ICIs May Induce Higher Risk for irAEs in Older Patients With Early Breast Cancer

According to a multi-institutional study, older patients with early-stage breast cancer treated with immune checkpoint inhibitors (ICIs) are at a higher risk for developing high-grade immune-related adverse effects (irAEs).¹

In particular, the study, which was presented at the 2024 San Antonio Breast Cancer Symposium (SABCS), showed that risk was highest among those with chronic kidney disease (CKD).

Findings from a multi-institutional study show an increased risk of developing high-grade immune-related AEs among select older patients with early-stage breast cancer.

Specifically, the incidence of any- and high-grade irAEs were 72.6% and 18.9%, respectively, in patients with early-stage breast cancer who received ICIs. Among the patient population, Alexis LeVee, MD and colleagues demonstrated the distribution of any- and high-grade irAEs. Regarding any-grade irAEs, 0 was observed in 27% of patients, 1 was observed in 43%, 2 was observed in 23%, 3 was observed in 5.7%, 4 was observed in 1.2%, and 5 was observed in 0.2%. Of note, 0 high-grade irAEs occurred in 81% of patients, 1 was observed in 17%, and 2 was observed in 1.7%. There were no high-grade irAEs reported after 2.

“The rate of irAE may be higher in our study compared [with] that observed in KEYNOTE-522 (NCT03036488) due to differences in patient populations, with our study consisting of older patients who may have had more comorbidities,” LeVee and colleagues wrote in the poster presentation.

LeVee is chief fellow, Hematology and Medical Oncology at City of Hope in Los Angeles, California.

Background and Baseline Patient Characteristics

KEYNOTE-522 was a phase 3 trial that evaluated previously untreated patients with stage II or III triple-negative breast cancer. Patients on the trial were either treated with neoadjuvant therapy with 4 cycles of pembrolizumab (Keytruda) at 200 mg every 3 weeks plus paclitaxel and carboplatin or placebo every 3 weeks plus paclitaxel and carboplatin. The incidence of treatment-related AEs of grade 3 or greater was 78.0% in the pembrolizumab arm vs 73.0% in the placebo arm.²

The multi-institutional study included patients with stages I-III breast cancer who received ICIs at 4 academic institutions between 2014 and 2024.¹ Charts were reviewed for irAE, which was defined as toxicity determined by an oncologist to be possibly or definitely related to ICI or treated with high dose steroids when the causative agent was not identified.

Baseline laboratory values were collected before ICI, which included albumin, thyroid stimulating hormone (TSH), white blood cell (WBC), hemoglobin, platelets, absolute neutrophil count (ANC), absolute eosinophil count (AEC), absolute lymphocyte count (ALC), and absolute monocyte count (AMC). Ratios were calculated for lymphocyte to monocyte ratio, neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (dNLR) as defined by ANC or WBC-ANC. Uni- and multivariate logistic regression was performed to determine factors predictive of high-grade irAE.

Regarding patient demographics, any-grade irAE occurred in 307 patients and did not occur in 116. Median age was 51.0 (range, 41.0-62.0) and 50.5 (range, 43.0-60.0) in patients from the any-grade irAE cohort who did and did not have any-grade irAEs, respectively (P = 0.8). Regarding menopausal status, 155 patients (50.5%) and 61 patients (52.6%) were premenopausal among those who experienced any-grade irAEs and those who did not, respectively (P = 0.6). Postmenopausal status was reported in 148 (48.2%) and 52 (44.8%) patients, and unknown menopausal status was reported in 4 (1.3%) and 3 (2.6%) patients. Body mass index (BMI) was 27.0 (range, 23.4-31.3) and 27.3 (range, 23.7-31.5) (P = 0.7). Among patients who did and did not experience any-grade irAEs, respectively, 199 (64.8%) and 56 (48.3%) of patients were White, 25 (8.1%) and 27 (23.3%) were Hispanic, 26 (8.5%) and 10 (8.6%) were Black, 30 (9.8%) and 12 (10.3%) were Asian or Pacific Islander, 1 (0.3%) and 2 (1.7%) were American Indian, and 26 (8.5%) and 9 (7.8%) were unknown (P = 0.0005).

Smoking status and comorbidities were also observed among patients in the study. In patients from the any-grade irAEs cohort who did and did not experience any-grade irAEs, respectively, 12 (3.9%) and 4 (3.4%) patients were current smokers, 101 (32.9%) and 32 (27.6%) were former smokers, 193 (62.9%) and 80 (69.0%) were never smokers, and 26 (8.5%) and 9 (7.8%) were unknown (P = 0.6). Autoimmune disease was reported in 30 (9.8%) and 5 (4.3%; P = 0.07), diabetes mellitus in 35 (11.4%) and 12 (10.3%; P = 0.8), hypertension in 85 (27.7%) and 34 (29.3%; P = 0.7), hyperlipidemia in 51 (16.6%) and 15 (12.9%; P = 0.4), CKD in 12 (3.9%) and 1 (0.9%; P = 0.1), preexisting lung disease 102 (33.2%) and 35 (30.2%; P = 0.6), and coronary artery disease in 12 (3.9%) and 2 (1.7%; P = 0.3).

Patient demographics for the high-grade irAE cohort included 80 patients who experienced high-grade irAEs and 343 who did not experience high-grade irAEs. Median age was 53.5 (range, 46.0-64.0) and 50.0 (range, 40.0-61.0) in patients from the high-grade irAE cohort who did and did not experience high-grade irAEs, respectively (P = 0.05). Regarding menopausal status, 34 patients (42.5%) and 182 patients (53.1%) were premenopausal among those who experienced high-grade irAEs and those who did not, respectively (P = 0.2). Postmenopausal status was reported in 45 (56.2%) and 155 (45.2%) patients, and unknown menopausal status was reported in 1 (1.3%) and 6 (1.7%) patients. BMI was 26.1 (range, 23.1-30.7) and 27.4 (range, 23.7-31.5) (P = 0.1). Among patients who did and did not experience high-grade irAEs, respectively, 54 (67.5%) and 201 (58.6%) of patients were White, 5 (6.3%) and 47 (13.7%) were Hispanic, 3 (3.8%) and 33 (9.6%) were Black, 11 (13.8%) and 31 (9.0%) were Asian or Pacific Islander, 1 (1.3%) and 2 (0.6%) were American Indian, and 6 (7.5%) and 29 (8.5%) were unknown (P = 0.1).

Smoking status and comorbidities were also observed among patients in the study. In patients from the high-grade irAEs cohort who did and did not experience high-grade irAEs, respectively, 3 (3.8%) and 13 (3.8%) patients were current smokers, 18 (22.5%) and 115 (33.5%) were former smokers, 58 (72.5%) and 215 (62.7%) were never smokers, and 1 (1.3%) and 0 (0%) were unknown (P = 0.5). Autoimmune disease was reported in 8 (10.0%) and 27 (7.9%; P = 0.5), diabetes mellitus in 12 (15.0%) and 35 (10.2%; P = 0.2), hypertension in 22 (27.7%) and 97 (28.3%; P = 0.9), hyperlipidemia in 10 (12.5%) and 56 (16.3%; P = 0.4), CKD in 7 (8.8%) and 6 (1.7%; P = 0.001), preexisting lung disease 32 (40.0%) and 105 (30.6%; P = 0.1), and coronary artery disease in 5 (6.3%) and 9 (2.6%; P = 0.10).

Additional Data and Safety Plus Next Steps

Univariate and multivariate logistic analysis was used to identify predictors for high-grade irAEs in the study. These included age of at least 50 or older vs age 50 or younger (univariate: OR, 1.80; 95% CI, 1.08-3.01; P = 0.03; multivariate: OR, 1.69; 95% CI, 1.00-2.86; P = 0.05), BMI (univariate: OR, 0.96; 95% CI, 0.92-1.01; P = 0.10), race of non-Hispanic White vs Hispanic White (univariate: OR, 2.23; 95% CI, 0.84-5.90; P = 0.10), all other races vs Hispanic White (univariate: OR, 1.79; 95% CI, 0.63-5.08; P = 0.60), menopausal status (post vs pre/peri) (univariate: OR, 1.59; 95% CI, 0.97-2.62; P = 0.07), stage II vs I breast cancer (univariate: OR, 1.65; 95% CI, 0.66-4.14; P = 0.14, stage III vs IV (univariate: OR, 1.13; 95% CI, 0.43-2.99; P = 0.70, and smoking status of never vs current/past (univariate: OR, 1.61; 95% CI, 0.93-2.78; P = 0.09).

Univariate and multivariate logistic analysis was also used to identify predictors for high-grade irAEs regarding comorbidities and laboratory values. Comorbidities included diabetes mellitus (univariate: OR, 1.61; 95% CI, 0.79-3.29; P = 0.18), hypertension (univariate: OR, 0.98; 95% CI, 0.57-1.70; P = 0.90), hyperlipidemia (univariate: OR, 0.74; 95% CI, 0.36-1.52; P = 0.40), CKD (univariate: OR, 5.42; 95% CI, 1.77-16.6); P < 0.001; multivariate: OR, 4.93; 95% CI, 1.59-15.3; P = 0.006), preexisting lung disease (univariate: OR, 1.49; 95% CI, 0.90-2.48; P = 0.12), coronary artery disease (univariate: OR, 2.49; 95% CI, 0.81-7.65; P = 0.11). Laboratory values included albumin (univariate: OR, 0.78; 95% CI, 0.42-1.42; P = 0.40), TSH (univariate: OR, 0.99; 95% CI, 0.82-1.21; P = 0.90), WBC (univariate: OR, 0.99; 95% CI, 0.89-1.11; P = 0.90), hemoglobin (univariate: OR, 1.04; 95% CI, 0.96-1.12; P = 0.90), platelet (univariate: OR, 1.00; 95% CI, 0.998-1.005; P = 0.40), ANC (univariate: OR, 0.96; 95% CI, 0.84-1.10; P = 0.50), AEC (univariate: OR, 2.49; 95% CI, 0.42-14.6; P = 0.30), ALC (univariate: OR, 0.91; 95% CI, 0.68-1.23; P = 0.60), AMC (univariate: OR, 0.69; 95% CI, 0.20-2.40; P = 0.60). The calculated ratios were also included as such: LMR (univariate: OR, 1.03; 95% CI, 0.94-1.12; P = 0.50), NLR (univariate: OR, 1.01; 95% CI, 0.96-1.06; P = 0.80), MLR (univariate: OR, 1.02; 95% CI, 0.59-1.77; P = 0.90), PLR (univariate: OR, 1.00; 95% CI, 0.999-1.001; P = 0.70), and dNLR (univariate: OR, 0.99; 95% CI, 0.92-1.06; P = 0.80).

“Univariate analysis showed that patients [who are White] and those with higher AEC values were at higher risk of any-grade irAE, while older patients, never smokers, and those with CKD were at higher risk of high-grade irAE,” LeVee and colleagues wrote on the poster. “Multivariate logistic regression demonstrated that older age and the presence of CKD were significant predictors of high-grade irAE.”

The most common types of any-grade irAEs included thyroiditis (n = 112), rash (n = 106), colitis (n = 67), hepatitis (n = 49), adrenal insufficiency/hypophysitis (n = 35), arthralgia (n = 25), pneumonitis (n = 11), and diabetes mellitus (n = 8). The majority of these irAEs were grades 1 or 2.

“Further research is necessary to identify risk factors and biomarkers of irAE in patients with early breast cancer to help prevent irAE and mitigate long-term complications,” LeVee and colleagues concluded.

References

1. LeVee A, Jacob S, Fisch S, et al. Incidence and Risk Factors of Immune-Related Adverse Events in Early-Stage Breast Cancer Patients: Findings from a Multi-Institutional Study. Presented at: 2024 San Antonio Breast Cancer Symposium; December 10-13, 2024; San Antonio, TX. Abstract PS5-03.
2. Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for Early Triple-Negative Breast Cancer. The New England Journal of Medicine. 2020;382(9):810-821. doi: 10.1056/NEJMoa1910549