Of the 357 patients that underwent breast MRI after the COVID-19 vaccination, 356 (> 99%) were female. The median age of the patients was 55 years (IQR 45, 65). There were 16/357 (4%) patients with current ipsilateral breast cancer and 57/357 (16%) patients with a history of treated ipsilateral breast cancer.

The clinical indication for the breast MRI exam was screening in 219/357 (61%) patients and diagnostic in 138/357 (39%) patients, including 44 to evaluate breast implants, 42 to evaluate a known cancer, 27 for short-term follow-up of a probably benign finding, 13 for a breast symptom, 9 to evaluate a breast lesion identified by another imaging modality, 2 to evaluate a surgical bed, and 1 for palpable axillary adenopathy.

The Pfizer-BioNTec vaccine was administered in 175/357 (49%) patients, the Moderna vaccine was administered in 137/357 (38%) patients, the Janssen vaccine was administered in 1/357 (< 1%) patients, and in 44/357 (12%) patients the type of vaccine was unknown. The vaccine was administered in the left arm in 268/357 (75%) patients, in the right arm in 69/357 (19%) patients, one dose in each arm in 12/357 (4%) patients, and in 8/357 (2%) patients the laterality was unknown. The number of doses received before the MRI exam was one dose for 173/357 (48%) patients and two doses for 184/357 (52%) patients. The median number of days was 26 days (IQR 14, 44) between the first dose of the vaccine and the exam date, 18 days (IQR 7, 29) between the second dose and the exam, and 15 days (IQR 8, 24) between the latest dose (either first or second dose) and the exam. Patient information is summarized in Table 1.

There were 104/357 (29%; 95% CI: 25, 34) patients with ipsilateral axillary adenopathy after COVID-19 vaccination. Of these, 13/104 (13%) presented with adenopathy beyond 4 weeks following completion of vaccination. Patients with only one dose of the vaccine presented adenopathy between 4 and 30 days after the vaccine, while patients with two doses presented adenopathy between 1 and 62 days after the second dose of the vaccine. Younger age was statistically significantly associated with the development of adenopathy (p = 0.002). This was confirmed after false discovery rate adjustment (adjusted p value = 0.012). The median age of patients with adenopathy was 51 years (IQR 41, 59) in comparison with 57 years (IQR 46, 67) of patients without adenopathy. The presence of either a current ipsilateral breast cancer or a history of a treated ipsilateral breast cancer was not associated with the development of adenopathy (p = 0.087 and 0.5, respectively). Of patients with concurrent ipsilateral breast cancer, 8/104 (8%) had adenopathy in comparison to 8/253 (3%) without adenopathy and among patients with a personal history of ipsilateral breast cancer, 14/104 (13%) developed adenopathy while 43/253 (17%) did not. Similarly, the clinical indication of the exam, either screening or diagnostic, was not associated with adenopathy (p = 0.9). Contralateral axillary adenopathy was not associated with COVID-19 vaccination and was only identified in patients with contralateral breast cancer and metastatic adenopathy.

A slightly larger proportion of patients who received the Moderna vaccine developed adenopathy 47/137 (34%) in comparison with the proportion of patients who received the Pfizer-BioNTec vaccine and who subsequently developed adenopathy 44/175 (25%), but this difference was not statistically significant (p = 0.12). Only one patient in our study population had received the Janssen vaccine, and this patient did not develop adenopathy. The comparison of types of vaccine between groups with and without adenopathy is demonstrated in Table 2.

Of the patients who received only one dose of the vaccine, 50/173 (29%) developed adenopathy; similarly, this finding was observed in 54/184 (29%) of patients who received two doses of the vaccine. No association was found between adenopathy and the number of doses of the vaccine prior to the MRI (p > 0.9). Likewise, the side where the vaccine was administered was not associated with adenopathy (p = 0.4).

The time between the second dose of the vaccine and the MRI exam was inversely associated with the development of adenopathy (p = 0.002). This was confirmed after false discovery rate adjustment (adjusted p value = 0.012). The time between the first dose or the latest dose and the exam were not associated with the development of adenopathy. The median time interval between vaccination doses and the presence of adenopathy on breast MRI exams are demonstrated in Table 3.

Of 104 patients with ipsilateral axillary adenopathy on breast MRI following COVID-19 vaccination, 74/104 (71%) had imaging follow-up performed at least 4 weeks after the second dose of the vaccine. All patients underwent an axillary ultrasound on follow-up, except for three who underwent PET-CT, two who underwent breast MRI, and one who underwent chest CT. In 8/74 (11%) patients, imaging follow-up performed at least 4 weeks after the second dose of the vaccine (ranging between 29 and 64 days after the last dose) demonstrated persistence of adenopathy. In 72/74 (97%) patients, the adenopathy was deemed resolved or decreased on follow-up, including 6 patients with persistent adenopathy on initial follow-up. Adenopathy was considered resolved in 57/74 (77%) patients (ranging between 31 and 130 days after vaccination) while in 15/74 (20%) patients, it was considered still abnormal but decreased in correlation with the breast MRI exam (ranging between 28 and 104 days after vaccination). Two patients were lost to follow-up after an axillary ultrasound performed more than 4 weeks after vaccination demonstrated persistence of adenopathy. The range of days after vaccination on imaging follow-up of patients with persistent, decreased, and resolved adenopathy is demonstrated in Table 4.

An interventional procedure was performed in 10/104 (10%) patients with adenopathy, including four FNAs, one CNB, and five sentinel lymph node biopsies (SLNBs). The results were positive for metastatic carcinoma in 3/104 (3%) patients, all of whom had concurrent ipsilateral breast cancer. These three patients represent 38% of a total of eight patients with axillary adenopathy and concurrent ipsilateral breast cancer. Two patients without ipsilateral breast cancer underwent FNA that yielded benign results. One additional patient had a previously diagnosed chronic lymphocytic leukemia and underwent an axillary CNB that yielded the same diagnosis.