Safety and efficacy of COVID‐19 convalescent plasma in severe pulmonary disease: A report of 17 patients

Abstract

To date, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has infected over 30 million people, resulting in over 900 000 deaths globally and counting.1 Investigational therapies, including hydroxychloroquine/azithromycin and lopinavir/ritonavir, have been disappointing. Remdesivir shortened median recovery time by 4 days but led to a non-significant decrease in mortality.2 Nevertheless, remdesivir is unlikely to be available on a large scale in the near future. Thus, immediate interventions to improve COVID-19 mortality represent a public health emergency. COVID-19 convalescent plasma (CCP) is a promising approach, whereby plasma carrying antibodies against SARS-CoV-2 from recently recovered donors is transfused into patients, conferring passive immunity in patients susceptible to having poor outcomes. Passive immunity was first applied in a pandemic during the 1918 influenza outbreak, where recovered patients' sera successfully treated acutely ill patients, reducing fatality from 37% to 16%. The benefit was even more pronounced when convalescent plasma was infused within 4 days of decompensation, reducing the fatality rate from 59% to 19%.3 In the SARS-CoV-1 epidemic, patients' overall survival improved from 12.5% to 17% when treated with convalescent plasma, with the highest benefit seen when plasma was infused within 14 days of symptom onset.4 Studies regarding the clinical use of CCP have been inconsistent due to the pressing need for immediate effective treatment during this pandemic. A Cochrane Review by Valk et al highlighted eight studies regarding the use of CCP.5 Based on the wide spectrum of patients, small total number of treated patients and lack of an ubiquitous endpoint, the authors were unable to draw any concrete conclusions about overall mortality or clinical improvement with regard to CCP use. However, despite various differences, many studies demonstrated some improvement. An early study by Ye et al from Wuhan showed an improvement in chest computed tomography findings post-transfusion of CCP in six patients. It is worth noting that, in this study, only four of the six patients required supplemental oxygen via nasal cannula prior to transfusion (highest need was 5 L/min), and none were intubated. Original data from China showed that only 41% of symptomatic people required supplemental oxygen, of which 6% required mechanical ventilation.6 Therefore, it is unclear whether those six patients treated with CCP would have progressed to more severe disease or recovered independently. Nonetheless, the Shen et al case series of five critically ill patients from the Shenzhen province highlighted successful extubation of all five patients following CCP infusion.7 A clinical trial by Li et al noted that patients with severe COVID-19 infection who had CCP added to standard treatment, compared with standard treatment alone, did not demonstrate a statistically significant improvement in both time to clinical improvement and overall mortality.8 However, this study had a significantly older patient population, as well as a delayed time from symptom onset to administration of CCP. In contrast, a recent clinical trial by Joyner et al demonstrated a mortality benefit between early transfusion of CCP, as well as higher antibody titres.9 Recently, a propensity score-matched case-control study of 39 patients by Liu et al demonstrated a benefit in both clinical symptoms and overall survival.10 Despite the uncertainty surrounding CCP, the food and drug administration (FDA) recently announced emergency authorisation use (EUA) for CCP.11 For COVID-19, the optimal timing and frequency of CCP infusion remains largely unknown. Similarly, the role of CCP in cancer patients, particularly those with haematological malignancies, remains unknown. Here, we describe the outcomes of 17 critically ill patients with COVID-19, including six with haematological malignancies, displaying varying ranges of severe illness and length of infection, who were treated with CCP with marked clinical improvement. Thirteen donors with blood types O, A and B donated two to four CCP units each (200 mL per unit) 18 to 56 days following full recovery from COVID-19. Ten men and seven women between the ages of 24 and 81 years (mean 56) received CCP following informed consent (Data S1). All patients were diagnosed by a reverse transcription polymerase chain reaction (RT-PCR)-based technique with the exception of patients 2, 4, 11 and 13, who were diagnosed using the highly sensitive clustered regularly interspaced short palindromic repeats (CRISPR)-based qualitative COVID-19 assay, as detailed in Table S1.12 Interestingly, these four patients had haematological malignancies and had multiple false negative RT-PCR results prior to the CRISPR diagnosis. Patients 1 and 12 also had haematological malignancies. Most patients had multiple medical comorbidities, and 14 of the 17 patients were treated in the intensive care unit (Table S1). The average time from illness to treatment with convalescent plasma was 12 days (range 4-41) (Table S1). Further patients' characteristics are summarised in Table S1. At the time of CCP infusion, all patients were either mechanically ventilated (six patients), on non-invasive support with high-flow nasal cannula (four patients), on bilevel ventilation (one patient) or on nasal cannula (six patients). Using enzyme-linked immunosorbent assay (ELISA), we were able to determine the Spike protein IgG titres on CCP units used to treat patients 5, 6, 9, 11, 12, 13, 14, 15 and 17 to be 1:1600, 1:3200, 1:3200, 1:800, 1:3200, 1:400, 1:1600, 1:6400 and 1:3200, respectively. The recent FDA EUA recommended an IgG titre of at least 1:250 in transfused CCP,11 whereas a study by Salazar et al noted a reduction in mortality with CCP IgG titres greater than 1:1350.13 Joyner et al noted a mortality gradient with better outcomes associated with higher titre and early administration.9 The nine CCP IgG titres obtained in our study were all above the FDA recommendation of 1:250, whereas seven of the nine were above the 1:1350 mortality benefit seen by Salazar et al. Treatment was with a single unit of 200 mL of CCP given over 1 to 2 hours, with the exception of patients 4, 10 and 11, who received two units roughly 8 days apart (Table S1) due to severe immunosuppression, continual hypoxia with exertion and goal to discharge without oxygen. Patient 4 was on rituximab and steroids for chronic graft vs host disease following haploidentical stem cell transplantation, whereas patient 11 had T-cell acute lymphoblastic leukaemia, receiving lymphodepleting induction chemotherapy. Patient 10 was critically ill and had not responded to any other treatment (Table S1). Nevertheless, steady improvement in oxygenation levels was observed following each CCP infusion. No adverse events were reported in patients with the exception of a fever during CCP transfusion in patient 7, resulting in infusion of only 100 mL. Details of the treatment can be found in Data S1. Patient disease progression and outcomes are summarised in Figure 1 and Table S2. Overall, of the six intubated patients, three were extubated between 1 and 13 days post-CCP infusion. The other 11 patients showed a dramatic decline in oxygen needs and did not require ventilatory support. Of the 17 patients included here, 2 patients, patients 3 and 10, died in the hospital (patient 3 died 2 days following extubation secondary to progression of medical comorbidities and the family's decision to transition to comfort care, whereas patient 10 died after developing an intraparenchymal haemorrhage resulting in complete brain herniation, after which the family transitioned to comfort care). Of the 15 survivors, 14 were discharged from the hospital, whereas 1 was extubated to tracheostomy. Two patients (patients 1 and 11) with advanced haematological malignancies died at home after being discharged off oxygen with home hospice. Interpretation of the data could potentially be affected by the concomitant clinical trial enrolment of some patients. Patients 12 and 15 were enrolled in ACTT-1, a randomised, double-blind, placebo-controlled trial to evaluate the safety and efficacy of remdesivir in hospitalised adults diagnosed with COVID-19 (ClinicalTrials.gov Identifier: NCT04280705); patients 3, 15 and 17 were enrolled in REGN88, a randomised, double-blind, placebo-controlled trial to evaluate the safety and efficacy of sarilumab in hospitalised adults diagnosed with COVID-19 (ClinicalTrials.gov Identifier: NCT04327388), whereas patients 1, 2, 5, 6 and 10 received remdesivir outside of a clinical trial context, and patients 5, 6, 8, 9, 10, 15 and 16 received dexamethasone. The blinded nature of the ACTT-1 and REGN88 trials make it impossible to determine whether patients 12 and 15 received remdesivir or placebo and whether patients 3, 15 and 17 received sarilumab or placebo. Thus, it is difficult to determine whether the evaluated medication played any role in the observed clinical improvement of these patients. Patients 4, 7, 11, 13 and 14 received no additional COVID-19-directed therapies of CCP infusion. The mechanism driving improvement in patients receiving passive antibody therapy is currently unclear. Antibodies are known to work by destroying viral particles via complement activation; opsonisation; or through neutralising the virus by blocking attachment, cell entry or uncoating inside the cell cytoplasm. These mechanisms imply that the optimal timing for CCP infusion is early in the disease process when the virus is still actively replicating. Later phases of COVID-19 are characterised by widespread tissue damage secondary to uncontrolled inflammation, with minimal to no viral detection, which challenges any role of CCP. A report from the Henan province challenged the benefit of CCP if administered late in the disease course of patients with severe COVID-19.14 Although all of our patients showed improvement following CCP, the most impressive effects were seen in patients 2, 4, 5, and 7 and 13 when CCP was administered early in their disease course (days 5, 5, 5, 17 and 5 of disease, respectively). Two CCP infusions were used in three patients, and in each case, we saw an incremental improvement in patient oxygenation. Two patients were considered lymphodepleted secondary to ongoing therapy related to their underlying haematological malignancy. Therefore, those patients were less likely to mount an appropriate humoral response to SARS-CoV-2 due to B-cell depletion. Thus, a second CCP infusion was likely beneficial. Patient 10 did not show improvement following the first unit; thus, a second unit was given, with rapid improvement in oxygenation following the second unit. Although a randomised controlled clinical trial is needed to determine with certainty the role of CCP in treating severe COVID-19, our limited data represent a sign that CCP is safe and may be efficacious in COVID-19; this underscores the potential role for passive immunity in this disease. We thank our patients for participating and donating the tissue samples that made this research possible. P.P., M.I., A.N., A.M. and N.S.S. collected and analysed data and wrote the manuscript; T.H. and B.N. performed the RT-PCR/CRISPR assay; K.Z. and J.R. performed ELISA; all other authors contributed to treatment, reviewing, editing, writing of the manuscript. The authors have no competing interests. Data S1. Supporting information. Table S1. Supporting information. Table S2. Supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts