Paradigm Shift Intervention Monitoring
twitter
Live feed of underlying pandemic map data here
Commentary
Community
Transmission of pH1N1 Tamiflu Resistance In Israel
Recombinomics
Commentary 12:50
May 18, 2010
The above comments are from the upcoming paper entitled “Community-acquired Oseltamivir-Resistant Pandemic (H1N1) 2009 in Child, Israel”. The child was subsequently treated with oseltamivir, but the sample collected prior to the start of treatment was H274Y positive. Since the brother was the likely index case in the familial cluster, and had not been treated, he likely also had been infected with pH1N1 containing H274Y.
The H274Y detected in the younger brother was in a mixture with wild type, and detected with a sensitive assay directed at H274Y. Since the confirmation was reported after the start of Tamiflu treatment and the patient’s condition improved, detection of H274Y did not change the antiviral treatment. Thus, although the detection of H274Y was not related to treatment, the only matching category for this patient in WHO and CDC reports would be detection of H274Y in a patient treated with Tamiflu.
The train passengers in Vietnam, who were treated with Tamiflu after sample collection, represented a similar situation. H274Y was detected through routine screening of the samples months after the patients recovered. It appears that these patients, who had H274Y prior to treatment, would also be categorized as H274Y in patients treated with Tamiflu.
The authors of the above paper, as well as the paper on the train passengers, noted the low number of patients who were H274Y positive and not linked to treatment. However, this low number appears to be based on a classification system that is factually correct, but misleading to most, including physicians and researchers in the field. The situations in Israel and Vietnam are common, since testing for H274Y is not done in real time, and even when in-house testing is in place, patients are still started on Tamiflu prior to receipt of test results. Thus, patients with H274Y and no treatment are largely limited to patients that are tested but not treated, which is rare because most hospitalized patients are treated and most non-hospitalized patients, like the index case in this cluster, are not tested.
Moreover, the detection of H274Y in treated patients is usually linked to a short time after the start of treatment, supporting the presence of a mixture, which becomes H274Y positive, when the ratio favoring H274Y is increased due to treatment. Thus, prophylactic patients develop symptoms 5-6 days after the start of treatment, and infected patients convert to H274Y positive after a few days of treatment.
Even closely monitored immuno-suppressed patients do not require long term treatment to develop detectable H274Y. The first two reported cases were in Washington state. Although the MMWR presented the cases as a warning for the development of resistance in immune-suppressed patients after extensive treatment with antivirals, the paper provided no data to support long term development. Although only wild type pH1N1 was detected prior to treatment and shortly after treatment, one patient was H274Y positive at day 10 and another was at day 18. However, for the first patient no sample was collected between day 4 and day 10, and for the second patient, samples were collected but not tested. Only results for day 1 and day 18 were reported. Thus, there was no evidence against development of detectable resistance at day 5 or 6, as had been seen in previously health contacts who had been prophylactically treated and developed symptoms on day 5 or 6. Moreover the H274Y positive sequences did not show a mixture, indicating detectable H274Y was present prior to the reported collection day, and the two sequences were from the same sub-clade, supporting circulation of H274Y in a local sub-clade at levels that were not detected by routine surveillance.
The cases in Washington state were followed by a report on immune-suppressed Maryland patients. Samples were collected and tested at positive for H274Y at 9 and 14 days post treatment, but samples were not collected between days 4 and 9. The Maryland paper emphasized rapid development of resistance, but did not address detection failures of a minor population or lack of sample collection at earlier dates.
Thus, there is no data arguing against a minor subpopulation that rises to the detectable level 5-6 days post treatment as was seen in prophylactic patients.
This scenario is supported by the Israeli and Hong Kong patients who had detectable H274Y in spite of a lack of treatment and the H274Y was found in association with wild type pH1N1.
The only argument for development of H274Y in response to treatment was made by Roche, who cited older examples in Japanese children treated with a sub-optimal dose. These cases were largely limited to H3N2 and involved multiple genetic changes that were not transmitted. In contrast, the pH1N1 cases, like the seasonal H1N1 cases only involve a single change, H274Y, which is transmitted and detected in patients not treated with anti-virals.
The vast majority of 2010 pH1N1 sequences in Japan now have H274Y and isolates cluster in phylogenetic analysis, indicating H274Y is widely transmitted in Japan. Thus, the spread of H274Y in untreated pH1N1 patients is following the pattern season during the fixing of H274Y in seasonal. Recombination facilitates the jump of H274Y from one genetic background to another, which eventually leads to increases through clonal expansion and genetic hitchhiking.
Media Links
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings