26 August 2021 Episode 165 Update on inactivated COVID vaccines

Dear colleagues,

Much has been written about the inactivated vaccines, but mostly in non-scientific papers.  Therefore, I tried to collect the most relevant scientific papers on the inactivated forerunners

• CoronaVac,  formally known as PicoVac from Chinese Sinovac Co;
• BBIBP-Corv from Chinese Sinopharm Co
• Covaxin, formally known as BBV-152 from the Indian Bharat Co; 

Some of the papers (when still in preprint form) have already been briefly discussed in previous episodes, but they are now (as peer-reviewed papers) put in perspective with all available published evidence, including recent preprints on “real world” data. The quality of many of those publications remains questionable, but I hope that you bear with me to see if my conclusions at the end make sense….

Note: Similar vaccines are in development in Cuba, Mexico, Iran, Egypt… and Europe (Valneva).  I will not discuss those here, since there isn’t any body of scientific papers on them.

CORONAVAC (Sinovac)

Ep 165-1: Gao et al were the first authors to publish very nice preclinical data already in April 2020 in bioRxiv and in July 2020 in Science.  Fig 1 shows that it is a very traditional cell grown and beta-propiolactone inactivated vaccine, formulated with Al(OH)3 . It induces neutralizing Ab against the prevalent “strains” early 2020 and protects well in a macaque model at doses 3 and 6 µg three times (0,7, 14 d).

Ep 165-2: Presentation at WHO on human trials. Some highlights

• Slide 7 shows variable protection against symptomatic COVID: from 50 % in Brazil, over 65-67 % in Chile and Indonesia to 80 % in Turkey.  Protection against hospitalization 85-100 %.
• Slide 12: very favorable side effect profile: mostly mild and local.

Ep 165-3: Preprint data from Brazil:  Clearly NOT a formal phase 3 study, rather observational.  In Sao Paolo, during a rising phase of the epidemic January-Feb 2021 with about half of the isolates belonging to P1 lineage, health-care workers (HCW), some with very high risk, were vaccinated twice with “600 SU (?)” 2-4 weeks apart.  Cases in HCW were compared with general population and showed about 50 % protection against symptomatic disease over about 5 weeks after vaccination.

Ep 165-4: Lancet published formal phase 3 study in Turkey, performed between Sept 2020 and Jan 2021.  Two doses of 3 µg 14 days apart with over 6500 participants in vaccine arm and over 3500 in placebo (AlOH). Only subjects below 60 years old. Median follow-up 43 days. Over 10,000 subjects in per-protocol analysis:

• 9 symptomatic cases in vaccine and 32 in placebo: 83 % efficacy
• 0 hospitalizations in vaccine and 6 in placebo: 100 % efficacy

Ep 165-5: Jara in NEJM: prospective cohort study on 10 million Chileans Feb-May 2021, (P1 or gamma and Lambda VOC were dominant, not delta). Dose not specified, but presumably 3 µg (see https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30987-7/fulltext)

Overall: the adjusted vaccine effectiveness (VE) was 65.9% for the prevention of Covid-19 and 87.5%

hospitalization, 90.3% ICU admission, and 86.3% Covid-19–related death

In the subgroup of fully immunized persons > 60 years: very similar VE

66.6% for the prevention of Covid-19 and 85.3% for hospitalization, 89.2% for ICU admission, and 86.5% for the prevention of Covid-19–related death

Ep 165-6: Hitchings in Lancet Regional Health, using a “test-negative case-control”* approach after one or two doses of CoronaVac in HCW during the gamma epidemic in Manaus, Brazil (Jan-April 2021): vaccine efficacy (VE) against symptomatic COVID 

• 49.6%, VE  during the period 14 days or more after the first dose.
• Very remarkably: VE only 36.8%,during the period 14 days or more after the second dose.
• And: vaccinated individuals were much more likely to be infected than unvaccinated individuals in the period 0-13 days after first dose (aOR 2.11, 95% CI 1.36-3.27) suggests  unmeasured confounding.

* Test-negative studies recruit cases who attend a healthcare facility and test positive for a particular disease; controls are patients undergoing the same tests for the same reasons at the same healthcare facility and who test negative

Ep 165-7: The same authors, using the same strategy in BMJ estimated the VE of CoronaVac in almost 44,000 older subjects (>70 yrs) of Sao Paolo during the same period (Jan-April 2021). In this case VE after 2nd dose was reported:

• Against symptomatic COVID: 25 % < 14 days after 2nd dose and 47 % > 14 days
• Against hospitalization: 55 % and death 66 % > 14 days after 2nd dose.

Clear age effect:

Ep 165-8: Medeiros-Ribeiro report from Sao Paolo in Nat Med on the safety and immunogenicity of CoronoVac in patients with auto-immune rheumatic diseases. As expected: 

- Seroconversion was lower (70.4 versus 95.5%) and neut rates are lower (56 vs 79 %).

- Subjects who are older or receive immunosuppressive treatment have lower titers.

Ep 165-9: Soysal from Turkey in Hum. Vacc Immunth. That previously infected HCW can safely receive CoronaVac and develop about double antibody titers than naïve HCW.  A finding consistent with earlier observations in mRNA recipients.

Ep 165-10: Wey Wen Lim from Hong Kong in a brief letter to Lancet reports 10X  lower immunogenicity of CoronaVac as compared to Pfizer in HCW.  The average 50 % titer of neutralization was 269 in 12 full Pfizer recipients and only 27 in CoronaVac recipients.

Conclusion on CoronaVac:

• Much less immunogenic than mRNA. 
• In “high risk” populations (HCW, institutionalized elderly Brazil) maybe about 50 % effective against COVID.  Higher efficiency against hospitalization and death.
• In general population (phase 3 Turkey, country-wide cohort Chile): more effective in 65-80 % range against COVID and 86-100 % against death.
• Clearly waning with age in institutionalized elderly (Brazil); but less clear in general population (Chile).   

BBIBP-CorV (Sinopharm): less info available on real world effectiveness

Ep 165-11: A very nice preclinical paper by Wang in Cell shows that this vaccine, based on the HB02 strain, is very similar to CoronaVac.  It was used at 2 µg twice in macaques.

Ep 165-12 A: Xia in Lancet Infect Dis Oct 2020 shows good results in phase 1 and 2 using inactivated HB02 strain with 100 % seroconversion and high neutralizing titers, especially with 2 doses of 4 µg separated by 21 or 28 days.

Ep 165-12 B: Xia in JAMA Sept 2020 presents results on the inactivated WIV04 strain in doses of The

• A longer interval (21 days and 28 days) between the first and second injections produced higher antibody responses compared with a shorter interval schedule (14-day group).
• Not much difference between low (2.5 µg ), medium (5 µg) and high (10 µg) dose
• Ab titers started to rise after 2nd dose and further increased after the third injection, suggesting the need for a booster injection. 

Ep 165-13:  Al Kaabi present data on phase 3 in the United Arab Emirates in JAMA.  In this study, 2 inactivated vaccines: 4 µg HB02 and 5 µg WIV04 were compared with placebo: 2 IM injections 21 days apart.   The participants were almost all < 60 yrs and 80 % male.  Median follow-up 73 days.

• VE against symptomatic COVID was similar: 73 % WIV04 and 78 % HB02
• Only 2 severe COVID in placebo group and none in the WIV04 and HB02 groups.

Ep 165-14: Ferenci from Hungary preprint on neutralizing Ab after 2 doses of Sinopharm vaccine:

• Present in about 90% of individuals below 50 years, decreasing with increasing age: above the age of 60, any protective antibody up to 50% at older ages
• Titres after Sinopharm vaccination were almost an order of magnitude lower than after the Pfizer/BioNTech vaccine; where all individuals had at least some protective Ab.

( = Very similar findings as with CoronaVac Ep 165-10).

Ep 165-15: Jeewadara from Sri Lanka preprint on Sinopharm

• High percentage (95%) of seroconversion, but also decreasing with age.
• Neut titers comparable with convalescent sera: some difference between VOC (but only recombinant receptor binding domain assay used; no real viral; no cell culture).
• Clear boosting effect of vaccine in previously infected subjects.

Ep 165-16: Maa in Cell Discovery: extensive profiling and compairson of Sinopharm vaccinated and convalescent sera by microarrays.

• Overall titers are lower in vaccinees
• Some difference in fine specificity

Ep 165-17: Tingting Liu in Science Bulletin investigates 400 HCW before and after Sinopharm: this vaccine does not influence the profile of antiphospholipid antibody and anti-PF4-heparin antibody nor increase the risk of thrombosis

(Well done, but with Adeno vaccines, the risk is only < 1/100,000, so can you prove anything with 400 subjects). 

Conclusions BBIPB-Corv: very similar observations as with CoronaVac:

• 10 times lower immunogenicity than mRNA
• Good results in general population
• Lower titers in elderly

BBV152 or Covaxin (Bharat)

Ep 165-18: Ganneru in iScience describes the production: similar to the previously mentioned Chinese vaccines: based on cell factory cultivated Whuan strain , beta-propiolactone inactivation, BUT use of Algel + Toll-like receptor 7/8 agonist IMDG or Imidazoquinoline (which effectively stimulates T helper 1 responses)

Ep 165-19: Mohandas in iScience shows that 3 doses of 3 µg BBV152 + Algel-IMDG protect hamsters from SARS-CoV-2 pneumonia, better than 6 µg BBV152+ Algel alone.

Ep 165-20: Yadav in Nat Comm shows that two doses of each vaccine formulations (either 6 µg + Algel alone; 3 µg + Algel-IMGD or 6 µg + Algel-IMDG)  were able to protect rhesus macaques from viral challenge, in the sense that no replicating virus could be found in the upper or lower respiratory tract 1 week later.

Ep 165-21: Ella in Lancet Infect Dis confirms in phase 1-2 human trial the superiority of 2 doses of 6 µg BBV152 + Algel-IMDG with close to 100 % seroconversion and neut Ab, that declined 5-fold 3 months later.  No problems with tolerability.

Ep 165-22: Ella in medRxiv July 2021 phase 3 between Nov 2020 and Jan 2021 in almost 25,000 Indians: 50-50  vaccine (6 µg +Algel-IMDG)  or placebo (Algel alone) in 2 doses with 28 days

• Incidence of adverse events observed over a  median of 146 days was lower than that observed with other COVID-19 vaccines.
• Protection against:
  • Asymptomatic 64 %
  • Symptomatic 78 %
    • < 60 yrs 79
    • > 60 yrs 68 %
    • With co-morbidities: 66 %
  • Against severe disease: 93 %
• Efficacy against Variants:
  • Alpha 75 %
  • Kappa (B.1.617.1) 90 %
  • Delta (B.1.617.2): 65 %
  • Other: 73 %

Ep 165-23:   Fernando in KIReports on Covaxin and Covishield (ChAdOx1-COVID-19 = Astra-Zeneca) in Indian hemodialysis patients:   60 patients received at least 1 dose, and 42 patients received 2 doses

• 92 % seroconvert with high titers (> 250  U)
• 5/60 patients developed COVID after first dose, with hospitalization but no mortality
• 2/42 developed COVID 2 weeks after 2nd dose, with hospitalization and 1 died.

The authors conclude that both vaccines are excellent  and protect > 90 % ……!?

→  Very weak evidence

Ep 165-24: Farinholt in medRxiv about delta breakthrough in Houston: a fully Covaxin-vaccinated and PCR(-)  couple attended a wedding and got sick.  They also infected 4 other family members, 2 fully Pfizer and 2 Moderna vaccinated, although wedding in tent in open air and 2 people died (1 vaccinated with Covaxin and 1 with Pfizer) ….

→ Rather frightening case report

Ep 165-25: Sharma in medRxiv on breakthrough infections in Indian HCW after 1 or 2 doses of Covaxin or Covishield during the delta epidemic by telephone interview:

• 37 of 325 (11 %) of HCW experienced  breakthrough > 14 days after 2nd dose (median 47 d)
• 20 (6 %) got infected between the first dose and before 14 D after 2nd dose
• Those who were NOT previously infected, only vaccinated, had a 3.8 X higher risk of breakthrough.  →Previous infection provides extra level of protection over vaccination.
• Slightly higher, but non-significant, infection risk with Covaxin (19%) as compared to Covishield (15 %). 

 → Purely observational study, based on telephone interviews, no control of unvaccinated.

Ep 165-26: Guha in medRxiv tries to find if there is a difference between Covishield and Covaxin vs unvaccinated subjects with regard to in hospital mortality.  The numbers are difficult to read, but there was no clear-cut different mortality according to vaccination status.  Obviously, this is a selected population, with various co-morbidities and after hospitalization.  Therefore the data are very difficult to interpret without a larger control group.

Ep 165-27:  A more interesting study by Jagadeesh Kumar in medRxiv on 1161 hospitalized patients with delta infections, of whom about 40 % vaccinated with either Covaxin or Covishield shows that fully vaccinated subjects had

• Higher levels of neut Ab lower ferritin and LDH (the latter two parameters biomarkers of severity)
• Clinical severity (3.2 vs 7.2 %), need for ventilation (2.8 vs 5.9 %) and mortality (1.5 vs 3.45 %) were all significantly lower in the vaccinated vs unvaccinated group, despite the fact that the vaccinated subjects were on average older and presented with more risk factors. 
• Comparing  Covaxin and Covishield: no clinical difference, but Covishield higher Neut Ab, lower ferritin and higher leukocyte count

Their conclusion: COVISHIELD and COVAXIN offer protection and have comparable efficacies against the Delta variant resulting in (about 50%) reduction of the disease severity and mortality among the hospitalized patients with two doses of vaccination.

Ep 165-28: Gosh in medRxiv reports on sequencing of 36 breakthrough (BT) infections between 29 March and 15 June, confirming that delta was found in 80 %. In fact all 9 Covaxin BT were delta and 20/27 Covishield BT.  The other 7 BT were 2 kappa, 1 B1, 1 AY3, 1 B 1.57 and 1 B 1.629. 

Clinical data are incomplete, but  only 1 hospitalization (with co-morbidities) was noted and there were no known cases of prior SARS-CoV-2 infection.

→ Again very observational and difficult to interpret without controls, but seems to follow the general trend of the Indian epidemic at that time. 

Ep 165-29: Kumar Singh compared the seroconversion and titers after full Covaxin vs Covishield vaccination in HCW:  

• Combined both vaccines showed 95% seropositivity to anti-spike antibody, 21-36 days after the second completed dose
• However both seropositivity rate (98 vs 80 %) and titer (127 vs 53 U/ml) were clearly higher in Covishield vs Covaxin
• Women higher titers than men; Age > 60 and co-morbidity (diabetes) associated with lower titers.

Ep 165-30: Rajni Kant in medRxiv reports on “accidental” heterologous prime with Covishield and boost with Covaxin.  As compared to both homologous schemes, the safety profile is similar, but the antibody responses, including neut versus alpha, beta and delta are clearly superior with the heterologous scheme.

Conclusions Covaxin:  

Covaxin has the theoretical advantage over BBIBP-Corv Sinopharm and CoronaVac Sinovac of a more powerful Th1 skewing adjuvant, which may engage more T cell responses that could increase protective capacity on the short run and improve immune memory on the long run.

The results of the phase 3 Covaxin trial, which was carried out before the delta epidemic, are, in fact very comparable with the Turkish CoronaVac and the UAE trial with the two different BBIBP Sinopharm vaccines, with about 75 % protection against symptomatic COVID and > 85 % protection against severe disease.

Data in real world are very difficult to interpret, but:

• Some evidence of about 50 % protection against severity and death in hospitalized patients
• Rather high levels of breakthrough infections.
• Decreasing effect with age and co-morbidities, as expected.

Covaxin seems less immunogenic than Covishield (= Indian Astra-Zeneca Chimp Adeno vaccine), which is known to be less immunogenic than mRNA.   There is also some evidence that Covishield offers clinical benefits in the real world.

The (accidental) observation of the much better immunological response of Covishield prime – Covaxin boost, also against variants, may  be important for the Indian subcontinent, where this “mixing” could improve protection at population level at no  additional cost.  

I hope you’re still with me and enjoyed this super-long episode.

Best wishes,

Guido