Patients’ characteristics
From 404 patients referred to our fast-track H&N clinics during the study period, 88 patients had cervical lymphadenopathy. A total of 13 patients (14.8%) had cCVAL and were consecutively included in the study (Figure 1 ). The mean age was 54.8 ± 16.1 years, and interestingly most of the patients were females (N=11, 84.6%),Table 1. All patients had the Pfizer/BioNTech vaccine, and the majority had the injection in the left deltoid (N=12, 92.3%). The study period mostly covered the early phases of vaccination rollout in the UK, and most patients (N=12) had only one dose by the time of presentation. All patients experienced a feeling of a lump in the ipsilateral neck, but pain was only reported in six patients (46.2%), Table 1 and 2. The swelling was first noticed by patients within a median of four days (IQR 2-7) from vaccination, and in seven patients (53.9%) a referral was made to the fast-track clinic within three weeks from the onset of symptoms (Table 2 ).
Lymphadenopathy features on neck ultrasound scans (USS)
All patients had USS, with median interval between the swelling onset and the scan was 25 days (IQR 10-49). Table 3 summarises the USS characteristics of the examined lymph nodes (LNs). Targeted USS confirmed the presence of one or more LNs, all in level IV or V of the neck. The average short axis diameter (SAD) of the most prominent nodes was 5.5 ± 1.4 mm, but five patients (38.5%) had more rounded nodes with short axis: long axis ratio (S:L) > 0.5. Scans performed ≤4 weeks from swelling onset showed significantly larger nodes (6.5 ± 1.4 mm) compared to scans performed >4 weeks (4.8 ± 0.8 mm),P =0.03. The overall scan impression was recorded as benign reactive lymphadenopathy in all patients, but two cases had increased vascularity on colour doppler (Figure 2, Table 3 ). Moreover, two patients had mildly hypoechoic LNs but none had fatty hilar abnormality. Two patients had fine-needle aspiration biopsy, but cytology demonstrated features of benign reactive lymphadenopathy. Moreover, two patients had follow-up USS that showed reduction in CVAL size.
Lymphadenopathy outcomes
The outcomes of lymphadenopathy were subjectively reported by patients during virtual follow-up (Table 2 ). Seven patients (53.9%) reported full resolution (FR) of their palpable swelling within an average of 3.1 ± 2.3 weeks from the onset of their symptoms. Five patients (38.5%) reported partial reduction (PR) in the size of their palpable lumps over an average period of 8.4 ± 3.1 weeks (Figure 1 ). When compared to the FR group, the PR group interestingly presented to their general practitioner significantly later (35.6 vs. 9.0 days, P = 0.001 ), and had significantly smaller LNs on USS (4.6 vs. 6.4 mm, P =0.024 ), Table 4.However, neither the patients’ age, nor the vaccination to lymphadenopathy interval were found to be significantly Impacting the clinical outcome.

Discussion

Post-vaccination lymphadenitis is an uncommon phenomenon that can be triggered by intramuscular vaccine injections in the deltoid muscle. It has been previously reported in adults following many viral vaccinations especially for human papillomavirus (HPV)18 and H1N1 influenza19. More recently, cases with CVAL have been described in reports from the United States, Spain, Israel and the UK.8,12-14,20-22
The available evidence has suggested that mRNA-based vaccines are likely more immunogenic than standard vaccines, and hence show higher incidence of CVAL.7,21,22 In clinical trials on the Pfizer/BioNTech mRNA vaccine, lymphadenopathy was only reported as an unsolicited ADR, with incidence in the vaccine group as high as ten times that in the placebo group (0.3% and 0.03% respectively).1 As a solicited ADR, axillary lymphadenopathy was the second most frequently reported reaction in the Moderna vaccine trials, with incidence of 10.2% and 14% after the first and second doses respectively.3 Reporting lymphadenopathy in these trials was only based on physical examination, and the true incidence rate was likely much higher.1,3,7 Interestingly, clinical trials on the Oxford/AstraZeneca adenovirus-vectored vaccine reported a lower incidence of lymphadenopathy in the vaccination group (0.3%) compared to the placebo group (0.4%).2 Our results show that all patients referred to us with cervical CVAL (cCVAL) had the Pfizer/BioNTech vaccine, and none had the Oxford/AstraZeneca vaccine despite it accounting to almost 53% of the total UK vaccine doses given during our study period.4 Moreover, real-life data from the MHRA (reported using the yellow card scheme for ADRs) has shown that the Pfizer/BioNTech vaccine had almost double the number of reports for lymphatic system disorders compared to the Oxford/AstraZeneca vaccine (22.4 vs . 11.7 per 100,000 doses given respectively).4-6
Most of the COVID-19 vaccinations during our study period were nationally prioritised to people ≥ 65 years old. The mean age of patients with cCVAL in our study was 54.8 years, with only three patients were 65 years or older. Data from the Pfizer/BioNTech trials also demonstrated higher incidence and severity of ADRs in younger participants, with lymphadenopathy reported five times more common in the 16-55 years age group (0.5%) compared to the >55 years age group (0.1%).1 Similarly, CVAL was more frequently reported in younger individuals (18-64 years) following the first and second doses of the Moderna vaccine (11.6% and 16% respectively), compared to individuals aged ≥65 years (6.1% and 8.4% respectively).3 In a case series of 20 female healthcare workers with cCVAL, Fernández-Prada et al. 12 reported that 75% of patients (N=15) had full resolution within 16 days from symptoms onset. In our study, full clinical resolution was reported within an average of 3.1 weeks from symptoms onset in more than half of our cohort, and partial improvement within an average of 8.4 weeks in 40% of patients. Interestingly, half of our patients were directly referred to our fast-track H&N cancer clinics within three weeks of symptoms onset.

USS features of COVID-19 vaccine-associated lymphadenopathy

The role of ultrasonography in the assessment of supraclavicular (level IV/V) lymphadenopathy is well established. All our patients had USS within a median of 8 days from referral, and a median of 25 days from onset of symptoms. There was a significant inverse association between the timing of the scan and the size of the imaged nodes, possibly highlighting a time-dependant reactive nature of the nodes. While the overall impression was of benign nature, some nodes in our cohort showed abnormal features. Large nodes with increased vascularity and a high S:L ratio (more rounded) usually indicate abnormality.23Ying et al. recommended that the optimum cut-off value of SAD (S:L ratio) in the lower neck should be 3-5 mm (0.4-0.5), with a high specificty and moderate sensitivity.23 Our data demonstrated that around half of our patients had nodes >5 mm and 40% had S:L ratio >0.5. These findings are in line with previous reports that demonstrated that CVALs may show abnormal morphology, and can appear enlarged, rounded, hypoechoeic and with loss of echogenic fatty hila.8,9,13,20