Efficacy of nab-paclitaxel combined with immune checkpoint inhibitors in solid tumors: a systematic review and meta-analysis of randomized controlled trials

Introduction

Over the past decade, immune checkpoint inhibitors (ICIs) targeting programmed cell death protein 1 (PD-1), programmed death-ligand 1 (PD-L1), and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), as well as emerging immunomodulatory targets such as ecto-5’-nucleotidase (CD73), have fundamentally transformed the therapeutic landscape for solid tumors (1,2). By releasing the inhibitory brakes on antitumor T-cell responses, ICIs have demonstrated durable clinical benefits across a wide spectrum of malignancies, including non-small cell lung cancer (NSCLC), triple-negative breast cancer (TNBC), and gastric cancer (1). Despite these advances, a substantial proportion of patients fails to respond to ICI monotherapy, largely owing to the immunosuppressive tumor microenvironment (TME), limited tumor immunogenicity, and insufficient T-cell infiltration (3,4). This therapeutic reality has driven intensive investigation into rational combination strategies that can augment the efficacy of immunotherapy.

Among these strategies, the combination of chemotherapy with ICIs has gained the most extensive clinical validation. Standard cytotoxic agents can enhance antitumor immunity through multiple mechanisms, including the induction of immunogenic cell death (ICD), the release of tumor-associated neoantigens, the depletion of immunosuppressive regulatory T cells (Tregs), and the promotion of dendritic cell maturation and antigen presentation (5,6). However, conventional solvent-based paclitaxel formulations require corticosteroid premedication to mitigate hypersensitivity reactions, which may attenuate ICI-mediated immune activation (3,7). Nanoparticle albumin-bound paclitaxel (nab-paclitaxel, Abraxane), a solvent-free formulation utilizing 130-nm albumin nanoparticles as a drug carrier, has been developed to overcome these pharmacological limitations (8). Nab-paclitaxel exploits the endogenous albumin transport pathway, binding to the gp60 receptor on vascular endothelial cells and the secreted protein acidic and rich in cysteine (SPARC) enriched in the TME, thereby achieving higher intratumoral drug concentrations and faster tissue penetration compared with solvent-based paclitaxel (8). Crucially, nab-paclitaxel does not require corticosteroid premedication, thereby preserving the immune milieu necessary for optimal ICI efficacy (3). Beyond PD-1/PD-L1 and CTLA-4, the CD73-adenosine pathway has emerged as another immunosuppressive axis within the TME, in which CD73 generates extracellular adenosine that suppresses T-cell antitumor activity (9).

In recent years, neoadjuvant immunotherapy has emerged as a rapidly evolving paradigm in oncology, grounded in the rationale that administering systemic therapy prior to surgical resection may elicit broader and more robust antitumor T-cell responses against the intact tumor and regional lymph nodes (10,11). Pathological complete response (pCR), defined as the absence of residual invasive tumor in the resected specimen, has been increasingly adopted as a key efficacy endpoint in neoadjuvant trials, serving as an early surrogate for long-term survival outcomes such as event-free survival and overall survival (OS) in certain tumor types (11,12). In early-stage TNBC, the GeparNuevo trial demonstrated that durvalumab combined with nab-paclitaxel-based neoadjuvant chemotherapy improved pCR and long-term survival outcomes (13), and the IMpassion031 trial confirmed that neoadjuvant atezolizumab plus nab-paclitaxel followed by anthracycline-based chemotherapy significantly increased pCR rates (14). In esophageal squamous cell carcinoma (ESCC) and gastric cancer, camrelizumab-based combinations with nab-paclitaxel have shown promising pCR improvements in recent phase II and III trials (15-17). These developments underscore the need for comprehensive evidence synthesis across tumor types and treatment settings.

Despite the expanding body of evidence, existing meta-analyses on nab-paclitaxel plus ICI combinations have notable limitations that constrain the generalizability of their conclusions. Some prior systematic reviews focused exclusively on a single cancer type, such as TNBC (18) or NSCLC, while others adopted broader scopes encompassing all taxane formulations rather than isolating the specific contribution of nab-paclitaxel (19). Importantly, most previous meta-analyses did not include pCR as a primary endpoint, thereby overlooking the growing body of neoadjuvant trial data. Furthermore, prior reviews were conducted before the publication of several recent large-scale phase II/III trials, which have substantially enriched the available evidence base. These gaps highlight the need for an updated, cross-tumor meta-analysis that integrates the most current high-quality randomized controlled trial (RCT) data and provides a focused assessment of nab-paclitaxel plus ICI efficacy across both the advanced and neoadjuvant treatment settings.

To address these research gaps, we conducted this comprehensive systematic review and meta-analysis of RCTs to evaluate the efficacy of nab-paclitaxel combined with ICIs compared with nab-paclitaxel-based regimens without immunotherapy across multiple solid tumor types. Our primary objectives were to quantify the pooled effects on OS, progression-free survival (PFS), and pCR, while performing subgroup analyses by cancer type and ICI class to identify potential differential treatment benefits. By synthesizing data from 24 RCTs enrolling 6,682 patients across five tumor types, this study provides the most up-to-date and comprehensive evidence to inform clinical decision-making regarding nab-paclitaxel plus ICI combination strategies in solid tumors. We present this article in accordance with the PRISMA reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0627/rc).

Methods

The study protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number: CRD420261307950).

Inclusion and exclusion criteria

Eligible studies were identified using the Population, Intervention, Comparator, Outcome, and Study design (PICOS) framework. The population comprised patients with pathologically confirmed solid tumors, including but not limited to NSCLC, TNBC, pancreatic ductal adenocarcinoma (PDAC), gastric cancer, and ESCC, with no restrictions on disease stage, age, sex, race, or PD-L1 expression status. The intervention was defined as nab-paclitaxel (albumin-bound paclitaxel; Abraxane) combined with any ICI, including anti-PD-1 agents (pembrolizumab, nivolumab, camrelizumab, serplulimab, toripalimab, tislelizumab), anti-PD-L1 agents (atezolizumab, durvalumab), and anti-CTLA-4 agents (ipilimumab, tremelimumab), and anti-CD73 agents (oleclumab), with or without additional chemotherapy agents. The comparator was nab-paclitaxel monotherapy or nab-paclitaxel-based chemotherapy without immunotherapy, placebo plus nab-paclitaxel, or other standard-of-care treatment regimens. Only RCTs were eligible for inclusion. The primary outcomes were OS, defined as the time from randomization to death from any cause; PFS, defined as the time from randomization to the first occurrence of disease progression or death from any cause, whichever came first; and pCR, defined as the absence of residual invasive tumor in the resected specimen at the time of surgery (applicable to neoadjuvant studies only). Secondary outcomes were objective response rate (ORR), defined as the proportion of patients achieving a complete response (CR) or partial response (PR) according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1; and disease control rate (DCR), defined as the proportion of patients achieving CR, PR, or stable disease per RECIST v1.1.

Studies were excluded if they met any of the following criteria: (I) non-randomized study designs, including single-arm trials, observational studies, case reports, reviews, and conference abstracts without sufficient data; (II) duplicate publications or overlapping patient populations (in such cases, the most recent or comprehensive report was retained); and (III) studies with insufficient data for quantitative synthesis that could not be obtained from the authors.

Search strategy

A comprehensive literature search was performed across four electronic databases: PubMed, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), and Web of Science, from inception through January 1, 2026. The Embase and CENTRAL searches were conducted simultaneously via the Ovid platform. The search strategy combined medical subject headings (MeSH) terms and free-text keywords. No language or publication date restrictions were applied. The complete search strategies for all databases are presented in Appendix 1. In addition to the electronic database searches, the reference lists of all included studies and relevant systematic reviews were manually screened to identify any additional eligible trials that may have been missed by the primary search.

Data extraction

Two reviewers independently performed study screening and data extraction using a pre-designed standardized data extraction form. Titles and abstracts were first screened for eligibility, followed by full-text review of potentially relevant articles. Any disagreements were resolved through discussion with a third reviewer.

For time-to-event outcomes (OS and PFS), hazard ratios (HRs) and their 95% confidence intervals (CIs) were extracted directly from the published reports. When HRs were not explicitly reported, they were estimated from Kaplan-Meier survival curves. For dichotomous outcomes (pCR, ORR, and DCR), the number of events and total participants in each treatment arm were extracted to calculate risk ratios (RRs) with 95% CIs.

Risk of bias assessment

The methodological quality of included RCTs was assessed independently by two reviewers using the Cochrane Collaboration’s Risk of Bias tool (20). Each study was evaluated across seven domains: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other potential sources of bias. Each domain was classified as “low risk”, “unclear risk”, or “high risk” of bias. Discrepancies between reviewers were resolved through consensus or adjudication by the third author. The results of the risk of bias assessment were presented graphically as both a summary figure and an individual study-level figure.

Statistical analysis

All statistical analyses were performed using R software (version 4.4.2) with the meta package. For time-to-event outcomes (OS and PFS), pooled HRs with 95% CIs were calculated using the generic inverse-variance method, where an HR <1 indicated a survival benefit favoring the nab-paclitaxel plus ICI group. For dichotomous outcomes (pCR, ORR, and DCR), pooled RRs with 95% CIs were computed using the Mantel-Haenszel method, where an RR >1 indicated superior efficacy in the combination immunotherapy group.

Both the common-effect (fixed-effect) model and the DerSimonian-Laird random-effects model were employed for all analyses. Statistical heterogeneity was evaluated using the Cochran’s Q test and the I² statistic. Based on the thresholds proposed by Higgins et al. (21), I² values of <25%, 25–50%, and >50% were interpreted as indicating low, moderate, and substantial heterogeneity, respectively. Subgroup analyses were performed by cancer type and by ICI class. For each outcome, subgroup analysis by cancer type was performed using all available studies that reported data on that specific outcome. Variation in the number of contributing studies across outcomes therefore reflects the differential reporting practices of the original RCTs rather than analytical exclusion. In particular, pCR was reported only in neoadjuvant studies, whereas OS, PFS, ORR, and DCR were available across a broader range of tumor types and treatment settings. All ICI-class subgroups present in the included RCTs were retained in the subgroup analyses for transparency and completeness of reporting. For ICI combinations represented by only a single trial (anti-PD-L1 plus anti-CTLA-4; anti-PD-L1 plus anti-CD73), the displayed estimates correspond to the original trial result rather than a pooled effect size; these are reported descriptively, and no formal heterogeneity or between-subgroup interaction inference was drawn from them. Differences between subgroups were assessed using the Chi-squared test for interaction, with a P value <0.05 considered statistically significant.

Leave-one-out sensitivity analyses were performed to evaluate the robustness of pooled estimates by sequentially omitting each individual study and recalculating the summary effect size. Publication bias was assessed through visual inspection of funnel plots as well as formal statistical testing using Egger’s linear regression test and Begg’s rank correlation test. A two-tailed P value <0.05 was considered indicative of potential publication bias for all statistical tests. All reported P values were two-sided.

Results

Characteristics of the included studies

The study selection process is detailed in the PRISMA flow diagram (Figure 1). A total of 24 RCTs met the eligibility criteria and were included in the quantitative synthesis, enrolling 6,682 patients across five solid tumor types. These studies were published between 2019 and 2025, encompassing a total of 6,682 patients, with sample sizes ranging from 36 to 899 per study. Ten studies (41.7%) were conducted in China, 8 (33.3%) were global multicenter trials, 3 (12.5%) were conducted in Japan, and one each was conducted in Canada, Germany, and as a multinational collaboration. The main characteristics of the 24 included RCTs are presented in Table 1.

Figure 1 PRISMA diagram of search results and selections. CENTRAL, Cochrane Central Register of Controlled Trials; ICI, immune checkpoint inhibitor; RCT, randomized controlled trial.

Risk of bias in studies

The risk of bias assessment for all 24 included RCTs was performed using the Cochrane Collaboration’s Risk of Bias tool, and the results are presented in Figure 2. Overall, the methodological quality of the included studies was considered satisfactory. All studies demonstrated adequate randomization and allocation concealment. The main sources of potential bias were related to performance bias in open-label trials and unclear risks in selective reporting and other biases. These findings suggest that the pooled estimates derived from these studies are generally reliable, although the open-label design of some trials warrants cautious interpretation of subjective endpoints such as ORR.

Figure 2 Risk of bias assessment (A) in individual studies and (B) across the included studies. +, low risk of bias; −, high risk of bias; ?, unclear risk of bias.

Overall efficacy of nab-paclitaxel combined with ICIs

Primary outcomes

OS

A total of 17 studies involving 5,436 patients reported OS data. The pooled analysis under both the common-effect model (HR =0.79; 95% CI: 0.73–0.84) and the random-effects model (HR =0.79; 95% CI: 0.73–0.84) demonstrated a statistically significant 21% reduction in the risk of death with nab-paclitaxel plus ICI therapy compared with control regimens (Figure 3). Heterogeneity across studies was negligible (I²=0.0%; τ²<0.0001; P=0.57), and the prediction interval (0.72–0.85) confirmed that the survival benefit would be expected to persist in future similar study settings.

Figure 3 Forest plot of OS. HR <1 favors nab-paclitaxel plus ICI. CI, confidence interval; HR, hazard ratio; ICI, immune checkpoint inhibitor; OS, overall survival.

PFS

The same 17 studies also reported PFS outcomes. The common-effect model yielded a pooled HR of 0.63 (95% CI: 0.59–0.68), and the random-effects model showed a consistent estimate (HR =0.63; 95% CI: 0.58–0.69), indicating a significant 37% reduction in the risk of disease progression or death (Figure 4). Low-to-moderate heterogeneity was observed (I²=23.4%; τ²=0.0050; P=0.18), and the prediction interval (0.53–0.75) supported the generalizability of the PFS benefit across diverse clinical settings.

Figure 4 Forest plot of PFS. HR <1 favors nab-paclitaxel plus ICI. CI, confidence interval; HR, hazard ratio; ICI, immune checkpoint inhibitor; PFS, progression-free survival.

pCR

Nine studies including 1,791 patients (894 in the intervention group and 897 in the control group) reported pCR data. The common-effect model yielded a pooled RR of 1.35 (95% CI: 1.21–1.50), and the random-effects model showed an RR of 1.28 (95% CI: 1.15–1.43), both indicating a statistically significant improvement in pCR rates with nab-paclitaxel combined with ICIs (Figure 5). Moderate heterogeneity was detected (I²=57.1%; τ²<0.0001; P=0.02), with a prediction interval (1.13–1.46) that nonetheless excluded 1.0, indicating the treatment advantage is likely to persist in future neoadjuvant settings.

Figure 5 Forest plot of pCR. RR >1 favors nab-paclitaxel plus ICI. CG, control group; CI, confidence interval; ICI, immune checkpoint inhibitor; IG, intervention group; pCR, pathological complete response; RR, risk ratio.

Secondary outcomes

ORR

Seventeen studies involving 5,436 patients reported ORR data. The pooled RR under the common-effect model was 1.30 (95% CI: 1.23–1.37), and the random-effects model yielded an RR of 1.33 (95% CI: 1.20–1.47), both demonstrating a statistically significant improvement in tumor response (Figure S1). Substantial heterogeneity was present (I²=68.5%; τ²=0.0218; P<0.001), and the prediction interval (0.95–1.85) encompassed 1.0, suggesting that the ORR benefit may not universally extend to all possible clinical contexts.

DCR

Eight studies encompassing 1,162 patients (663 in the intervention group and 499 in the control group) reported DCR data. The common-effect model generated a pooled RR of 1.13 (95% CI: 1.07–1.20), and the random-effects model yielded an RR of 1.15 (95% CI: 1.02–1.29), demonstrating a statistically significant advantage in disease control with the combination regimen (Figure S2). Moderate to substantial heterogeneity was observed (I²=68.2%; τ²=0.0193; P=0.003), and the prediction interval (0.80–1.64) encompassed 1.0, reflecting variability in DCR benefit across different study populations.

Subgroup analyses

Subgroup analysis by cancer type

The composition of cancer-type subgroups varied across outcomes according to data availability in the original trials, with the number of contributing studies for each subgroup indicated within the corresponding forest plots.

For OS (Figure S3), subgroup analysis by cancer type revealed consistent survival benefits across all tumor subtypes. In NSCLC (10 studies), the pooled HR was 0.78 (95% CI: 0.71–0.86); in TNBC (5 studies), the HR was 0.73 (95% CI: 0.60–0.89); in PDAC (3 studies), the HR was 0.84 (95% CI: 0.67–1.06); and in gastric cancer (1 study), the HR was 0.62 (95% CI: 0.35–1.09). No significant subgroup differences were detected under either the common-effect model [χ²=1.00, degrees of freedom (df) =3, P=0.80] or the random-effects model (χ²=1.44, df=3, P=0.70), indicating that the OS benefit was consistent across cancer types.

For PFS (Figure S4), the most pronounced benefit was observed in NSCLC (HR =0.62; 95% CI: 0.57–0.68), followed by TNBC (HR =0.61; 95% CI: 0.53–0.71) and gastric cancer (HR = 0.79, 95% CI: 0.60–1.05). In PDAC, the trend favored treatment but did not reach statistical significance (HR =0.81; 95% CI: 0.65–1.01). A borderline difference across cancer type subgroups was observed under the common-effect model (χ²=7.12, df=3, P=0.07), though this was not significant under the random-effects model (χ²=5.11, df=3, P=0.16).

For pCR (Figure S5), subgroup analysis was conducted among TNBC (6 studies), ESCC (2 studies), and gastric cancer (1 study). In TNBC, the pooled RR was 1.25 (95% CI: 1.12–1.40), with no observed heterogeneity (I²=0%, P=0.70). The ESCC subgroup showed a strong trend toward benefit (RR =2.04; 95% CI: 1.41–2.97 under the common-effect model), though the random-effects estimate was attenuated by substantial inter-study variability (I²=85.2%). The test for subgroup differences was significant under the common-effect model (χ²=7.45, df=2, P=0.02) but not under the random-effects model (χ²=2.16, df=2, P=0.34).

For ORR (Figure S6), the greatest improvement was seen in gastric cancer (RR =1.69; 95% CI: 1.18–2.42) and PDAC (RR =1.42; 95% CI: 1.04–1.94). NSCLC showed a significant benefit (RR =1.43; 95% CI: 1.31–1.56), as did TNBC (RR =1.15; 95% CI: 1.07–1.23). The test for subgroup differences was statistically significant under both models (common-effect: χ²=18.67, df=3, P<0.001; random-effects: χ²=8.47, df=3, P=0.04), confirming that the magnitude of ORR benefit varied by tumor type.

For DCR (Figure S7), the pooled RR was 1.18 (95% CI: 1.04–1.33) for gastric cancer, 1.14 (95% CI: 1.01–1.29) for PDAC, 1.10 (95% CI: 1.00–1.22) for NSCLC, and 1.30 (95% CI: 0.73–2.30) for TNBC. No significant subgroup differences were identified (random-effects: χ²=0.73, df=3, P=0.87).

Subgroup analysis by ICI type

For the anti-PD-L1 plus anti-CTLA-4 and anti-PD-L1 plus anti-CD73 combinations, only one RCT contributed to each subgroup; the corresponding estimates therefore reflect single-trial results and should be interpreted as descriptive rather than as pooled meta-analytic estimates.

For OS (Figure S8), anti-PD-1 agents (HR =0.72; 95% CI: 0.64–0.80) and anti-PD-L1 agents (HR =0.84; 95% CI: 0.76–0.93) were both associated with improved survival. The anti-PD-L1 plus anti-CTLA-4 combination subgroup showed a non-significant trend (HR =0.84; 95% CI: 0.67–1.06). No statistically significant subgroup differences were observed (χ²=5.29, df=3, P=0.15). For PFS (Figure S9), anti-PD-1-based regimens achieved a pooled HR of 0.59 (95% CI: 0.53–0.65), whereas anti-PD-L1-based regimens yielded an HR of 0.65 (95% CI: 0.59–0.73). The test for subgroup differences was statistically significant under both models (χ²=9.90, df=3, P=0.02), suggesting a potential differential PFS benefit favoring anti-PD-1 agents across ICI categories.

For ORR (Figure S10), anti-PD-1 agents yielded an RR of 1.31 (95% CI: 1.22–1.39), and anti-PD-L1 agents achieved an RR of 1.29 (95% CI: 1.18–1.41). No statistically significant subgroup differences were identified (random-effects: χ²=0.75, df=3, P=0.861). For DCR (Figure S11), anti-PD-1 agents yielded an RR of 1.11 (95% CI: 1.05–1.18) under the common-effect model. No significant subgroup differences were observed (χ²=0.19, df=2, P=0.91).

It should be noted that the anti-PD-L1 plus anti-CD73 combination subgroup, present in Figures S8-S11, was based on a single trial in metastatic PDAC and is therefore not interpreted further

Sensitivity analyses

Leave-one-out sensitivity analyses confirmed the robustness of all primary and secondary outcomes (Figures S12-S16). For OS, the pooled HR remained between 0.76 and 0.80 with I²=0.0% across all iterations. For PFS, the pooled HR ranged from 0.62 to 0.65 (I²=0.0–28.2%), with heterogeneity lowest upon omitting Renouf et al. [2022]. For pCR, the pooled RR ranged from 1.25 to 1.42 (all P≤0.02); omitting Loibl et al. [2019] eliminated all heterogeneity (I²=0%), indicating this study as the primary source of between-study variability. For ORR, the pooled RR ranged from 1.30 to 1.37 (I²=53.5–74.9%), with the greatest reduction in heterogeneity upon excluding Chen et al. [2025]. For DCR, the pooled RR ranged from 1.09 to 1.19; omitting Fan et al. [2024] or Sun et al. [2025] yielded borderline non-significant results, suggesting these two small-sample camrelizumab studies contributed meaningfully to the overall effect. Overall, no individual study disproportionately altered the direction or magnitude of any pooled estimate.

Publication bias assessment

Publication bias was assessed using funnel plots, Egger’s regression test, and Begg’s rank correlation test (Figure 6; Figure S17). For OS, the funnel plot was approximately symmetric; Egger’s test was non-significant (intercept =−1.178, P=0.13), whereas Begg’s test was significant (τ =−0.478, P=0.008), likely reflecting the influence of small studies with extreme effects (e.g., Loibl et al., HR =0.24) rather than true publication bias. For PFS, Egger’s test indicated significant asymmetry (intercept =−0.684, P=0.001) despite a non-significant Begg’s test (τ=−0.193, P=0.28), suggesting potential small-study effects, although the low overall heterogeneity (I²=23.4%) mitigates this concern. For pCR, neither test suggested significant bias (Egger’s P=0.62; Begg’s P=0.36), though statistical power was limited (n=9). For ORR, Egger’s test was non-significant (P=0.59) whereas Begg’s test was significant (τ=0.437; P=0.02), likely attributable to genuine heterogeneity across cancer types. For DCR, both tests indicated significant asymmetry (Egger’s P=0.043; Begg’s P=0.002); however, with only 8 studies, the observed asymmetry may reflect clinical heterogeneity or small-study effects rather than selective publication.

Figure 6 Funnel plot for (A) OS, (B) pCR, and (C) PFS. OS, overall survival; pCR, pathological complete response; PFS, progression-free survival.

Discussion

This systematic review and meta-analysis comprehensively synthesized data from 24 RCTs enrolling 6,682 patients to evaluate the efficacy of nab-paclitaxel combined with ICIs across five solid tumor types. Our findings demonstrate that this combination regimen confers statistically significant and clinically meaningful improvements across all assessed efficacy endpoints. Specifically, nab-paclitaxel plus ICI therapy substantially reduced the risks of death and disease progression compared with nab-paclitaxel-based regimens without immunotherapy, and in the neoadjuvant setting it markedly increased the likelihood of achieving pCR. These benefits were consistently observed across cancer types and ICI classes, supporting the broad clinical applicability of this therapeutic strategy. The negligible between-study heterogeneity for OS and a prediction interval excluding the null further indicate that the survival benefit is robust and likely generalizable to similar future trial populations.

The consistent efficacy observed across diverse tumor types can be attributed to the complementary mechanisms of action between nab-paclitaxel and ICIs. Unlike conventional solvent-based paclitaxel, nab-paclitaxel exploits the albumin-SPARC interaction to achieve preferential intratumoral accumulation, resulting in higher local drug concentrations (8). This enhanced tumor exposure not only amplifies direct cytotoxic effects but also augments the immunogenic consequences of tumor cell death. Preclinical studies have demonstrated that paclitaxel induces ICD through the TLR4-mediated pathway, leading to calreticulin surface exposure, extracellular adenosine triphosphate (ATP) release, and high mobility group box 1 protein (HMGB1) secretion—key damage-associated molecular patterns (DAMPs) that activate dendritic cells and promote antigen cross-presentation to CD8⁺ T cells (41,42). Crucially, nab-paclitaxel preserves these immune-activating effects more effectively than solvent-based formulations, as the albumin nanoparticle carrier is internalized by tumor-associated macrophages via macropinocytosis, promoting their polarization toward the immunostimulatory M1 phenotype in a Toll-like receptor 4 (TLR4)-dependent manner (43). A recent study further revealed that nab-paclitaxel, unlike conventional paclitaxel, does not expand immunosuppressive triggering receptor expressed on myeloid cells 2 (TREM2)⁺ macrophages in the TME, which may contribute to its superior clinical efficacy when combined with PD-L1 blockade (44).

Furthermore, paclitaxel-based chemotherapy has been shown to upregulate PD-L1 expression on tumor cells (42), which, while potentially facilitating immune evasion, simultaneously creates a therapeutic vulnerability to PD-1/PD-L1 blockade. This bidirectional mechanism—whereby nab-paclitaxel both primes antitumor immunity and sensitizes tumors to ICI-mediated checkpoint blockade—provides a compelling mechanistic foundation for the synergistic efficacy observed in our meta-analysis. The elimination of corticosteroid premedication in nab-paclitaxel regimens further avoids the immunosuppressive effects that might otherwise dampen ICI activity (3,7).

Our findings are broadly consistent with, yet substantially extend, the conclusions of prior meta-analyses. Sharmni et al. [2022] (18) conducted a meta-analysis restricted to atezolizumab plus nab-paclitaxel in TNBC, pooling data from a limited number of trials and concluding that the combination improved PFS and ORR but not OS in the intention-to-treat population. Our analysis confirmed the PFS benefit (HR =0.61) and additionally demonstrated a significant OS improvement (HR =0.73) in the TNBC subgroup—likely reflecting the inclusion of newer anti-PD-1-based studies that showed clearer survival gains. Hao et al. [2023] (19) performed a network meta-analysis focused on immune-related adverse events associated with nab-paclitaxel/paclitaxel plus ICI combinations but did not assess efficacy endpoints or distinguish between taxane formulations, limiting the clinical utility of their findings for guiding treatment selection. Quan et al. [2023] (45) evaluated neoadjuvant immunochemotherapy across solid tumors but included only six RCTs and did not specifically focus on nab-paclitaxel-based regimens, resulting in insufficient power to detect treatment effects in specific tumor subgroups.

Importantly, our study is the first meta-analysis to incorporate pCR as a primary endpoint across multiple tumor types in the nab-paclitaxel plus ICI context, pooling data from nine neoadjuvant RCTs encompassing TNBC, ESCC, and gastric cancer. The significant improvement in pCR (RR =1.28) and the prediction interval excluding 1.0 (1.13–1.46) provide high-quality evidence supporting the integration of ICI therapy into nab-paclitaxel-based neoadjuvant regimens. This finding is particularly timely and clinically relevant given the growing acceptance of pCR as a regulatory endpoint and an early surrogate for long-term survival in neoadjuvant settings, particularly in TNBC and NSCLC (11,12).

The subgroup analyses by cancer type yielded several clinically informative observations. The consistent OS benefit across NSCLC, TNBC, PDAC, and gastric cancer (P_interaction=0.70) suggests that the immunomodulatory properties of nab-paclitaxel synergize with ICI-mediated immune activation regardless of the tumor-specific immune microenvironment. This finding is particularly noteworthy for PDAC, a malignancy traditionally regarded as resistant to immunotherapy owing to its densely fibrotic, immunosuppressive TME (46). Although the PDAC subgroup did not achieve statistical significance individually for OS (HR =0.84; 95% CI: 0.67–1.06), the direction and magnitude of effect were consistent with other tumor types, suggesting that nab-paclitaxel may partially overcome the immunosuppressive barriers in pancreatic cancer through macrophage reprogramming and enhanced antigen release (43).

The statistically significant subgroup difference favoring anti-PD-1 over anti-PD-L1 agents for PFS (P_interaction=0.02) is an intriguing finding that warrants careful interpretation. Several factors may contribute to this differential. First, anti-PD-1 antibodies block the interaction of PD-1 with both PD-L1 and programmed death-ligand 2 (PD-L2), whereas anti-PD-L1 agents only disrupt the PD-1/PD-L1 axis, leaving the PD-1/PD-L2 pathway intact (1). Second, the anti-PD-1 subgroup in our analysis included a higher proportion of newer-generation agents (e.g., camrelizumab, tislelizumab, toripalimab, retifanlimab) evaluated in more recent trials with potentially refined patient selection and optimized combination regimens. Third, cross-trial confounding—including differences in concurrent chemotherapy partners, patient populations, and PD-L1 testing methodologies—precludes a definitive conclusion regarding the intrinsic superiority of one ICI class over another. Nevertheless, this observation supports ongoing clinical interest in comparing PD-1 versus PD-L1 inhibitors within standardized combination frameworks.

The heterogeneity in pCR across TNBC, ESCC, and gastric cancer subgroups, which differed in significance between the common-effect and random-effects models, suggests that tumor-specific biology may modulate the depth of pathological response to nab-paclitaxel plus ICI combinations. The ESCC subgroup showed particularly large pCR improvements, consistent with the known immunogenicity and high PD-L1 expression characteristic of squamous cell histologies (15,17). The substantial between-study variability in this subgroup reflects differences in ICI agents, combination partners, and patient selection criteria across the two included ESCC trials, highlighting the need for additional phase III data to confirm the magnitude of pCR benefit in this setting.

Several subgroups in our analyses were represented by only a single eligible RCT, notably gastric cancer in the OS and pCR analyses and the anti-PD-L1 plus anti-CD73 combination across all endpoints. The corresponding estimates are descriptive rather than meta-analytic and warrant cautious interpretation. Nevertheless, these preliminary observations carry hypothesis-generating value: the activity of CD73-pathway inhibition combined with PD-L1 blockade in metastatic PDAC, and of nab-paclitaxel plus ICI in gastric cancer, point to specific therapeutic combinations and tumor contexts that warrant dedicated randomized evaluation. Such considerations are characteristic of an evolving therapeutic landscape, in which early-phase signals and limited subgroup analyses inform the design and prioritization of subsequent confirmatory trials.

The findings of this meta-analysis have several implications for clinical practice and future research. First, the robust and consistent survival benefits observed across multiple tumor types support the continued use of nab-paclitaxel as a preferred chemotherapy backbone for ICI-based combination regimens, particularly in settings where corticosteroid avoidance is desirable. The absence of corticosteroid premedication is not merely a pharmacological convenience but may represent a clinically meaningful advantage, given accumulating evidence that baseline steroid use is associated with inferior outcomes in patients receiving ICI therapy (7). Second, the significant improvement in pCR in the neoadjuvant context suggests that nab-paclitaxel plus ICI should be considered a viable and promising option for patients with resectable TNBC, ESCC, and potentially other solid tumors amenable to neoadjuvant approaches. The prediction interval for pCR (1.13–1.46) entirely excluding 1.0 provides reassurance that the benefit is likely to be observed in future similar trial populations, strengthening the evidence base for the adoption of this strategy in clinical guidelines and institutional treatment protocols. Third, the signal favoring anti-PD-1 over anti-PD-L1 agents for PFS may inform the selection of ICI partners in future combination trials, although this finding requires validation in head-to-head comparisons.

From a health management and policy perspective, these findings also carry implications for treatment pathway optimization and resource allocation. The demonstration of consistent benefits across five tumor types supports the development of unified treatment frameworks for nab-paclitaxel plus ICI combinations rather than entirely tumor-specific protocols, potentially streamlining clinical decision-making and guideline development. The significant pCR improvements in the neoadjuvant setting may also translate into downstream benefits, including the potential for organ preservation, de-escalated adjuvant therapy, and reduced long-term healthcare utilization among patients who achieve complete pathological responses.

Several areas warrant further investigation. Future RCTs should explore biomarker-driven patient selection strategies—including PD-L1 expression, tumor mutational burden, and circulating tumor DNA dynamics (47)—to identify patients most likely to achieve pCR and long-term benefit from nab-paclitaxel plus ICI combinations. Head-to-head comparisons between nab-paclitaxel and conventional paclitaxel within the same ICI combination framework would help clarify whether the immunomodulatory advantages of the albumin-bound formulation translate into clinically meaningful differences in efficacy. This question is particularly relevant given the cost differential between the two formulations and the need for pharmacoeconomic justification in resource-constrained healthcare systems. Additionally, the role of nab-paclitaxel plus ICI in traditionally immunoresistant tumors such as PDAC deserves continued exploration, particularly in combination with novel immunomodulatory approaches targeting the myeloid compartment. Long-term safety data, including the incidence and management of late-onset immune-related adverse events in the neoadjuvant and adjuvant settings, also remain important knowledge gaps to address. The integration of patient-reported outcome measures and health-related quality-of-life assessments into future trials will be essential for capturing the full clinical value of these combination regimens beyond traditional survival endpoints.

This study has several limitations that should be acknowledged. First, despite the inclusion of 24 RCTs, some tumor-type and ICI-type subgroups contained relatively few studies—notably, gastric cancer in the OS analysis and the anti-PD-L1 plus anti-CD73 combination across all endpoints—limiting the statistical power for subgroup-specific conclusions. Second, the included studies employed heterogeneous ICI agents, dosing schedules, concurrent chemotherapy combinations, and lines of therapy, which may contribute to the observed heterogeneity for endpoints such as ORR and DCR. Third, individual patient data were not available, precluding analyses stratified by important clinical variables such as PD-L1 expression status, age, performance status, or prior treatment history. Fourth, although most included trials were phase III RCTs, several were phase II studies with smaller sample sizes and less stringent blinding protocols, which may introduce performance and detection bias for subjective endpoints. Fifth, the pooled pCR analysis combined data from trials enrolling patients with TNBC, ESCC, and gastric cancer—tumor types with distinct biological characteristics—and caution is warranted in extrapolating the overall pCR estimate to any individual tumor type. Sixth, publication bias assessments yielded mixed signals across endpoints, and while no consistent evidence of selective publication was identified, the possibility cannot be entirely excluded, especially for DCR where the number of available studies was limited. Finally, our analysis focused exclusively on efficacy outcomes and did not incorporate safety or quality-of-life data, which are important considerations in the risk-benefit assessment of combination immunotherapy. In addition to this, several ICI combinations, including anti-PD-L1 plus anti-CTLA-4 and anti-PD-L1 plus anti-CD73, were each represented by only a single eligible RCT. The corresponding subgroup estimates therefore reflect the underlying trial results rather than meta-analytic pooled effects and should be interpreted with caution; confirmation in additional RCTs is needed before robust conclusions can be drawn for these specific combinations.

Conclusions

This comprehensive meta-analysis of 24 RCTs enrolling 6,682 patients demonstrates that nab-paclitaxel combined with ICIs provides statistically significant and clinically meaningful improvements in OS, PFS, pCR, ORR, and DCR across multiple solid tumors. The survival benefit is consistent across cancer types—including NSCLC, TNBC, PDAC, gastric cancer, and ESCC—and ICI classes, with negligible between-study heterogeneity for OS. In the neoadjuvant setting, the combination significantly enhances pCR rates, supporting its integration into perioperative treatment strategies. Anti-PD-1-based regimens appear to confer a greater PFS advantage compared with anti-PD-L1-based regimens, although this finding requires further validation. These results provide the most up-to-date and comprehensive evidence supporting the broad clinical applicability of nab-paclitaxel plus ICI combination therapy in solid tumors.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the PRISMA reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0627/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0627/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0627/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chen H, Yang H, Guo L, et al. The Role of Immune Checkpoint Inhibitors in Cancer Therapy: Mechanism and Therapeutic Advances. MedComm (2020) 2025;6:e70412.
2. Ramos-Casals M, Brahmer JR, Callahan MK, et al. Immune-related adverse events of checkpoint inhibitors. Nat Rev Dis Primers 2020;6:38. [Crossref] [PubMed]
3. Soliman HH. nab-Paclitaxel as a potential partner with checkpoint inhibitors in solid tumors. Onco Targets Ther 2017;10:101-12. [Crossref] [PubMed]
4. Emens LA, Middleton G. The interplay of immunotherapy and chemotherapy: harnessing potential synergies. Cancer Immunol Res 2015;3:436-43. [Crossref] [PubMed]
5. Bracci L, Schiavoni G, Sistigu A, et al. Immune-based mechanisms of cytotoxic chemotherapy: implications for the design of novel and rationale-based combined treatments against cancer. Cell Death Differ 2014;21:15-25. [Crossref] [PubMed]
6. Arimoto KI, Miyauchi S, Liu M, et al. Emerging role of immunogenic cell death in cancer immunotherapy. Front Immunol 2024;15:1390263. [Crossref] [PubMed]
7. Verheijden RJ, de Groot JS, Fabriek BO, et al. Corticosteroids for Immune-Related Adverse Events and Checkpoint Inhibitor Efficacy: Analysis of Six Clinical Trials. J Clin Oncol 2024;42:3713-24. [Crossref] [PubMed]
8. Desai N, Trieu V, Yao Z, et al. Increased antitumor activity, intratumor paclitaxel concentrations, and endothelial cell transport of cremophor-free, albumin-bound paclitaxel, ABI-007, compared with cremophor-based paclitaxel. Clin Cancer Res 2006;12:1317-24. [Crossref] [PubMed]
9. Vijayan D, Young A, Teng MWL, et al. Targeting immunosuppressive adenosine in cancer. Nat Rev Cancer 2017;17:709-24. [Crossref] [PubMed]
10. Forde PM, Chaft JE, Smith KN. etal. Neoadjuvant PD-1 Blockade in Resectable Lung Cancer. N Engl J Med 2018;378:1976-86. [Crossref] [PubMed]
11. Sugiyama K, Gordon A, Popat S, et al. Is pathological response an adequate surrogate marker for survival in neoadjuvant therapy with immune checkpoint inhibitors? ESMO Open 2025;10:104122. [Crossref] [PubMed]
12. Feng Y, Guo K, Jin H, et al. Efficacy and safety of neoadjuvant combination immunotherapy in surgically resectable malignant solid tumors: a systematic review and meta-analysis. Expert Rev Anticancer Ther 2024;24:169-81. [Crossref] [PubMed]
13. Loibl S, Untch M, Burchardi N, et al. A randomised phase II study investigating durvalumab in addition to an anthracycline taxane-based neoadjuvant therapy in early triple-negative breast cancer: clinical results and biomarker analysis of GeparNuevo study. Ann Oncol 2019;30:1279-88. [Crossref] [PubMed]
14. Mittendorf EA, Zhang H, Barrios CH, et al. Neoadjuvant atezolizumab in combination with sequential nab-paclitaxel and anthracycline-based chemotherapy versus placebo and chemotherapy in patients with early-stage triple-negative breast cancer (IMpassion031): a randomised, double-blind, phase 3 trial. Lancet 2020;396:1090-100. [Crossref] [PubMed]
15. Li Y, Zhou A, Liu S, et al. Comparing a PD-L1 inhibitor plus chemotherapy to chemotherapy alone in neoadjuvant therapy for locally advanced ESCC: a randomized Phase II clinical trial : A randomized clinical trial of neoadjuvant therapy for ESCC. BMC Med 2023;21:86. [Crossref] [PubMed]
16. Lin JX, Tang YH, Zheng HL, et al. Neoadjuvant camrelizumab and apatinib combined with chemotherapy versus chemotherapy alone for locally advanced gastric cancer: a multicenter randomized phase 2 trial. Nat Commun 2024;15:41. [Crossref] [PubMed]
17. Qin J, Xue L, Hao A, et al. Neoadjuvant chemotherapy with or without camrelizumab in resectable esophageal squamous cell carcinoma: the randomized phase 3 ESCORT-NEO/NCCES01 trial. Nat Med 2024;30:2549-57. [Crossref] [PubMed]
18. Sharmni Vishnu K, Win TT, Aye SN, et al. Combined atezolizumab and nab-paclitaxel in the treatment of triple negative breast cancer: a meta-analysis on their efficacy and safety. BMC Cancer 2022;22:1139. [Crossref] [PubMed]
19. Hao W, Zhang J, Wang Y, et al. Immune-related adverse events associated with nab-paclitaxel/paclitaxel combined with immune checkpoint inhibitors: a systematic review and network meta-analysis. Front Immunol 2023;14:1175809. [Crossref] [PubMed]
20. Higgins JP, Altman DG, Gøtzsche PC, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. [Crossref] [PubMed]
21. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60. [Crossref] [PubMed]
22. Jia R, Si HY, Fan MJ, et al. First-line treatment with chemotherapy, surufatinib (an angio-immuno kinase inhibitor), and camrelizumab (an anti-PD-1 antibody) for locally advanced or metastatic pancreatic ductal adenocarcinoma: a phase Ib/II randomized study. Signal Transduct Target Ther 2025;10:339. [Crossref] [PubMed]
23. Renouf DJ, Loree JM, Knox JJ, et al. The CCTG PA.7 phase II trial of gemcitabine and nab-paclitaxel with or without durvalumab and tremelimumab as initial therapy in metastatic pancreatic ductal adenocarcinoma. Nat Commun 2022;13:5020. [Crossref] [PubMed]
24. Wang P, Chen Y, Wang F, et al. Camrelizumab plus chemotherapy versus chemoradiotherapy as neoadjuvant therapy for resectable esophageal squamous cell carcinoma: Phase 2 randomized trial (REVO). Nat Commun 2025;16:9676. [Crossref] [PubMed]
25. Chen L, Li H, Zhang H, et al. Camrelizumab vs Placebo in Combination With Chemotherapy as Neoadjuvant Treatment in Patients With Early or Locally Advanced Triple-Negative Breast Cancer: The CamRelief Randomized Clinical Trial. JAMA 2025;333:673-81. [Crossref] [PubMed]
26. Sugawara S, Tanaka K, Imamura F, et al. Pembrolizumab plus chemotherapy in Japanese patients with metastatic squamous non-small-cell lung cancer in KEYNOTE-407. Cancer Sci 2023;114:3330-41. [Crossref] [PubMed]
27. Coveler AL, Reilley MJ, Zalupski M, et al. A Phase Ib/II Randomized Clinical Trial of Oleclumab with or without Durvalumab plus Chemotherapy in Patients with Metastatic Pancreatic Ductal Adenocarcinoma. Clin Cancer Res 2024;30:4609-17. [Crossref] [PubMed]
28. Saji S, Ohsumi S, Ito M, et al. Subgroup analysis of Japanese patients in a phase III randomized, controlled study of neoadjuvant atezolizumab or placebo, combined with nab-paclitaxel and anthracycline-based chemotherapy in early triple-negative breast cancer (IMpassion031). Jpn J Clin Oncol 2022;52:1124-33. [Crossref] [PubMed]
29. Iwata H, Inoue K, Kaneko K, et al. Subgroup analysis of Japanese patients in a Phase 3 study of atezolizumab in advanced triple-negative breast cancer (IMpassion130). Jpn J Clin Oncol 2019;49:1083-91. [Crossref] [PubMed]
30. Emens LA, Adams S, Barrios CH, et al. First-line atezolizumab plus nab-paclitaxel for unresectable, locally advanced, or metastatic triple-negative breast cancer: IMpassion130 final overall survival analysis. Ann Oncol 2021;32:983-93. [Crossref] [PubMed]
31. Jotte R, Cappuzzo F, Vynnychenko I, et al. Atezolizumab in Combination With Carboplatin and Nab-Paclitaxel in Advanced Squamous NSCLC (IMpower131): Results From a Randomized Phase III Trial. J Thorac Oncol 2020;15:1351-60. [Crossref] [PubMed]
32. Paz-Ares L, Vicente D, Tafreshi A, et al. A Randomized, Placebo-Controlled Trial of Pembrolizumab Plus Chemotherapy in Patients With Metastatic Squamous NSCLC: Protocol-Specified Final Analysis of KEYNOTE-407. J Thorac Oncol 2020;15:1657-69. [Crossref] [PubMed]
33. Wang J, Lu S, Yu X, et al. Tislelizumab plus chemotherapy versus chemotherapy alone as first-line treatment for advanced squamous non-small-cell lung cancer: final analysis of the randomized, phase III RATIONALE-307 trial. ESMO Open 2024;9:103727. [Crossref] [PubMed]
34. Cortes J, Rugo HS, Cescon DW, et al. Pembrolizumab plus Chemotherapy in Advanced Triple-Negative Breast Cancer. N Engl J Med 2022;387:217-26. [Crossref] [PubMed]
35. Fan L, Wang ZH, Ma LX, et al. Optimising first-line subtyping-based therapy in triple-negative breast cancer (FUTURE-SUPER): a multi-cohort, randomised, phase 2 trial. Lancet Oncol 2024;25:184-97. [Crossref] [PubMed]
36. Sun L, Gong Z, Wang L, et al. Multi-center phase II study of nab-paclitaxel plus camrelizumab versus nab-paclitaxel alone as second-line treatment for advanced gastric cancer. Oncologist 2025;30:oyaf189. [Crossref] [PubMed]
37. Gianni L, Huang CS, Egle D, et al. Pathologic complete response (pCR) to neoadjuvant treatment with or without atezolizumab in triple-negative, early high-risk and locally advanced breast cancer: NeoTRIP Michelangelo randomized study. Ann Oncol 2022;33:534-43. [Crossref] [PubMed]
38. Lu S, Vynnychenko O, Kulyaba Y, et al. Retifanlimab versus placebo in combination with platinum-based chemotherapy in patients with first-line non-squamous or squamous metastatic non-small-cell lung cancer (POD1UM-304): a phase 3, multiregional, placebo-controlled, double-blind, randomised study. Lancet Respir Med 2025;13:1096-107. [Crossref] [PubMed]
39. West H, McCleod M, Hussein M, et al. Atezolizumab in combination with carboplatin plus nab-paclitaxel chemotherapy compared with chemotherapy alone as first-line treatment for metastatic non-squamous non-small-cell lung cancer (IMpower130): a multicentre, randomised, open-label, phase 3 trial. Lancet Oncol 2019;20:924-37. [Crossref] [PubMed]
40. Jiang Z, Ouyang Q, Sun T, et al. Toripalimab plus nab-paclitaxel in metastatic or recurrent triple-negative breast cancer: a randomized phase 3 trial. Nat Med 2024;30:249-56. [Crossref] [PubMed]
41. Lau TS, Chan LKY, Man GCW, et al. Paclitaxel Induces Immunogenic Cell Death in Ovarian Cancer via TLR4/IKK2/SNARE-Dependent Exocytosis. Cancer Immunol Res 2020;8:1099-111. [Crossref] [PubMed]
42. Yang Q, Shi G, Chen X, et al. Nanomicelle protects the immune activation effects of Paclitaxel and sensitizes tumors to anti-PD-1 Immunotherapy. Theranostics 2020;10:8382-99. [Crossref] [PubMed]
43. Cullis J, Siolas D, Avanzi A, et al. Macropinocytosis of Nab-paclitaxel Drives Macrophage Activation in Pancreatic Cancer. Cancer Immunol Res 2017;5:182-90. [Crossref] [PubMed]
44. Xing Y, Zhong R, Li Q, et al. Paclitaxel drives TREM2(+) macrophage expansion underlying its inferior therapeutic efficacy compared to Nab-paclitaxel. Nat Commun 2026;17:2272. [Crossref] [PubMed]
45. Quan R, Li Z, Zhang H, et al. Do patients receive any benefit from the addition of perioperative immunotherapy-chemotherapy for solid tumors? Medicine (Baltimore) 2023;102:e35603. [Crossref] [PubMed]
46. Beatty GL, Werba G, Lyssiotis CA, et al. The biological underpinnings of therapeutic resistance in pancreatic cancer. Genes Dev 2021;35:940-62. [Crossref] [PubMed]
47. Hui X, Tian X, Ding S, et al. Reprogramming the tumor microenvironment to overcome immunotherapy resistance in pancreatic cancer. Front Immunol 2025;16:1717062. [Crossref] [PubMed]