Weak immunogenicity after a single dose of SARS-CoV-2 mRNA vaccine in treated cancer patients.

Active cancer and ongoing antineoplastic treatments are major factors for severe coronavirus disease 2019 (COVID-19) and death; reasons why the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccination remains a priority in cancer patients (CPs). However, immunocompromized patients were excluded from major studies on mRNA vaccines, , and could have a decreased response to vaccination, as recently demonstrated in solid organ transplant recipients. Herein, we aimed to assess the proportion of antibody response 4 weeks after the first injection of the BNT162b2 (Pfizer-BioNTech) vaccine in CPs and health care workers (HCWs) as the control population.

All consecutive patients with cancer on active treatment or with treatment in the last 2 years and HCWs who underwent SARS-CoV-2 vaccination between 17 February 2021 and 18 March 2021 at the Pitié Salpêtrière Hospital, Paris, France, were selected for analysis. The titration of SARS-CoV-2 antibodies was proposed just before the second injection of BNT162b2 vaccine. Serum anti-nucleoprotein (N) immunoglobulin G (IgG) and anti-spike protein (S) IgG against the receptor binding domain (RBD) of the S1 domain were detected using the Abbott SARS-CoV-2 IgG chemiluminescent microparticle immunoassay (CMIA), according to the manufacturer's instructions. The presence of anti-N IgG was used as a surrogate marker of prior COVID-19.

Statistical analysis consisted of univariable analysis (Chi-square tests) and then multivariable analysis (binary logistic regression, including all variables with P value < 0.1 in univariable analysis) to determine the factors associated with the lack of seroconversion in CPs. Median titers of anti-S IgG were compared between CPs and HCWs, using a Mood's test. This study was approved by the Commission Nationale de l'Informatique et des Libertés (MR004, registration number: 2221945).

SARS-CoV-2 antibodies were measured in 110 CPs and 25 HCWs (Table 1 ). In CPs who did not have COVID-19 before vaccination, the seroconversion rate was only 55%, while it reached 100% in HCWs. Titers of anti-S IgG were significantly higher in HCWs in comparison with seropositive CPs (680 versus 315 UA/ml, P = 0.04). Sex, cancer locations and metastatic status were similar in seroconverters and non-seroconverter CPs (Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2021.04.020). After adjustment for potential confounders, two factors were strongly associated with no seroconversion: age >65 years [odds ratio 3.58, 95% confidence interval (CI) 1.40-9.15, P = 0.008] and treatment by chemotherapy (odds ratio 4.34, 95% CI 1.67-11.30, P = 0.003).

Table 1

Characteristics of cancer patients and health care workers with SARS-CoV-2 serological outcome

Cancer patients (N = 110)
Sex, n (%)
Women	66 (60)
Men	44 (40)
Age, years, median (IQR)	66 (54-74)
Cancer location, n (%)a
Breast	37 (34)
Lung	15 (14)
Gynecological	15 (14)
Prostate	11 (10)
Digestive	8 (7.3)
Kidney	7 (6.4)
Bladder	5 (4.5)
Upper aero-digestive tract	6 (5.5)
Thyroid	5 (4.5)
Others	3 (2.7)
Cancer staging, n (%)
Local	47 (43)
Metastatic	63 (57)
Cancer treatment, n (%)b
Chemotherapy	38 (35)
Targeted therapy	26 (24)
Immunotherapy	17 (16)
Hormonotherapy	16 (15)
Radiotherapy	6 (5.5)
Clinical surveillance	18 (16)
Time between first vaccine injection and SARS-CoV-2 serology, days, median (IQR)	27 (26-28)
Positive anti-N IgG, n (%)c	15 (14)
Positive anti-S IgG, n (%)c
In all patients	64 (58)
Among patients with positive anti-N IgG (N = 15)	12 (80)
Among patients with negative anti-N IgG (N = 95)	52 (55)
Titer of anti-S IgG, UA/mL, median (IQR)
In all anti-S positive patients (N = 64)	359 (178-998)
Among patients with positive anti-N IgG (N = 12)	657 (366-14, 112)
Among patients with negative anti-N IgG (N = 52)	315 (140-748)

IQR, interquartile range; N, nucleoprotein; S, spike protein; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Two patients had synchronous cancers (prostate + lung and prostate + colon).

Non-exclusive categories.

Abbott SARS-CoV-2 IgG chemiluminescent microparticle immunoassay (CMIA), with detection threshold: 0.8 UA/ml for anti-N IgG, and detection threshold: 50 UA/ml for anti-S IgG.

No symptomatic COVID-19 occurred between the two injections of vaccine in CPs and HCWs.

In summary, almost half of CPs showed no anti-spike antibody response after the first injection of BNT162b2 vaccine, and this low seroconversion rate could be much worse in elderly patients and in patients under chemotherapy. In comparison, 100% of the HCWs had anti-spike seroconversion. Moreover, even in CPs with seroconversion, the level of antibody response could be lower than expected.

In conclusion, our findings argue for not extending the 21-day period between the two SARS-CoV-2 vaccine injections in CPs, and for performing serological monitoring to assess antibody response in this particular population, which could lead to adapting this vaccine strategy. We would also recommend a vaccine strategy including family and friendship circles.