Although no direct evidence of pharmacogenomics data in patients with COVID-19 was available at the time of writing this manuscript, there are plausible mechanisms by which genetic determinants may play a role in adverse drug responses. Having diverse population genetic information and genetic databases, could help clinicians avoid additional risks for treating COVID-19 patients. While NSAIDs, including ibuprofen, may be prescribed for the management of pain and fever, controversy arose on the use of ibuprofen due to the possibility of a worse COVID-19 prognosis. In this work, several genetic markers were analyzed across diverse ethnic backgrounds to identify population differences in drug responses and toxicity events associated with ibuprofen treatment. Results from this study showed that pharmacogenomics studies can be leveraged to enhance the understanding of adverse reactions to the treatment of COVID-19 symptoms and support advancement of drug development pipelines.
Based on the CPIC updated report of the CYP2C9 guidelines [11], seven geographically-defined groups (American, Central/South Asian, East Asian, European, Near Eastern, Oceanian, and Sub-Saharan African) and two admixed groups (African American/Afro-Caribbean and Latino) are described, defined by global autosomal genetic structure based on data from large-scale sequencing initiatives [11, 24]. Geographic grouping pattern was selected as geography has historically been the greatest predictor of genetic variation between human populations, with genetic distance increasing as a function of geographic distance [7]. We also intended these groups to represent people with a predominance of ancestors who were in the region pre-diaspora and pre-colonization [24]. Pharmacogenomics studies are typically not large, from a single country or ethnic group, therefore difficult to implement or incorporate into broader research goals. Therefore, frequencies from larger groups of subjects from multiple different sampled populations are more likely to result in a more accurate estimate. However, it is noted that broadly grouping global populations is an over simplification of human diversity and does not capture complex social and cultural identities. As such, geographic grouping pattern with population group identifiers is an important component of knowledge extraction from curated literature [24].
In this study, 18 genetic markers for the CYP2C9 gene were investigated within 101 individuals of Jordanian Arab descent. The minor alleles (T) and (C) of alleles *2 and *3 (defined by rs1799853 and rs1057910) were 0.094 and 0.084, respectively, which are significantly different from frequencies observed across the additional studied populations. These variant genotypes (CT and AC) were correlated with reduced enzyme function, and therefore are associated with PharmGKB recommendations for changes to ibuprofen dosing [7]. Approximately 33% of these individuals were either intermediate (IM) or poor metabolizers (PM) of ibuprofen based on the sequence variant analysis of CYP2C9. The rs1057911 marker was found to be in LD with the variant of CYP2C9*3 (rs1057910; Fig. 1A, Table 3), which is consistent with the recently published PharmVar change adding c.1425A > T (rs1057911) to the *3 haplotype definition [44]. The rs67807361 was also significantly different from other populations, and the nucleotide BLAT search revealed a 100% sequence identity with CYP2C19. Further analysis indicated that the array probe used for genotyping was not able to bind specifically with the target SNPs, due to non-specific binding to another genomic region. Awareness of problematic regions is critical during test design and reporting to guide decisions regarding exclusion of regions and/or whether alternative assays must be used. This is particularly the case for CYP2C9*7_5080C > A(L19I) (rs67807361), where both statistical and genetic tests revealed a homologous sequence that may result in false positive or false negative variant calls.
MDS analysis showed that the Jordanian Arab population clustered with multiple regions within European and Near Eastern, particularly with Turkish, Israeli, Caucasian, Italian, Romanian, Iranian and Lebanese populations. Interestingly, this cluster showed three Turkish populations and two Iranian populations. However, all three Turkish populations were the most similar to Jordanian populations based on MDS analysis. MDS results were further validated by the pairwise Fst values, where the lowest level of differentiation was observed between the Jordanian Arab population and Saudi Arabian, followed by the Italian and Turkish. Collectively these results validated that the current Jordanian population today falls into two main groups: one sharing more genetic characteristics with modern-day Europeans and Central Asians, and the other with closer genetic affinities to Arabia [45]. In addition, the autosomal analyses are in agreement with recent studies using large-scale genomics that indicated three major genetic events related to Levant populations. During the late neolithic, gene pools across Anatolia and the Southern Caucasus mixed, resulting in an admixture cline [46]. The second event occurred during the Early Bronze Age, where Northern Levant populations, a region flanked by the Middle East and Europe, experienced gene flow in a process that likely involved a yet to-be-sampled neighboring population from Mesopotamia [46]. The most recent event for the modern Levant was largely determined by subsequent repopulations and mass movements associated with multiple cultural changes within the last two millennia. This appeared to have facilitated and maintained admixture between culturally different populations [47]. Conversion of the region's populations to Islam, appeared to have also introduced major rearrangements in the populations genetic relations with an admixture of culturally similar populations [45]. In general, the Jordanian population was not significantly different from their Levantine neighbours, and fit consistently into a Middle East-Anatolia-Balkan-Caucasus geographic and genetic continuum [48].
PharmGKB meta-analysis of several annotation studies supported by a number of clinical trials [7], showed a very strong correlation between CYP2C9 genotypes and plasma levels of ibuprofen [12–16]. PharmGKB level 1 (1A) represents CYP2C9 *2 (rs1799853) and *3 (rs1057910) alleles with a strong evidence of PK/PD alteration [7]. Mapped frequencies of these alleles showed that Central/South Asian, Near Eastern, and European populations were 7.9x more likely to show impaired CYP2C9 metabolism than Sub-Saharan populations, and 4.9x more likely than East Asian ancestry populations (Table 8). Inference that a higher proportion of East Asian and African ancestry populations have normal ibuprofen metabolism, and therefore are less susceptible to complications for ibuprofen-based treatment of COVID-19 related symptoms should be done with caution. Figure 5 shows genotype frequencies of global populations of CYP2C9*2 and *3 for more than 100,000 subjects within 412 reports (Table 8). Group boundaries for the seven geographical groups fall predominantly along national boundaries to aid the assignment of group membership. The two admixed groups of African American/Afro-Caribbean and Latino were not represented on this figure as the geographical grouping based on the location of genetic ancestors pre-diaspora and pre-colonization could not be applied to the two admixed groups.
These findings are supported by recent European reports of potential harm with ibuprofen usage in patients with COVID-19 symptoms [3]. Further supported by multiple reports, the National Agency for the Safety of Medicines and Health Products (ANSM) of France, issued a warning in April 2019 about the use of NSAIDs for patients with infectious diseases based on an analysis of 20 years of safety data of ibuprofen and ketoprofen. Consequently, the French regulatory body was concerned that existing infections might be worsened by the use of NSAIDs [49]. Furthermore, one large case-control study found an association between NSAIDs and respiratory complications, regardless of whether the NSAIDs were taken long term or as a treatment for acute illness, suggesting that the association was not simply a result of increased prescription in response to acute illness [50].
However, additional research is necessary to clarify whether further variants should be incorporated into clinical decision making. Collectively, this work demonstrates the capability and application of large-scale pharmacogenomics studies to elucidate genetic variation effects on NSAID efficacy in COVID-19 patients. Ultimately, the implementation of pharmacogenetics in clinical settings can leads to more efficient, safer and cost-effective treatments.