Stereotactic body radiotherapy combined with immunotherapy: a systematic review focus on timing and toxicity profile.

Introduction
Immune checkpoint inhibitors (ICIs) have revolutionized the treatment landscape for a broad range of solid malignancies by unleashing antitumor immune responses that were previously suppressed by tumor-mediated immune evasion [1]. Despite their remarkable efficacy in a subset of patients, a significant proportion fails to achieve durable responses [2]. In this context, Stereotactic Body Radiotherapy (SBRT) has garnered interest as a potential synergistic partner due to its ability to induce immunogenic cell death [3], promote antigen presentation, and modulate the tumor microenvironment in ways that may enhance systemic immune activation—a phenomenon commonly referred to as the “abscopal effect” [4].
Preclinical models and early-phase clinical data suggest that combining SBRT with ICIs may significantly amplify immune-mediated tumor control [5].
The radiobiological rationale supporting the combination of SBRT with immune checkpoint inhibitors is grounded in the capacity of ablative radiation to promote immunogenic cell death and to potentiate antitumor immunity. High-dose per fraction irradiation triggers calreticulin exposure, ATP release, and HMGB1 translocation, facilitating dendritic-cell activation and antigen cross-presentation, thereby enhancing the priming of tumor-specific cytotoxic T cells [6]. Fractionated regimens in the range of 8–12 Gy per fraction have been shown to optimally stimulate immune activation, whereas very high single-fraction doses induce excessive TREX1 exonuclease activity, leading to cytosolic DNA degradation and suppression of type I interferon signaling [7], [8]. In parallel, SBRT modulates the tumor microenvironment by reducing immunosuppressive populations such as regulatory T cells and myeloid-derived suppressor cells and by promoting infiltration of both resident and circulating T-cell clones, especially when combined with PD-1 blockade [6]. Radiation also induces transient upregulation of PD-L1 on tumor and stromal cells, creating a state of adaptive immune resistance that becomes particularly susceptible to checkpoint inhibition [3], [9]. Altogether, these mechanisms support the concept of SBRT as an in situ vaccine, capable of broadening tumor antigenicity, enhancing systemic immune activation, and providing a strong biological basis for synergy with ICIs [10].
The clinical translation of this strategy remains challenging, largely due to considerable heterogeneity in study design, including variations in dose prescription, planning, timing, and treated tumor types [11]. Furthermore, the toxicity profile of this combined approach remains overall poorly characterized, with inconsistent reporting of adverse events and a lack of standardized definitions for immune-related toxicity.
To address these uncertainties, we conducted a systematic review with a structured search strategy and performed data extraction focused on toxicity outcomes from clinical trials and cohort studies investigating SBRT in combination with ICIs in both primary and metastatic settings, with most studies enrolling patients with advanced or metastatic disease. The aim of this systematic review was to identify patterns of grade ≥3 toxicity and assess whether the addition of SBRT alters the immune toxicity landscape.

Methods
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [12]. The objective was to evaluate the safety profile of stereotactic body radiotherapy (SBRT) in combination with immune checkpoint inhibitors (ICIs), with a specific focus on the incidence and characterization of grade ≥3 treatment-related toxicities.
A comprehensive literature search was performed in PubMed/MEDLINE, Embase, and the Cochrane Library. The search covered all records published up to the date of the final search and was restricted to studies involving human subjects. The search strategy combined controlled vocabulary terms and free-text keywords related to radiotherapy and immunotherapy, including: “stereotactic body radiotherapy”, “SBRT”, “stereotactic ablative radiotherapy”, “SABR”, “immunotherapy”, “immune checkpoint inhibitors”, “PD-1”, “PD-L1”, “CTLA-4”, and “toxicity”. Reference lists of relevant reviews and included articles were manually screened to identify additional eligible studies.
After removal of duplicates, titles and abstracts were independently screened by two reviewers to exclude clearly irrelevant studies. Full-text articles were then assessed for eligibility. Any disagreements between reviewers were resolved through consensus discussion. Studies were included if they met all of the following criteria: Clinical studies (prospective or retrospective) involving adult patients with solid tumors; Use of SBRT in combination with ICIs; Reporting of treatment-related toxicity, with explicit documentation of grade ≥3 adverse events according to standardized criteria CTCAE; Original research articles published in peer-reviewed journals. Studies were excluded if they: Did not report toxicity outcomes; Investigated tumor settings outside the scope of this review; Included intracranial-only radiotherapy; Were preclinical studies, reviews, editorials, or case reports.
From each eligible study, the following data were systematically extracted: study design, tumor type, number of patients, disease stage, prior systemic therapies, type of ICI, SBRT dose and fractionation, timing of SBRT relative to immunotherapy (concurrent or sequential), and incidence and type of grade ≥3 treatment-related toxicities. Where available, toxicity data were recorded separately for concurrent and sequential treatment strategies. Survival outcomes (overall survival, progression-free survival, and local control) were collected as secondary endpoints when reported, but were not the primary focus of this review Table 1. For the purpose of this review, treatment sequencing was classified as concurrent when SBRT was delivered during ongoing immunotherapy or within the same treatment cycle. Sequential treatment was defined as SBRT delivered before or after immunotherapy with a clear temporal separation between modalities. When studies used non-standard terminology or did not explicitly define sequencing, classification was determined based on the reported timing of SBRT relative to initiation of immune checkpoint inhibitors.
As this study was based exclusively on previously published data and did not involve human participants or identifiable patient information, ethical approval was not required according to institutional policies.
A prospectively developed and registered review protocol was not available prior to study initiation. The absence of a predefined protocol may have introduced a potential risk of bias in study selection, data extraction, and outcome reporting. However, the review was conducted following a predefined methodological approach based on PRISMA recommendations.

Results
Starting from n = 69 screened records, n = 22 studies were excluded due to the absence of toxicity data and n = 16 were excluded as they addressed tumor settings not relevant to the scope of this review. Overall, n = 31 (44.9%) studies met the inclusion criteria and included Fig. 1.
The final cohort comprised n = 2,355 patients presenting diverse malignancies, encompassing n = 31 studies: n = 11 NSCLC, five melanoma, four breast cancer, two renal cell carcinoma, prostate cancer, three prostate cancer, two pancreatic cancer, one hepatocellular cancer, one colorectal cancer, one urothelial cancer and one HNSCC.
Eligibility criteria were generally homogeneous across the included studies, primarily enrolling patients with advanced or metastatic disease. Specifically, n = 27 out of n = 31 studies (87%) enrolled patients with stage IV or recurrent disease. Oligometastatic presentation—defined as the presence of one to five metastatic lesions [13]—was an explicit inclusion criterion in n = 18 studies (58%), while the remaining nine studies (29%) allowed broader metastatic involvement. The remaining two studies did not clearly specify metastatic burden criteria.
Prior systemic therapy, including chemotherapy or immunotherapy, was permitted in n = 23 studies (74%), and n = 14 studies (45%) specifically allowed inclusion of patients who had previously received ICIs. Adequate performance status, generally described as ECOG 0–1, was required in n = 29 studies (94%), and basic laboratory thresholds to ensure sufficient hematologic, renal, and hepatic function were described in n = 25 studies (81%), although specifics varied.
Brain metastases were an exclusion criterion in nine studies (29%), whereas the remaining studies either permitted stable brain lesions or did not specify.
In terms of study design, eight were randomized trials (25.8%), n = 15 were prospective observational studies (48.4%), and eight were retrospective analyses (25.8%).
Twelve studies (38.7%) were designed as Phase I/II investigations with safety and tolerability as primary endpoints, specifically evaluating toxicity profiles and determining maximum tolerated doses for various SBRT-immunotherapy regimens.
Eleven trials (35.5%) were Phase II/III studies with efficacy as their principal focus, measuring hard clinical endpoints including overall survival, progression-free survival, and objective response rates.
Four studies (12.8%) specifically addressed treatment sequencing by directly comparing concurrent versus sequential administration approaches.
Two trials (6.5%) were dedicated to mechanistic investigations of immunological biomarkers and tumor microenvironment modifications, while the remaining two studies (6.5%) explored specialized clinical applications in neoadjuvant and oligometastatic settings.
Regarding treatment sequencing, n = 15 studies (48.4%) used concurrent SBRT and immunotherapy, n = 13 (41.9%) used a sequential approach, and three (9.7%) included both strategies. Patient allocation mirrored this distribution, with n = 1,032 patients (43.8%) treated concurrently, 1,239 (52.7%) sequentially, and n = 84 (3.5%) either untreated or whose treatment sequence was unspecified.
A wide range of SBRT dose and fractionation regimens was utilized across the included studies. The most commonly adopted schedule was 24 Gy in three fractions (8 Gy per fraction), corresponding to a BED10 of 43.2 Gy, frequently used in lung, breast, and melanoma trials.
Higher-dose regimens included 30 Gy in three fractions (10 Gy per fraction, BED10 = 60 Gy) and 50 Gy in four fractions (12.5 Gy per fraction, BED10 = 112,5 Gy), applied particularly in lung and melanoma cohorts.
Single-fraction treatments, such as 20 Gy in one fraction (BED10 = 60 Gy), were also reported, especially for renal cell carcinoma metastases. Intermediate regimens like 36 Gy in three fractions (BED10 = 57.6 Gy) and 45 Gy in five fractions (BED10 = 85.5 Gy) were observed in upper abdominal tumors.
When the target’s proximity to critical organs discouraged higher doses, conservative protocols were employed, including regimens of 30–51 Gy in six to ten fractions, such as 30 Gy in five fractions (BED10 =43.68 Gy), particularly in gastrointestinal and hepatobiliary malignancies.
In a limited number of prostate and lung cancer studies, traditional palliative regimens like 45 Gy in 15 fractions (3 Gy per fraction, BED10 = 58.5 Gy) were used, though these regimens constituted exceptions to the predominantly ablative SBRT approaches.
Toxicity data showed that 126 (12%) patients in the ICI concurrent group and 131 (11%) in the sequential group experienced grade ≥3 treatment-related toxicities, with no significant difference between the two strategies (p = 0.22). Pneumonitis was markedly more frequent in the concurrent setting (6.1% vs. 1.9%), representing the only toxicity with a statistically significant difference (p < 0.001). Gastrointestinal toxicities, such as colitis, were slightly more common in the sequential group (2.3% vs. 2.0%; p = 0.71). Other grade ≥3 events showed no significant differences between concurrent and sequential treatment, including hepatitis (1.2% vs. 1.5%; p = 0.55), rash (0.9% vs. 1.7%; p = 0.087), diarrhea (0.8% vs. 1.4%; p = 0.17), fatigue (0.7% vs. 1.0%; p = 0.35), and hematologic toxicities (0.6% vs. 0.8%; p = 0.52). Across the studies, five cases of grade 5 toxicity were reported: two cases of pneumonitis, one case of multi-organ failure, one case of upper gastrointestinal bleeding, and one case of abdominal infection. These fatal events were considered potentially attributable to the combination of SBRT and systemic therapy (<1%), although their rarity precluded any meaningful statistical analysis Table 2. These comparisons represent unadjusted pooled estimates derived from trial-level data rather than patient-level meta-analytic comparisons, and should therefore be interpreted with caution.
The analysis of survival outcomes across multiple tumor types did not disclose any trend when comparing concurrent versus sequential administration of SBRT with ICIs.
Concurrent treatment appears to offer modest advantages in NSCLC. In the trials reporting survival outcomes, median OS was 9.1 months with concurrent therapy compared with 7.6 months with sequential therapy, while PFS similarly favored concurrent approaches (6.9 vs 4.8 months) [9], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23].
Melanoma stands out as particularly responsive to concurrent therapy, demonstrating a striking median OS of 27.5 months with combined treatment versus 10.6 months without radiotherapy, though PFS outcomes remain variable across studies [24], [25], [26], [27], [28].
Breast cancer presents a more complex picture, where sequential strategies show promising 1-year OS rates of 85%, median PFS remains limited to 1.9–2.6 months regardless of treatment timing [29], [30], [31], [32].
Prostate cancer data suggest that while concurrent therapy yields a median OS of 14.1 months, sequential approaches with ipilimumab still demonstrate meaningful benefit (11.0 vs. 10.0 months for placebo) [33], [34], [35].
In pancreatic cancer concurrent SBRT with pembrolizumab and trametinib achieves a median OS of 24.9 months compared to 22.4 months for controls, accompanied by substantially improved PFS (18.3 vs. 15.6 months) [36], [37].
Head and neck squamous cell carcinoma (HNSCC) shows minimal benefit from SBRT addition with concurrent immunotherapy (OS 13.9–14.2 months; PFS 1.9–2.6 months) [38].
Renal cell carcinoma demonstrates appreciable outcomes with sequential therapy (20 months OS; 5.6 months PFS), while urothelial cancer shows a clear advantage for concurrent timing (12.1 vs. 4.5 months OS; 3.5 vs. 3.3 months PFS) [39], [40].
Aggregating data across all tumor types reveals that concurrent therapy maintains a slight advantage, with weighted average median OS of 12.4 months versus 10.2 months for sequential approaches, and PFS of 6.8 versus 4.5 months. These values represent descriptive cross-trial averages rather than adjusted comparative analyses and should therefore be interpreted with caution.
A further methodological limitation concerns the absence of standardized criteria for target volume selection. Among the 31 included studies, only 21 (68%) explicitly defined eligibility based on the number of metastatic lesions, typically requiring at least two measurable sites amenable to SBRT. However, even within these studies, the rationale for selecting which lesions to irradiate was poorly detailed. None of the trials provided systematic guidance on whether all metastatic sites, only extracranial lesions, the largest lesions, or the most biologically active targets should be treated. The remaining 10 studies (32%) did not report any information regarding lesion selection, extent of metastatic burden treated, or criteria for prioritizing SBRT targets. This variability in target definition and the lack of explicit volume-selection protocols introduce substantial heterogeneity in treatment exposure, potentially influencing both clinical outcomes and the interpretation of synergistic effects between SBRT and immunotherapy Table 3.

Discussion
This review analyzes the role of combining SBRT with ICIs in more than 2,300 patients across 31 studies. Despite increasing interest, the available evidence is limited by substantial heterogeneity in trial design, including tumor types, patient characteristics, SBRT dose and fractionation, and, most notably, the timing of SBRT relative to immunotherapy. “Concurrent” treatment was indeed variably defined—from days to weeks after ICI initiation—while “sequential” strategies ranged from weeks to months, precluding identification of an optimal schedule. Preclinical studies suggest that delivering SBRT shortly before or during ICIs may enhance immune priming, but this has not been consistently demonstrated in clinical settings.
Dose and fractionation varied widely, from palliative regimens (8 Gy in single fraction) to ablative schedules (e.g., 50 Gy in 4 fractions), with BED10 values ranging from ∼ 40 Gy to over 110 Gy. In vitro data indicate that fractionated regimens of 8–12 Gy/fraction over three to five sessions may best balance immune stimulation and toxicity, while very high single-fraction doses (>20 Gy) may activate immunosuppressive pathways [41], [42]. However, such biological considerations were inconsistently applied in the clinical protocols analyzed. Notably, 36% of studies failed to report dose and fractionation in sufficient detail for BED calculation, further complicating cross-study comparison.
Toxicity reporting was inconsistent as well. Some trials only provided aggregated rates, others entirely omitted data, and only few attempted to distinguish radiation-related from immune-mediated adverse events. This is particularly problematic in the thoracic setting, where pneumonitis is a potential consequence of both SBRT and ICIs. In our analysis, the overall incidence of grade ≥ 3 toxicity was similar between concurrent (12–14%) and sequential (11–15%) strategies. Still, pneumonitis was more frequent with concurrent treatment, whereas systemic immune-related adverse events such as colitis, rash, and fatigue appeared more common in sequential regimens. These differences suggest possible interactions between timing and immune response, but data quality remains insufficient for firm conclusions.
Despite these limitations, prospective studies increasingly integrate translational endpoints such as tumor mutational burden, PD-L1 expression, immune profiling, and radiomics, reflecting a shift toward mechanism-driven trial design. An additional methodological limitation is the absence of a prospectively defined and registered review protocol. Without a predefined protocol, there is an increased risk of selection bias, data-driven decisions during study inclusion, and selective outcome reporting. This may have influenced the consistency and transparency of the review process. Although efforts were made to apply systematic and reproducible methods in line with PRISMA recommendations, the lack of protocol registration should be considered when interpreting the findings.
A further evident practical advantage of integrating SBRT with immunotherapy lies in the ability to quickly deliver effective treatment courses, enabling the irradiation also of multiple metastatic lesions within a very short timeframe. Modern SBRT platforms allow for precise, image-guided treatment of different targets without prolonging patients’ overall treatment time (OTT). This rapid scheduling can be particularly valuable when SBRT is administered between immunotherapy cycles, minimizing delays in systemic therapy and reducing the risk of disrupting immune activation. Treating multiple lesions in close temporal proximity may also enhance the systemic immunologic effect by simultaneously increasing tumor antigen release from different metastatic sites, thereby amplifying the in situ vaccination effect [43].
These operational and biological considerations underscore the potential of SBRT to be efficiently integrated into the immunotherapy workflow without compromising treatment continuity or systemic therapeutic intensity.
Nevertheless, key unanswered questions still remain open, such as the optimal SBRT regimen to enhance ICIs efficacy, the sequencing of therapies to maximize synergy, and the tumor types most likely to benefit. Breast cancer, for instance, is underrepresented in literature with early data suggesting higher toxicity, while prostate studies often use lower dose palliative SBRT, unlikely to trigger immune effects.
Future research should focus on standardizing treatment parameters, harmonizing toxicity attribution, and incorporating translational biomarkers.

Conclusion
Current evidence does not support a systematic increase in severe toxicity when SBRT is combined with ICIs. Grade ≥ 3 adverse events remain relatively common but are largely immune-mediated and consistent with known ICIs safety profiles, particularly in NSCLC, the most extensively studied setting.
The current barriers to progress are mainly methodological: heterogeneous SBRT schedules, variable definitions of treatment sequencing, and inadequate reporting or attribution of adverse events. Breast cancer raises specific concerns due to limited data and early signals of increased toxicity, while hepatobiliary, pancreatic, and prostate studies remain exploratory and underpowered.
Future trials should adopt harmonized protocols for SBRT dosing and timing, systematic toxicity attribution, and integrated translational endpoints. These efforts are essential to define the therapeutic window, refine patient selection, and identify tumor contexts where SBRT-ICI combinations may offer the greatest benefit. For now, clinical application should remain cautious and individualized, but available data suggest that SBRT can be safely integrated with ICIs in selected patients.

CRediT authorship contribution statement
Carlo Gugliemo Cattaneo: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing. Giuditta Chiloiro: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Writing – original draft, Writing – review & editing. Angela Romano: Formal analysis, Methodology, Writing – review & editing. Giulia Panza: Visualization. Matteo Galetto: Visualization. Matteo Nardini: Visualization. Lorenzo Placidi: Supervision, Validation. Maria Antonietta Gambacorta: Resources, Supervision. Luca Boldrini: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.