Astra Zeneca clinical trial data review

Source

Assessment report

29 January 2021 EMA/94907/2021 Committee for Medicinal Products for Human Use (CHMP)

COVID-19 Vaccine AstraZeneca

Common name: COVID-19 Vaccine (ChAdOx1-S [recombinant])

Placebo

Brazil, UK: MenACWY (meningococcal group a, c, w-135, and y conjugate vaccine)

South Africa: Saline

Efficacy

Prevention, against hospitalization, against serious disease and death

Abbreviations: CI = Confidence Interval; ICU = Intensive Care Unit; NE = Not Evaluable; VE = Vaccine Efficacy. Source: Supplemental Tables IEMT 207.1, IEMT 207.2, IEMT 207.3, IEMT 207.4, IEMT 207.5.

COVID-19 events are adjudicated events based on virologically- confirmed results from RT-PCR or other nucleic acid amplification test. The 4 to 12 weeks dosing interval range corresponds to ≥ 28 days to ≤ 84 days.

COVID19: VE of AZD1222 versus control, the 95% CI and p value were estimated based on Poisson regression with robust variance including the term of study code, treatment, age group at screening (18-55 years, 56-69 years, and >=70 years) as covariates as well as the log of the follow-up time as an offset.

Hospitalization: The maximum likelihood estimate of VE of AZD1222 versus control, the exact 95% CI (or 97.5% onesided) and p value were estimated based on stratified Poisson regression with Exact Conditional Method including treatment as factor, study code and age group at screening (18-55 years, 56-69 years, and >=70 years) as strata factors as well as the log of total number of participants for each combination of treatment and strata VE is defined as 1-(incidence from the AZD1222 arm / incidence from the control arm) expressed as a percentage, where the risk ratio is derived from Poisson regression with robust variance.

Efficacy against hospitalization

1-(hospitalization incidence from the AZD1222 arm / hospitalization incidence from the control arm) expressed as a percentage,

(1- (0/0.15)) = 100%

Vaccine Efficacy

The 95% CI for the VE is obtained by taking 1 minus the 95% CI of the risk ratio derived from the model. The observation period for the endpoint was 15 days post second dose up to 1 year in study.

Benefits

Preventing Covid-19

Base on positive on PCR test with primary symptoms (similar to a common cold) from incidences of 2,96 % (154⁄5210) in the placebo group versus 1.22% (64⁄5258 individuals) in the inoculated arm.

Preventing hospitalization

Based hospitalization incidence of 0.15 % (8⁄5210 individuals) in the placebo group versus o incidence in the inoculated arm

Preventing Severe disease

Cannot be assessed (Not applicable due to low incidence)

Preventing ICU

Cannot be assessed ( Not applicable due to low incidence)

Preventing death

Cannot ne assessed (Not applicable due to low incidence)

Safety

Risk assessment through incidences of adverse events.

Adversity

Adversity is calculated as 1-(AE incidence from the control arm / AE incidence from the AZD1222 arm) expressed as a percentage.

Incidence values: N values of the adverse event were used instead of actual incidence since the N value for the AZD122 arm is almost equal to the N values of the control arm.

We set the threshold of incidence to less than 0.1%:  Adversity for Adverse event with incidence of less than 0.1% is insignificant.

	AZD1222 Inoculated N=12021	Control N=11724	Adversity %
Participants with any investigational product Related Unsolicited AE	3570	2172	39.16
*System organ class uncoded*	65	42	35.38
Infections and infestations	68	70	–2.94
Blood and lymphatic system disorders	28	28	0.00
Immune system disorders	3	1	66.67
Metabolism and nutrition disorders	30	13	56.67
Psychiatric disorders	28	14	50.00
Nervous system disorders	1117	644	42.35
Eye disorders	32	19	40.63
Ear and labyrinth disorders	8	18	–125.00
Cardiac Disorders	13	3	76.92
Vascular disorders	34	32	5.88
Respiratory, thoracic, and mediastinal disorders	128	121	5.47
Gastrointestinal disorders	323	184	43.03
Skin and subcutaneous tissue disorders	118	73	38.14
Musculoskeletal and connective tissue disorders	1081	448	58.56
Renal and urinary disorders	3	2	33.33
Reproductive system and breast disorders	5	4	20.00
General disorders and administration site conditions	2813	1505	46.50
Investigations	96	31	67.71
Injury, poisoning and procedural complications	11	10	9.09

The risks

Mostly adverse in terms of:

Cardiac Disorders (76.92%)

Musculoskeletal and connective tissue disorders (58.56%)

Metabolism and nutrition disorders (56.76%)

Psychiatric disorders (50%)

Uncertainties and limitations about unfavorable effects

Citations of main limitations

1

Long-term safety is considered a missing information in the Safety specification in the RMP, and will be characterised as part of the continuation of the pivotal clinical trial, other trials and a PASS.

2-

Safety data in participants with severe immunodeficiency, or participants with severe underlying disease
(including autoimmune or inflammatory disorders) are lacking, as all these populations were excluded from
the studies. The safety of AZD1222 in immunocompromised subjects will be evaluated post-authorization.

3-

Further, there is only very limited clinical experience in pregnant women, with 14 pregnant women in the
safety database who were exposed to AZD1222. Data from non-clinical studies do not indicate any harm
during pregnancy.* In the absence of clinical data to confirm lack of risks, risks during pregnancy remain,
albeit theoretical.

*there is no indication about the source of this affirmation; what non-clinical studies are they referring to?

Conclusion

It is questionable weather the benefit outweighs the risk since, to prevent SARS-CoV-2 infection with primary symptoms, like the common cold (100% efficacy) and possible hospitalization, one would take the risk of adversity in terms of cardiac disorders ( 76.92% ), musculoskeletal and connective tissue disorders (58.56% adversity ), metabolism and nutrition disorders (56.76% a) and psychiatric disorders (50%). Additionally, individuals with severe immunodeficiency, or with severe underlying disease (including autoimmune or inflammatory disorders) were excluded from the trial and a insignificant number of pregnant women in the database who were exposed to AZD1222. For these demographics, safety data are missing; safety and risks remain albeit theoretical. Overall, there is no telling what the future holds, these results being based on short-term clinical trial data.