Introduction
Triple-negative breast cancer (TNBC) is a biologically aggressive subtype of breast cancer defined by the absence of estrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 (HER2) expression. It accounts for 15–20% of all breast cancers worldwide and is disproportionately diagnosed in younger women, with a median age at diagnosis of ~53 years and a particularly high incidence among premenopausal women and those under 40 years of age (1). Clinically, TNBC often presents as a rapidly growing palpable mass and, due to its high proliferative rate, it is more likely to manifest as an interval cancer between screening mammograms (1). The disease is characterized by substantial biological heterogeneity, with the basal-like molecular subtype predominating, and is associated with poor histological differentiation, aggressive clinical behavior, and an elevated risk of early recurrence and visceral metastasis (1,2). Unlike hormone receptor-positive or HER2-positive breast cancers, TNBC lacks established molecular targets, and until recently, cytotoxic chemotherapy remained the only systemic treatment option. This therapeutic limitation, combined with its aggressive natural history, has historically resulted in a poorer prognosis, with a 5-year overall survival rate of 60–65% for patients with locally advanced disease (1).
The treatment landscape for early-stage TNBC has been transformed by the introduction of immune checkpoint inhibitors (ICIs) targeting the programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) axis. The landmark KEYNOTE-522 phase III trial demonstrated that the addition of pembrolizumab to standard neoadjuvant chemotherapy significantly improved the pathological complete response (pCR) rate from 51.2 to 64.8% and prolonged event-free survival (3). Based on these results, neoadjuvant pembrolizumab plus chemotherapy followed by adjuvant pembrolizumab has become the standard of care for high-risk, early-stage TNBC. In patients who achieve pCR after neoadjuvant therapy, the decision to administer postoperative radiotherapy (RT) is typically guided by the initial clinical stage and other risk factors, including tumor size and lymph node involvement at presentation (4). For patients with residual disease after neoadjuvant therapy, postoperative RT is routinely indicated. Consequently, an increasing number of patients with TNBC are receiving adjuvant ICIs concurrently with postoperative RT, raising important clinical questions regarding the safety of this combined approach.
Radiotherapy exerts its antitumor effects not only through direct DNA damage but also through the induction of immunogenic cell death, which can prime systemic antitumor immune responses (5). Preclinical studies have demonstrated that RT upregulates PD-L1 expression on tumor cells and enhances the presentation of tumor-associated antigens, thereby providing a mechanistic rationale for combining RT with ICIs (6,7). This synergy has been shown to enhance local tumor control and, in some instances, mediate abscopal effects, where distant, non-irradiated tumors also regress (7). However, this immunostimulatory activity also raises concerns that the combination of RT and ICIs could potentiate immune-mediated injury in normal tissues within the radiation field. Evidence from multiple studies in lung cancer has shown that the addition of ICIs to thoracic RT is associated with an increased risk of serious adverse events, particularly radiation pneumonitis (8,9). In the PACIFIC trial of durvalumab following chemoradiotherapy for stage III non-small-cell lung cancer, the incidence of any-grade pneumonitis was 33.9% in the durvalumab group compared with 24.8% in the placebo group, and grade 3–4 pneumonitis occurred in 3.4% of patients receiving durvalumab (9). In severe cases, radiation pneumonitis can be life-threatening, and this toxicity has become a major consideration in the clinical implementation of combined RT-ICI regimens.
Despite these concerns, data on the safety of concurrent RT and ICIs in patients with breast cancer, and particularly in those with TNBC, remain limited. Compared with thoracic RT for lung cancer, RT fields for breast cancer typically involve smaller lung volumes and lower doses to critical organs. Furthermore, the patient population differs substantially in terms of baseline pulmonary function and comorbidities. A recent real-world study specifically evaluating the safety of concurrent pembrolizumab and RT in patients with TNBC reported that the combination was generally well tolerated, but the evidence base remains limited by small sample sizes and short follow-up (10). Therefore, the safety of concurrent postoperative RT and maintenance ICIs in patients with TNBC remains poorly understood.
To address this uncertainty, the present study retrospectively evaluated the safety and tolerability of concurrent postoperative RT and ICIs following neoadjuvant ICIs in patients with TNBC treated at Hubei Cancer Hospital (Wuhan, China). The findings of the present study may help guide the concurrent use of ICIs and postoperative RT in this patient population.
Discussion
The present study revealed that concurrent immunotherapy was not associated with a statistically significant increase in the risk of radiation pneumonitis during postoperative RT for TNBC; however, a clinically meaningful trend toward a higher incidence of grade 1 radiation pneumonitis cannot be excluded. This finding is consistent with the results of the KEYNOTE-522 trial, which reported a low incidence of radiation pneumonitis in patients receiving immunotherapy (3). However, hypothyroidism was observed more frequently in patients treated with concurrent immunotherapy, likely attributable to ICI-associated adverse events. Therefore, ICI-related immune thyroiditis remains the most likely explanation for the increased incidence of hypothyroidism.
RT has traditionally been considered a localized cytotoxic treatment. However, emerging evidence suggests that RT can induce immunogenic cell death, particularly with hypofractionated regimens (14). Irradiated tumor cells release damage-associated molecular patterns (for example, high mobility group box 1, ATP and calreticulin), which act as danger signals to promote dendritic cell maturation and antigen presentation (15). When combined with immune checkpoint blockade, these newly exposed antigens stimulate tumor-specific T-cell responses, converting bystander T cells into active effector cells and ultimately triggering systemic antitumor immunity; the abscopal effect (16). In postoperative breast cancer RT, local elimination of microscopic residual disease may be accompanied by enhanced systemic immune surveillance, which could reduce the risk of recurrence and metastasis (17).
RT can reduce the infiltration of myeloid-derived suppressor cells and M2-polarized tumor-associated macrophages within the tumor microenvironment. Concurrently, RT upregulates programmed death-ligand 1 (PD-L1) expression on tumor cells, thereby providing a direct target for ICIs (18). RT also improves tumor vascular structure and perfusion, facilitating the infiltration and persistence of effector T cells within tumors. In the setting of postoperative RT for breast cancer, in which treatment is typically administered following chemotherapy and/or neoadjuvant immunotherapy, immune memory may already be established. Postoperative RT may further reshape the tumor microenvironment, thereby enhancing the local execution of immune memory responses (15).
Although the present study did not observe a statistically significant increase in radiation pneumonitis, RT and immunotherapy may exert additive toxic effects on normal tissues from a mechanistic perspective. Radiation-induced lung injury is mediated by inflammatory cytokines such as transforming growth factor-β and tumor necrosis factor-α, whereas immune checkpoint blockade may activate T cells which also target normal tissues (5,19,20). Clinically, the relatively low incidence of pneumonitis in patients with breast cancer may be attributable to smaller radiation field sizes and specific fractionation schedules. Notably, the relatively high incidence of grade 1 radiation pneumonitis observed in the present study (63.3% in the ICIs-RT group and 46.1% in the RT-alone group) may partially reflect the sensitivity of routine contrast-enhanced CT imaging performed every 3 months, which could detect subtle, asymptomatic radiological changes that might otherwise remain unnoticed. By contrast, the risk of pneumonitis is substantially higher in patients with lung or esophageal cancer, warranting careful monitoring in these settings. In patients with left-sided breast cancer, postoperative RT may expose the heart to radiation. Although immune-mediated myocarditis is rare, it can be life-threatening; therefore, stricter cardiac dose constraints should be considered when combining RT with immunotherapy (21).
The choice of PD-1 inhibitor (pembrolizumab, tislelizumab, camrelizumab or toripalimab) was based on physician preference and local drug availability, and was not determined by patient or disease-specific factors. Given the small number of patients receiving each agent, formal subgroup analyses were not feasible. No obvious differences in toxicity profiles were observed across agents; however, agent-specific effects cannot be excluded and should be examined in larger cohorts.
The present study has several limitations. First, as a single-center retrospective analysis, potential selection bias cannot be excluded. Notably, the significantly higher proportion of patients receiving platinum-based chemotherapy in the ICIs-RT group (90.0 vs. 56.8%; P<0.001) reflects real-world clinical practice, as ICIs in the neoadjuvant setting for TNBC are predominantly combined with platinum-containing regimens according to the KEYNOTE-522 protocol. This imbalance may have contributed to the higher pCR rate observed in the ICIs-RT group, and the relative contributions of platinum agents and ICIs cannot be disentangled within the current retrospective study design. Due to the small number of patients in the ICIs-RT group (n=30), propensity score matching and multivariable regression analyses were not feasible for adjusting baseline imbalances. Second, the follow-up duration was limited (median, 12 months); therefore, the data from the current study support conclusions regarding acute and early subacute toxicities only. Late effects, such as radiation fibrosis, cardiac toxicity and long-term survival outcomes, require extended follow-up. Furthermore, the small sample size resulted in insufficient statistical power. With only 30 patients in the intervention group and a low event count, the study was underpowered to detect uncommon but clinically meaningful toxicities, such as grade ≥2 radiation pneumonitis. Therefore, the absence of a statistically significant difference in grade 1 radiation pneumonitis should not be interpreted as evidence of equivalent safety. Future prospective studies with larger sample sizes and longer follow-up are warranted to validate these findings and to employ appropriate methodologies for adjusting baseline confounding. Multiple ongoing trials are currently evaluating combined RT and immunotherapy strategies in the neoadjuvant, adjuvant and metastatic settings. Radiation dose and fractionation schedules may modulate immunologic synergy (22). Notably, hypofractionated RT may enhance immune activation (23). However, the optimal timing and sequencing; whether concurrent, sequential or sandwich, remain undefined. Finally, the identification of predictive biomarkers such as PD-L1 expression and tumor-infiltrating lymphocyte levels, will be crucial for identifying patients most likely to benefit from combined therapy (24,25).
In summary, concurrent immunotherapy during postoperative RT was not associated with a significant increase in radiation dermatitis or a statistically significant increase in radiation pneumonitis; however, a clinically meaningful trend toward a higher incidence of grade 1 radiation pneumonitis cannot be excluded. Concurrent immunotherapy did not appear to adversely affect short-term clinical outcomes; however, the limited follow-up duration and small number of events preclude definitive conclusions. These findings support the safety of the combined approach but highlight the need for continued vigilance.