Primary relevance to vitamin D Vitamin D is a stated subject of this paper.
Vitamin D is a stated subject of this paper; the supporting source evidence is shown below.
Why this paper is included
Supporting source evidence:
The aim was to investigate the association between pretreatment vitamin D level and response to NAC and subsequent survival outcomes in BC patients.
Source: Abstract. View original paper
Relevance classification does not establish study quality, causation or editorial approval.
Study design Review / evidence synthesis.
Review / evidence synthesis
Reviews existing research; the exact review methods need further assessment.
Meta-analysis reported · Blinding not applicable to this review · Research setting awaiting verification
How this paper was classified
Automatic source-based classification; awaiting editorial verification. This is not a quality score.
Given the inherent heterogeneity cross included studies, we performed meta-analysis using a random-effects model with the DerSimonian and Laird method to calculate pooled odds or hazard ratios and their 95% CIs.
Source: Methods.
Publisher article type: review-article
Source: Publisher publication metadata.
What the study types mean
- Systematic review
- Uses explicit methods to identify, assess and summarise research on a defined question.
- Scoping review
- Maps the extent and characteristics of research on a topic.
- Narrative review
- Discusses existing research; systematic-review methods have not been established by this classification.
- Review / evidence synthesis
- Reviews existing research; the exact review methods need further assessment.
- Randomised controlled trial
- Assigns participants or groups to comparison interventions by chance.
- Non-randomised intervention study
- Examines an intervention without random allocation.
- Cohort study
- Examines exposures and outcomes over time within a defined group.
- Case-control study
- Compares people with an outcome against people without it, examining exposures.
- Cross-sectional study
- Examines a population at a particular point in time.
- Observational study
- Observes exposures and outcomes without assigning the exposure; the subtype is not established here.
- Case report
- Describes an individual case.
- Case series
- Describes a group of individual cases.
- Secondary analysis
- Analyses existing study data; this paper is not automatically another independent trial.
- Laboratory study
- Investigates biological processes in a laboratory setting.
- Qualitative study
- Examines experiences or meanings using qualitative methods.
- Genetic instrumental-variable study
- Uses genetic variation as an instrument to investigate relationships; this is not a randomised treatment trial.
- Study protocol
- Describes planned methods and is not evidence that the planned study has reported results.
- Multiple reported designs
- More than one design is explicitly reported; examine the supporting methods.
- Study design awaiting verification
- The available source does not establish a sufficiently clear study design.
Meta-analysis combines numerical results from studies. Blinding describes who was unaware of treatment allocation. Neither is a separate quality grade. Study design, vitamin D relevance and document format are independent filters.
Full studies
Pretreatment plasma vitamin D and response to neoadjuvant chemotherapy in breast cancer: evidence from pooled analysis of cohort studies
Review / evidence synthesis · International Journal of Surgery (London, England) · 2024
Original publication title
Pretreatment plasma vitamin D and response to neoadjuvant chemotherapy in breast cancer: evidence from pooled analysis of cohort studies
Copyright © 2024 The Author(s). Published by Wolters Kluwer Health, Inc.
Original publication · Publisher record · CC BY 4.0
Source wording retained. Formatting, section navigation and display-title punctuation adapted for this research record.
At a glance
Study background
What background does the paper provide?
Highlights
Pretreatment vitamin D deficiency was associated with 50% increased odds of non-pCR to neoadjuvant chemotherapy in breast cancer patients.
Combining estimates from studies using varied criteria for pathological response, lower vitamin D level was also associated with inferior overall response, and potential worse event-free survival.
Our findings laid the foundation for future prospective investigations and randomized controlled trials exploring the effects of vitamin D supplementation on improving neoadjuvant chemotherapy response.
Breast cancer (BC) is the most prevalent female malignancy and the fifth leading cause of cancer-related deaths worldwide1. In recent years, neoadjuvant chemotherapy (NAC) has become a standard care for locally advanced BC, in pursuit of an increased breast conservation rate by reducing tumor size2Not reported4. Response to NAC has been robustly linked to higher R0 resection rate and favorable subsequent survival outcomes5,6. However, tumor heterogeneity and variations in patient characteristics impose major challenges in predicting patients’ response to NAC7Not reported9.
Identifying early predictors of response to NAC has important clinical implications in informing treatment efficacy and, therefore, improving prognosis.
Vitamin D is an essential nutrient that plays key roles in a wide spectrum of biological processes. In particular, the antitumor effect of vitamin D has been increasingly recognized, including regulating proliferation, differentiation, apoptosis, angiogenesis, invasion, and metastasis of tumor cells10Not reported12. Preclinical studies have indicated that vitamin D could enhance the efficacy of BC chemotherapy by synergizing or amplifying the cytotoxic activity of antineoplastic agents, such as anthracyclines and taxanes13Not reported15. Specifically, it acts on the transporter to regulate the distribution of the drug in the body, thereby increasing the concentration of the drug in cancer cells16.
Adding vitamin D to paclitaxel could reduce the expression of chemo-resistance marker genes such as multidrug resistance complex 117. Evidence also reported that vitamin D deficiency was associated with a higher risk of side effects such as chemotherapy-induced peripheral neuropathy18. These findings shed light upon the possible role of vitamin D in augmenting the NAC effectiveness within a limited timeframe of a neoadjuvant setting.
Whilst aforementioned biological evidence supported a possible beneficial effect of vitamin D on response to NAC for BC patients, findings of population-level investigations in the association between pretreatment circulating vitamin D and pathological response to NAC remained inconclusive, which might be attributed to the relatively small sample sizes for single clinical investigations regarding NAC for BC patients. In addition, there has been a dearth of a comprehensive review on these evidence despite the potential clinical relevance. We therefore set out to systematically review published evidence and to estimate the association between pretreatment plasma vitamin D level and response to NAC in BC patients with a markedly improved statistical power by performing a meta-analysis.
Source: Introduction Source: Introduction Source: Introduction Source: IntroductionStudy methods
What methods does the paper describe?
Study design and search strategies
We searched the Medline, Embase, Cochrane Library, and Web of Science databases and clinical trial registries from inception to 8 October 2024, using combined keywords including ‘breast cancer’ AND (‘vitamin D’ OR ‘25-hydroxyvitamin D’) AND ‘neoadjuvant’. We described the search strategies in detail in the Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The references of relevant articles were manually screened to identify additional eligible studies. Also, the investigators of relevant conference abstracts were traced separately to identify any subsequent potential publications.
This work has been reported in line with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)19 checklist presented in Supplemental Digital Content 2, http://links.lww.com/JS9/D557 and flow diagram in Supplemental Digital Content 3, http://links.lww.com/JS9/D558 and Assessing the methodological quality of systematic reviews (AMSTAR)-2 guidelines20 (Supplemental Digital Content 4, http://links.lww.com/JS9/D559). The protocol was registered on the International Prospective Register of Systematic Reviews (PROSPERO, CRD42024532891).
Inclusion and exclusion criteria
Eligible studies were expected to meet the following inclusion criteria: 1) studies focusing on females with histologically confirmed BC; 2) investigations with sample sizes no less than 30; 3) studies that assayed pretreatment circulating vitamin D level at BC diagnosis as the primary exposure; 4) studies that adopted measures for NAC response, such as pCR, or long-term (≥12 months) survival outcomes for particular events, for example, tumor recurrence or death from BC after NAC as the outcomes; 5) investigations with study designs including clinical trials, prospective or retrospective cohort studies, and case: control studies.
Studies were excluded if they were published as conference abstracts, comments, case reports, letters, or editorials. During the screening process, two reviewers (C.S. and Q.Y.) independently screened titles and abstracts, with disagreements resolved by consensus or consultation with a third reviewer (Y.H.). We documented the source of included patients in each study, namely the patient database, registry, or study cohort, along with the time-span of patient enrollment to ascertain whether there were overlapping patients included by other eligible studies. Where the name of the data source was not provided, we noted the hospital or institute of patient enrollment instead.
If multiple studies included patients from the same data source or institution within an overlapped time span, only the latest publication was included.
Data extraction and quality assessment
Data items regarding basic characteristics were extracted from each eligible study by two independent reviewers (C.S. and Q.Y.), including the first author, publication year, data source, country, study design, molecular subtypes, clinical tumor stage, vitamin D assessment measure, pretreatment vitamin D level, vitamin D deficiency cut-off value, outcome measures, and sample sizes. The details of data extraction items are presented in the Supplemental Table 2 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). We also documented effect estimates reported by each study including odds ratios (ORs) or hazard ratios (HRs) with corresponding 95% CI (or event numbers where effect estimates could be derived).
HRs with 95% CIs were not obtained from the survival curves.
We evaluated the methodological quality of the included studies using the Newcastle: Ottawa Scale (NOS)21,22. The scoring criteria are presented in the Supplemental Method 1 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). Studies with the NOS scores no less than 7 were deemed high quality. Any discrepancies were solved by discussion with a senior clinician and epidemiologist (Y.H. and C.X.).
Outcome definition
We employed the pCR as the primary outcome, which was defined as the absence of invasive tumor cellularity in the resected breast tumor specimen and axillary lymph nodes after NAC (residual in situ cancer without invasive cancer was considered to have a pCR23). Previous evidence has established robust associations between achieving pCR and improved long-term survival outcomes in BC5,6; and subsequently pCR has been officially listed as a recommended surrogate outcome by the Food and Drug Administration (FDA) and European Medicines Agency (EMA)24,25.
We also defined an overall pathological response to NAC by combining treatment responses using a combined criteria, that is patients showing responses determined by achieving pCR or response in any of other criteria including residual cancer burden (RCB)26 score (RCB ≤1 was defined as responders) and Miller-Payne grading (MPG, grade ≥4 was defined as responders)27. The detailed definitions for each outcome are summarized in the Supplemental Method 2 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). As for survival outcomes, an event-free survival (EFS), where events included local recurrence, distant metastasis, or death from cancer, was adopted to evaluate any associations with long-term prognosis.
Statistical analysis
All statistical analysis was conducted using the ‘meta’ package of R statistical software (version 4.1.3; https://www.R-project.org/). We conducted evidence synthesis to estimate the pooled effect estimates of vitamin D deficiency (defined as pretreatment serum vitamin D<20 ng/ml based on the recommendation from Institute of Medicine [IOM]28) on each outcome. The between-study heterogeneity in the reported effect estimates was estimated using the I 2 statistic29. Given the inherent heterogeneity cross included studies, we performed meta-analysis using a random-effects model with the DerSimonian and Laird method to calculate pooled odds or hazard ratios and their 95% CIs.
Forest plots were used to visualize each individual and the pooled effect estimates with their 95% CIs. An OR or HR >1 indicated a detrimental effect of vitamin D deficiency on the odds of response to NAC.
Subgroup analyses were preplanned based on prior evidence that suggested possible varied vitamin D effects across different patient strata, including biological relevance (tumor stage30, molecular subtype31,32), treatment regimen33, patient ethnicity34, and cut-off thresholds of vitamin D level35. We assessed publication bias by visual evaluation of the funnel plot symmetry and by conducting the Egger’s36 tests where a P-value<0.10 indicated the presence of a significant small study effect. Sensitivity analyses were conducted by excluding studies of lower quality (NOS<7) from the meta-analysis.
All statistical tests were 2-sided, with a P-value <0.05 deemed statistically significant.
Results: Study selection and quality assessment
Our initial literature search yielded 189 citations. After excluding 46 duplicates, 132 studies were removed by reviewing the titles and abstracts based on our inclusion criteria. After that, a total of eleven studies were reviewed in full text5,37Not reported46. Of which, one study was excluded due to a duplicated study population5, one study was excluded for a small sample size (<30)43, three studies were excluded for lacking available outcome data44Not reported46 (details of the flow diagram presented in Fig. 1).
Flowchart of the studies selection process.
Six studies with a total of 1291 BC patients were included37Not reported42. All studies were retrospective in design. Agents used in the NAC regimens adopted by these studies included anthracycline-based drugs, taxane, and trastuzumab. All included studies examined serum vitamin D levels prior to NAC. The basic characteristics of eligible studies are presented in Table 1 and the Supplemental Table 3 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556), where all six included studies enrolled patients from independent data sources with no duplicate.
Three studies were assigned with high methodological quality (NOS≥7)37,38,40. The detailed results of the quality assessment are presented in the Supplemental Tables 4 to 6 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The risk of biases for the included studies was likely to occur in aspects of patient selection and comparability across groups (Fig. 2 and Supplemental Fig. 1A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The quality of evidence for included studies (AMSTAR-2) is presented in Supplemental Digital Content 4, http://links.lww.com/JS9/D559.
Main characteristic of included studies.
| Author | Year | Country | Data source | Study design | Molecular subtypes | Tumor stage | Vitamin D assessment | Vitamin D deficiency cut-off value, ng/ml | Outcomes | OR or HR (95% CI) | Sample sizes |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Clark37 | 2014 | America | I-SPY1 Trial | R | HER-2 negative | II: III | RIA | <20 vs. ≥20 or per 1-unit decreasea | RCB EFS | ORa: 1.01 (0.96 to 1.05) OR: 0.75 (0.14 to 2.19) HR: 1.30 (0.57 to 2.94) | 82 |
| Charehbili38 | 2016 | Netherlands | NEOZOTAC Trial | R | HER-2 negative | II: III | NR | Per 1-unit decreasea | pCR | ORa: 1.00 (0.99 to 1.01) | 169 |
| Kim39 | 2018 | Korea | Severance Hospital | R | NS | I: III | RIA | <20 vs. ≥20 | pCR EFS | OR: 1.54 (0.92 to 2.60) HR: 1.00 (0.46 to 2.17) | 374 |
| Viala40 | 2018 | America, France | Cancer Center in Montpellier and Lowa | R | NS | I: III | ECLIA, MFA | <20 vs. ≥20 | pCR EFS | OR: 1.64 (1.02 to 2.66) HR: 1.11 (0.67 to 1.67) | 327 |
| Atci41 | 2021 | Turkey | Istanbul City Hospital | R | NS | III | ECLIA | <20 vs. ≥20 | MPG | OR: 0.63 (0.27 to 1.47) | 89 |
| Tokunaga42 | 2022 | Japan | Kyushu Cancer Center | R | NS | I: III | ELISA | <20 vs. ≥20 | pCR EFS | OR: 1.27 (0.71 to 2.28) HR: 2.28 (1.12 to 5.03) | 250 |
ECLIA, electrochemiluminescence immunoassay; EFS, event-free survival, including recurrence-free survival (RFS), disease-free survival (DFS), progression-free survival (PFS) and time to distant recurrence (TTDR); ELISA, Enzyme linked immunosorbent assay; HER-2, human epidermal growth factor receptor 2; HR, hazard ratio; MFA, multiplex flow immunoassay; MPG, Miller-Payne grading (MPG ≥ 4 was defined as responders); NR, not reported; NS, not specified; OR, odds ratio; pCR, pathological complete response (No invasive tumor cell in the resected breast tumor specimen and axillary lymph nodes after neoadjuvant chemotherapy); R, retrospective; RCB, residual cancer burden (RCB ≤ 1 was defined as responders); RIA, radioimmunoassay.
Vitamin D as a continuous variable.
Risk of bias plot on pathological complete response (pCR).
Associations between vitamin D and NAC response
Results: Associations between vitamin D and NAC response
Among the included studies, three included stage I: III tumors and used pCR as the primary outcome to measure the pathological response to NAC. As shown in Fig. 3A, by conducting meta-analysis, we observed a significant association between pretreatment vitamin D deficiency and an increased odd of non-pCR (OR=1.50, 95% CI: 1.11 to 2.03, P=0.008). Of note, effect estimates reported by these three studies showed good consistency where no between-study heterogeneity was detected (I 2=0%). As for overall pathological response (Fig. 3B) with combined measures of pCR, RCB and MPG, five studies were included and the meta-analysis identified a significant association of vitamin D deficiency with inferior response to NAC (OR=1.33, 95% CI: 1.01 to 1.75, P=0.046).
Of note, we observed a moderate level of heterogeneity (I2=15%) for this outcome with a broader definition of response. Visual inspection of the contour-enhanced funnel plot (Supplemental Fig. 2A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) did not identify any evident asymmetry, and the Egger’s test did not find any significant small study effects or publication bias (pCR: P=0.104; overall pathological response: P=0.819).
A Forest plot of the effects of vitamin D deficiency on not achieving pathological complete response (pCR); B Forest plot of the effects of vitamin D deficiency on no overall pathological response; C Forest plot of the effects of vitamin D deficiency on event-free survival (EFS).
Four studies reported EFS in BC patients receiving NAC. As shown in the forest plot (Fig. 3C), with mild heterogeneity revealed (I 2=1%), BC patients with pretreatment vitamin D deficiency tended to experience inferior EFS, although the association was not statistically significant (HR=1.27, 95% CI: 0.92 to 1.75, P=0.139). The funnel plot for EFS (Supplemental Fig. 2C of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) was also deemed symmetrical, with an Egger’s test presenting no evidence of small study effects (P=0.585).
Source: Associations between vitamin D and NAC response Source: Associations between vitamin D and NAC response Source: Associations between vitamin D and NAC responseResults: Subgroup and sensitivity analysis
We conducted a subgroup meta-analysis using a random-effects model based on different study characteristics, including molecular subtype (hormone receptor, HR), treatment regimen, patient ethnicity, and clinical tumor stage. Outside meta-analysis on studies including stages I: III patients where pCR was adopted as the primary outcome, we did not find any significant associations between vitamin D deficiency and response to NAC, although all the pooled estimates showed concordant directions of effect that indicated potential harmful roles of vitamin D deficiency on NAC response (vitamin D as a dichotomous variable, with forest plots presented in Fig. 4A: F).
Similar results were also found in terms of overall pathological response (vitamin D as a continuous variable, Supplemental Fig. 3 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556), as well as EFS (regimen, ethnicity, and tumor stage, Supplemental Fig. 4A: D of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Subgroup analysis on the overall pathological response (vitamin D as a dichotomous variable). A Subgroup plot with HR-positive; B Subgroup plot with HR negative; C Subgroup plot with Chemotherapy including trastuzumab; D Subgroup plot with Asian; E Subgroup plot with European; F subgroup plot with stages I: III; HR-positive: Either estrogen receptor (ER) or progesterone receptor (PR) is positive; HR negative: Both estrogen and progesterone receptors are negative.
We conducted sensitivity analysis on response to NAC (Supplemental Fig. 5A of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) and EFS (Supplemental Fig. 5B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) outcomes by including only studies with high methodological quality (NOS≥7), and the direction of effects remained consistent (response: OR=1.46, 95% CI: 0.85 to 2.52, P=0.170; EFS: OR=1.27, 95% CI: 0.92 to 1.75, P=0.139). In addition, two studies analyzed the association using vitamin D level of a continuous scale37,38, our meta-analysis failed to observe a significant association between lower vitamin D (per 1 unit) and response to NAC (OR=1.00, 95% CI: 0.99 to 1.01, P=0.924, Supplemental Fig. 6 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Source: Subgroup and sensitivity analysis Source: Subgroup and sensitivity analysis Source: Subgroup and sensitivity analysisDiscussion: Strengths and limitations of the study
We have performed, to our best knowledge, the first systematic review and meta-analysis on the association between pretreatment vitamin D and response to NAC in BC patients. However, there were several limitations that should be noted for current evidence. Firstly, although our meta-analysis has provided pooled effect estimates with much improved statistical power, relatively small sample sizes, especially for specific subgroups, could limit the credibility of our findings. Given the unavailable individual-level data, we were unable to adjust the possible confounders such as comorbidities and lifestyles (e.g. obesity and physical activity), which were associated with vitamin D level and could influence response to NAC.
Moreover, there has been a lack of data for patients of particular molecular subtypes such as triple-negative BC. Secondly, the retrospective nature of included studies incurred a risk of biases, and the observed association could not be interpreted as causality. Our study calls for more prospective studies with larger sample sizes to further validate the association. Thirdly, the included studies used the same threshold to dichotomize vitamin D, although the measured level could vary across diverse populations and different measuring techniques, which remained to be further investigated with more granular data accumulated in the future.
Fourthly, although the funnel plots and Egger’s test did not identify significant biases from small study effects, the statistical tests were underpowered due to the small number of included studies (n<10) and the results should be interpreted with caution. Finally, due to limited data, current evidence was unable to address the possible dose-response effect of vitamin D on NAC response. As such, future efforts are imperative to conduct large-scale prospective cohort studies with well-defined subgroups of various molecular subtypes to reassess this association.
Source: Strengths and limitations of the studyClinical implications and future directions
Discussion: Clinical implications and future directions
Vitamin D deficiency is highly prevalent worldwide and persists as a potential shared risk factor for many conditions including cancer, posing a major threat to human health63. BC patients are more susceptible to vitamin D deficiency in comparison to the general population64,65. Moreover, a significant decrease of the vitamin D level has been widely observed after the completion of NAC in BC patients66,67. Previous evidence has observed an adverse impact of lower plasma vitamin D level on survival outcomes of BC patients, on top of which our study provided further insights that vitamin D could potentially sensitize patients to NAC.
Our meta-analysis of retrospective cohorts calls for more prospective studies to investigate the potential causality between vitamin D and response to NAC. On the basis of a robust association, advanced approaches such as mendelian randomization could be performed to explore the potential causality between vitamin D level and clinical outcomes in BC patients, and randomized controlled trials (RCTs) that investigate the effects of vitamin D supplementation with varied doses and time schedules on improving response to NAC in BC patients are also expected in the future.
Also, as circulating vitamin D can be readily modified at a low cost, our findings highlighted the possibility of improving response to NAC for BC patients via vitamin D supplementation and laid the groundwork for future well-designed clinical investigations exploring varied doses of vitamin D supplementation for BC patients receiving NAC. Given the limited sample size of current studies68, whether intense pretreatment vitamin D supplementation could compensate for the adverse impact on response to NAC or whether it would be too late still needs to be verified by future population-based prospective studies.
Source: Clinical implications and future directions Source: Clinical implications and future directionsSource of funding
This work was supported by the National Key Research and Development Program of China (grant no. 2023YFC3402100), the National Natural Science Foundation of China (grant no. 92259102 and 82103918), the Natural Science Foundation of Sichuan Province (grant no. 2024NSFSC0057 and 2025NSFSC0554).
Source: Source of fundingResults
Study selection and quality assessment
Our initial literature search yielded 189 citations. After excluding 46 duplicates, 132 studies were removed by reviewing the titles and abstracts based on our inclusion criteria. After that, a total of eleven studies were reviewed in full text5,37Not reported46. Of which, one study was excluded due to a duplicated study population5, one study was excluded for a small sample size (<30)43, three studies were excluded for lacking available outcome data44Not reported46 (details of the flow diagram presented in Fig. 1).
Flowchart of the studies selection process.
Six studies with a total of 1291 BC patients were included37Not reported42. All studies were retrospective in design. Agents used in the NAC regimens adopted by these studies included anthracycline-based drugs, taxane, and trastuzumab. All included studies examined serum vitamin D levels prior to NAC. The basic characteristics of eligible studies are presented in Table 1 and the Supplemental Table 3 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556), where all six included studies enrolled patients from independent data sources with no duplicate.
Three studies were assigned with high methodological quality (NOS≥7)37,38,40. The detailed results of the quality assessment are presented in the Supplemental Tables 4 to 6 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The risk of biases for the included studies was likely to occur in aspects of patient selection and comparability across groups (Fig. 2 and Supplemental Fig. 1A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The quality of evidence for included studies (AMSTAR-2) is presented in Supplemental Digital Content 4, http://links.lww.com/JS9/D559.
Main characteristic of included studies.
| Author | Year | Country | Data source | Study design | Molecular subtypes | Tumor stage | Vitamin D assessment | Vitamin D deficiency cut-off value, ng/ml | Outcomes | OR or HR (95% CI) | Sample sizes |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Clark37 | 2014 | America | I-SPY1 Trial | R | HER-2 negative | II: III | RIA | <20 vs. ≥20 or per 1-unit decreasea | RCB EFS | ORa: 1.01 (0.96 to 1.05) OR: 0.75 (0.14 to 2.19) HR: 1.30 (0.57 to 2.94) | 82 |
| Charehbili38 | 2016 | Netherlands | NEOZOTAC Trial | R | HER-2 negative | II: III | NR | Per 1-unit decreasea | pCR | ORa: 1.00 (0.99 to 1.01) | 169 |
| Kim39 | 2018 | Korea | Severance Hospital | R | NS | I: III | RIA | <20 vs. ≥20 | pCR EFS | OR: 1.54 (0.92 to 2.60) HR: 1.00 (0.46 to 2.17) | 374 |
| Viala40 | 2018 | America, France | Cancer Center in Montpellier and Lowa | R | NS | I: III | ECLIA, MFA | <20 vs. ≥20 | pCR EFS | OR: 1.64 (1.02 to 2.66) HR: 1.11 (0.67 to 1.67) | 327 |
| Atci41 | 2021 | Turkey | Istanbul City Hospital | R | NS | III | ECLIA | <20 vs. ≥20 | MPG | OR: 0.63 (0.27 to 1.47) | 89 |
| Tokunaga42 | 2022 | Japan | Kyushu Cancer Center | R | NS | I: III | ELISA | <20 vs. ≥20 | pCR EFS | OR: 1.27 (0.71 to 2.28) HR: 2.28 (1.12 to 5.03) | 250 |
ECLIA, electrochemiluminescence immunoassay; EFS, event-free survival, including recurrence-free survival (RFS), disease-free survival (DFS), progression-free survival (PFS) and time to distant recurrence (TTDR); ELISA, Enzyme linked immunosorbent assay; HER-2, human epidermal growth factor receptor 2; HR, hazard ratio; MFA, multiplex flow immunoassay; MPG, Miller-Payne grading (MPG ≥ 4 was defined as responders); NR, not reported; NS, not specified; OR, odds ratio; pCR, pathological complete response (No invasive tumor cell in the resected breast tumor specimen and axillary lymph nodes after neoadjuvant chemotherapy); R, retrospective; RCB, residual cancer burden (RCB ≤ 1 was defined as responders); RIA, radioimmunoassay.
Vitamin D as a continuous variable.
Risk of bias plot on pathological complete response (pCR).
Associations between vitamin D and NAC response
Among the included studies, three included stage I: III tumors and used pCR as the primary outcome to measure the pathological response to NAC. As shown in Fig. 3A, by conducting meta-analysis, we observed a significant association between pretreatment vitamin D deficiency and an increased odd of non-pCR (OR=1.50, 95% CI: 1.11 to 2.03, P=0.008). Of note, effect estimates reported by these three studies showed good consistency where no between-study heterogeneity was detected (I 2=0%). As for overall pathological response (Fig. 3B) with combined measures of pCR, RCB and MPG, five studies were included and the meta-analysis identified a significant association of vitamin D deficiency with inferior response to NAC (OR=1.33, 95% CI: 1.01 to 1.75, P=0.046).
Of note, we observed a moderate level of heterogeneity (I2=15%) for this outcome with a broader definition of response. Visual inspection of the contour-enhanced funnel plot (Supplemental Fig. 2A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) did not identify any evident asymmetry, and the Egger’s test did not find any significant small study effects or publication bias (pCR: P=0.104; overall pathological response: P=0.819).
A Forest plot of the effects of vitamin D deficiency on not achieving pathological complete response (pCR); B Forest plot of the effects of vitamin D deficiency on no overall pathological response; C Forest plot of the effects of vitamin D deficiency on event-free survival (EFS).
Four studies reported EFS in BC patients receiving NAC. As shown in the forest plot (Fig. 3C), with mild heterogeneity revealed (I 2=1%), BC patients with pretreatment vitamin D deficiency tended to experience inferior EFS, although the association was not statistically significant (HR=1.27, 95% CI: 0.92 to 1.75, P=0.139). The funnel plot for EFS (Supplemental Fig. 2C of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) was also deemed symmetrical, with an Egger’s test presenting no evidence of small study effects (P=0.585).
Subgroup and sensitivity analysis
We conducted a subgroup meta-analysis using a random-effects model based on different study characteristics, including molecular subtype (hormone receptor, HR), treatment regimen, patient ethnicity, and clinical tumor stage. Outside meta-analysis on studies including stages I: III patients where pCR was adopted as the primary outcome, we did not find any significant associations between vitamin D deficiency and response to NAC, although all the pooled estimates showed concordant directions of effect that indicated potential harmful roles of vitamin D deficiency on NAC response (vitamin D as a dichotomous variable, with forest plots presented in Fig. 4A: F).
Similar results were also found in terms of overall pathological response (vitamin D as a continuous variable, Supplemental Fig. 3 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556), as well as EFS (regimen, ethnicity, and tumor stage, Supplemental Fig. 4A: D of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Subgroup analysis on the overall pathological response (vitamin D as a dichotomous variable). A Subgroup plot with HR-positive; B Subgroup plot with HR negative; C Subgroup plot with Chemotherapy including trastuzumab; D Subgroup plot with Asian; E Subgroup plot with European; F subgroup plot with stages I: III; HR-positive: Either estrogen receptor (ER) or progesterone receptor (PR) is positive; HR negative: Both estrogen and progesterone receptors are negative.
We conducted sensitivity analysis on response to NAC (Supplemental Fig. 5A of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) and EFS (Supplemental Fig. 5B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) outcomes by including only studies with high methodological quality (NOS≥7), and the direction of effects remained consistent (response: OR=1.46, 95% CI: 0.85 to 2.52, P=0.170; EFS: OR=1.27, 95% CI: 0.92 to 1.75, P=0.139). In addition, two studies analyzed the association using vitamin D level of a continuous scale37,38, our meta-analysis failed to observe a significant association between lower vitamin D (per 1 unit) and response to NAC (OR=1.00, 95% CI: 0.99 to 1.01, P=0.924, Supplemental Fig. 6 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Conclusion
Our study suggested an association between pretreatment vitamin D deficiency and reduced response to NAC in BC patients, although causality could not be inferred from current evidence with observational design. Therefore, active surveillance on vitamin D level as a biomarker before and during NAC were warranted for BC patients. However, as a safe intervention at a low cost, the causal effect of vitamin D supplementation to improve treatment response could only be proved by future RCTs. Also, our meta-analysis supports a compelling need for future prospective studies with larger sample sizes and well-designed subgroups to strengthen current evidence.
Source: ConclusionAuthors’ abstract
Background:
Biological evidence has revealed antitumor effect of vitamin D, but whether it could predict the response to neoadjuvant chemotherapy (NAC) in breast cancer (BC) patients remains inconclusive. The aim was to investigate the association between pretreatment vitamin D level and response to NAC and subsequent survival outcomes in BC patients.
Materials and methods:
The authors systematically searched the Medline, Embase, Cochrane Library, and Web of Science databases and clinical trial registries to identify relevant articles from inception to 8 October 2024. Eligible studies investigating the associations between pretreatment plasma vitamin D and response to NAC in BC patients were selected according to the predefined criteria, with the study characteristics extracted by two reviewers. The primary outcome was pathological complete response (pCR), while overall pathological response and event-free survival (EFS) were adopted as secondary outcomes.
Summary effect estimates of odds ratios (ORs) or hazard ratios (HRs) with 95% CIs were pooled using a random-effects model. Subgroup and sensitivity analyses were performed based on study characteristics and methodological quality.
Results:
Six retrospective cohort studies involving 1291 BC patients were included. The authors observed a significant association between pretreatment vitamin D deficiency and 50% increased odds of non-pCR after NAC (OR=1.50, 95% CI: 1.11 to 2.03, P=0.008) with no heterogeneity (I 2=0%). The authors also identified a significant association of vitamin D with the overall pathological response (OR=1.33, 95% CI: 1.01 to 1.75, P=0.046). A similar association with EFS (HR=1.27, 95% CI: 0.92 to 1.75, P=0.139) was also noted although the effect estimate was not statistically significant.
Sensitivity analyses based on methodological quality showed consistent findings.
Conclusion:
Pretreatment vitamin D deficiency is associated with an inferior response to NAC in BC patients. Our meta-analysis advocates further prospective studies with large sample sizes before vitamin D supplementation could be administered to improve NAC response and subsequent prognosis of BC patients.
Read the full paper
Introduction
Highlights
Pretreatment vitamin D deficiency was associated with 50% increased odds of non-pCR to neoadjuvant chemotherapy in breast cancer patients.
Combining estimates from studies using varied criteria for pathological response, lower vitamin D level was also associated with inferior overall response, and potential worse event-free survival.
Our findings laid the foundation for future prospective investigations and randomized controlled trials exploring the effects of vitamin D supplementation on improving neoadjuvant chemotherapy response.
Breast cancer (BC) is the most prevalent female malignancy and the fifth leading cause of cancer-related deaths worldwide1. In recent years, neoadjuvant chemotherapy (NAC) has become a standard care for locally advanced BC, in pursuit of an increased breast conservation rate by reducing tumor size2Not reported4. Response to NAC has been robustly linked to higher R0 resection rate and favorable subsequent survival outcomes5,6. However, tumor heterogeneity and variations in patient characteristics impose major challenges in predicting patients’ response to NAC7Not reported9.
Identifying early predictors of response to NAC has important clinical implications in informing treatment efficacy and, therefore, improving prognosis.
Vitamin D is an essential nutrient that plays key roles in a wide spectrum of biological processes. In particular, the antitumor effect of vitamin D has been increasingly recognized, including regulating proliferation, differentiation, apoptosis, angiogenesis, invasion, and metastasis of tumor cells10Not reported12. Preclinical studies have indicated that vitamin D could enhance the efficacy of BC chemotherapy by synergizing or amplifying the cytotoxic activity of antineoplastic agents, such as anthracyclines and taxanes13Not reported15. Specifically, it acts on the transporter to regulate the distribution of the drug in the body, thereby increasing the concentration of the drug in cancer cells16.
Adding vitamin D to paclitaxel could reduce the expression of chemo-resistance marker genes such as multidrug resistance complex 117. Evidence also reported that vitamin D deficiency was associated with a higher risk of side effects such as chemotherapy-induced peripheral neuropathy18. These findings shed light upon the possible role of vitamin D in augmenting the NAC effectiveness within a limited timeframe of a neoadjuvant setting.
Whilst aforementioned biological evidence supported a possible beneficial effect of vitamin D on response to NAC for BC patients, findings of population-level investigations in the association between pretreatment circulating vitamin D and pathological response to NAC remained inconclusive, which might be attributed to the relatively small sample sizes for single clinical investigations regarding NAC for BC patients. In addition, there has been a dearth of a comprehensive review on these evidence despite the potential clinical relevance. We therefore set out to systematically review published evidence and to estimate the association between pretreatment plasma vitamin D level and response to NAC in BC patients with a markedly improved statistical power by performing a meta-analysis.
Materials and methods
Study design and search strategies
We searched the Medline, Embase, Cochrane Library, and Web of Science databases and clinical trial registries from inception to 8 October 2024, using combined keywords including ‘breast cancer’ AND (‘vitamin D’ OR ‘25-hydroxyvitamin D’) AND ‘neoadjuvant’. We described the search strategies in detail in the Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The references of relevant articles were manually screened to identify additional eligible studies. Also, the investigators of relevant conference abstracts were traced separately to identify any subsequent potential publications.
This work has been reported in line with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)19 checklist presented in Supplemental Digital Content 2, http://links.lww.com/JS9/D557 and flow diagram in Supplemental Digital Content 3, http://links.lww.com/JS9/D558 and Assessing the methodological quality of systematic reviews (AMSTAR)-2 guidelines20 (Supplemental Digital Content 4, http://links.lww.com/JS9/D559). The protocol was registered on the International Prospective Register of Systematic Reviews (PROSPERO, CRD42024532891).
Inclusion and exclusion criteria
Eligible studies were expected to meet the following inclusion criteria: 1) studies focusing on females with histologically confirmed BC; 2) investigations with sample sizes no less than 30; 3) studies that assayed pretreatment circulating vitamin D level at BC diagnosis as the primary exposure; 4) studies that adopted measures for NAC response, such as pCR, or long-term (≥12 months) survival outcomes for particular events, for example, tumor recurrence or death from BC after NAC as the outcomes; 5) investigations with study designs including clinical trials, prospective or retrospective cohort studies, and case: control studies.
Studies were excluded if they were published as conference abstracts, comments, case reports, letters, or editorials. During the screening process, two reviewers (C.S. and Q.Y.) independently screened titles and abstracts, with disagreements resolved by consensus or consultation with a third reviewer (Y.H.). We documented the source of included patients in each study, namely the patient database, registry, or study cohort, along with the time-span of patient enrollment to ascertain whether there were overlapping patients included by other eligible studies. Where the name of the data source was not provided, we noted the hospital or institute of patient enrollment instead.
If multiple studies included patients from the same data source or institution within an overlapped time span, only the latest publication was included.
Data extraction and quality assessment
Data items regarding basic characteristics were extracted from each eligible study by two independent reviewers (C.S. and Q.Y.), including the first author, publication year, data source, country, study design, molecular subtypes, clinical tumor stage, vitamin D assessment measure, pretreatment vitamin D level, vitamin D deficiency cut-off value, outcome measures, and sample sizes. The details of data extraction items are presented in the Supplemental Table 2 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). We also documented effect estimates reported by each study including odds ratios (ORs) or hazard ratios (HRs) with corresponding 95% CI (or event numbers where effect estimates could be derived).
HRs with 95% CIs were not obtained from the survival curves.
We evaluated the methodological quality of the included studies using the Newcastle: Ottawa Scale (NOS)21,22. The scoring criteria are presented in the Supplemental Method 1 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). Studies with the NOS scores no less than 7 were deemed high quality. Any discrepancies were solved by discussion with a senior clinician and epidemiologist (Y.H. and C.X.).
Outcome definition
We employed the pCR as the primary outcome, which was defined as the absence of invasive tumor cellularity in the resected breast tumor specimen and axillary lymph nodes after NAC (residual in situ cancer without invasive cancer was considered to have a pCR23). Previous evidence has established robust associations between achieving pCR and improved long-term survival outcomes in BC5,6; and subsequently pCR has been officially listed as a recommended surrogate outcome by the Food and Drug Administration (FDA) and European Medicines Agency (EMA)24,25.
We also defined an overall pathological response to NAC by combining treatment responses using a combined criteria, that is patients showing responses determined by achieving pCR or response in any of other criteria including residual cancer burden (RCB)26 score (RCB ≤1 was defined as responders) and Miller-Payne grading (MPG, grade ≥4 was defined as responders)27. The detailed definitions for each outcome are summarized in the Supplemental Method 2 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). As for survival outcomes, an event-free survival (EFS), where events included local recurrence, distant metastasis, or death from cancer, was adopted to evaluate any associations with long-term prognosis.
Statistical analysis
All statistical analysis was conducted using the ‘meta’ package of R statistical software (version 4.1.3; https://www.R-project.org/). We conducted evidence synthesis to estimate the pooled effect estimates of vitamin D deficiency (defined as pretreatment serum vitamin D<20 ng/ml based on the recommendation from Institute of Medicine [IOM]28) on each outcome. The between-study heterogeneity in the reported effect estimates was estimated using the I 2 statistic29. Given the inherent heterogeneity cross included studies, we performed meta-analysis using a random-effects model with the DerSimonian and Laird method to calculate pooled odds or hazard ratios and their 95% CIs.
Forest plots were used to visualize each individual and the pooled effect estimates with their 95% CIs. An OR or HR >1 indicated a detrimental effect of vitamin D deficiency on the odds of response to NAC.
Subgroup analyses were preplanned based on prior evidence that suggested possible varied vitamin D effects across different patient strata, including biological relevance (tumor stage30, molecular subtype31,32), treatment regimen33, patient ethnicity34, and cut-off thresholds of vitamin D level35. We assessed publication bias by visual evaluation of the funnel plot symmetry and by conducting the Egger’s36 tests where a P-value<0.10 indicated the presence of a significant small study effect. Sensitivity analyses were conducted by excluding studies of lower quality (NOS<7) from the meta-analysis.
All statistical tests were 2-sided, with a P-value <0.05 deemed statistically significant.
Results
Study selection and quality assessment
Our initial literature search yielded 189 citations. After excluding 46 duplicates, 132 studies were removed by reviewing the titles and abstracts based on our inclusion criteria. After that, a total of eleven studies were reviewed in full text5,37Not reported46. Of which, one study was excluded due to a duplicated study population5, one study was excluded for a small sample size (<30)43, three studies were excluded for lacking available outcome data44Not reported46 (details of the flow diagram presented in Fig. 1).
Flowchart of the studies selection process.
Six studies with a total of 1291 BC patients were included37Not reported42. All studies were retrospective in design. Agents used in the NAC regimens adopted by these studies included anthracycline-based drugs, taxane, and trastuzumab. All included studies examined serum vitamin D levels prior to NAC. The basic characteristics of eligible studies are presented in Table 1 and the Supplemental Table 3 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556), where all six included studies enrolled patients from independent data sources with no duplicate.
Three studies were assigned with high methodological quality (NOS≥7)37,38,40. The detailed results of the quality assessment are presented in the Supplemental Tables 4 to 6 (Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The risk of biases for the included studies was likely to occur in aspects of patient selection and comparability across groups (Fig. 2 and Supplemental Fig. 1A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556). The quality of evidence for included studies (AMSTAR-2) is presented in Supplemental Digital Content 4, http://links.lww.com/JS9/D559.
Main characteristic of included studies.
| Author | Year | Country | Data source | Study design | Molecular subtypes | Tumor stage | Vitamin D assessment | Vitamin D deficiency cut-off value, ng/ml | Outcomes | OR or HR (95% CI) | Sample sizes |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Clark37 | 2014 | America | I-SPY1 Trial | R | HER-2 negative | II: III | RIA | <20 vs. ≥20 or per 1-unit decreasea | RCB EFS | ORa: 1.01 (0.96 to 1.05) OR: 0.75 (0.14 to 2.19) HR: 1.30 (0.57 to 2.94) | 82 |
| Charehbili38 | 2016 | Netherlands | NEOZOTAC Trial | R | HER-2 negative | II: III | NR | Per 1-unit decreasea | pCR | ORa: 1.00 (0.99 to 1.01) | 169 |
| Kim39 | 2018 | Korea | Severance Hospital | R | NS | I: III | RIA | <20 vs. ≥20 | pCR EFS | OR: 1.54 (0.92 to 2.60) HR: 1.00 (0.46 to 2.17) | 374 |
| Viala40 | 2018 | America, France | Cancer Center in Montpellier and Lowa | R | NS | I: III | ECLIA, MFA | <20 vs. ≥20 | pCR EFS | OR: 1.64 (1.02 to 2.66) HR: 1.11 (0.67 to 1.67) | 327 |
| Atci41 | 2021 | Turkey | Istanbul City Hospital | R | NS | III | ECLIA | <20 vs. ≥20 | MPG | OR: 0.63 (0.27 to 1.47) | 89 |
| Tokunaga42 | 2022 | Japan | Kyushu Cancer Center | R | NS | I: III | ELISA | <20 vs. ≥20 | pCR EFS | OR: 1.27 (0.71 to 2.28) HR: 2.28 (1.12 to 5.03) | 250 |
ECLIA, electrochemiluminescence immunoassay; EFS, event-free survival, including recurrence-free survival (RFS), disease-free survival (DFS), progression-free survival (PFS) and time to distant recurrence (TTDR); ELISA, Enzyme linked immunosorbent assay; HER-2, human epidermal growth factor receptor 2; HR, hazard ratio; MFA, multiplex flow immunoassay; MPG, Miller-Payne grading (MPG ≥ 4 was defined as responders); NR, not reported; NS, not specified; OR, odds ratio; pCR, pathological complete response (No invasive tumor cell in the resected breast tumor specimen and axillary lymph nodes after neoadjuvant chemotherapy); R, retrospective; RCB, residual cancer burden (RCB ≤ 1 was defined as responders); RIA, radioimmunoassay.
Vitamin D as a continuous variable.
Risk of bias plot on pathological complete response (pCR).
Associations between vitamin D and NAC response
Among the included studies, three included stage I: III tumors and used pCR as the primary outcome to measure the pathological response to NAC. As shown in Fig. 3A, by conducting meta-analysis, we observed a significant association between pretreatment vitamin D deficiency and an increased odd of non-pCR (OR=1.50, 95% CI: 1.11 to 2.03, P=0.008). Of note, effect estimates reported by these three studies showed good consistency where no between-study heterogeneity was detected (I 2=0%). As for overall pathological response (Fig. 3B) with combined measures of pCR, RCB and MPG, five studies were included and the meta-analysis identified a significant association of vitamin D deficiency with inferior response to NAC (OR=1.33, 95% CI: 1.01 to 1.75, P=0.046).
Of note, we observed a moderate level of heterogeneity (I2=15%) for this outcome with a broader definition of response. Visual inspection of the contour-enhanced funnel plot (Supplemental Fig. 2A,B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) did not identify any evident asymmetry, and the Egger’s test did not find any significant small study effects or publication bias (pCR: P=0.104; overall pathological response: P=0.819).
A Forest plot of the effects of vitamin D deficiency on not achieving pathological complete response (pCR); B Forest plot of the effects of vitamin D deficiency on no overall pathological response; C Forest plot of the effects of vitamin D deficiency on event-free survival (EFS).
Four studies reported EFS in BC patients receiving NAC. As shown in the forest plot (Fig. 3C), with mild heterogeneity revealed (I 2=1%), BC patients with pretreatment vitamin D deficiency tended to experience inferior EFS, although the association was not statistically significant (HR=1.27, 95% CI: 0.92 to 1.75, P=0.139). The funnel plot for EFS (Supplemental Fig. 2C of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) was also deemed symmetrical, with an Egger’s test presenting no evidence of small study effects (P=0.585).
Subgroup and sensitivity analysis
We conducted a subgroup meta-analysis using a random-effects model based on different study characteristics, including molecular subtype (hormone receptor, HR), treatment regimen, patient ethnicity, and clinical tumor stage. Outside meta-analysis on studies including stages I: III patients where pCR was adopted as the primary outcome, we did not find any significant associations between vitamin D deficiency and response to NAC, although all the pooled estimates showed concordant directions of effect that indicated potential harmful roles of vitamin D deficiency on NAC response (vitamin D as a dichotomous variable, with forest plots presented in Fig. 4A: F).
Similar results were also found in terms of overall pathological response (vitamin D as a continuous variable, Supplemental Fig. 3 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556), as well as EFS (regimen, ethnicity, and tumor stage, Supplemental Fig. 4A: D of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Subgroup analysis on the overall pathological response (vitamin D as a dichotomous variable). A Subgroup plot with HR-positive; B Subgroup plot with HR negative; C Subgroup plot with Chemotherapy including trastuzumab; D Subgroup plot with Asian; E Subgroup plot with European; F subgroup plot with stages I: III; HR-positive: Either estrogen receptor (ER) or progesterone receptor (PR) is positive; HR negative: Both estrogen and progesterone receptors are negative.
We conducted sensitivity analysis on response to NAC (Supplemental Fig. 5A of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) and EFS (Supplemental Fig. 5B of Supplemental Digital Content 1, http://links.lww.com/JS9/D556) outcomes by including only studies with high methodological quality (NOS≥7), and the direction of effects remained consistent (response: OR=1.46, 95% CI: 0.85 to 2.52, P=0.170; EFS: OR=1.27, 95% CI: 0.92 to 1.75, P=0.139). In addition, two studies analyzed the association using vitamin D level of a continuous scale37,38, our meta-analysis failed to observe a significant association between lower vitamin D (per 1 unit) and response to NAC (OR=1.00, 95% CI: 0.99 to 1.01, P=0.924, Supplemental Fig. 6 of Supplemental Digital Content 1, http://links.lww.com/JS9/D556).
Discussion
Summary of main findings
To our best knowledge, this is the first systematic review and meta-analysis investigating the association between pretreatment vitamin D and response to NAC in BC patients. Our findings revealed a significant association of vitamin D deficiency with inferior response to NAC. Subgroup and sensitivity analysis found generally consistent evidence, though causality cannot be inferred from these observational data.
Comparison with existing literature and interpretation
Previous epidemiological evidence has robustly linked low circulating vitamin D level to escalated BC risk47Not reported49. In addition, vitamin D deficiency was more prevalent among patients with advanced stages and unfavorable clinicopathological characteristics, and could also influence overall survival for BC patients50,51. These findings were consistent with biological evidence underpinning the potential antitumor functions of vitamin D. Recent evidence has shed light upon the potential synergetic effect of vitamin D on chemotherapy drugs and its protective effect against chemotherapy side effects.
For example, vitamin D could intensify the combined effect of 5-aminolevulinic acid-based photodynamic therapy (ALA-PDT) and paclitaxel on tumor growth and apoptosis in BC cells14. Also, studies showed that vitamin D supplementation provided protective effects against doxorubicin (DOX)-induced cardiotoxicity in BC patients52,53. Vitamin D insufficiency, on the other hand, was associated with chemotherapy-induced peripheral neuropathy from paclitaxel for early-stage BC18. Therefore, current biological evidence linked vitamin D level to response to NAC of BC and laid the foundation for our study to further investigate its predictive role in clinical practice.
Our meta-analysis provided the first population-level evidence supporting the linkage between pretreatment vitamin D deficiency and inferior response to NAC in BC patients. By pooling effect estimates, our study overcame the primary limitation of previous studies: insufficient statistical power from limited sample sizes: in the setting of NAC. We employed pCR as the primary outcome measuring pathological response to NAC given that it has been officially listed as a surrogate outcome by FDA and EMA24,25 for accelerated drug approval, and has been widely adopted as the primary outcome by many published trials.
Notably, two of the included studies37,41 utilized MPG or RCB criteria to evaluate tumor response, where formulas were applied to created scores to measure the residual tumor burden. These two studies included relatively small sample sizes and adding their estimates to our meta-analysis yielded a significant association between vitamin D deficiency and poor overall response, although an expected elevated moderate heterogeneity was observed for this secondary outcome. With a broader definition of treatment response using various criteria such as RCB and MPG, the pooled estimate still indicated a significant association between vitamin D deficiency and inferior response, which further strengthened the robustness of our findings.
We also noticed that these two studies reported point estimates of opposite direction, although the effects were nonsignificant. In addition to the different criteria used to define pathological response, they included BC patients with more advanced tumor stages (II: III), leading to possible lower response rates indicated by previous evidence40,42, and subsequently contributing to the heterogeneous effect estimates. The small sample sizes (<100) also yielded a greater extend of variability to their estimates, and their point estimates should be interpreted with caution.
EFS was also investigated as a secondary outcome in our study. Due to a limited number of outcome events, we failed to identify a significant impact of vitamin D on EFS, although a consistent direction of association was detected. Previous observational evidence had uncovered a favorable impact of vitamin D level on overall and disease-free survival for all BC patients54, and this was expected to be further validated among BC patients receiving NAC by conducting more prospective studies with larger sample sizes.
Another point worth mentioning was the dichotomized definition and arbitrary threshold selection of vitamin D deficiency. Although all the included studies chose a widely recommended threshold (<20 ng/ml), vitamin D levels and deficiency thresholds can vary depending on the assay used, geography, and population characteristics55,56. Moreover, epidemiological evidence demonstrated that dichotomy could lead to loss of information with subsequent impaired statistical power and possible biases57. Our systematic review only found two studies37,38 that reported the regression coefficients where vitamin D level was modeled in a continuous scale, and null effects were identified from the underpowered meta-analysis.
Future efforts were merited to explore the possible dose-response relationship between vitamin D level and response to NAC.
Of note, our extensive literature search identified no prospective studies yet on this topic, leading to all the included studies in our meta-analysis being retrospective in design, which introduced a risk of bias, particularly in patient selection and comparability, across different studies. Among the eligible studies, the primary source for risk of bias based on the NOS score was the lack of comparability, where crude ORs were used as effect estimates with covariates unadjusted. Consequently, one of the three studies with high risk (NOS<7)39,41,42 reported an reversed direction of effect41, although the result was not statistically significant.
To mitigate the risk of selection bias, the included high-quality (NOS≥7) studies enrolled patients from multiple centers, while adjusting for common covariates in multivariable analyses37,38,40. Based on that, we conducted sensitivity analysis by excluding studies with low NOS scores, and observed a consistent direction of effect, which further strengthened the robustness of our findings. However, due to the retrospective nature, our findings could provide initial evidence on the association between vitamin D and NAC response, and support a pressing need for prospective studies in the future to confirm this finding.
Anthracycline-based or taxane-based therapies are commonly prescribed in the current regimen of NAC58, which were also adopted by studies included in this meta-analysis. For human epidermal growth factor receptor 2 (HER-2) positive patients, trastuzumab would also be added. We conducted subgroup analysis based on treatment regimen and found the nonsignificant effect of vitamin D among studies including trastuzumab for HER-2 positive patients. However, due to data availability, we were unable to perform separate analysis on HER-2 positive patients, which merited future investigation in whether vitamin D was less predictive of response to targeted therapies as current biological evidence has been primarily focusing on chemotherapeutic agents.
Scientific evidence indicated that vitamin D intensified the effects of paclitaxel through down-regulated B-cell lymphoma-2 (Bcl-2) expression and upregulated the expression of Bcl-2 associated X (Bax) and cleaved caspase-3 in BC cells, thereby promoting the apoptosis of BC tumor cells14,17. With respect to other molecular subtypes, we conducted stratified analyses by hormone receptor (HR) status, but with only two studies eligible, we did not identify any significant effects of vitamin D among either HR+ or HR− patients. There has been evidence showing that lower vitamin D level was associated with unfavorable BC molecular subtypes, that is triple-negative tumors35,59.
Vitamin D might also influence the therapeutic effect in BC patients with certain molecular subtypes, such as estrogen receptor (ER) and HR-positive tumors60Not reported62. These findings pointed to a potential modifying effect of molecular subtypes on the association between vitamin D and treatment response. Owing to limited data, however, we were unable to assess possible effect modification by conducting subgroup analyses. More studies with granular subgroup data are warranted in the future to further clarify these effects.
Strengths and limitations of the study
We have performed, to our best knowledge, the first systematic review and meta-analysis on the association between pretreatment vitamin D and response to NAC in BC patients. However, there were several limitations that should be noted for current evidence. Firstly, although our meta-analysis has provided pooled effect estimates with much improved statistical power, relatively small sample sizes, especially for specific subgroups, could limit the credibility of our findings. Given the unavailable individual-level data, we were unable to adjust the possible confounders such as comorbidities and lifestyles (e.g. obesity and physical activity), which were associated with vitamin D level and could influence response to NAC.
Moreover, there has been a lack of data for patients of particular molecular subtypes such as triple-negative BC. Secondly, the retrospective nature of included studies incurred a risk of biases, and the observed association could not be interpreted as causality. Our study calls for more prospective studies with larger sample sizes to further validate the association. Thirdly, the included studies used the same threshold to dichotomize vitamin D, although the measured level could vary across diverse populations and different measuring techniques, which remained to be further investigated with more granular data accumulated in the future.
Fourthly, although the funnel plots and Egger’s test did not identify significant biases from small study effects, the statistical tests were underpowered due to the small number of included studies (n<10) and the results should be interpreted with caution. Finally, due to limited data, current evidence was unable to address the possible dose-response effect of vitamin D on NAC response. As such, future efforts are imperative to conduct large-scale prospective cohort studies with well-defined subgroups of various molecular subtypes to reassess this association.
Clinical implications and future directions
Vitamin D deficiency is highly prevalent worldwide and persists as a potential shared risk factor for many conditions including cancer, posing a major threat to human health63. BC patients are more susceptible to vitamin D deficiency in comparison to the general population64,65. Moreover, a significant decrease of the vitamin D level has been widely observed after the completion of NAC in BC patients66,67. Previous evidence has observed an adverse impact of lower plasma vitamin D level on survival outcomes of BC patients, on top of which our study provided further insights that vitamin D could potentially sensitize patients to NAC.
Our meta-analysis of retrospective cohorts calls for more prospective studies to investigate the potential causality between vitamin D and response to NAC. On the basis of a robust association, advanced approaches such as mendelian randomization could be performed to explore the potential causality between vitamin D level and clinical outcomes in BC patients, and randomized controlled trials (RCTs) that investigate the effects of vitamin D supplementation with varied doses and time schedules on improving response to NAC in BC patients are also expected in the future.
Also, as circulating vitamin D can be readily modified at a low cost, our findings highlighted the possibility of improving response to NAC for BC patients via vitamin D supplementation and laid the groundwork for future well-designed clinical investigations exploring varied doses of vitamin D supplementation for BC patients receiving NAC. Given the limited sample size of current studies68, whether intense pretreatment vitamin D supplementation could compensate for the adverse impact on response to NAC or whether it would be too late still needs to be verified by future population-based prospective studies.
Conclusion
Our study suggested an association between pretreatment vitamin D deficiency and reduced response to NAC in BC patients, although causality could not be inferred from current evidence with observational design. Therefore, active surveillance on vitamin D level as a biomarker before and during NAC were warranted for BC patients. However, as a safe intervention at a low cost, the causal effect of vitamin D supplementation to improve treatment response could only be proved by future RCTs. Also, our meta-analysis supports a compelling need for future prospective studies with larger sample sizes and well-designed subgroups to strengthen current evidence.
Ethical approval
Not applicable.
Consent
Not applicable.
Source of funding
This work was supported by the National Key Research and Development Program of China (grant no. 2023YFC3402100), the National Natural Science Foundation of China (grant no. 92259102 and 82103918), the Natural Science Foundation of Sichuan Province (grant no. 2024NSFSC0057 and 2025NSFSC0554).
Author contribution
Q.Y., C.S., J.H., X.X., H.L., X.W., X.C., Y.X., Y.Z., C.X., and Y.H.: acquisition, analysis, or interpretation of data; Q.Y., C.S., Y.H., and C.X.: drafting of manuscript; C.S. and X.X.: statistical analysis; Y.H. and C.X.: study supervision. All co-authors contributed in critical revision of the manuscript for important intellectual content.
Conflicts of interest disclosure
The authors declare no conflicts of interest.
Research registration unique identifying number (UIN)
Name of the registry: International Prospective Register of Systematic Reviews (PROSPERO).
Unique identifying registration number: CRD42024532891.
Hyperlink to the registration: https://www.crd.york.ac.uk/prospero/#recordDetails.
Guarantor
Yazhou He.
Data availability statement
All data in this article can be shared upon reasonable request to the corresponding author.
Provenance and peer review
Not commissioned, externally peer-reviewed.
Supplementary Material
SUPPLEMENTARY MATERIAL
Acknowledgements
Assistance with the study: None declared.
Presentation: None.
Chi Shu, Qian Yang, and Jun Huang contributed equally to this manuscript.
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.lww.com/international-journal-of-surgery.
Published online 18 November 2024
References
Publication and source details
Original title: Pretreatment plasma vitamin D and response to neoadjuvant chemotherapy in breast cancer: evidence from pooled analysis of cohort studies
Chi Shu; Qian Yang; Jun Huang; Xuan Xie; Hong Li; Hong Wu; Xin Wang; Xin Chen; Yuping Xie; Yanhong Zhou; Yazhou He; Chuan Xu
International Journal of Surgery (London, England) · 2024
* Corresponding author. Address: Department of Oncology and Department of Epidemiology and Medical Statistics, West China School of Public Health and West China Fourth Hospital, Sichuan University, Chengdu 610041, China. Tel.: +86 288 550 1630. E-mail: yazhou.he@scu.edu.cn (Y. He); Department of Oncology and Cancer Institute, Sichuan Academy of Medical Sciences, Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu 610072, China. Tel.: +86 288 739 4677. E-mail: xuchuan100@uestc.edu.cn (C. Xu).Copyright © 2024 The Author(s). Published by Wolters Kluwer Health, Inc.2024https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the Creative Commons Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. http://creativecommons.org/licenses/by/4.0/Repository JATS snapshot; version as supplied
Source retrieved: 2026-11-11T03:11:04.276242+00:00
DOI: 10.1097/js9.0000000000002142 · PMC11634150