Systematic review of the clinical effectiveness of Tixagevimab/Cilgavimab for prophylaxis of COVID-19 in immunocompromised patients

Study design

This was a systematic review of clinical studies in peer-reviewed journal articles and pre-print articles (Prospero registration 348513).

Search strategy

Two independent reviewers conducted an electronic search strategy of two online databases, PubMed and medRxiv, including studies published in the period of 1^st January 2021 to 1^st October 2022. The initial search was performed with search terms including “Evusheld”, “tixagevimab”, “cilgavimab”, “AZD7442”, “AZD8895” and “AZD1061”. The reviewers then assessed each paper generated from the search and excluded articles firstly based on title, then abstract, then following review of the full text. References of the filtered papers were searched for additional studies. Additional studies were supplemented on the 1^st of October following the same search strategy at the request senior policy makers. Any disagreements between the reviewers were resolved by consulting a separate adjudicator and a discussion between all three parties.

Eligibility and exclusion criteria

Eligible studies had to meet the following criteria: (1) involved a clinical study-randomised controlled trials, observational studies or cohort studies, (2) prophylaxis treatment with Tixagevimab/Cilgavimab, (3) report of breakthrough SARS-CoV-2 infections. Other descriptions of clinical outcomes, including hospitalisation, ITU admission or deaths, and sera SARS-CoV-2 neutralising ability were reported as secondary outcomes.

Data extraction

Once all papers from the search had been identified the two independent reviewers reviewed the full text of all identified papers. Descriptive data for each article were identified including; study design-date accessed, title, year of publication, author, author contact details, funding sources; methodology-study aim, design, setting, start and end date, duration of participation; population-population description, setting, total number randomised, baseline imbalances, withdrawals and exclusions, age, sex (percent female), race/ ethnicity, severity of illness, co-morbidities, sociodemographic data, reason for being high risk for COVID-19; intervention-number randomised to group, description, duration of treatment period, timing, delivery method, providers, co-interventions, economic information, resource requirements, integrity of delivery; comparator(s)-number randomised to group, description, duration of treatment period, timing, delivery method, providers, co-interventions, economic information, resource requirements, integrity of delivery; outcome measures-number of COVID-19 infections, infection complications, hospitalisations, deaths and the neutralising ability of sera against SARS-CoV-19. For each outcome, the time point and duration, analysis scale and unit, number of missing participants and why, statistical methods used, and appropriateness was noted and Information for risk of bias and GRADE assessments.

Once all papers from the search had been identified the two independent reviewers reviewed the full text of all identified papers. Descriptive data for each article were identified including study design, methodology, population, intervention, comparator(s) and outcome measures. For each outcome, the time point and duration, analysis scale and unit, number of missing participants and why, statistical methods used, and appropriateness was noted and Information for risk of bias and GRADE assessments.

Systematic review analysis

The risk of bias was conducted for each randomised controlled trial by two independent reviewers. A study is deemed at overall high risk if it is rated at high risk for any domain, at low risk if it is deemed low risk for all domains, and all other studies deemed at unclear risk. The GRADE tool was used to assess the certainty of the evidence for the primary outcome, judged against the following domains: participants follow up, imprecision, inconsistency, indirectness and publication bias¹⁴. GRADE levels were assigned as follows: (-)(-)(-)(-) = Level 0, (+)(-)(-)(-) = Level 1, (+)(+)(-)(-) = Level 2, (+)(+)(+)(-) = Level 3 and (+)(+)(+)(+) = Level 4.

Data analysis

The outcomes assessed in this study included: COVID-19 infection (confirmed through any approved method of detecting COVID-19 infection including polymerase chain reaction tests (PCR) and lateral flow tests (LFTs)), admissions to hospital, including ICU, due to COVID-19 infection, mortality due to COVID-19 infection and all-cause mortality, and participant serum SARS-CoV-2 neutralising ability.

The clinical effectiveness of Tixagevimab/Cilgavimab was determined using the case-control clinical effectiveness methodology, similar to test-negative case control (TNCC) as previously reported for determining clinical effectiveness of SARS-CoV-2 vaccines¹⁵. Odds Ratios and 95% confidence intervals (CI) between intervention and control patient groups were calculated using a two-sided Fisher’s Exact test, subtracted from 1 and multiplied by 100 to calculate each clinical effectiveness percentage. Aggregated clinical effectiveness percentages of multiple studies were generated where corresponding intervention and control data was available^8,16–20. 95% confidence intervals for infection and case-outcome percentages were calculated using the Wilson score method without continuity correction²¹. Statistical analyses and data visualisation were performed within R version 4.1.2.

Study quality

The GRADE tool was used to assess the certainty of the evidence for the primary outcome in each of the nine included studies. Supplementary Table 1 provide a summary of the GRADE scores of each study, and the reasons for each negative scoring.

The PROVENT study, the only included randomised controlled trial, was assessed according to the Cochrane risk of bias assessment and deemed to have a low risk of bias (full summary in Appendix 2)^8,32.

Clinical effectiveness against breakthrough COVID-19 infections

Tixagevimab/Cilgavimab is a combination of two human monoclonal antibodies targeted against the surface spike protein of SARS-CoV-2. The antibodies have a modification to their Fc receptor to ensure a longer duration of action and enable use as a prophylactic antibody therapy⁶. As such, the primary clinical effectiveness measure is the prevention of breakthrough COVID-19 infections. We identified 17 studies that reported on this measure.

The original phase 3 licensing study was the PROVENT study. This was a randomised controlled study, randomising 5197 patients between Tixagevimab/Cilgavimab or placebo. This demonstrated a relative risk reduction of 76.7% (95% confidence interval [CI], 46.0-90.0; p<0.001) after a median of 83 days from Tixagevimab/Cilgavimab administration (150/150mg) compared to placebo. The secondary outcome findings after a median of 196 days revealed a symptomatic COVID-19 infection risk reduction of 82.8% (CI: 65.8-91.4%) ⁸.

Subsequent to the phase 3 study, sixteen real-world post-licensing studies have been performed assessing clinical effectiveness against breakthrough infections. Five were performed in reference to a control group^16–20, and reported clinical outcomes from propensity-matched U.S. army veterens¹⁶, solid organ transplant recipients¹⁷, kidney transplant recipients¹⁹ and from generally immunocompromised patients^18,20. Four studies showed a reduction in COVID-19 infections in the Tixagevimab/Cilgavimab group compared to control groups ^{16– 19}. Chen et al compared the same participant population before and after Tixagevimab/Cilgavimab, finding a reduced infection rate post intervention²⁰.

It is important to note that three studies^8,18,19 included 150mg/150mg Tixagevimab/Cilgavimab dosages, while three studies also included participants who received 300/300mg dosages^16,17. Al Jurdi et al. also included a 450/450mg Tixagevimab/Cilgavimab recipient¹⁷. Al Jurdi et al. found that, during a mean follow-up of 87 ±30 days, the rate of COVID-19 infection breakthrough was higher amongst those who received the 150/150mg Tixagevimab/Cilgavimab dose compared to the 300/300mg dose (p=0.025)¹⁷.

Of the eleven cohort studies without controls, six were prospective^26–31 and five were retrospective^12,22–25. Infection rates between 0-19% over a 0.5-4-month period were reported in these cohort studies ^12,22–31. Nguyen et al. reported a low infection rate of 4.4% (49/1112) after a median follow-up of 63 days. Of the 29 cases with viral sequencing from Nguyen’s paper, all the cases were found to be Omicron infections. At the time of the study, the mean weekly incidence rate in Ile-de-France was 1,669 infections in 100,000 inhabitants, whereas amongst the study population the incidence rate was 530 infections in 100,000 inhabitants²⁷. Breakthrough infection rates reported across each study are summarised in Table 1.

Overall, the clinical effectiveness of COVID-19 breakthrough infections in patients who received group Tixagevimab/Cilgavimab compared to a control group was 40.47% (CI 29.82-49.67) (p<0.0001).

Clinical effectiveness against COVID-19 hospitalisation

Tixagevimab/Cilgavimab is believed to reduce the severe sequelae of hospitalisation in a similar fashion to coronavirus vaccines^15,33. In the original PROVENT trial, one patient in the control group was hospitalised with a severe infection (needing non-invasive ventilation or high-flow oxygen) and four patients in the control group were hospitalised with a severe infection. There were no COVID-19 hospitalisations in the Tixagevimab/Cilgavimab group⁸.

Of the seventeen studies included in this review, sixteen have reported on hospitalisation of patients who received Tixagevimab/Cilgavimab. Six of these studies included corresponding controls^8,16–20. All of these studies demonstrated reduced rates of hospitalisation in patients receiving Tixagevimab/Cilgavimab. The rates of hospitalisations in the ten cohort studies ranged between 0-9% over a 0.5-4-month period^{12,22–25,27–31}. Hospitalisation rates reported across each study are summarised in Table 2.

Overall, the clinical effectiveness of Tixagevimab/Cilgavimab against COVID-19 hospital admissions was 69.23% (CI: 50.78-81.64) (p<0.00001).

Three papers also reported numbers of intensive care hospitalisation^12,19,20, two of which were comparative studies^19,20. Kaminski et al. reported a statistically significant reduced number in the Tixagevimab/Cilgavimab intervention group of 0.6% (2/333 cases) compared to 6.2% (6/97) in the control group¹⁹. Chen et al. reported a decreased rate of intensive care hospitalisations after Tixagevimab/Cilgavimab administration (0.6% before, 0.0% afterwards)²⁰. The clinical effectiveness against COVID-19 intensive care admission from these two studies was 87.89% (CI: 47.12-98.66) (p=0.0008).

Clinical effectiveness in preventing COVID-19 mortality

Prior to this systematic review, there was limited evidence about the clinical effectiveness of Tixagevimab/Cilgavimab in preventing COVID-19 mortality.

In the original PROVENT study, there were no coronavirus deaths in the patients treated with Tixagevimab/Cilgavimab, and two deaths in the control group (2/3461, 0.06%)⁸.

Of the real-world evaluation studies reviewed, sixteen reported on mortality in patients who received Tixagevimab/Cilgavimab, with six including controls ^8,16–20. Of these, in five studies, the rates of both COVID-19-specific deaths^8,17,19, and all-cause mortality^8,16,18 were lower in the Tixagevimab/Cilgavimab-treated groups compared to controls. Chen et al. found no difference in COVID-19-related mortality before and after Tixagevimab/Cilgavimab²⁰. In the ten cohort studies without controls, rates of COVID-19-specific deaths were very low, ranging between 0.00-0.48% over a 0.5-4-month period^{12,22–25,27–31}. Eight of these studies reported mortality rates of 0.00% ^{22–25,28–31}. Al-Obaidi et al reported an all-cause mortality rate of 0.86% after a median of 53 days and was the only non-comparative cohort study which included all-cause mortality²³. Mortality rates reported across each study are summarised in Table 3.

The overall clinical effectiveness percentages of Tixagevimab/Cilgavimab for patients in studies reporting all-cause mortality ^8,16,18 and COVID-19-specific mortality ^8,17,19,20 were 81.29% (66.93-90.28) p<0.0001 and 86.36% (CI: 6.21-99.70) (p=0.0351), respectively.

Neutralising ability of patient sera

Five included studies analysed the neutralising ability of patient sera after Tixagevimab/Cilgavimab administration^{8,12,26,28,29}. The phase III PROVENT trial showed that the neutralising antibody titres to the wild-type SARS-CoV-2 receptor-binding domain (RBD) after Tixagevimab/Cilgavimab was significantly higher than the convalescent plasma antibody titres of healthy patients from the phase I trial⁸.

However, analysis from different papers shows varying amounts of neutralisation of different SARS-CoV-2 variants. Bruel et al. found that, 15-30 days post Tixagevimab/Cilgavimab (dosage not provided), sera from 8 of the 11 participants had neutralising activity against Omicron variant BA.1, and all 11 were able to neutralise BA.2. Neutralisation of SARS-CoV-2 delta variant was nine-fold higher than BA.2 neutralisation in the 11 Tixagevimab/Cilgavimab-treated participants²⁶. After a median of 33 and 29 days of 150/150mg Tixagevimab/Cilgavimab, respectively, both Stuver’s and Benotmane’s studies showed minimal Omicron-RBD neutralisation^28,29. Stuver found that, although participant sera had high anti-spike IgG levels and effective wildtype-RBD neutralisation, the median Omicron-RBD neutralisation failed to reach the positive cut-off value²⁹. Similarly, Benotmane found that only 9.5% of patient plasma (6/63) was able to neutralise Omicron variant BA.1 RBD, compared to 71% of patients who were infected naturally with SARS-CoV-2²⁸. Benotmane et al. (2) also found that, of the five kidney transplant recipients with breakthrough COVID-19 infection tested, all 5 had minimal anti-RBD IgG levels against SARS-CoV-2 BA.1 (<3500BAU/mL)¹².

Of note, Stuver’s study showed that those who were administered two 150/150mg or a 300/300mg Tixagevimab/Cilgavimab dose had significantly higher Omicron-RBD neutralisation (p=0.003), with 90% of patient sera neutralisation levels above the positive cut-off value²⁹.