Pathologic complete response may be strong surrogate endpoint for breast cancer survival

ByCharles L. Shapiro, MD, FASCO

Douglas Yee

SAN ANTONIO — Pathological complete response appeared to be a viable surrogate outcome for EFS and distant RFS, according to long-term efficacy results from the I-SPY2 TRIAL presented at the San Antonio Breast Cancer Symposium.

Perspective from Charles Shapiro, MD

“Many neoadjuvant trials have established a relationship between pathologic complete response and EFS,” Douglas Yee, MD, director of the Masonic Cancer Center, professor of medicine and pharmacology at University of Minnesota, and a HemOnc Today Editorial Board Member, said during his presentation, noting that this relationship has been shown across breast cancer subtypes. “The definition of pathological complete response has been consistent across all studies.”

The adaptively randomized I-SPY2 platform trial, which began in 2010, uses pathologic complete response as the primary endpoint to identify investigational agents that improve outcomes among women with stage II to stage III breast cancer who have a high risk for early recurrence, based on hormone receptor, HER-2 and 70-gene (MammaPrint) status. The trial has accrued more than 1,000 patients, using paclitaxel followed by doxorubicin and cyclophosphamide as the control arm, compared with outcomes for 11 other agents and combinations.

“Multiple drugs can exist at any time, and can exit the trial for various reasons,” Yee said

The current dataset included 746 patients. Thirty-five percent of patients achieved pathologic complete response, whereas 65% did not.

After a median follow-up of 2.7 years (range, 0.02-7.2) 126 EFS events and 109 distant RFS events had occurred.

The analysis included a cross-section of patients with different subtypes, including hormone receptor negative and HER-2 negative (n = 245), hormone receptor positive and HER-2 negative (n = 275), hormone receptor negative and HER-2 positive (n = 77), and hormone receptor positive and HER-2 positive (n = 149).

“The pathologic complete response rate varied across subgroups, as did the number of patients enrolled,” Yee said.

Overall 3-year EFS reached 94% among patients achieving a pathologic complete response and 76% for those who did not (HR = 0.2; 95% CI, 0.11-0.36).

Ninety-five of complete responders achieved 3-year distant RFS compared with 79% of nonresponders (HR = 0.2; 95% CI, 0.11-0.37).

“I want to make sure I’m clear on explaining who these patients are,” Yee said. “We are not comparing one arm vs. another arm. We are comparing all patients who achieved a pathologic complete response. They might have been on the control arm, or they might have been on arm that did not graduate.”

PAGE BREAK

Broken down by subgroup, a similar trend held for patients with triple-negative disease, with 3-year EFS rates at 92% in the complete response arm and 67% among those who did not reach this endpoint (HR = 0.17; 95% CI, 0.07-0.39). Distant RFS rates also were higher in the pathologic complete response group (94% vs. 70%; HR = 0.16; 95% CI, 0.06-0.4).

Among hormone receptor-positive, HER-2-negative patients, pathologic complete response was associated with higher 3-year EFS (94% vs. 79%; HR = 0.21; 95% CI, 0.05-0.85) and distant RFS (94% vs. 80%; HR = 0.22; 95% CI, 0.05-0.93).

For hormone receptor-negative, HER-2-positive patients, pathologic complete response also was associated with improved 3-year EFS (93% vs. 53%; HR = 0.1; 95% CI, 0.03-0.37) and distant RFS (93% vs. 62%; HR = 0.14; 95% CI, 0.04-0.51).

The trend held true in hormone receptor-positive, HER-2-positive disease, with a 96% 3-year EFS rate in the complete response group and an 87% rate for those who did not reach a complete response (HR = 0.26; 95% CI, 0.06-1.14). For distant RFS in this cohort, the rates were 97% for complete responders and 92% for noncomplete responders (HR = 0.19; 95% CI, 0.02-1.51).

“All the subgroups who achieve a pathologic complete response enjoy a very similar EFS,” Yee said. “It is very clear that patients who achieve the pathologic complete response enjoy a better EFS than any of the patients who did not.” – by Rob Volansky

Reference:

Yee D, et al. Abstract GS3-08. Presented at: San Antonio Breast Cancer Symposium; Dec. 5-9, 2017; San Antonio.

Disclosures: Yee reports no relevant financial disclosures. Please see the abstract for all other authors’ relevant financial disclosures.

Perspective

Charles Shapiro, MD

For a long time, we knew that pathologic complete response rate was a good endpoint, meaning that the higher the pathologic complete response rate with neoadjuvant therapy, the better patients did. What we didn’t know was whether it correlated with long-term favorable outcomes like EFS and distant RFS. The question was: Did achieving a pathologic complete response mean anything for long-term clinical outcomes?

This trial was ideally suited to answer this question. What researchers showed was that, indeed, complete pathologic response rate is a surrogate for EFS and distant RFS. It can’t yet be inferred that it can be surrogate for OS, but it stands to reason that it will.

The I-SPY trial is a masterful trial in terms of taking pains to use uniform treatment, imaging and pathology, and these insure the results are reliable. The other way that this trial is novel is that it uses a Bayesian design. It uses real-time information from patients on the trial to decide whether to go forward or not. This is not the way clinical trials are traditionally designed. This Bayesian trial design minimizes the number of patients who are treated with ineffective drugs and maximizes the number of patients treated with effective drugs. It is another way of streamlining the clinical trial process. The I-SPY is a beautiful clinical trial program.

Charles Shapiro, MD

HemOnc Today Editorial Board Member

Icahn School of Medicine

Disclosures: Shapiro reports no relevant financial disclosures.