Racial and ethnic disparities in survival outcomes among patients with inflammatory breast cancer: a systematic review and meta-analysis Authors: Chalothorn Wannaphut, Ben Ponvilawan, Toshiaki Iwase, Yoko Takahashi, Xiaoping Wang, Naoto T Ueno Published: 2025 ABSTRACT AbstractInflammatory breast cancer (IBC) is a rare and aggressive subtype of breast cancer characterized by rapid progression, early metastasis, and poor prognosis. While racial and ethnic disparities in IBC outcomes have been reported, existing findings remain inconsistent. We conducted the first systematic review and meta-analysis to comprehensively evaluate survival disparities by race and ethnicity among patients with IBC. A systematic search of MEDLINE and EMBASE identified 15 eligible studies. Compared to non-Hispanic White (NHW) patients, non-Hispanic Black patients had higher mortality (HR: 1.45; 95% CI: 1.34–1.57), while Asian or Asian Pacific Islander (API) patients had lower mortality (HR: 0.78; 95% CI: 0.65–0.94). Hispanic patients did not show statistically different survival outcomes compared to NHW patients (HR: 1.06, 95% CI 0.88–1.28). These findings underscore persistent racial disparities in IBC outcomes, although inherent biases and confounding in observational studies warrant cautious interpretation.Subject terms: Cancer, Oncology Introduction Inflammatory breast cancer (IBC) is a rare but highly aggressive type of breast cancer, making up roughly 1–5% of breast cancer diagnoses in the United States1–3. Even though its occurrence is low, IBC has a disproportionately high mortality rate due to its swift progression, early metastatic capabilities, and being diagnosed at advanced stages4–6. Clinically, IBC is characterized by widespread redness, swelling (peau d’orange), and a rapid onset of symptoms, which can lead to delays in both diagnosis and treatment initiation2. The usual treatment approach for IBC generally includes a combination of therapies such as primary chemotherapy, mastectomy, and radiation; however, survival rates are still significantly lower than those for non-IBC breast cancer types7,8. In recent years, researchers have increasingly focused on examining disparities in cancer outcomes across racial and ethnic groups. Numerous studies have documented disparities in breast cancer incidence, treatment access, and survival outcomes among racial and ethnic minorities in the U.S.9,10. Non-Hispanic Black (NHB) women with inflammatory breast cancer (IBC) have been shown to experience significantly worse survival outcomes compared to their non-Hispanic White (NHW) counterparts, with a large SEER-based study reporting a 2-year IBC-specific survival rate of 53% in NHB women versus 69% in NHW women (P < 0.0001)11,12. The reasons for these disparities are multifactorial, encompassing differences in tumor biology, delayed diagnosis, limited access to standard care, and socioeconomic inequities13. In contrast, Asian patients have often demonstrated more favorable outcomes, while Pacific Islander patients—despite being frequently grouped with Asians—represent a distinct and underrepresented population with potentially different risk profiles and outcomes14,15. Recent studies examining racial disparities in IBC have produced mixed results regarding survival outcomes. Some research indicates that NHB patients with IBC experience considerably poorer survival outcomes9,10,16, while another study finds no significant survival differences once clinical and treatment factors are considered17. Additionally, treatment pattern disparities—such as delays in chemotherapy, lower surgical intervention rates, and limited access to clinical trials—may play a role in these discrepancies10,18. Furthermore, the data on outcomes for Hispanic and Asian Pacific Islander (API) patients with IBC is sparse and inconsistent, mainly due to small sample sizes and underreporting in national databases15,17,19. Considering the aggressive characteristics of IBC and the current evidence regarding racial disparities in breast cancer overall, it is crucial to systematically evaluate how race and ethnicity affect survival outcomes in IBC. Gaining insight into these disparities is crucial not only for patient care but also for public health and policy initiatives aimed at improving equity in cancer treatment and outcomes. This systematic review and meta-analysis seek to consolidate existing evidence regarding racial and ethnic disparities in outcomes for IBC survival. By assessing hazard ratios among various racial and ethnic groups—including NHB, Hispanic, and API populations in comparison to NHW patients—this study aims to elucidate the extent of these disparities and examine possible contributing factors. These findings are crucial for informing equitable clinical care strategies, improving early detection, and guiding future research in this rare yet severe disease. Results Study selection The initial electronic search yielded a total of 922 articles from all databases. After removing duplicates, 916 articles remained for title and abstract screening. Of these, 50 studies met the preliminary screening criteria and were subsequently reviewed in full. Based on the pre-specified eligibility criteria, 15 studies were ultimately included in the final systematic review and meta-analysis. However, one study was excluded from the final analysis due to potential data overlap from the same institution (19). Therefore, 14 studies were included in the final analysis in the meta-analysis. Thirty-five studies were excluded after full-text review for the following reasons: 18 did not report differences in survival outcomes by race or ethnicity, 15 did not report outcomes specifically for patients with IBC, and two did not provide a 95% confidence interval for survival data. The study selection process is summarized in the PRISMA flow diagram (Fig. 1). PRISMA diagram summary for the literature search. Study characteristics All 15 included studies were retrospective cohort studies published between 2012 and 2025. The majority of studies were conducted in the United States (n = 14), with one study involving populations from both the United States and Japan (n = 1). These studies evaluated survival outcomes in patients diagnosed with IBC and stratified results based on racial and ethnic groups. Baseline patient and study characteristics are summarized in Table 1. Table 1Baseline characteristics of all studies included in the systematic review and meta-analysisAuthorStudy designCountrySample size IBC (N)Racial distributionMean age (years)Stage at diagnosisMolecular SubtypesTreatment modalitiesMedian Follow up. Duration (mo)Total deaths N (%)Adjustment for covariateNewcastle-Ottawa scoreDawood et al.11Retrospective cohort(SEER 2004–2007)USA2384NHW 2018, NHB 35656(21–98)IIIB 1267IIIC 325III Nos 70IV 722ER+ 1052.PR+ 780M 1534SegM 144RT 117315(0–47)684(28.7)Age, race, marital status, grade, stage, Sx, LN Sx,RTSelection: 4 starscomparability: 2 starsoutcome: 3 starsSchlichting et al.35Retrospective cohort(SEER 2000–2008)USA4866NHW 3799, NHB 729, API 289, AIAN 4957NAHR+ 3065Sx 4645RT 3126NANAAge, race, HR status, grade, Sx, RTSelection: 4 starscomparability: 2 starsoutcome: 3 starsNatori et al.15Retrospective cohort(2003–2009)Japan and USA406NHW 290. NHB 40, H 38, Asian 3350.5NAER− 208PR− 262HER2+ 127CMT 295Sx 375RT 30733.6NAAge, ER status, grade, BMI, raceSelection: 4 starscomparability: 2 starsoutcome: 3 starsSchinkel et al.14Retrospective cohort(SEER 1990–2004)USA7748NHW 5332, NHB 1160, H 874, API 382NAIII 6038IV 1710ER and PR+ 2904.ER and PR– 3463Sx 6169RT 4139NANAAge, race, marital status, grade, stage, Sx, RT, ER/PR statusSelection: 4 starscomparability: 2 starsoutcome: 3 starsWecsler et al.36Retrospective cohort(SEER 2004–2008)USA761NHW 512, NHB 97, H 109, Asian 34, AI 6Unknown 3NANAER+ 343PR+ 257Sx 305Sx&RT 456NANAAge, raceSelection: 4 starscomparability: 2 starsoutcome: 3 starsDenu et al.21Retrospective cohort(POC-BP 2004–2012)USA170NHW 81,NHB 68, H 14, Other 757.7NANANANA121(71.2)Age, race, metastasis, ER/PR status, treatmentSelection: 4 starscomparability: 2 starsoutcome: 3 starsFouad et al.19Retrospective cohort(2006–2013)USA248NHW 193, NHB 23, H 2651.6IIIB 83IIIC 73IV 91HR+/HER2− 89HR−/HER2+ 54HR+/HER2+ 42TN 63NANANAage, race, number of pregnancies, menopause status, smoking status, age of menarche, breastfeeding, stage, HRT, molecularSelection: 3 starscomparability: 1 staroutcome: 3 starsMuzaffar et al.37Retrospective cohort(SEER 1988–2013)USA11,604NHW 5871, NHB 1003, Other 43056(21–103)NAER+ 3458PR+ 2671HER2+ 535TN 374PM 409TM 6895NANAAge, node status, year dx, Sx, race, RT ER/PR statusSelection: 4 starscomparability: 2 starsoutcome: 3 starsDiao et al.16Retrospective cohort(SEER 2004–2013)USA1651NHW 1290, NHB 27056(22–98)M1 498HR+ 798HER2+ 167PM 72Simple mastectomy 177RM 926CMT 231RT 82131(0–131)NAAge, race, marital status, grade, N stageM stage, tumor extension, HR statusHER2 status, Sx, CMTRTSelection: 3 starscomparability: 2 starsoutcome: 3 starsGudina et al.31Retrospective cohort(MCSP 1998–2014)USA1324NHW 984NHB 227,H 86,Other 2759Localized 776distant 444ER+ 577NA13.6(0.12–198)876(66.2)Age, race, tumor stagemarital status.ER statusSelection: 4 starscomparability: 2 starsoutcome: 3 starsLi et al.20Retrospective cohort(SEER 2010–2013)USA293NHW 233, NHB 37,Other 14NAIIIA 186IIIB 48IV 59HR+/HER2− 102HR−/HER2+ 49HR+/HER2+ 60TN 66PMRT27(5–59)NAAge, race, marital status, gradestage, molecular, Sx, CMT, PM, RTSelection: 4 starscomparability: 2 starsoutcome: 3 starsAdesoye et al.17Retrospective cohort(2010–2019)USA523NHW, 417NHB 38,H 48Asian 12NA 2Other 6NAIV 43HR+/HER2− 193HR−/HER2+ 106HR+/HER2 +80TN 144MRMNACTXRTNANAAge, race, molecular, node status, pCRSelection: 3 starscomparability: 2 starsoutcome: 3 starsOchoa et al.10Retrospective cohort(SEER 2010–2016)USA1652NHW 1652NHB 371NANAHR+/HER2− 708HR−/HER2+ 335HR+/HER2+ 357TN 492BCS 70M 156439NAAge, race, marital status, county-level SES, rurality, histology, grade, molecular, Sx, RT, CMTSelection: 4 starscomparability: 2 starsoutcome: 3 starsYoshikawa et al.38Retrospective cohort(2000–2018)USA60NHW 12Asian 20PI 26Other 259III 36IV 20HR+/HER2− 23HER2+ 22TN 6Unk 9NA63NAAge, race, insurance, histology, molecular, stageSelection: 4 starscomparability: 2 starsoutcome: 3 starsTadros et al.9Retrospective cohort(NCD 2010–2018)USA6945NHW 4808NHB 1153API 161H 522Other 82Unk 21957III 3988Unk 72HR+ 1212TN 1754Unk 404TM 3564RT 4395NANAAge, race, grade, node status, insurance, facility, molecular, GCCSelection: 4 starscomparability: 2 starsoutcome: 3 starsAI American Indian, API AsianPacific Islander, BCS Brease conservative surgery, CMT chemotherapy, ER estrogen receptor, H Hispanic, HER-2 human epidermal growth factor-2, HR hormonal receptor, LN lymph node, M mastectomy, NA Native American, NHB non-Hispanic black, NHW non-Hispanic white, PM partial mastectomy, PMRT postmastectomy radiation, PR progesterone receptor, RM Radical mastectomy, RT radiation, SegM Segmental mastectomy, SM Simple mastectomy, Sx Surgery, TM total mastectomy, TN triple negative, Unk unknown. Baseline characteristics of all studies included in the systematic review and meta-analysis Retrospective cohort (SEER 2004–2007) 56 (21–98) IIIB 1267 IIIC 325 III Nos 70 IV 722 ER+ 1052. PR+ 780 M 1534 SegM 144 RT 1173 15 (0–47) 684 (28.7) Age, race, marital status, grade, stage, Sx, LN Sx, RT Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 2000–2008) Sx 4645 RT 3126 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (2003–2009) ER− 208 PR− 262 HER2+ 127 CMT 295 Sx 375 RT 307 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 1990–2004) III 6038 IV 1710 ER and PR+ 2904. ER and PR– 3463 Sx 6169 RT 4139 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 2004–2008) NHW 512, NHB 97, H 109, Asian 34, AI 6 Unknown 3 ER+ 343 PR+ 257 Sx 305 Sx&RT 456 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (POC-BP 2004–2012) 121 (71.2) Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (2006–2013) IIIB 83 IIIC 73 IV 91 HR+/HER2− 89 HR−/HER2+ 54 HR+/HER2+ 42 TN 63 Selection: 3 stars comparability: 1 star outcome: 3 stars Retrospective cohort (SEER 1988–2013) 56 (21–103) ER+ 3458 PR+ 2671 HER2+ 535 TN 374 PM 409 TM 6895 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 2004–2013) 56 (22–98) HR+ 798 HER2+ 167 PM 72 Simple mastectomy 177 RM 926 CMT 231 RT 821 31 (0–131) Age, race, marital status, grade, N stage M stage, tumor extension, HR status HER2 status, Sx, CMT RT Selection: 3 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (MCSP 1998–2014) NHW 984 NHB 227, H 86, Other 27 Localized 776 distant 444 13.6 (0.12–198) 876 (66.2) Age, race, tumor stage marital status.ER status Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 2010–2013) NHW 233, NHB 37, Other 14 IIIA 186 IIIB 48 IV 59 HR+/HER2− 102 HR−/HER2+ 49 HR+/HER2+ 60 TN 66 27 (5–59) Age, race, marital status, grade stage, molecular, Sx, CMT, PM, RT Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (2010–2019) NHW, 417 NHB 38, H 48 Asian 12 NA 2 Other 6 HR+/HER2− 193 HR−/HER2+ 106 HR+/HER2 +80 TN 144 MRM NACT XRT Selection: 3 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (SEER 2010–2016) NHW 1652 NHB 371 HR+/HER2− 708 HR−/HER2+ 335 HR+/HER2+ 357 TN 492 BCS 70 M 1564 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (2000–2018) NHW 12 Asian 20 PI 26 Other 2 III 36 IV 20 HR+/HER2− 23 HER2+ 22 TN 6 Unk 9 Selection: 4 stars comparability: 2 stars outcome: 3 stars Retrospective cohort (NCD 2010–2018) NHW 4808 NHB 1153 API 161 H 522 Other 82 Unk 219 III 3988 Unk 72 HR+ 1212 TN 1754 Unk 404 TM 3564 RT 4395 Selection: 4 stars comparability: 2 stars outcome: 3 stars AI American Indian, API AsianPacific Islander, BCS Brease conservative surgery, CMT chemotherapy, ER estrogen receptor, H Hispanic, HER-2 human epidermal growth factor-2, HR hormonal receptor, LN lymph node, M mastectomy, NA Native American, NHB non-Hispanic black, NHW non-Hispanic white, PM partial mastectomy, PMRT postmastectomy radiation, PR progesterone receptor, RM Radical mastectomy, RT radiation, SegM Segmental mastectomy, SM Simple mastectomy, Sx Surgery, TM total mastectomy, TN triple negative, Unk unknown. NHB vs NHW patients with IBC This forest plot in Fig. 2 presents a meta-analysis comparing survival outcomes between NHB and NHW patients, with NHW as the reference group. The pooled hazard ratio across all 10 studies is 1.45 (95% CI: 1.34–1.57), suggesting that NHB patients had a 45% higher risk of adverse survival outcomes compared to NHW patients. Forest plots of the meta-analyses in NHB vs NHW patients with IBC. Hispanic vs NHW patients with IBC This forest plot in Fig. 3 presents a meta-analysis comparing survival outcomes between Hispanic and NHW patients. The pooled HR across seven studies is 1.06 (95% CI: 0.88–1.28), suggesting no significant difference. Forest plots of the meta-analyses in Hispanic vs NHW patients with IBC. API vs NHW patients with IBC The forest plot in Fig. 4A presents a meta-analysis comparing survival outcomes between Asian or API and NHW patients with IBC from four studies with a pooled HR 0.78 (95% CI 0.65–0.94), indicating significantly better survival among Asian or API patients with IBC compared to NHW patients with IBC. Subgroup in Fig. 4B, which compares between Asian and NHW patients with IBC, showed Asian patients had a statistically significant survival outcome compared to NHW patients (HR 0.46, 95% CI 0.23–0.92), similar to the subgroup API vs NHW patients with IBC in Fig. 4C, which showed API patients had higher survival outcomes compared to NHW patients (HR 0.82, 95% CI 0.75–0.95). Pacific Islander patients compared to NHW patients from one study showed an HR of 1.17 (0.26–5.24) with no statistical significance (Fig. 4D). Forest plots of the meta-analyses API vs NHW patients with IBC A Asian or API vs NHW, B Asian vs NHW, C API vs NHW, D PI vs NHW. Sensitivity analysis To address potential duplication of data from the SEER database, we conducted a sensitivity analysis by excluding five overlapping cohorts (11, 14, 16, 21, 22). One study was removed due to the presence of data from the same institution in the dataset (19). A summary of the included and excluded studies is provided in Supplementary Table S1. The results remained consistent with the primary analysis. Specifically, based on six studies (9–10, 17, 23–25), the hazard ratio (HR) for NHB versus NHW patients was 1.43 (95% CI: 1.34–1.53); based on six studies, the HR for Hispanic versus NHW patients was 1.06 (95% CI: 0.88–1.28); and based on three studies, the HR for Asian or API versus NHW patients was 0.65 (95% CI: 0.45–0.94), as shown in Supplementary Figs. S1–S3, respectively. Discussion This is the first comprehensive meta-analysis focused specifically on racial and ethnic disparities in survival outcomes among patients with IBC, revealing significant differences in survival rates across racial and ethnic groups. NHB patients face a 45% greater risk of mortality than NHW patients, underscoring a persistent survival gap. This observation supports previous studies showing a worse survival outcome for NHB patients with IBC, reported at HR 1.86 (1.19–2.91)20,21. A previous study reported a similar trend in patients with non-IBC, showing that NHB patients with non-IBC had a 21% increased risk of death compared to their NHW counterparts, with a hazard ratio (HR) of 1.21 (95% CI: 1.16–1.27) after adjusting for age and year of diagnosis14. Several factors may contribute to this disparity. Socioeconomic inequities, variations in healthcare access, and delays in both diagnosis and the initiation of treatment have been identified as significant contributors to worse outcomes among NHB patients22–24. NHB patients tend to present with more advanced-stage disease, receive less effective treatment, and encounter longer delays from diagnosis to therapy than NHW patients7,25,26. Additionally, NHB patients with IBC are noted to have more aggressive tumor characteristics, including a higher prevalence of triple-negative IBC27, which correlates with a poorer prognosis and fewer targeted therapies available. These clinical and systemic hurdles likely combine to increase survival disparities further. Previous studies reported that API patients with IBC have a better survival outcome compared to NHW patients9,14. Similarly, our study also showed a 12% reduced risk of death with statistically significant results in Asian or API patients with IBC. Several possible explanations have been proposed for this trend. Asian patients tend to have a lower prevalence of aggressive breast cancer subtypes, such as triple-negative breast cancer (TNBC), compared to other racial and ethnic groups28, and are more likely to present with hormone receptor-positive tumors29, which have better prognoses30. Cultural factors, lifestyle behaviors, and potentially favorable genetic or tumor biological characteristics may contribute to their survival advantage (12). It is also important to acknowledge that there is a study included in the meta-analysis that comprised Asian populations outside the United States, particularly from Japan. The inclusion of these non-U.S.-based cohorts within the broader API category may partly account for the observed survival advantage and should be taken into consideration when interpreting these findings in the context of U.S.-based populations. In our analysis, Hispanic patients did not demonstrate statistically significant differences in survival compared to NHW patients. While Fouad et al. study reported worse outcomes among Hispanic patients with IBC19, others have found no difference or even improved survival, possibly due to variations in study design, population characteristics, and adjustment for confounders such as insurance status and treatment received9,17,31. A population-based study from Puerto Rico described the unique characteristics and poor outcomes of Hispanic women with IBC, with a 3-year survival of only 39%, highlighting the vulnerability of this subgroup and the need for disaggregated analyses within Hispanic populations32. This variability underscores the complexity of disentangling the effects of race, ethnicity, socioeconomic factors, and biology in survival analyses, and highlights the need for more granular, disaggregated data. Our study boasts several strengths, including a large cumulative sample size and the inclusion of studies representing diverse geographic regions and populations, which improves external validity. Nonetheless, there are significant limitations to consider when interpreting the findings. While some included studies adjusted for disease stage, others did not, and our meta-analysis was unable to harmonize these differences. As a result, residual confounding by stage is likely. Moreover, we could not account for other critical factors, such as socioeconomic status, healthcare access, or system-level differences, which are known to influence IBC outcomes and are likely drivers of the observed disparities. Differences in how race and ethnicity were classified and reported, along with variability in study design, patient populations, treatment patterns, and disease staging, may also have introduced heterogeneity into the pooled results. Several studies utilized data from large national databases with overlapping timeframes, which may have resulted in some cohort duplication; although the extent of this overlap could not be fully quantified, we carefully considered this possibility in our interpretation. Additionally, survival outcomes for Native American, Alaska Native, and specific Pacific Islander subgroups remain underreported, highlighting a critical gap in the literature. Addressing this limitation will require future research that prioritizes comprehensive data collection and disaggregated reporting of Asian and Pacific Islander subgroups. Finally, while this meta-analysis aimed to evaluate the association between race and ethnicity with survival outcomes, we could not fully account for unmeasured confounders such as comorbidities, genetic factors, and differences in healthcare access. Taken together, these limitations underscore the need for cautious interpretation of our findings and reflect the complexity of disentangling race and ethnicity-related disparities in survival outcomes from observational data. Our meta-analysis suggests potential racial and ethnic differences in survival among patients with IBC. Specifically, NHB patients appeared to have lower survival rates compared to NHW patients, whereas API patients demonstrated better outcomes. These findings contribute to the growing understanding of survival disparities in IBC but should be interpreted with caution, given the inherent biases and confounding associated with observational studies. Future research using more granular racial and ethnic categories and rigorous adjustment for socioeconomic and biological factors is warranted to better characterize these differences. Methods Study design The current systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline33. Literature search Two investigators (C.W., B.P.) independently searched for published articles indexed in Embase and Ovid Medline. The electronic search was conducted separately, utilizing keywords to search for all articles published before March 16, 2025. The keywords, which included terms related to IBC and racial disparity, were combined using the Boolean values ‘AND’ and ‘OR.’ The complete list of search terms used in the systematic search is presented in Supplementary Data 1. After the eligible articles were identified, a manual review of the reference lists of the selected studies was conducted to ensure that all relevant articles were included in the analysis. Eligibility criteria The reviewers utilized eligibility criteria to include studies in the analysis. According to the pre-determined eligibility criterion, articles were included in the analysis if they met the following inclusion criteria: Articles published in the English language. This criterion ensured that the original meaning of the data collected was not lost during translation and that biased results were not reported. Articles published in the English language. This criterion ensured that the original meaning of the data collected was not lost during translation and that biased results were not reported. Studies in patients with IBC. Studies in patients with IBC. Studies have reported differences in survival outcomes based on race/ethnic background. Studies have reported differences in survival outcomes based on race/ethnic background. Studies that included a total sample size greater than 50 patients. This criterion helped enhance the statistical power of our meta-analysis. Studies that included a total sample size greater than 50 patients. This criterion helped enhance the statistical power of our meta-analysis. Similarly, studies were excluded from the current analysis if they met the following exclusion criteria: Articles not published in the English language. Articles not published in the English language. Studies that did not investigate survival outcomes among patients with IBC by race. Studies that did not investigate survival outcomes among patients with IBC by race. Data extraction Two independent reviewers (C.W., B.P.) evaluated all the articles that had been accepted and selected based on the inclusion criteria. They then gathered all the relevant data from the articles used in the current analysis. The data obtained from each study included: Author (the first author’s last name and the year of publication of the article), study design, characteristics of participants (sample size, age, gender distributions, racial distribution), median follow-up period, OS of the patients, and primary outcomes. Quality assessment The Newcastle-Ottawa Scale (NOS) was used to assess the methodological quality of the included studies. This scale assesses the quality of the studies using three domains: comparability, participant selection, and outcomes reporting. Supplementary Table S2 presents the quality of each study. Statistical analysis All statistical analyses were performed using R version 4.3.2 (Vienna, Austria) and ‘meta’ package version 7.0-0. The hazard ratio (HR) and its associated 95% confidence interval (CI) from each study were obtained and combined using the generic inverse variance method per DerSimonian and Laird34. The random effects model was applied due to the different study designs and patient populations of the included studies. Study heterogeneity was evaluated using Cochran’s Q test and I2 2 statistic, with the I2 value of more than 75% being high heterogeneity, 50–75% moderate heterogeneity, and less than 50% low heterogeneity. Sensitivity analysis To address the possibility of overlapping data across studies, particularly those derived from large national databases with similar timeframes, we conducted a sensitivity analysis excluding studies suspected of using overlapping datasets. The impact on pooled hazard ratios was then re-evaluated to ensure the robustness of the overall findings. Supplementary information Supplementary data Acknowledgements This work was supported by NIH grant 3R01CA258523-03S1 (NTU) and Breast Cancer Research Foundation grant BCRF-24-164 (NTU). The funder had no role in the study design, data collection, analysis, interpretation, or manuscript writing. We gratefully acknowledge Dr. Yu Shen from the Department of Biostatistics, Division of Discovery Science, The University of Texas MD Anderson Cancer Center, for the valuable input on the biostatistical and methodological review of this work. Author contributions Conceptualization, C.W., B.P., and N.U.; methodology, C.W. and B.P.; formal analysis, C.W. and B.P.; writing—original draft preparation, C.W., B.P., T.I., Y.T, X.W.; writing—review and editing, C.W., T.I., Y.T., X.W., N.U., Supervision, N.U. Data availability All data generated or analyzed during this study are included in this published article and its supplementary information files. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at 10.1038/s41523-025-00879-9. References Associated Data Supplementary Materials Supplementary data Data Availability Statement All data generated or analyzed during this study are included in this published article and its supplementary information files.