Back Back
Breakthrough Infections and Rebound
New Data on COVID-19 Treatment and Risk Factors for Severe Breakthrough Infections Post Vaccination

Released: November 17, 2022

Gregory Huhn
Gregory Huhn, MD, MPHTM
Activity Information

Activity

Progress
1
Course Completed

Key Takeaway

  • Risk factors for severe COVID-19 breakthrough infections after vaccination include being 80 years of age or older, immunocompromising medications and conditions, and chronic comorbid conditions.

Among the many interesting COVID-19 presentations delivered at IDWeek 2022, I found 2 particularly relevant when considering their potential impact on clinical practice. These included an analysis of risk factors for severe breakthrough infections and a secondary endpoint analysis of the randomized EPIC-HR study comparing nirmatrelvir plus ritonavir vs placebo in nonhospitalized individuals at high risk for severe disease.

Risk Factors for Severe Breakthrough COVID-19
Branch-Elliman and colleagues reported results from a nationwide retrospective cohort of 110,760 veterans who received the primary COVID-19 vaccination series, as well as a subset with an additional booster vaccine, and subsequently developed laboratory-confirmed SARS-CoV-2 between December 2020 and February 2022. The aim of the study was to identify factors associated with severe breakthrough infections. This type of analysis can improve patient care by helping healthcare professionals (HCPs) determine which patients are at high risk for progression to severe disease so that they can select appropriate interventions in a timely manner.

The results showed that older age had the strongest association with severe breakthrough infection, with an adjusted odds ratio of 16.1 (95% CI: 13.1-19.9) for those aged 80 years or older vs patients aged 45-50 years. Immunocompromising medications and conditions (cytotoxic chemotherapy or glucocorticoid use after vaccination and leukemia or lymphoma) also were associated with an increased risk of severe breakthrough infection, as were chronic comorbid conditions (heart failure, dementia, chronic kidney disease). Of note, receipt of a vaccine booster was associated with a 50% lower risk of severe breakthrough infection.

When investigators applied the identified factors to a predictive model, the area under the receiver operating characteristic curve approached 0.84, demonstrating a high probability of prediction accuracy. Investigators plan to use the model to develop prediction tools for web-based application to guide HCPs and patients toward personalized management strategies in the vaccination era. Automation will enable machine learning to continually update the tool with new data that emerge over time, for example, as data accrue with use of the new bivalent boosters.

Some of the attributes that made this a particularly relevant study were that it was based on the large Veterans Affairs cohort dataset, which includes both immunocompetent and immunocompromised individuals, as well as a considerable proportion of individuals older than 65 years of age, allowing for reliable identification of risk factors for severe breakthrough infections. For me, the most exciting aspect of this study is that investigators will be rolling out a predictive tool informed by machine learning that will allow both HCPs and patients to assess risk profiles and make determinations on the best intervention according to that risk.

An additional interesting finding comes from the fact that the analysis included infections that occurred during both the delta and omicron periods, with ~29,000 during the delta period and ~78,000 during the omicron period. The numbers of severe breakthrough infections and hospitalizations were equivalent in both periods—736 during delta and 728 during omicron—despite the much larger number of omicron infections overall, supporting the previously recognized lower disease severity of omicron vs delta in this very large database.

EPIC-HR: Nirmatrelvir Plus Ritonavir vs Placebo for Nonhospitalized Patients With COVID-19
The EPIC-HR study is a double-blind phase II/III trial comparing nirmatrelvir plus ritonavir vs placebo in unvaccinated nonhospitalized adults with confirmed SARS-CoV-2 with symptom onset ≤5 days before randomization and >1 characteristic or condition associated with high risk of severe COVID-19. The analysis presented by Hammond and colleagues reported on secondary endpoints through Week 24, including COVID-19–related ICU admission, oxygen support, COVID-19–related medical visits, safety, and mortality.

Results showed an 86% relative risk reduction (RRR) in COVID-19–related hospitalizations or all-cause mortality through Day 28, a 73% RRR in any COVID-19–related medical visit through Day 34, and an 81.5% RRR in requiring oxygen support with nirmatrelvir plus ritonavir vs placebo. In addition, nirmatrelvir plus ritonavir reduced symptom severity and shortened symptom duration by 2-3 days vs placebo. Of importance, at Week 24, there were no deaths in the nirmatrelvir plus ritonavir arm (n = 1039) vs 15 in the placebo arm (n = 1046; P <.0001).

Another interesting finding from this study is that the rates of rebound as defined by detectable polymerase chain reaction upon treatment completion and after reduction in viral load were 2.3% in the nirmatrelvir plus ritonavir arm vs 1.8% in the placebo arm. The rate seems low compared with the many anecdotal reports that have emerged of people experiencing rebound with nirmatrelvir plus ritonavir therapy. In a National Institutes of Health report by Epling and colleagues evaluating 8 individuals with symptom rebound (6 after receiving nirmatrelvir plus ritonavir, 2 without prior antiviral therapy) and 7 with early acute omicron infection, those with rebound had more robust SARS-CoV-2–specific T-cell responses vs those with early acute COVID-19. Based on this information, it is possible that vaccinated individuals with robust T-cell responses may have a higher rebound rate than was observed among the unvaccinated population in EPIC-HR. 

To me, the relevance of these EPIC-HR findings is that they extend outcomes from the previously reported primary endpoint analysis of EPIC-HR at Day 34 to longer-term follow-up to Week 24. For limitations, the study was conducted during the delta period, as well as in unvaccinated individuals.

Data for another critical area of unmet need that requires long-term endpoints, namely Long COVID, were not captured in the EPIC-HR study, and enrollment in the follow-up study in patients at standard risk for severe disease, EPIC-SR, was ceased because of a low rate of hospitalization or death in the standard-risk population. We may therefore need to look to large retrospective cohorts in clinical care to evaluate the effectiveness of antiviral interventions in lowering the risk of Long COVID symptoms. Encouragingly, recent prepublication data from the US Department of Veterans Affairs (VA) investigators suggest that among people at high risk for severe COVID-19 who received nirmatrelvir plus ritonavir within 5 days of a positive SARS-CoV-2 test (N = 9217) from March 1, 2022 to June 30, 2022 during the omicron era, the risk of Long COVID symptoms was reduced by 26% (hazard ratio: 0.74; 95% CI: 0.69-0.81) compared to a control group of patients at the VA who received no antiviral or monoclonal antibody therapy (N = 47,123), regardless of vaccination or previous COVID-19 infection status.

Finally, multiple omicron BA5 subvariants will soon exceed 50% of all new SARS-CoV-2 infections in the United States, and available monoclonal antibodies are not effective against these sublevel variants. Until further drug discovery again catches up with new monoclonal antibody constructs capable of neutralizing the expanding evolution of these SARS-CoV-2 subvariants, we have moved now at this stage of the pandemic to rely primarily on antiviral therapies for COVID-19 ambulatory treatment, further underscoring the importance of studies of antivirals.  

Your Thoughts?
How will new COVID-19 data presented at IDWeek 2022 affect your clinical practice? Join the discussion and share your thoughts by posting a comment.