximately 80% of patients had PD‑L1–positive breast cancer, and 51% of patients were clinically node‑positive at study entry.
Joyce O’Shaughnessy, MD:
As you would expect given the pCR outcomes, there was an increase in the prevalence of RCB‑0 with pembrolizumab, 63.4% vs 56.2% with placebo/chemotherapy. RCB-0 means no invasive cancer remains in the breast or axilla. Each of the other RCB categories was slightly less prevalent among the patients who received pembrolizumab. RCB-1, which indicates a very small amount of cancer left in breast or lymph node, was seen in 8.8% of patients who received pembrolizumab vs 11.5% with placebo; RCB‑2 was seen in 18.5% of patients who received pembrolizumab vs 20.3% with placebo. Thankfully, rates of RCB-3, which represents a substantial amount of residual disease, was low in both treatment groups, 5.1% with pembrolizumab and 6.7% with placebo.
Joyce O’Shaughnessy, MD:
The analysis looked at the impact of RCB on EFS. Of interest, among patients with RCB-0, both treatment groups had very high 3-year EFS and pembrolizumab was associated with a 2% absolute improvement with a hazard ratio of 0.70 (95% CI: 0.38-1.31), which was not statistically significant. In patients with RCB‑1, there was no difference in 3-year EFS, with both treatment groups approximately 84%.
In patients with RCB-2, however, pembrolizumab had a large effect on EFS, with 75.7% of patients receiving pembrolizumab event-free at 3 years vs 55.9% with placebo (HR 0.52; 95% CI: 0.32-0.82).
Among the small number of patients with RCB‑3, there was no difference in EFS between treatment arms. At 3 years, 26.2% of patients receiving pembrolizumab and 34.6% of those receiving placebo were event-free, but as noted, this is a very small subgroup.
As expected, EFS events increased along with higher RCB. Distant recurrence was the most common EFS event across all groups.
Joyce O’Shaughnessy, MD:
When I see the very good EFS outcomes in KEYNOTE-522, particularly the very strong effect in the sizeable RCB‑2 population, my interpretation is that pembrolizumab had a very favorable effect in patients with clinically significant residual disease. We did not see that effect in patients with RCB‑3, where the EFS curves are very similar between treatment arms, but this was a small patient subset. It is possible that patients with a high burden of residual disease or primary refractory disease do not benefit from pembrolizumab, but the numbers are too small to be definitive, and these results do not detract from the value of pembrolizumab in patients with stage II/III TNBC. We want to offer preoperative as well as adjuvant pembrolizumab to patients with stage II/III TNBC, even those patients who have a pCR—because having a pCR is not tantamount to cure. In addition, the trial has a median follow-up of just 3 years so additional recurrences are expected in the pCR group. Also, there is synergy between checkpoint inhibitors and radiation therapy, which increases tumor immunogenicity.
I was last author on a presentation on the NeoPACT trial, a single‑arm study in 117 patients with stage I-III TNBC.18 A small subset of patients had very low ER or PgR expression, 1% to 10% positive, but most patients matched the KEYNOTE‑522 population: stage II/III TNBC with ER and PgR <1%. The study design was like KEYNOTE-522, except patients received neoadjuvant docetaxel (75 mg/m2), carboplatin (AUC 6), and pembrolizumab (200 mg) every 3 weeks for 6 cycles, then went to surgery. NeoPACT patients did not receive adjuvant pembrolizumab because this trial was enrolled before we had results and FDA approval of the KEYNOTE‑522 regimen. The primary endpoint was pCR rate.
In NeoPACT, the overall pCR rate was 58%, and the 95% CI of 51% to 70% overlaps that of the 64.8% pCR rate reported from KEYNOTE‑522. The 2‑year EFS was 89% overall, matching KEYNOTE‑522 results; it was 98% in patients with a pCR and only 78% in those who did not have a pCR. These are very encouraging results and I think this regimen can be utilized as an alternative regimen in patients with stage II/III TNBC in whom an anthracycline is contraindicated. It also supports a planned SWOG trial that will be a head‑to‑head comparison of the regimen used in KEYNOTE-522 vs docetaxel/carboplatin/pembrolizumab, again in patients with stage II/III TNBC.
Joyce O’Shaughnessy, MD:
The LUMINA trial is a very interesting local‑regional control trial for patients with stage I, node‑negative, luminal A, HR‑positive, HER2‑negative breast cancer. We know that patients with luminal A breast cancer—defined as having low Ki-67 expression (in this trial, ≤13.25% by central assessment)—treated with a lumpectomy, radiation therapy and endocrine therapy have an in‑breast recurrence risk of approximately 2.5% at 5 years. Knowing that the in‑breast recurrence risk is so low in these patients with favorable, endocrine therapy–sensitive, indolent biology, can radiation therapy be safely eliminated? This is the question the LUMINA trial set out to answer, assessing the use of clinical pathologic factors and luminal A subtype to identify patients at very low risk of local recurrence after breast-conserving surgery for whom radiotherapy might be omitted.19
Joyce O’Shaughnessy, MD:
LUMINA is a single‑arm trial in 500 patients, 55 years of age or older, with ductal, not lobular, cancer that was not multicentric or multifocal. Eligible patients could not have extensive ductal carcinoma in situ or lymphovascular invasion. They had to have stage I, node-negative breast cancer with ER ≥1%, PgR >20%, HER2 negative, and luminal A subtype as defined by centrally confirmed Ki-67 ≤13.25%. Patients had to have breast‑conserving surgery with clear margins and they had to be planning to take endocrine therapy for at least 5 years, and radiation therapy was not administered.
After surgery and initiating endocrine therapy, patients were followed up with yearly mammograms and seen clinically every 6 months for 2 years and then annually. The primary endpoint was local recurrence, including any invasive or noninvasive event. Secondary endpoints were contralateral breast cancer, any recurrence, and disease‑free survival and OS. The ITT analysis was planned at a median follow-up of 5 years.
Joyce O’Shaughnessy, MD:
Sixty percent of these patients were 65 years of age or older, 50% had T1a/T1b disease, and the other one half had T1c disease. All patients had grade 1 or grade 2 tumors. For their endocrine therapy, 40% received tamoxifen and approximately 60% received an aromatase inhibitor.
Joyce O’Shaughnessy, MD:
The trial met its primary endpoint with a 5‑year local recurrence rate of 2.3%, with 10 events. The 90% CI was 1.3-3.8, below the 5% prespecified boundary. Incidence of contralateral breast cancer was 1.9% (90% CI: 1.1-3.2) and rate of any recurrence was 2.7% (90% CI: 1.6-4.1). Survival outcomes showed 5-year disease-free survival of 89.9% and OS of 97.2%.
Joyce O’Shaughnessy, MD:
Based on these results, the investigators concluded that patients with node-negative, luminal A breast cancer at low risk of local recurrence could be considered candidates for the omission of radiation therapy after breast‑conserving surgery and endocrine therapy. These results could apply to an estimated 30,000‑40,000 patients with breast cancer in North America per year.
At ASCO, this presentation was discussed by Jennifer Bellon, MD, a radiation oncologist at the Dana-Farber Cancer Institute. Dr. Bellon reminded us of the CALGB 9343 trial, which evaluated outcomes with or without radiation therapy in women 70 years of age or older with ER-positive, node-negative, early-stage breast cancer who had a lumpectomy and received adjuvant endocrine therapy.20 The local recurrence risk in this population was 4% at 5 years and 9% at 10 years, consistent with the expected 1% per year risk of in‑breast recurrence with surgery and endocrine therapy alone in a favorable population.
Based on these previous results, Dr. Bellon cautioned that the 2.3% risk seen at 5 years in the LUMINA trial would likely increase with 5 and 10 more years of follow-up. Even so, this low level of risk over time is unlikely to have a meaningful impact on patients’ OS. However, as in-breast recurrence risk continues for many years to decades, omitting radiation therapy might be more appropriate to discuss with older patients. Dr. Bellon indicated that we probably are not ready to adopt this as a new standard of care, especially for the younger subset of women with luminal A breast cancer, those in their 50s and 60s, who may have decades of life ahead of them. I agree with that assessment. She advocated for longer follow-up of LUMINA and also mentioned several other ongoing single‑arm or randomized trials that are asking this same question in a biologically defined low‑risk group of stage I, HR‑positive, HER2‑negative patients receiving lumpectomy along with endocrine therapy without radiation therapy. These additional data will be very important, particularly in younger patients.
Back
