The effectiveness of control interventions against COVID-19 is threatened by the emergence of SARS-CoV-2 variants of concern. We present a mathematical model for studying the transmission dynamics of two of these variants (Delta and Omicron) in the United States, in the presence of vaccination, treatment of individuals with clinical symptoms of the disease and the use of face masks. Using current daily case data for COVID-19, we showed that the predominant Omicron variant can be eliminated if current control measures are maintained at their baseline levels. Vaccine-derived herd immunity can be achieved (so that the pandemic will be eliminated) if at least 68% of the population is fully-vaccinated. We showed that elimination is feasible by June 2022 if current baseline level of full vaccination coverage is increased by about 20%. The prospect of pandemic elimination is significantly improved if vaccination is combined with a face mask strategy that prioritizes moderately effective and high-quality masks. Having a high percentage of the populace wearing the moderately-effective surgical mask is more beneficial to the community than having low percentage of the populace wearing the highly-effective N95 masks. We showed that waning natural and vaccine-derived immunity (if considered individually) offer marginal impact on disease burden, except for the case when they wane at a much faster rate (e.g., within three months), in comparison to the baseline (estimated to be within 9 months to a year). Treatment of symptomatic individuals has marginal effect in reducing daily cases of SARS-CoV-2, in comparison to the baseline, but it has significant impact in reducing daily hospitalizations. Further, while treatment significantly reduces hospitalization, the prospects of COVID-19 elimination in the United States is more significantly enhanced if investments in control resources are focused on mask usage and vaccination rather than on treatment.