Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infections and the resulting disease, COVID-19, first emerged in 2019, before being declared pandemic by the WHO in March 2020. Those at highest risk of severe disease and death included the elderly, and those with pre-existing health conditions including cancers1–3. Patients with blood cancers were defined as extremely clinically vulnerable to COVDI-19 and advised to ‘shield’ during first and subsequent waves of infection. In the first wave, patients with haematological malignancies had poor outcomes with mortality rates of 20-40%4–7.
UK data during the most recent ‘Omicron wave’ (December 2021-February 2022) has shown that despite high levels of community infections, hospital admissions and deaths are proportionally lower than previous waves. This change is likely multifactorial; largely driven by viral variant characteristics, population-level immune protection conferred by vaccination, and availability of antiviral and monoclonal antibody therapies, particularly in the non-hospitalised setting.
Prophylactic vaccines focus on immunisation with the spike (S) protein, the main target for neutralising antibodies. Neutralising antibodies block viral entry into host cells by preventing interaction between the spike protein receptor binding motif and the host cell angiotensin-converting enzyme-2, and vaccines were expected to be protective for alpha strains8,9. Commercial quantitative anti-S antibody assays are widely available, although there remains uncertainty as to what threshold of anti-S IgG titre correlates with effective viral neutralisation. Results from the RECOVERY trial confirm a higher mortality in patients hospitalised with COVID-19 where no antibody response was detectable upon admission, and show that treatment with casirivimab and imdevimab, a neutralising monoclonal antibody (nMAB) cocktail, reduced the relative risk of death by 20%10. In immunocompromised groups, NHS-England extended the use of casirivimab and imedevimab to those with very low antibody responses, defined as anti-S titres in the bottom 10% of the assay’s detection range, on the premise that these were likely inadequate responses11 . Subsequent emergence of omicron, which is resistant to casirivimab and imedevima, limited it use to those infected with the delta variant12. However other, omicron-active, nMAB have shown efficacy in reducing rates of hospitalisation and death when delivered in the community, early in the disease course13.
The degree to which patients with blood cancers are afforded protection by vaccination is less clear, and likely to be heterogenous. Whilst some with chronic haematological malignancies in remission on long-term treatment appear to have near-normal responses14, the OCTAVE study has reported the first 600 patients, including some with haematological malignancies, or stem cell transplant (n=80), and has reported no measurable anti- Spike antibody responses at 4 weeks following second primary vaccine in 11.1% and 11.9% respectively15.
Patients with blood cancers were amongst the highest priority group to receive SARS-CoV-2 vaccinations in the UK, with most patients receiving either the mRNA-based Pfizer-BioNTech-BNT162b2 or adenoviral-based AstraZeneca-ChAdOx1-S/nCoV-19. Both require two vaccines to complete the primary course, and in the UK, as an urgent public health measure, the second primary vaccine was delayed to a 12-week interval for many. Based on the early evidence of suboptimal antibody responses to two vaccines in some patients, the MHRA approved the use of a third primary mRNA-based vaccine as a reinforcing dose, given at least eight weeks after completion of the second primary dose, and distinct from the ‘booster’ dose, which was additionally available 8 weeks after completion of the primary course.
Reflecting the emerging real-world data of suboptimal vaccine responses and continued vulnerability of high-risk groups, a national outpatient treatment programme was established in the UK to deliver anti-viral or nMAB to at risk individuals, including those with haematological malignancies, in central hubs. Currently, the program currently prioritises oral anti-virals over nMABs, and makes no distinction of treatment preference based on vaccine status or serological responses. In this study we report real-world post vaccination SARS-CoV-2 antibody responses in patients routinely attending haematology outpatient clinics at a single centre, with the aim of quantifying longitudinal serological responses utilising a commercial assay. This has the potential to identify the most at-risk groups of patients with haematological malignancies, which could guide the prioritisation of nMAB over anti-virals.
Three hundred and eighty-one patients were included in this study, with the demographics shown in Table 1. Their first two vaccinations were AstraZeneca-ChAdOx1-S/nCoV-19 in 149 patients (39.1%) and Pfizer-BioNTech-BNT162b2 in 228 (59.8%), with 1 receiving moderna-mRNA1273, and 3 unknown. Almost all patients (n=160, 98.3%) receive Pfizer-BioNTech-BNT162b2 for their third primary dose. The median time between first and second vaccines was 75 [8-181] days, and between second and third was 187 [114-300] days. Antibody results were available after one, two and three vaccines for 171, 327 and 162 patients respectively, with a median time between vaccination and antibody test of 56 [11-160] days, 48.5 [8-250] and 51 [7-117] days post first, second and third doses respectively. At each of the three time-points evaluated, there were no significant differences in the representation of patient diseases or treatment groups (Table 1), although there was a small but significant difference in median age (62, 65, 66 years post first, second and third respectively).
While all healthy controls had detectable anti-S antibodies after the first vaccination, this was not true in patients with haematological malignancies. There was however an incremental increase in the proportion of patients with detectable responses after the first, second and third doses (52.6% 72.8% and 87.0% respectively p<0.001, table 1, figure 1a). Responses were not uniform however, and whilst overall, detectable responses were seen in 87% of patients after 3 doses, only 57.9% of those with CLL and 60.9% of those with B-cell lymphoma had measurable anti-S antibodies (Figure 1b). There were also significant differences when patients were stratified by treatment, with detectable responses in only 37.5% of those treated with Bruton’s tyrosine kinase (BTK) inhibitors, 55.0% of those receiving anti-CD20 monoclonal antibody-containing regimens (rituximab, obinutuzumab) and 75.0% of those on bcl-2 inhibitor venetoclax. There was a clear divergence in response to the third dose, with some groups of patients with previously reported poor responses to 2 doses, such as those on JAK-STAT inhibitors 16 being partially rescued by the 3rd dose, (detectable anti-S antibodies after 1, 2 and 3 doses of 18.2%, 52.4% and 90.9% respectively) compared to some groups, such as those on BTK inhibitors, showing only modest increase in antibody detection with increasing number of doses (anti-S antibodies after 0%, 15.4% and 37.5% after 1, 2 and 3 doses respectively) (Figure 1c).
For patients with detectable anti-S antibodies after the second vaccine, the median titre was significantly lower than healthy controls (patients 214.0 [7.8-5680] BAU/ml, controls 603.6 [82.8-5531] BAU/ml, p<0.001, figure 1a). However anti-S titres following a third primary vaccine in patients was not significantly different from the post second anti-S titre observed in the controls (median for patients after three doses 1026 [7.89-5680] BAU/ml and controls after two doses 603.6 [82.81-5531] BAU/ml, p0.696, Figure 1a).
While only 19.3% of patients had anti-S tires above 568 BAU/ml (bottom 10% of antibody detection range) following two vaccinations, compared to 56.0% of healthy controls, 58.6% of patients achieve this after the third primary dose (Figure 1b). Again, these responses were not uniform, with only 21.1% of those with CLL, 21.7% of those with B-cell lymphoma, and 28.6% of those with plasma cell dyscrasias crossing this threshold. Importantly no patients treated with a BTK inhibitor mounted responses above 568 BAU/ml, while only 15.0% of those receiving rituximab- or Obinutuzumab-containing regimens, 25.0% of those receiving venetoclax, 36.4% receiving other cytotoxic chemotherapy and 36.4% receiving JAK-STAT inhibitors achieved titres over 568 BAU/ml.
The MARCH study was designed to report real-world serological responses to SARS-CoV-2 vaccination in a large number of patients with haematological malignancies. There are of course biases in the nature of this study design, including the heterogeneity of underlying disease, treatments and potentially selection bias for those patients who continued to attend for appointments during pandemic, nevertheless it captures the most at-risk patients and provides insights into the quantitative nature of vaccine responses in a specific vulnerable patient group. Whilst the healthy controls and the patients with malignancies were not perfectly matched, they provide context to the stark difference in antibody responses.
Following the first two vaccines, a substantial proportion of patients with haematological malignancies fail to mount any detectable serological response, and many of those who do, have only low-level titres with unknown degree of protection. Importantly, a large proportion of patients were effectively salvaged following the third primary vaccine, with antibody levels similar to those of the healthy controls who had received two doses. While some of the enhanced responses to the third dose may be explained by a longer period since treatment completion, for others there is no obvious clinical explanation. There does however remains a significant number of vulnerable patients, primarily those with CLL, B-cell lymphomas, myelofibrosis, and those on specific therapies (B-cell depleting monoclonal antibodies, BTK-, bcl2- and JAK STAT- inhibitors) who persist with sub-optimal vaccine responses even after three doses who likely remain vulnerable to severe infection.
Current NHS-England and CDC guidance tends to group patients with haematological malignancies together, and this data suggests this is not a homogenous cohort, and whilst oral antivirals may be reasonable treatment in some groups, suggests that others with no, or low-level, sero-protection following vaccination may benefit from neutralising monoclonal antibodies in first-line therapy in the event of SARS-CoV-2 infection, a provision that in the UK has already been made following solid organ transplantation. The emerging trends also suggest that sup-optimal responses can be enhanced with additional primary vaccine doses, supporting the ongoing strategy for 4th doses in vulnerable patients groups