Questions concerning the New Haven Coronavirus

To the Editor—Esper et al. present the discovery of a novel human coronavirus (HCoV) in young children and infants with respiratory tract disease in New Haven: HCoV-NH [1]. However, they also mention that the virus is very similar to HCoV-NL63, a virus that was identified previously in Amsterdam, The Netherlands [2]. Despite this, the 2 studies by Esper et al. [1, 3] and an Editorial Commentary by McIntosh [4] avoid usage of the name “HCoV-NL63” while repeatedly claiming the discovery of a novel virus

To judge whether HCoV-NH is really a novel HCoV, a comparative analysis of HCoV-NH with a number of different features of established HCoVs should be performed. Examining the relatedness of genome sequences is one facet of such an analysis. Unfortunately, limited data on the genome sequence for HCoV-NH are available, but inspection of a 126-bp fragment clearly shows that all HCoV-NH isolates cluster together with the HCoV-NL63 Amsterdam-1 strain (figure 3 in [1]). This result strongly suggests that the viruses found by Esper et al. are New Haven isolates of HCoV-NL63. Moreover, the actual nucleotide difference between the New Haven isolates (GenBank accession nos. AY870943–AY871008) and the HCoV-NL63 isolate Amsterdam-1 (GenBank accession no. NC_005831) is 0%–6%. This degree of difference falls well within the range of genetic variation observed among different HCoV-NL63 isolates from Amsterdam [2]. We reported the presence of distinct HCoV-NL63 variants that apparently are cocirculating, as has been confirmed recently by Arden et al. [5] and Bastien et al. [6]

Esper et al. do not seem to dispute that HCoV-NH is very similar to HCoV-NL63. What then made them decide to claim the identification of a novel virus?. The only argument mentioned is that the research project was initiated before the first article on HCoV-NL63 was published. Is that how it works in science?. No—only the first report can claim a novel scientific finding. In fact, for HCoV-NL63, Esper et al.’s is the third article that claims its discovery. The identification of HCoV-NL63 was first announced in an article in Nature Medicine (which was published electronically on 21 March 2004 [2]), and an article by Fouchier et al. in the Proceedings of the National Academy of Sciences of the United States of America described the same virus [7]. Esper et al. submitted their manuscript to the Journal of Infectious Diseases on 7 September 2004, about 6 months after the initial description of HCoV-NL63 was published. Other studies on HCoV-NL63 were submitted for publication in 2004 (and some were published earlier than the one by Esper et al.), but all appropriately acknowledged the discovery of HCoV-NL63 and used its nomenclature, and none claimed the discovery of a novel virus [5–9 ]. In fact, the issue of the Journal of Infectious Diseases in which the study by Esper et al. was published also contains a survey on the prevalence of HCoV-NL63—and indeed, it appropriately names the virus—in Canada [6]

The Esper et al., Fouchier et al., and Arden et al. articles claim in their titles that HCoV-NL63 is associated with respiratory tract disease [1, 5, 7]. However, this conclusion can be drawn only when an appropriate patient control group without respiratory tract disease is included. These studies did not include such a control group, and it is therefore premature to conclude that HCoV-NL63 is associated with disease. The suggestion may be strong, but much work remains ahead of us to accurately define the spectrum of respiratory tract disease caused by HCoV-NL63. The study by Esper et al. on the association between HCoV-NH and Kawasaki disease did include the proper control group, and this finding is very intriguing indeed [3]

The detection of HCoV-NL63 in The Netherlands, Canada, Australia, Japan, and the United States convincingly demonstrates the worldwide distribution of HCoV-NL63. Proposing a different name for basically the same virus each time it is “rediscovered” needlessly complicates the HCoV literature