Our study design was a prospective longitudinal cohort study. Patients with early breast cancer who were 20 years or older and received systemic anti-cancer treatments for breast cancer at Seoul National University Hospital (SNUH) were recruited as in previous study (23). Controls were healthcare workers at SNUH who did not have any systemic diseases. Participants with confirmed COVID-19 infections were excluded. The Institutional Review Board approved the study (IRB No. 2103-121-1206 and 2102-032-1193). Written informed consents were obtained from all participants before the initial vaccination. All the data from the patients were anonymized and de-identified prior to analysis.

Definition of early breast cancer was confined to breast disease with or without the involvement of regional lymph nodes but without distant metastasis. Medical anti-cancer treatments comprised cytotoxic chemotherapy, molecular targeted therapy, and endocrinal treatment. The patients were stratified by the treatments they received within 28 days of vaccination: cytotoxic chemotherapy group, non-cytotoxic anticancer group, and post-treatment group. The cytotoxic drugs included anthracyclines, cyclophosphamide, docetaxel, paclitaxel, and carboplatin. Non-cytotoxic treatments were molecular targeted therapy and endocrinal therapy. Targeted antibodies were trastuzumab and pertuzumab. Endocrinal therapy was chosen among tamoxifen, letrozole, anastrozole, and gonadotropin-releasing hormone agonists in the context of the patient’s menopausal status and clinical conditions. Post-treatment group did not receive any anti-cancer treatments within 28 days of vaccination. Corticosteroid doses equivalent to or more than 10 mg of prednisolone within 14 days of vaccination were analyzed for immune-modulatory effect.

Primary vaccinations were defined as either two doses of PZ, Moderna, AZ, or a single dose of Janssen. A booster injection was defined as the additional dose of PZ or Moderna to primary vaccinations. Cross-injection among different vaccines was allowed. The patients who received a single dose of the Janssen vaccine and booster dose were only included in the first dose and the booster dose group for analysis. The serum of participants was collected at baseline before the vaccination, three weeks after a first and second dose of primary vaccinations, before the booster dose, and three weeks after the booster dose.

Our primary endpoint was the level of antibody response of the COVID-19 vaccinations and boosters in patients with early breast cancer compared to healthy controls. Secondary outcomes were the effect of age and corticosteroid use on the immunogenicity of vaccines and the efficacy against variants of COVID-19. Blood samples were collected at baseline and within 21 to 28 days after the vaccinations and booster injections.

pcDNA3.1 SARS-CoV-2 S D614G was a gift from Jeremy Luban (Addgene plasmid # 158075; http://n2t.net/addgene:158075; RRID: Addgene_158075). SARS-CoV-2 S D614G protein, RBD_wt, RBD_δ, and RBD_ο were produced in Expi293 cells (Thermo Fisher Scientific) and were purified using Ni-NTA agarose resin (Thermo Fisher Scientific) affinity chromatography, as described previously (24, 25).

Briefly, Expi293 cells were cultured at 37 °C with 5% CO₂ for five days after transfection of each plasmid encoding SARS-CoV-2 S D614G protein, RBD_wt, RBD_δ, or RBD_ο (BA1). The supernatant was collected and passed over the Ni-NTA agarose resin column three times. After washing with 100 mL of phosphate-buffered saline (PBS), the his-tagged protein was eluted by elution buffer (pH8.0, 50 mM sodium phosphate, 300 mM NaCl, and 250 mM imidazole). Finally, samples were buffer-exchanged into pH 7.4 PBS using Amicon Ultra-4 (Merck Millipore, Burlington, MA, USA) spin columns with a 10 kDa cutoff. The purity of purified samples was assessed by 14% SDS-PAGE gel.

The binding activity of serum antibody (Ab) level against each SARS-CoV2 antigen (spike protein, RBD_wt, RBD_δ, RBD_ο, and nucleocapsid) was measured by enzyme-linked immunosorbent assay (ELISA), as described previously (24–26). The 96-well polystyrene ELISA plate (Thermo Fisher Scientific) was coated with 100 ng of each antigen per well overnight at 4 °C. Each well was blocked with 100 μl of PBS (pH 7.4) containing 3% bovine serum albumin (BSA) for an hour at room temperature, and then the plate was washed four times with the PBST buffer (PBS with 0.05% Tween 20). The diluted serum samples (1:10, 1:50, 1:250) were added into wells, and each sample was diluted five-fold serially. After incubating at room temperature for an hour, wells were washed with PBST four times. Then, goat anti-human IgG Fc Ab-conjugated with HRP (1:12,000, Arigobio, Hsinchum Taiwan) was added and incubated at room temperature for an hour. Alternatively, mouse anti-human IgA (1:100) Ab was added and incubated at room temperature for an hour, and then anti-mouse IgG (H+L)-conjugated with HRP (Thermo Fisher Scientific) was added and incubated at room temperature for an hour for IgA detection. After washing with the PBST four times, 50 μl of 3,3’,5,5’-tetramethylbenzidine was added per well as chromogen substrate. The plate was kept at room temperature for 20 min, and the reaction was terminated by adding 50 μl of 2M H₂SO₄. Finally, absorbance was measured at 450nm with an Infinite 200 PRO-Nano Quant microplate reader (Tecan Trading AG, Mannedorf, Switzerland). Relative binding activity was calculated as the ratio of the binding activity at a certain time point to the binding activity at pre-vaccination in the same donor.

Data are presented as mean ± standard deviation of the mean (s.d.) and as dot plots. Multiple unpaired T-test was performed to compare levels of immune responses between two groups. P < 0.05 was considered indicative of statistical significance. All statistical analyses were two-tailed and used GraphPad Prism 9 (GraphPad Software, La Jolla, CA, USA). All graphs were generated using GraphPad Prism 9.

Initially, 41 patients with early breast cancer and 20 healthy controls were recruited. One participant with breast cancer was excluded due to confirmed COVID-19 infection prior to the study. All breast cancer patients were female, and the median age was 51.5 years. During the primary vaccinations, PZ vaccines were most commonly injected in 32 (80.0%), followed by Moderna (4 of 40, 10.0%), AZ (3 of 40, 7.5%), and Janssen (1 of 40, 2.5%). Three-quarters of patients received anti-cancer treatments within 28 days of initial vaccinations. Of 15 patients who received chemotherapy regimens with cytotoxic drugs, two patients received concurrent HER2 targeted treatment. The non-cytotoxic treatment group comprised 15 patients who received hormonal therapies. Post-treatment group comprised 10 patients who were not under active anti-cancer treatments. Corticosteroids were only used in adjunct to cytotoxic chemotherapy.

Later at the time of receiving booster doses, however, all patients had completed cytotoxic chemotherapy and were re-categorized into a non-cytotoxic group or post-treatment group. Most patients (28 of 40, 70.0%) were on endocrinal treatments or targeted therapies, while 12 patients (30.0%) did not have further anti-cancer treatments. All except three patients received booster vaccination with either PZ (33 of 40, 82.5%) or Moderna (4 of 40, 10.0%). One patient in cytotoxic group had symptomatic COVID-19 infection after the second vaccination and did not receive booster dose. The other two patients, also in cytotoxic group, refused booster doses. In the healthy control group, all were female, and the median age was 31.0 years. All participants received PZ vaccines The patient characteristics are shown in Table 1 .

Prior to analysis, the SARS-CoV-2 nucleocapsid (NC) protein reactivity was examined to detect the participants who experienced asymptomatic COVID-19 infections. As a result, two additional breast cancer patients with NC-positivity were excluded from the study ( Figure 1 ).

The binding activity of serum IgG against recombinant D614G spike protein ( Supplementary Figure 1 ) was examined in the serially diluted serum samples by enzyme-linked immunosorbent assay (ELISA) ( Supplementary Figure 2 ). For more detailed analysis, the ELISA results of 10-fold diluted serum samples were used. Consistent with the previous report (24, 26), the healthy control group showed 1.9-fold, 3.6-fold, and 4.0-fold increase in IgG response against SARS-CoV-2 spike D614G compared to the baseline after the first, second, and booster doses of vaccinations, respectively ( Figure 2A ). On the other hand, the breast cancer group showed weaker humoral immune responses than the healthy control. The breast cancer group showed 1.5-fold, 2.3-fold, and 2.5-fold increased antibody response against SARS-CoV-2 spike D614G compared to the baseline after the first, second, and booster doses of vaccinations, respectively ( Figure 2A ).

Next, we stratified the breast cancer patients by their age, types of anti-cancer therapy, and use of adjunct steroids, and analyzed the antibody responses in each subgroup. In our study, most patients with breast cancer were under 65 years old. The patients were divided into an elderly group (>55 years old) and a younger group (≤55 years old), and the antibody response was equivalent in the two groups ( Figure 2B ). We also confirmed using Pearson’s correlation analysis that there was no significant correlation between age and humoral immune responses in the BC group ( Supplementary Figure 3 ).

In the analysis of antibody response among patients treated with different anticancer therapies, the breast cancer patients who had concurrent chemotherapy showed significantly lower antibody responses against SARS-CoV-2 Spike D614G protein than the non-cytotoxic treatment group but equivalent to the post-treatment group after the first vaccination ( Figure 2C ). All groups showed increased binding activity after the second vaccination. At the time point of the pre-booster, all patients had completed cytotoxic chemotherapy and were classified into the non-cytotoxic or post-treatment group. The non-cytotoxic treatment group showed a higher antibody response than the post-treatment groups throughout the study ( Figure 2C ).

Next, we compared the antibody responses of breast cancer patients according to the use of adjunct steroids. Steroids were used only in the chemotherapy group as pretreatment to ameliorate hypersensitivity reaction and emesis. The antibody responses, however, did not show a significant difference with the use of adjunct steroid among patients who received cytotoxic treatment during the primary vaccinations ( Figure 2D ). No patients received cytotoxic chemotherapy and adjunct steroids during pre-booster and booster.

Similar to IgG responses, IgA responses against SARS-CoV-2 spike D614G in the healthy control group showed a higher fold increment than in the breast cancer group. The healthy control group showed a 4.2-fold, 10.4-fold, and 5.2-fold increase in IgA response against SARS-CoV-2 spike D614G compared to the baseline after the first, second, and booster doses of vaccinations, respectively ( Figure 2E ). In comparison, the breast cancer group showed 2.1-fold, 3.0-fold, and 3.6-fold increased antibody responses, respectively ( Figure 2E ). The younger group (≤55 years old) showed higher binding activities to SARS-CoV-2 spike D614G, but there was no statistical significance ( Figure 2F ). In addition, the types of anticancer therapies and adjunct steroids did not affect IgA response in BC patients ( Figures 2G, H ).

The binding activities of the serum antibodies produced by the vaccination to the SARS-CoV-2 VOCs were analyzed using recombinant RBD_wt, RBD_δ, and RBD_ο proteins ( Supplementary Figure 1 ). Serum IgG and IgA antibodies of breast cancer patients generally showed weaker binding activities for RBD_wt and RBD_δ than those of the healthy control group ( Figures 3A, B, D, E ). On the other hand, breast cancer patients had significantly weaker IgG binding activity to RBD_ο after the second vaccination than healthy control. After the booster shot, however, the IgG binding properties of breast cancer patients became equivalent to those of healthy control ( Figure 3C ). There was no statistical difference in IgA response between the healthy control group and breast cancer patients ( Figure 3F ).