Safety and efficacy of docosahexaenoic acid supplementation during neoadjuvant breast cancer therapy: Findings from the phase II, double‐blind, randomized controlled DHA‐WIN trial

Abstract

There is limited clinical evidence of docosahexaenoic acid (DHA) efficacy during breast cancer neoadjuvant chemotherapy (NAC). This randomized, double‐blind, placebo‐controlled trial aimed to investigate the safety and efficacy of DHA supplementation in breast cancer patients undergoing NAC. Participants (n = 49) were assigned to receive either DHA 4.4 g/day orally (algae triacylglycerol) or a placebo (corn/soy oil) over six cycles (18 weeks) of NAC. The primary outcome was the evaluation of changes in the percentage of Ki‐67 expression, assessed by immunohistochemistry analysis from pre‐ to post‐treatment. Secondary outcomes included pathological complete response, incidence of adverse effects, and 3‐year survival analysis. Compliance was evaluated by fatty acid analysis of plasma phospholipids and erythrocyte total lipids quantified by gas–liquid chromatography. The expression of Ki‐67 significantly decreased in both groups, with no significant effects of the DHA intervention (p = 0.38). When stratified by breast cancer subtype, there was a trend of greater reduction in Ki‐67 expression in the human epidermal growth receptor 2 (HER2+++) subtype in the DHA group compared to placebo (p = 0.1). The % of DHA in erythrocytes and plasma phospholipids was increased by two‐fold at 9 and 15 weeks of therapy in the DHA group, while it remained unchanged in the placebo group (p‐interaction <0.001). There was no reported incidence of adverse effects related to the intervention, and no significant effects were found in the other secondary outcomes. NAC significantly decreased the expression of Ki‐67, with no additional beneficial effects observed by DHA supplementation. Further research is necessary to confirm these findings.

What's New?

Omega‐3 docosahexaenoic acid (DHA) has shown pleiotropic anti‐cancer effects both in vitro and in animal models. However, clinical evidence of DHA efficacy is limited. This randomized, double‐blind, placebo‐controlled trial tested the efficacy of supplementation with DHA 4.4 g/day during 18 weeks of breast cancer neoadjuvant chemotherapy. Despite trends toward greater reduction in subgroup analyses, no additional effect of DHA supplementation was observed in the reduction of the proliferation marker Ki‐67 by the end of neoadjuvant chemotherapy. The rate of pathological complete response following surgery also remained unchanged. The supplementation was considered safe and well tolerated.

Abbreviations

EPA

eicosapentaenoic acid

HER2

human epidermal growth factor receptor 2

HR

hormone receptors

LCPUFAs

long‐chain polyunsaturated fatty acids

NAC

neoadjuvant chemotherapy

pCR

pathological complete response

TNBC

triple negative breast cancer

DHA

docosahexaenoic acid

1. INTRODUCTION

Breast cancer is the most prevalent and one of the leading causes of cancer mortality among women worldwide. ¹ Considered a heterogeneous disease, molecular classification is widely used to tailor treatment decisions and predict patients' prognoses. Among the most aggressive subtypes of breast cancer, the human epidermal growth factor receptor 2 positive (HER2+++) subtype accounts for 15%–20% of breast cancer diagnoses, characterized by high invasiveness and generally poor prognosis. ² While the introduction of the HER2 monoclonal antibodies trastuzumab and pertuzumab has significantly improved treatment outcomes for HER2+++ patients, 4%–23% of patients still experience disease recurrence and metastasis post‐treatment. ² Several challenges affect therapy success rates, including the limitations of the current definition of HER2+++, which does not fully account for the diversity of tumor characteristics, variations in clinical prognosis, and potential differences in treatment response. ³ It is estimated that 60%–70% of HER2+++ tumors co‐express hormone receptors (HR). ⁴ HR status impacts the response to neoadjuvant chemotherapy (NAC). Moreover, tumors lacking HR expression are associated with lower 5‐year survival rates (85.6%) compared to triple‐positive tumors (91%). ⁵ Triple negative breast cancer (TNBC) is traditionally defined by the lack of expression of the estrogen receptor (ER), progesterone receptor (PR), and HER2, and is characterized by poor prognosis, with higher disease recurrence and shorter survival rates compared to other breast cancer subtypes. ⁶ Apart from the addition of immunotherapy, there is currently no other targeted therapy for neoadjuvant TNBC, a subtype that represents 10%–15% of breast cancer diagnoses. ⁷ , ⁸ Therefore, new therapeutic strategies in both TNBC and HER2+++ are warranted.

NAC for breast cancer, administered prior to surgery, aims to reduce the risk of distant recurrence, downstage the tumor, and inform subsequent adjuvant therapy by evaluating treatment response. ⁹ As endorsed by The Canadian National Neoadjuvant Breast Cancer Consortium, the recommendation for NAC applies to all patients with locally advanced breast cancers where systemic chemotherapy is recommended, and for tumor downstaging to increase the chances of breast‐conserving surgery. ¹⁰ One important end point of NAC is the pathological complete response (pCR), achieved when there is the absence of invasive tumor cells in the breast tissue and lymph nodes. ¹¹ , ¹² Cell proliferation is a fundamental feature of tumoral cells, and changes in the expression of Ki‐67 from pre‐treatment to post‐treatment, a cell‐cycle specific nuclear antigen identified by immunohistochemistry, have been defined as an independent prognostic factor for NAC response. ¹³ , ¹⁴ , ¹⁵ However, the clinical use of Ki‐67 is limited by the absence of standardized detection methods and defined reference values. ¹⁶

The pleiotropic anticancer effects of the long‐chain polyunsaturated fatty acid (LCPUFA) docosahexaenoic acid (DHA) in vivo and in vitro have been documented by our group and others (reviewed by Newell et al., ¹⁷ ). Studies on mice implanted with TNBC patient‐derived xenografts and fed a high‐DHA diet (3.9% w/w of total fat) demonstrate DHA's synergistic effects with the chemotherapy drug docetaxel, as evidenced by reductions in tumor size, decreases in Ki‐67 expression, and increases in markers of apoptosis and necroptosis in the excised tumors. ¹⁸ , ¹⁹ While the synergetic mechanisms of DHA and cytotoxic drugs on tumor cells are not fully understood, in vitro and in vivo studies suggest that tumor cells incorporate DHA into lipid rafts microdomains, disrupting the composition of the phospholipid membrane. This change affects membrane fluidity and receptor activity, leading to cellular death through direct membrane effects or the disruption of receptor signaling, ²⁰ and are associated with cell cycle arrest, decreased proliferation, and the activation of apoptosis ¹⁸ and necroptosis pathways. ¹⁹

In addition to the n‐3 LCPUFAs' direct cytotoxic effects on tumor cells, some clinical evidence suggests that n‐3 LCPUFA supplementation can reduce chemotherapy side effects in breast cancer patients, such as the incidence of paclitaxel‐induced peripheral neuropathy. ²¹ However, clinical studies with breast cancer patients have mainly focused on the supplementation of DHA in combination with the n‐3 LCPUFA eicosapentaenoic acid (EPA). The rationale for using DHA alone in breast cancer chemotherapy is based on early findings by our group and others, which indicate that DHA has greater cytotoxicity to TNBC cell lines at the same concentrations compared to EPA or their combination. ²² , ²³ , ²⁴ Further, DHA but not EPA improved the efficacy of doxorubicin in TNBC cell lines. ²⁵ There is also some in vitro evidence showing that DHA in combination with trastuzumab promotes cellular death by the inhibition of the lipogenic genes that are required for HER2 overexpression in normal breast cell lines. ²⁶ The only clinical evidence that DHA supplementation alone may improve chemotherapy efficacy in breast cancer treatment comes from an open‐label trial (n = 25). ²⁷ This study found that DHA supplementation (1.8 g/day) during chemotherapy for breast cancer patients with visceral metastasis improved overall survival when highly incorporated into plasma phospholipids (defined as an increase ≥2.5%). ²⁷ Despite the promising results in metastatic breast cancer, these findings have not been tested in other studies, and there is not sufficient clinical evidence supporting the translation to breast cancer treatment regimens.

Therefore, the purpose of this study was to investigate whether DHA supplementation during NAC increases the efficacy of the systemic chemotherapy treatment. The primary outcome was to compare the changes in the percentage of expression from the end of treatment to the diagnosis of the proliferative markers Ki‐67 with a methodology that had been previously clinically validated. ²⁸ Secondary outcomes included the analysis of safety, adverse effects, pCR, surgery outcomes, and long‐term survival.

2. METHODS

2.1. Study design

The Docosahexaenoic acid (DHA) for Women with Breast Cancer in the Neoadjuvant Setting (DHA WIN) trial was a two‐arm parallel, double‐blind, phase II, randomised controlled trial conducted at the Cross Cancer Institute (CCI, University of Alberta, Edmonton, AB). The trial compared the supplementation of 4.4 g/day of DHA‐enriched algae triacylglycerol form (11 capsules/day, each capsule containing 400 mg of DHA, Life's DHA S40‐O400, a gift from DSM Nutritional Products, Columbia, MD) to a placebo (11 capsules/day a mixture of corn/soy oil, from DSM Nutritional Products). Given the immediate initiation of treatment following diagnosis, it was neither clinically nor ethically feasible to establish a DHA‐treatment group without NAC. To account for this, the baseline (pre‐chemotherapy) measurement was obtained for each patient. The composition of the capsules can be found in the Table S1. The composition of the placebo was designed to balance total polyunsaturated content. Linoleic acid comprised 53.5% of total fatty acids in the placebo supplement. This proportion represented 6.7% of total diet fat intake and was not expected to influence clinical outcomes during treatment. The capsules did not physically differ in characteristics between placebo and DHA, including texture or taste.

2.2. Study population

The complete description of the protocol can be found elsewhere. ²⁹ Briefly, oncologists and clinical nurses recruited newly diagnosed women with invasive breast cancer (any subtype) in clinical stages I, II, or III for whom neoadjuvant chemotherapy was recommended. Women were considered eligible if they met the following criteria: Eastern Cooperative Oncology Group performance status (ECOG) 0 or 1, hematology and biochemistry assessments within normal range; and adequate tissue specimen for the diagnosis, biomarkers and Ki‐67 assays. The exclusion criteria included: use of supplements containing omega‐3, fish oil, DHA (doses >200 mg), vitamin C, vitamin E, ß‐carotene or other antioxidant exceeding the dietary reference intakes, known allergy to components of the capsules (corn or soy), symptomatic but untreated cholelithiasis, history of any of the following diseases: deep venous thrombosis, active thrombophlebitis, pulmonary embolism, stroke, acute myocardial infarction, congestive cardiac failure, untreated hypertension or known inherited hypercoagulable disorder, diagnosis of any other malignancy within the previous year except for adequately treated basal cell or squamous cell skin cancer, medically documented history of a psychiatric disorder that would preclude consent, and partial or complete loss of vision or diplopia, from ophthalmic vascular disease. The histological classification of breast cancer subtypes was based on physician's decision at the screening following the St Gallen breast Cancer Conference. ³⁰ Tumors were classified as luminal A (ER+ and/or PR+, Ki67 low [<20%] and HER2−), luminal B (ER+ and/or PR+, Ki67 high [>20%] and HER2−), HER2‐positive (HER2+, with or without ER+, and/or PR+) and TNBC (ER−, PR−, and HER2−). ³¹ , ³² In this study, all participants with HER2 positive status were classified as the HER2+++ subtype. The classifications of tumor size, stage, and lymph node status were defined according to the American Joint Committee on Cancer. ³³ , ³⁴ Briefly, tumor size (T) was classified as follows: “T1, tumor ≤20 mm in greatest dimension; T2, tumor >20 mm but ≤50 mm in greatest dimension; T3, tumor >50 mm in greatest dimension; T4, tumor of any size with direct extension to the chest wall and/or skin”. ³³ , ³⁴ Lymph node status (N) was classified as follows: “N0, no regional lymph node metastases; N1, metastases to movable ipsilateral level I or II axillary lymph node(s); N2, metastases in ipsilateral level I, II axillary lymph nodes that are clinically fixed or matted; N3, metastases in ipsilateral infraclavicular (level III axillary) lymph node(s) with or without level I or II axillary lymph node involvement”. ³³ , ³⁴ The disease or anatomic stage was determined based on the combination of T, N, and distant metastases status. Stage IV was assigned when distant detectable metastases were diagnosed through clinical and radiographic assessment.

2.3. Randomization, blinding and intervention

Randomization was performed by a biostatistician, who generated a patient randomization list and randomized bottle numbers using covariate‐adaptive randomization (block randomization). The randomized bottle numbers were provided to DSM for labeling for both the DHA and placebo groups, as well as to the unblinded Clinical Trials Coordinator (CTC, Clinical Trials Unit) and pharmacist. Unique study identifiers and information from participants were entered by the study coordinator into the REDCap database, an electronic data capture tool hosted by the Women & Children's Health Research Institute at the University of Alberta. ³⁵ The key to the study arm was kept in password‐protected computers and were not shared with blinded patients, pathologists, physicians, or researchers. Blinding was only removed after the completion of the final study participant. The study participants received standard‐of‐care systemic chemotherapy, administered every 3 weeks for a total of six cycles. This resulted in a complete chemotherapy regimen lasting 18 weeks. The chemotherapy regimens were developed in a guideline‐coordinated system by a single team residing at the Cross Cancer Institute. Every cycle of chemotherapy was composed of one of two different regimens depending on the HER2 status. For HER2‐negative disease, patients received the FEC‐D regimen (fluorouracil 500 mg/m² on day 1, epirubicin 100 mg/m² on day 1, cyclophosphamide 500 mg/m² on day 1, repeated for 3 cycles and followed by 3 consecutive cycles of docetaxel 100 mg/m²), while HER2‐positive patients received the TCH regimen (docetaxel 75 mg/m² IV day 1, carboplatin AUC 6 IV day 1, repeated for 6 cycles with trastuzumab 8 mg/kg IV week 1, and followed by 6 mg/kg IV in the other cycles). Participants received the supplementation from the start of the first cycle of chemotherapy until surgery (21–35 days after the last administration of cytotoxic chemotherapy).

2.4. Primary outcome: Immunohistochemistry analysis of Ki‐67

The primary outcome of this study was to determine the changes in Ki‐67 from core needle biopsy to surgical excision. Ki‐67 quantification was performed by immunohistochemistry (IHC) analysis followed by image analysis at the Diagnostic Biomarker Laboratory at the Cross Cancer Institute, using a method previously described by Torlakovic et al. ²⁸ The protocol followed was based on the MonarchE clinical trial, ³⁶ which was a pioneering effort in establishing Ki‐67 as a Class II biomarker assay. This was achieved through the development of an optimized assay, as described by Polewski et al. ³⁷ Briefly, coded unstained slides from formalin‐fixed/paraffin‐embedded needle core biopsy surgical specimens were received by the pathologist, and all tumoral material was fixed in 10% neutral buffered formalin, adhering to the American Society of Clinical Oncology‐College of American Pathologists (ASCO‐CAP) guidelines. Ki‐67 testing was performed using the Dako Ki‐67 IHC MIB‐1 clone. The stained slides were digitized using an Aperio GT‐450 scanner at 40x magnification and analyzed with QuPath (Version 4.3), ³⁸ an open‐source software application for bioimage analysis. Any definitive brown nuclear staining was used as the threshold to differentiate positive from negative nuclei, corresponding to a 1+ staining intensity. For the Ki‐67 IHC assay, the unitary optical density vectors for the constituent stains—hematoxylin and 3,3′‐Diaminobenzidine (DAB)—were calculated using QuPath's “Estimate stain vectors” functionality. ³⁹ Cellular segmentation was performed using the StarDist plugin ⁴⁰ for QuPath. The lowest level of detection, defined as the minimal observable brown DAB staining, was visually identified, and the corresponding numerical optical density threshold was consistently applied to all Ki‐67‐stained samples. Tumoral and non‐tumoral areas were annotated on each sample and used to train a Random Forest classifier. This classifier was then applied across all samples. The features used to train the classifier included staining intensity, nuclear area, perimeter, circularity, and other morphometric parameters. The final quantification was reported as the percentage of positive cells in tumoral area. To explore additional potential prognostic factors, the tumor‐stroma ratio (TSR) was calculated using the same slides from tumor biopsies taken before NAC treatment by dividing the number of tumor cells by the number of stromal cells. The distinction between tumor and non‐tumor components is made possible by the QuPath random forest machine learning algorithm. This algorithm is trained by a pathologist who annotates representative tumor and stromal cells. Training is performed in real‐time, with annotations being added until the overall classification meets the trainer's satisfaction. ⁴¹

Of the 49 patients who completed the trial (Figure 1), tumor samples post‐NAC were not available for seven participants (n = 4 in the placebo group and n = 3 in the DHA group) due to their apparent successful response to therapy. For the subsequent analysis of this study, these patients were considered to have %Ki‐67 expression equal to zero. Additionally, four participants (n = 1 in the placebo group and n = 3 in the DHA group) did not have %Ki‐67 data available from the pre‐treatment biopsy. For these missing samples, single imputations were performed using the mean value of all participants with the same subtype of breast cancer. Analyses were also conducted without these subjects, and this did not significantly change the results.

FIGURE 1.

CONSORT flow diagram. Adverse events included (n = 13): Nausea (n = 9), severe pain (n = 2), elevation of liver enzymes (n = 1), and critically low hemoglobin (n = 1). Other reasons for discontinuation included (n = 12): Disease progression; early surgery requested (n = 2); metastatic disease (n = 4); refusal to take the study intervention or difficulty cooperating with the study (n = 4); gallbladder disease followed by physician's decision to remove the patient from the study (n = 1); and dose‐dense chemotherapy (n = 1).

2.5. Secondary outcomes: safety, efficacy, and survival assessments

The type of surgery, volume of blood loss, pCR, adverse events, and survival assessments were documented by the clinical team and made available in REDCap. pCR was assessed in resected breast tissue and all sampled axillary nodes and was defined as the absence of residual invasive cancer (ypT0/Tis ypN0), evaluated as part of standard‐of‐care using hematoxylin and eosin (H&E) staining. The disease‐free survival and overall survival were assessed approximately 3 years after the randomization and were calculated by counting the number of months from randomization to the diagnosis of any type of cancer or death. The adverse events were graded according to the International Common Terminology Criteria for Adverse Events V.5.0 for adverse event reporting and were recorded at the time of the participant's first dose of the treatment until the end of the study, including 28 calendar days after the last administration of the study agent.

2.6. Dietary assessment

The food frequency questionnaire (FFQ) was completed by participants at the baseline, before the initiation of the first cycle of chemotherapy, to assess dietary intake as a potential confounding factor. The FFQ used was the Canadian Diet History Questionnaire II (C‐DHQ II), a self‐administered and semi‐quantitative questionnaire comprised of 165 questions on foods consumed over the past year and includes questions about portion size. ⁴² Each item on the C‐DHQ II corresponds to a nutrient profile derived from the Canadian Community Health Survey nutrient database, which includes a list of 33 nutrients and is used to estimate participants' daily nutrient intake. ⁴³ The data from paper copies were manually entered into the C‐DHQ II website, ⁴⁴ and data were analyzed using Diet*Calc software. ⁴⁵

2.7. Compliance and fatty acid profile

Compliance was evaluated by the clinical research team at the CCI by counting the number of pills returned during each chemotherapy cycle visit and comparing it to the number of pills dispensed. Overall compliance was calculated as the average compliance across all chemotherapy cycles.

Fatty acid composition was assessed to evaluate compliance and to control for potential confounding factors. Venous blood was collected at the beginning of each cycle of chemotherapy. Plasma and erythrocytes were isolated by centrifugation (3000×g, 10 min; 4°C), frozen, and stored at −80°C until analysis. To determine the fatty acid composition, plasma phospholipids and erythrocyte total lipids were extracted by a modified Folch procedure, previously described. ⁴⁶ , ⁴⁷ Total lipids were extracted from the erythrocyte samples, and plasma phospholipids were separated from other major lipid classes by thin‐layer chromatography. Plasma phospholipids and total lipids extracted from red blood cells were methylated with boron trifluoride and hexane at 100°C and fatty acids were separated and quantified by automated GLC 7890A (Agilent Technologies) on a CP‐Sil 88 column (100 m × 0.25 mm; Agilent). ⁴⁸ Fatty acids are reported as relative percentages of total fatty acids identified and validated against fatty acid standards (GLC‐502 and GLC‐643) from NuChek Prep, Inc.

2.8. Sample size and statistical analysis

The sample size calculation was based on the primary outcome, aimed to determine the efficacy of DHA supplementation during NAC by change in the % Ki‐67 from biopsy to surgical excision. ²⁹ Group sample sizes of 23 patients in each group were required to achieve 81% power to detect a difference between the group proportions of 0.4 (effect size). The proportion in group 1 was assumed to be 0.3 under the null hypothesis and 0.7 under the alternate hypothesis. The proportion of group 2 was assumed to be 0.3 under the alternate hypothesis. The test statistic used was the two‐sided t‐test, targeting a significance level of 0.05.

Normality was assessed for each variable, and parametric or nonparametric statistical tests were applied accordingly. Continuous variables were reported as mean ± SD; frequency and proportions were reported for categorical variables. Changes in the % Ki‐67 from pre‐ to post‐intervention were calculated for each participant, and within‐group and between‐group differences were assessed by paired and unpaired t‐test, respectively. Categorical variables were analyzed using the Chi‐squared test or Fisher's exact test (for cell frequency <5%). Univariate binary logistic regression analysis was performed to evaluate the association of predictor variables for pCR (“yes” or “no”). Odds ratio (OR) and the corresponding 95% confidence intervals were reported. Kaplan–Meier estimates and the corresponding 95% confidence intervals were reported for overall survival and disease‐free analysis. SPSS (IBM Corp. IBM SPSS Statistics, Version 28.0. Armonk, NY) was used to perform all statistical analyses, and GraphPad Prism (GraphPad Software Version 10.0.3, Boston, MA) was used to generate the graphics. A p‐value <0.05 (two‐tailed) was used for statistical significance.

3. RESULTS

3.1. Participant characteristics, tolerability, incidence of adverse effects, and surgery outcomes

The sociodemographic and clinicopathological characteristics of participants are described in Table 1. The primary reasons for discontinuation of the trial were mostly due to adverse effects from chemotherapy, such as nausea and dyspnea, or physician decisions due to disease progression or metastasis (Figure 1). None of the adverse effects were determined to be associated with the treatment. The first and last participants were randomised on September 27, 2019, and on May 31, 2022, respectively. The last participant completed the trial in December 2022, resulting in a total of 49 participants who completed the intervention period (Figure 1). Forty‐three patients remain in the survival follow‐up, and the last patient survival follow‐up is expected in December 2032. The participants had a mean (SD) age of 50.8 (10.7) years, with the majority identifying as Caucasian (65.3%) (Table 1). At baseline, 73.4% of participants were classified as either overweight (36.7%) or obese (36.7%), and 51.9% were premenopausal. Among the different subtypes of breast cancer, 51% were classified as HER2+++, 24.5% as TNBC, 20.4% luminal A, and 4.1% luminal B (Table 1). Despite the high discontinuation rate (~36%), the characteristics of the participants who completed the trial did not differ significantly between the two intervention arms (Table 1). Of the participants who completed the trial, none were diagnosed with COVID‐19 from enrollment to the end of NAC.

TABLE 1.

Sociodemographic and clinicopathological characteristics of participants in the placebo and DHA groups.

	Total (n = 49)	Placebo (n = 26)	DHA (n = 23)	p value
Age (years) a	50.8 ± 10.7	51.2 ± 12.0	50.4 ± 9.3	0.80
BMI [kg/m2] a	28.8 ± 6.7	27.5 ± 6.0	30.3 ± 7.3
Underweight (<18.5) b	1 (2.0)	1 (3.8)	0 (0.0)	0.60
Healthy weight (18.5–24.9) b	13 (26.5)	8 (30.8)	5 (21.7)
Overweight (25–29.9) b	18 (36.7)	10 (38.5)	8 (34.8)
Obese (≥30) b	17 (34.7)	7 (26.9)	10 (43.5)
Ethnicity [n (%)] b
Not Hispanic or Latino	32 (65.3)	17 (65.4)	15 (65.2)	0.89
Asian	8 (16.3)	5 (19.2)	3 (13.0)
Black or African American	4 (8.2)	2 (7.7)	2 (8.7)
American Indian or Alaska Native	5 (10.2)	2 (7.7)	3 (13.0)
Menopausal status b
No	25 (51.0)	13 (50.0)	12 (52.2)	0.31
Yes	22 (44.9)	13 (50.0)	9 (39.1)
Missing	2 (4.1)	0 (0.0)	2 (8.7)
Diabetes b
No	47 (95.9)	26 (100.0)	21 (91.3)	0.22
Yes	2 (4.1)	0 (0.0)	2 (8.7)
Ethanol abuse b
No	48 (98.0)	26 (100.0)	22 (95.7)	0.47
Yes	1 (2.0)	0 (0.0)	1 (4.3)
Smokers b
No	41 (83.7)	21 (80.8)	20 (87.0)	0.42
Yes	8 (16.3)	5 (19.2)	3 (13.0)
Use of recreational drugs b
No	44 (89.8)	24 (92.3)	20 (87.0)	0.44
Yes	5 (10.2)	2 (7.7)	3 (13.0)
ECOG performance status b , c
Baseline
0	44 (89.8)	24 (92.3)	20 (87.0)	0.05
1	2 (4.1)	2 (7.7)	0 (0.0)
Missing	3 (6.1)	0 (0.0)	3 (13.0)
End of treatment
0	34 (69.4)	21 (80.8)	13 (56.5)	0.36
1	11 (22.4)	3 (11.5)	8 (34.8)
2	2 (4.1)	1 (3.8)	1 (4.3)
Missing	2 (4.1)	1 (3.8)	1 (4.3)
Histology
HER2+++	24 (51.0)	13 (50.0)	11 (47.8)	0.87
TNBC	13 (24.5)	6 (23.1)	7 (30.4)
Luminal A	10 (20.4)	6 (23.1)	4 (17.4)
Luminal B	2 (4.1)	1 (3.8)	1 (4.3)
Estrogen receptor status
Positive	26 (53.1)	16 (61.5)	10 (43.5)	0.21
Negative	23 (46.9)	10 (38.5)	13 (56.5)
Progesterone receptor status
Positive	17 (34.7)	11 (42.3)	6 (26.1)	0.24
Negative	31 (63.3)	14 (53.8)	17 (73.9)
Missing	1 (2.0)	1 (3.8)	0 (0.0)
HER2 Status
Positive	24 (49.0)	13 (50.0)	11 (47.8)	0.99
Negative	25 (51.0)	13 (50.0)	12 (52.2)
Disease stage
IIA	13 (26.5)	6 (23.1)	7 (30.4)	0.63
IIB	10 (20.4)	4 (15.4)	6 (26.1)
IIIA	14 (28.6)	9 (34.6)	5 (21.7)
IIIB	4 (8.2)	2 (7.7)	2 (8.7)
IIIC	1 (2.0)	0 (0.0)	1 (4.3)
Missing	7 (14.3)	5 (19.2)	2 (8.7)
Tumor size
T1	1 (2.0)	1 (3.8)	0 (0.0)	0.92
T2	27 (55.1)	15 (57.7)	12 (52.2)
T3	11 (22.4)	5 (19.2)	6 (26.1)
T4	5 (10.2)	3 (11.5)	2 (8.7)
Missing	5 (10.2)	2 (7.7)	3 (13.0)
Axillary node status
N0	12 (24.5)	6 (23.1)	6 (26.1)	0.99
N1	21 (42.9)	11 (42.3)	10 (43.5)
N2	5 (10.2)	3 (11.5)	2 (8.7)
N3	2 (4.1)	1 (3.8)	1 (4.3)
Missing	9 (18.4)	5 (19.2)	4 (17.4)
%Ki‐67 positive cells a	48.9 (24.0)	41.8 (24.7)	57.8 (20.5)	0.03
Low Ki‐67 (<20%), [n (%)]	6 (12.2)	6 (23.1)	0	0.03
High Ki‐67 (≥20%), [n (%)]	37 (75.5)	18 (69.2)	19 (82.6)
Missing, [n (%)]	6 (12.2)	2 (7.7)	4 (17.4)
Tumor stroma ratio a	0.92 (0.94)	1.02 (1.16)	0.81 (0.58)	0.940

^a Mean ± SD.

^b Count (% of total or given group). Pearson Chi‐Square test was used to compare categorical variables (exact test if cell count <5) and independent t‐test was used to compare continuous variables (mean ± SD).

^c Grades were defined as 0: Fully active, able to carry on all pre‐disease performance without restriction; 1: restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature (Oken et al., ⁶¹ ). BMI, body mass index; ECOG, Eastern Cooperative Oncology Group.

There were no significant differences between the groups at baseline in the estimated daily intake of energy, total carbohydrates, total protein, total fat, fatty acid groups (monounsaturated, polyunsaturated, and saturated), dietary fiber, cholesterol, sodium, or salt as obtained from the C‐DHQ II (Table S2).

The supplementation was well tolerated by participants. The mean overall trial compliance was 81.6% (SD, 21.7%), with no significant difference between groups: 81.0% in the placebo group and 82.4% in the DHA group (p = 0.970, Mann–Whitney test).

The incidence of adverse effects for participants who completed the trial and the type of surgery received by each individual are described in Tables S3 and S4, respectively. None of the adverse effects were directly attributed to the supplement, which was determined by the oncologists and nurses. Overall, there were no significant differences in the incidence, type, or severity of adverse effects reported in the placebo or DHA group.

The type of surgery received by each participant did not significantly differ between groups. Mastectomy was the most prevalent surgery (46.9%), followed by lumpectomy (40.8%), sentinel node dissection (6.12%), and full axillary dissection (4.08%) (Table S4). The volume of blood loss at the time of surgical removal after neoadjuvant chemotherapy was <50 mL for all participants and did not differ between arms of interventions (data not shown).

3.2. Primary outcome: changes in the expression of Ki‐67

There was a significant reduction in the expression of %Ki‐67 post‐NAC compared to pre‐treatment values in both treatment groups (p < 0.001, Figure 2A,B). Furthermore, there were no significant differences in the changes between the groups (Figure 2C). However, when stratified by breast cancer subtypes as determined by histopathological classification, there was a trend toward a greater reduction in the HER2+++ DHA group compared to the placebo group (p = 0.1, Figure 2D). Due to the small sample size for the other breast cancer subtypes, no significant differences were observed when comparing treatment groups (data not shown).

FIGURE 2.

Changes in the expression of the proliferation marker Ki‐67. (A) Changes from pre‐ to post‐intervention considering all participants with complete data. (B) Comparison of the change in Ki‐67 expression between pre‐ and post‐intervention within each subgroup. (C) Comparison of the change in Ki‐67 expression from pre‐ to post‐intervention between the groups. (D) Comparison of the change in Ki‐67 expression between pre‐ and post‐intervention within each subgroup in participants classified as HER2+++ subtype as classified by histopathological analysis. (B–D) Bars represent mean and 95% confidence interval. HER2+++, human epidermal growth receptor positive; NAC, neoadjuvant chemotherapy.

3.3. Fatty acids status

The fatty acid composition of erythrocytes is detailed in Table 2. The DHA‐supplemented group had a two‐fold increase in the relative % of DHA at 9 and 15 weeks of therapy, while it did not significantly change in the placebo (p‐interaction <0.001, Table 2). The n‐6: n‐3 ratio decreased by approximately 50% at 15 weeks compared to the beginning of treatment in the DHA group (Table 2). Overall, the fatty acid profile of erythrocytes in the placebo group remained constant during the treatment, except for a slight increase in linoleic acid and dihomo‐γ‐linolenic acid (DGLA) at 9 weeks, which returned to baseline levels by the end of the study, and a decrease in EPA at 15 weeks (Table 2). In contrast, DHA supplementation led to significant changes in the erythrocyte fatty acid composition, including an increase in the relative percentage of total SFA (p‐interaction = 0.009), DPA n‐6 (P‐interaction <0.001), total n‐3 (P‐interaction <0.001), and EPA (p‐interaction <0.001) (Table 2). Furthermore, the DHA group experienced a reduction in the percentage of total n‐6 (p‐interaction <0.001), linoleic acid (p‐interaction <0.001), DGLA (P‐interaction <0.001), and DPA n‐3 (p‐interaction = 0.004) (Table 2). There are no significant changes in the composition of ARA between groups of intervention (p‐interaction = 0.154). Consistent with the erythrocyte lipid changes, the fatty acid composition of plasma phospholipids, collected every 3 weeks, indicated a two‐fold increase in DHA from week 3 remaining constant through the 15 weeks of treatment (Figure S1).

TABLE 2.

Total lipids composition (relative % of total fatty acids) of red blood cells.

	Placebo (mean ± SD)			DHA (mean ± SD)			Group × Time	Group	Time
	0 weeks (n = 25)	9 weeks (n = 24)	15 weeks (n = 25)	0 weeks (n = 22)	9 weeks (n = 23)	15 weeks (n = 22)	p value6	p value7	p value8
Total SFA 1	50.8 ± 2.6	50.2 ± 2.5	50.2 ± 2.5	51.5 ± 2.6 a	53.2 ± 3.9 b *	52.6 ± 3.8 a *	0.009	0.004	0.444
14:0 (Myristic acid)	0.34 ± 0.1	0.35 ± 0.09	0.36 ± 0.09	0.44 ± 0.33 ab	0.42 ± 0.09 a *	0.36 ± 0.11 b	0.008	0.094	0.285
15:0 (Pentadecanoic acid)	0.42 ± 0.26	0.35 ± 0.2	0.31 ± 0.12	0.43 ± 0.3	0.39 ± 0.23	0.31 ± 0.1	0.762	0.645	0.017
16:0 (Palmitic acid)	27.9 ± 2.2	28.4 ± 1.8	28.3 ± 1.5	28.5 ± 1.5	29.8 ± 1.9	29.8 ± 1.7	0.411	0.003	0.031
17:0 (Heptadecanoic acid)	0.47 ± 0.12	0.49 ± 0.08	0.51 ± 0.11	0.48 ± 0.14	0.54 ± 0.1	0.52 ± 0.1	0.611	0.216	0.012
18:0 (Stearic acid)	16.6 ± 1.4	16.1 ± 1.5	16.4 ± 1.5	16.9 ± 1.5	17.2 ± 1.9	17.3 ± 2.1	0.139	0.050	0.779
20:0 (Arachidic Acid)	0.49 ± 0.09	0.46 ± 0.12	0.43 ± 0.13	0.48 ± 0.1	0.49 ± 0.08	0.40 ± 0.09	0.090	0.735	<0.001
24:0 (Lignoceric acid)	4.61 ± 0.63 a	4.09 ± 0.62 b	3.94 ± 0.72 b	4.33 ± 0.63 a	4.38 ± 0.81 a	3.95 ± 0.75 b	0.005	0.986	<0.001
Total MUFA 2	19.9 ± 3.1	21.3 ± 2.1	21.3 ± 2.3	20.1 ± 2.4	21.6 ± 2.0	21.2 ± 2.8	0.868	0.834	0.001
16:1 n‐7 (Palmitoleic acid)	0.36 ± 0.13	0.42 ± 0.18	0.41 ± 0.14	0.42 ± 0.14	0.37 ± 0.1	0.37 ± 0.11	0.005	0.648	0.702
18:1 c11 (Vaccenic acid)	0.37 ± 0.19	0.43 ± 0.24	0.36 ± 0.18	0.42 ± 0.24	0.49 ± 0.21	0.38 ± 0.18	0.768	0.265	0.012
18:1 n‐9 (Oleic acid)	14.2 ± 1.5	14.7 ± 1.0	14.5 ± 1.0	14.1 ± 1.2	14.6 ± 0.9	14.5 ± 1.4	0.759	0.917	0.003
18:1 n‐7 (Octadecenoic acid)	1.32 ± 0.25	1.31 ± 0.17	1.26 ± 0.26	1.39 ± 0.25	1.28 ± 0.23	1.38 ± 0.22	0.118	0.322	0.281
24:1 n‐9 (Nervonic acid)	3.69 ± 1.88	4.38 ± 1.52	4.79 ± 1.69	3.79 ± 1.37	4.87 ± 1.42	4.49 ± 1.89	0.307	0.635	<0.001
Total PUFA 3	29.3 ± 4.1	28.5 ± 3.2	28.5 ± 3.6	28.4 ± 4.0	25.2 ± 4.2	26.2 ± 4.9	0.085	0.019	0.001
Total n‐6 4	24.6 ± 3.9	24.1 ± 2.7	24.0 ± 2.9	24.4 ± 3.4 a	19.8 ± 3.1 b *	19.8 ± 3.4 b *	<0.001	<0.001	<0.001
18:2 n‐6 (Linoleic acid)	9.14 ± 1.48 ab	9.85 ± 1.25 a	9.34 ± 1.06 b	9.05 ± 1.13 a	7.15 ± 1.25 b *	7.01 ± 1.32 b *	<0.001	<0.001	<0.001
20:2 n‐6 (Eicosadienoic acid)	0.26 ± 0.12	0.29 ± 0.18	0.29 ± 0.12	0.29 ± 0.1	0.21 ± 0.07	0.22 ± 0.08	0.829	0.026	0.291
20:3 n‐6 (DGLA)	2.54 ± 0.36 a	2.64 ± 0.45 ab	2.7 ± 0.39 b	2.76 ± 0.37 a *	2.5 ± 0.4 b	2.31 ± 0.27 b	<0.001	0.587	0.286
20:4 n‐6 (ARA)	9.2 ± 1.86	8.62 ± 1.56	8.81 ± 1.66	8.95 ± 1.91	7.57 ± 1.65	7.67 ± 1.76	0.154	0.034	<0.001
22:4 n‐6 (Adrenic acid)	3.15 ± 1.61	2.39 ± 1.21	2.65 ± 1.37	3.08 ± 1.79	1.58 ± 1.15	1.81 ± 1.57	0.425	0.035	<0.001
22:5 n‐6 (DPA n‐6)	0.28 ± 0.2	0.28 ± 0.22	0.22 ± 0.05	0.27 ± 0.11 a	0.75 ± 0.26 b *	0.92 ± 0.39 c *	<0.001	<0.001	<0.001
Total n‐3 5	4.71 ± 1.54	4.45 ± 1.24	4.51 ± 1.39	4.00 ± 1.02 a	5.44 ± 1.64 b *	6.26 ± 2.14 b *	<0.001	0.038	<0.001
18:3 n‐3 (ALA)	0.44 ± 0.23	0.52 ± 0.32	0.46 ± 0.15	0.43 ± 0.09 a	0.38 ± 0.08 b *	0.38 ± 0.11 b	0.044	0.107	0.572
20:4 n‐3 (ETA)	0.22 ± 0.07	0.21 ± 0.06	0.21 ± 0.07	0.2 ± 0.06	0.23 ± 0.07	0.21 ± 0.09	0.555	0.777	0.252
20:5 n‐3 (EPA)	0.54 ± 0.43 a	0.44 ± 0.22 a	0.38 ± 0.16 b	0.37 ± 0.15 a	0.62 ± 0.28 b *	0.62 ± 0.26 b *	<0.001	0.161	0.102
22:5 n‐3 (DPA n‐3)	1.12 ± 0.43	0.94 ± 0.45	1.25 ± 0.51	0.98 ± 0.42 a	0.78 ± 0.25 b *	0.79 ± 0.28 ab *	0.004	<0.001	0.422
22:6 n‐3 (DHA)	2.39 ± 1.17 a	2.09 ± 0.87 b	2.22 ± 1.01 ab	2.01 ± 0.75 a	3.43 ± 1.34 b *	4.27 ± 1.99 b *	<0.001	<0.001	<0.001
Total n‐6: Total n‐3	5.76 ± 1.97	5.78 ± 1.57	5.67 ± 1.44	6.38 ± 1.36 a	3.89 ± 1.08 b *	3.43 ± 1.03 b *	<0.001	<0.001	<0.001

Note: Significant differences are in bold. Values represent the mean ± SD. Data were analyzed using Generalized Estimating Equations (GEE) to assess the interaction of DHA supplementation and changes over time. ^a,bLabelled means without a common letter differ (p < 0.05) based on post hoc with Bonferroni adjustment comparing changes within each group. *Indicates significant difference (p < 0.05) comparing placebo and DHA groups at the same time point, based on post hoc with Bonferroni adjustment. ¹SFA = 14:0, 15:0, 16:0, 17:0, 18:0, 20:0, 24:0. ²MUFA = 16:1 n‐7, 18:1 c11, 18:1 n‐9, 18:1 n‐7, 24:1 n‐9. ³Total PUFA = 18:2 n‐6, 20:2 n‐6, 20:3 n‐6, 20:4 n‐6, 22:4 n‐6, 22:5 n‐6, 18:3 n‐3, 20:4 n‐3, 20:5 n‐3, 22:5 n‐3, 22:6 n‐3. ⁴Total n‐6 = 18:2 n‐6, 20:2 n‐6, 20:3 n‐6, 20:4 n‐6, 22:4 n‐6, 22:5 n‐6. ⁵Total n‐3 = 18:3 n‐3, 20:4 n‐3, 20:5 n‐3, 22:5 n‐3, 22:6 n‐3. ⁶P value for the interaction between supplementation and time in the GEE. ⁷ p value for the main effect of time (0, 9 and 15 weeks) in the GEE. ⁸ p value for the main effect of the group (DHA or placebo) in the GEE.

Abbreviations: ALA, alpha‐linolenic acid; ARA, arachidonic acid; DGLA, dihomo‐γ‐linolenic acid; DHA, docosahexaenoic acid; DPA, docosapentaenoic acid; EPA, eicosapentaenoic acid; ETA, eicosatetraenoic acid; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids.

[Correction added on 23 July 2025, after first online publication: The values in the following rows have been revised: Total MUFA, Total PUFA, Total n‐3, Total n‐6: Total n‐3].

3.4. Pathological complete response

The percentage of participants who achieved pCR at the end of the trial did not differ between the placebo and DHA groups (26.9% vs. 43.4%, p = 0.220, Table S5). Among the 49 participants, 17 (34.7%) achieved pCR, with rates of 45.8% (11 out of 24) and 38.5% (5 out of 13) for participants classified as HER2+++ and TNBC subtypes, respectively. Irrespective of treatment arm, estrogen receptor‐positive status (ER status) was associated with lower chances of achieving pCR (ß = −2.07, OR (95% CI) = 0.13 (0.03, 0.49), p‐value = 0.003) (Figure 3). Additionally, changes in the %Ki‐67 expression were negatively associated with pCR (ß = −0.04; OR = 0.96; 95% CI: 0.93, 0.99), meaning that more negative changes or greater reductions were associated with higher chances of achieving pCR (Figure 3). There were no significant associations with the achievement of pCR and the other clinico‐histopathological characteristics evaluated (Table S6).

FIGURE 3.

Forest plot of univariate analysis of associations between different predictors and pathological complete response (pCR) irrespective of treatment arm. The continuous variables were age, BMI, change from baseline to end of chemotherapy (delta) %Ki‐67, TSR, while categorical variables included ER status (negative vs. positive), PR status (negative vs. positive), HER2 status (negative vs. positive), menopausal status (negative vs. positive), and disease stage (low vs. high). BMI, body mass index; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; TSR, tumor‐stroma ratio.

3.5. Long‐term follow‐up at 3‐years

Out of the 49 participants that completed the trial, 42 participants had reached the 36‐month mark since randomization. Overall, there were no significant differences between groups of intervention in disease‐free survival or overall survival (Figure 4). The number of participants who experienced disease recurrence did not significantly differ between groups (n = 4 in the placebo and n = 6 in the DHA group, p = 0.380). The most common sites of recurrence were the breast, abdomen, brain, and lung. Among the participants who experienced disease recurrence, the number of deaths also did not significantly differ between groups (n = 2 in the placebo and n = 4 in the DHA group, p = 0.353). All causes of death were attributed to metastases.

FIGURE 4.

Survival analysis at approximately 3 years after randomization (n = 21 in each group). Disease‐free survival (A) and overall survival (B) in months were analyzed using the Kaplan–Meier method. The mean survival was undetermined due to the small incidence of events (disease recurrence or death).

4. DISCUSSION

In this randomized clinical trial, supplementation with 4.4 g/day of DHA during 18 weeks of neoadjuvant chemotherapy in women with breast cancer (stages II–III) did not significantly change the tumor proliferation marker Ki‐67, nor did it improve rates of pCR or the incidence of adverse events and surgery outcomes when compared to a placebo. However, there was a trend (p = 0.1) of greater reduction in the % of Ki‐67 in the DHA group compared to the placebo within participants with the HER2+++ subtype. The supplementation was considered safe and well tolerated. Compliance was confirmed by the number of capsules returned at each cycle of chemotherapy and by the two‐fold increase in the relative % of DHA in erythrocytes total lipids and plasma phospholipids. This incorporation was found to reach its maximum by 3 weeks of supplementation. The dose 4.4 g/day of DHA was chosen because it was previously shown to be safe in other patient groups. ⁴⁹ , ⁵⁰ , ⁵¹ This dose as a proportion of dietary fat is similar to the level fed in our preclinical studies (3.9% w/w fat, ~4 g/day). ¹⁸ , ¹⁹

There is limited research on interventional studies with DHA in breast cancer or other types of cancer during chemotherapy treatment. In line with our findings, supplementation with 1 g/ day with DHA (in an enriched algae triacylglycerol form) over a period of 12 weeks to overweight/obese women (n = 32) with a history of breast cancer was well tolerated and significantly increased the % of DHA in erythrocytes at the end of treatment compared to baseline values (p < 0.001). ⁵² However, the study did not evaluate DHA levels at other time points. The dose‐dependent effect of n‐3 LCPUFAs and the safety of supplementation were previously demonstrated in an open‐label study. ⁵³ Doses ranging from 0.84 g to 7.56 g DHA + EPA were found to be well tolerated and resulted in a significant increase in serum DHA concentration, achieving maximum incorporation at two‐fold with a 5.04 g dose after 1 month, compared to concentrations observed after 6 months of supplementation in women at high risk of breast cancer (definition of high risk was not provided). ⁵³ However, these findings were not tested during NAC. Our study is the first to demonstrate the feasibility of administering 4.4 g/day (11 capsules) of DHA during NAC and to show that plasma phospholipid DHA levels increased two‐fold, reaching their maximum incorporation after 3 weeks of therapy.

In contrast to our findings, supplementation in patients with stage III breast cancer over three cycles of NAC (51 days) with fish oil (1 g/day, n = 24) significantly decreased Ki‐67 expression compared to a placebo (composition unknown; 42.4 ± 4.8 vs. 39.2 ± 5.3, p = 0.032). ⁵⁴ Consistent with our findings, a study by de la Rosa Oliva et al., ⁵⁵ found no significant differences in the incidence of adverse events following supplementation with 2.4 g/day of n‐3 PUFA (n = 26) for approximately 20–24 weeks of NAC, compared to a placebo containing sunflower oil (n = 23). However, this study observed a decrease in the xerostomia scale in the n‐3 PUFA group compared to the placebo, as measured by the Edmonton Symptom Assessment Scale. The discrepancies between our findings and those of these studies could be attributed to multiple factors. First, differences in patient populations, as the majority of participants in the study by Darwito et al., ⁵⁴ (45 out of 48) were HER2‐negative, while in the study by de la Rosa Oliva et al., ⁵⁵ the majority (29 out of 52) were classified as luminal B. Additionally, variations in Ki‐67 detection methodology, supplementation, dose, duration, type (fish oil), and the composition of the placebo may have also contributed to results. While our study did not observe a significant effect of supplementation on Ki‐67 expression, pCR rates, or the incidence of adverse events, our findings suggest a potential effect when groups were stratified by HER2‐positive status.

The rapid incorporation of DHA into tumor cell membranes is a key component of its cytotoxic mechanism and its ability to enhance the efficacy of chemotherapy drugs. This change affects membrane fluidity and receptor activity, leading to cellular death through direct membrane effects or the disruption of receptor signaling. ²⁰ These cytotoxic synergistic effects appear to be particularly selective to transformed and more aggressive cell lines, such as those derived from TNBC. In vitro studies show lower or no efficacy of DHA treatment in cell lines derived from luminal A subtypes, with no significant toxicity to non‐malignant cells. ⁵⁶ , ⁵⁷ Our study was not designed to specifically assess the effects of DHA supplementation across different breast cancer subtypes. Therefore, it remains unclear whether DHA supplementation may be more effective in a particular breast cancer subtype or a more homogeneous study population is needed to evaluate its efficacy. The DHA‐WIN trial was designed to evaluate secondary outcomes not reported here, including quality of life, effects on systemic immune function, and long‐term follow‐up at 5 and 7 years. ²⁹ These future analyses will provide a more comprehensive understanding of the impact of DHA supplementation during NAC.

Strengths of our study include its novelty. This is the first double‐blind RCT to test the efficacy of supplementing DHA in an algae‐enriched triacylglycerol form in breast cancer patients undergoing NAC and to evaluate changes in tumor expression of Ki‐67 and rates of pCR. Furthermore, the study not only assessed biomarkers of efficacy but also included an analysis of long‐term survival and disease recurrence 3 years after enrollment. The methodology had additional strengths, such as the verification of DHA incorporation through two distinct biomarkers (erythrocytes and plasma phospholipids) using well‐established analytical methods, as well as the achievement of the initially calculated statistical power.

The study has some limitations. Despite the success of the randomization, characterized by the well‐balanced clinical characteristics, the study included a heterogenous population of breast cancer subtypes and disease stages that could have influenced the lack of significance in our main primary outcome. It is worth noting that our study did not establish a minimum compliance threshold during the intervention period. Compliance was monitored throughout the trial based on capsule counts and fatty acid status, with a two‐fold increase in the %DHA confirming adherence in the DHA treatment. As fatty acid status did not change during the study in the control group, this was defined as compliance. As reported by Mantha et al. ⁵⁸ , compliance can vary greatly during trials, and differences in capsule consumption may have contributed to the lack of significant findings. However, we found no positive correlation between compliance and DHA levels in plasma or erythrocytes within the DHA group, suggesting either saturation of incorporation or insufficient power to detect a dose–response relationship. One strength of our trial is that we achieved the pre‐calculated sample size. However, the relatively small sample size still limited our ability to stratify results by various clinical characteristics beyond HER2+++ subtype and body mass index, and age, factors known to influence NAC outcomes. ⁵⁹ , ⁶⁰ Furthermore, due to the success of the treatment, few patients did not have a tumor sample at the end of treatment to evaluate the expression of Ki‐67 due to the achievement of therapeutic responses. The calculation of the sample size excluded this scenario, representing a limitation of the methodology in using changes in Ki‐67 expression as the primary outcome. Another limitation of our study is the difference in Ki‐67 between groups at study entry, suggesting that tumor heterogeneity may have contributed to the inconclusive findings. Additionally, some of the pre‐treatment values for the Ki‐67 expression were missing and the data for the follow‐up analysis were not available for all patients. However, the missing values for Ki‐67 were not different between groups, and the follow‐up analyses are still ongoing. Lastly, our study was not powered to evaluate the efficacy of supplementation by tumor subtype. Although our study found a potential effect in the group classified as HER2+++, the results should be interpreted as exploratory and not definitive, as this classification does not fully capture the complexity of molecular subtypes and may not reflect the heterogeneity in clinical characteristics. Our study did not examine the effects of supplementation on the fatty acid composition of tumors. While previous research has shown that DHA supplementation can be incorporated into breast tissue, ⁵³ the dynamics and profile of DHA incorporation in tumors of breast cancer patients undergoing NAC remain unclear.

5. CONCLUSION

This is the first study to evaluate the efficacy of DHA supplementation during neoadjuvant breast cancer therapy. The results suggest that DHA supplementation does not confer additional benefits to NAC efficacy, as evidenced by the lack of effects on the nuclear proliferation marker Ki‐67 and rates of pCR. However, the study suggests a potential effect of DHA supplementation when groups were stratified by HER2+++ subtype. The study also supports the feasibility of supplementing with high doses of DHA during NAC, as well as its successful incorporation into plasma phospholipids and the membranes of erythrocytes. Additional secondary outcomes of this study will help to understand the effects of this intervention on systemic immune function, quality of life, and long‐term survival at 5 and 7 years. Future research is necessary to confirm these findings and investigate the impact of DHA and n‐3 LCPUFAs supplementation in different breast cancer subtypes during NAC.

AUTHOR CONTRIBUTIONS

Jaqueline Munhoz: Writing – original draft; methodology; writing – review and editing; formal analysis; visualization; validation; investigation. Marnie Newell: Conceptualization; methodology; writing – review and editing; funding acquisition; validation; investigation. Gilbert Bigras: Writing – review and editing; conceptualization; investigation; methodology; validation. Susan Goruk: Writing – review and editing; conceptualization; methodology; validation. Anil Abraham Joy: Writing – review and editing; conceptualization; methodology. Sunita Ghosh: Writing – review and editing; formal analysis; conceptualization. Kerry S. Courneya: Writing – review and editing; conceptualization. Vera Mazurak: Writing – review and editing; conceptualization. Claire M. Douglas: Writing – review and editing; methodology. Xiaofu Zhu: Writing – review and editing; methodology. Bohdarianna Zorniak: Writing – review and editing; methodology. John Mackey: Writing – review and editing; conceptualization; methodology. Judith Meza Junco: Writing – review and editing; methodology. Julie Price Hiller: Writing – review and editing; methodology. Karen King: Writing – review and editing; methodology. Sanraj K. Basi: Methodology; writing – review and editing. Catherine J. Field: Conceptualization; investigation; funding acquisition; writing – review and editing; project administration; supervision; writing – original draft.

FUNDING INFORMATION

This research was funded by the Canadian Institutes of Health Research (grant number: RES0037745), Cross Cancer Institute Investigator Initiated Trials (grant number: IIT‐0005) and a gift from the Butler Family Foundation, Edmonton, Alberta. C.J.F. holds a Canadian Research Chair Tier 1 in Human Nutrition and Metabolism. J.M. received scholarships from the Cancer Research Institute of Northern Alberta (CRINA), Alberta Graduate Excellence Award, Hazel McIntyre Summer Research Award, and the Dr. Elizabeth A. Donald MSc and Anthony Fellowship in Human Nutrition from the Department of Agricultural, Food & Nutritional Science (University of Alberta, Edmonton, Alberta).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests in relation to the study.

ETHICS STATEMENT

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Health Research Board of Alberta—Cancer Committee (HREBA.CC‐18‐0381). Written informed consent was required before screening and was obtained from all subjects involved in the study. ClinicalTrials.gov Identifier: NCT03831178.

Supporting information

DATA S1. Supporting information.

Munhoz J, Newell M, Bigras G, et al. Safety and efficacy of docosahexaenoic acid supplementation during neoadjuvant breast cancer therapy: Findings from the phase II, double‐blind, randomized controlled DHA‐WIN trial. Int J Cancer. 2025;157(7):1405‐1419. doi: 10.1002/ijc.35517

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.