Background. Influenza A virus subtype H1N1 with the H274Y mutation emerged and spread worldwide. However, the clinical effectiveness of the neuraminidase inhibitors, oseltamivir and zanamivir, has not been adequately reevaluated.
Methods. Data from 164 patients with H1N1 virus infection and 59 patients with H3N2 virus infection during the 2008–2009 influenza season and 68 patients with H1N1 virus infection during the 2007–2008 influenza season who received a neuraminidase inhibitor were analyzed. The duration of fever (body temperature ⩾37.5°C) after the first dose of oseltamivir or zanamivir and from onset of symptoms was calculated from patient reports. The influenza virus was isolated, and its subtype was determined by hemagglutinin inhibition assay and polymerase chain reaction. The H274Y neuraminidase mutation status was determined by sequencing the neuraminidase segment.
Results. Of 68 patients with H1N1 virus infection during the 2007–2008 season, 41 were treated with oseltamivir, and 27 were treated with zanamivir. During the 2008–2009 season, 77 patients with H1N1 virus infection were treated with oseltamivir, and 87 were treated with zanamivir; 31 and 28 patients with H3N2 virus infection were treated with oseltamivir and zanamivir, respectively. All 49 analyzed H1N1 virus isolates obtained during the 2008–2009 season, but none of the isolates obtained during the 2007–2008 season, contained the H274Y mutation. The mean ± standard deviation duration of fever after the start of oseltamivir therapy was significantly longer for patients with H1N1 virus infection (49.1±30.2h) than it was for patients with H3N2 virus infection (33.7±20.1h; P<.01) during the 2008–2009 season and patients with H1N1 virus infection during the 2007–2008 season (32.0±18.9h; P<.001). The duration of fever was significantly longer after the first dose of oseltamivir than it was after the first dose of zanamivir for patients with H1N1 virus infection during the 2008–2009 season (P<.001). The duration of fever from onset of H1N1 virus infection was significantly longer for children ⩽15 years of age during 2008–2009 (70.6±34.5h) than it was for such children during 2007–2008 (48.4±21.2h).
Conclusion. The effectiveness of oseltamivir, but not that of zanamivir, decreased significantly for H1N1 virus infection during the 2008–2009 season.
The World Health Organization announced that the incidence of infection due to influenza A virus subtype H1N1 with the oseltamivir-resistant N1 neuraminidase mutation H274Y has increased in Japan, from 3% during the 2007–2008 influenza season to 97% (34 of 35 isolates) on 22 January 2009 [1]. The Infectious Disease Surveillance Center of Japan announced on 14 February 2009 that 99.5% of H1N1 isolates in Japan (420 of 422 isolates) were resistant to oseltamivir.
In enzyme assay testing of neuraminidase activity in the presence of oseltamivir, the H1N1 virus with the H274Y mutation showed increased 50% inhibitory concentration (IC50) values that were 265–400-fold higher, suggesting a reduction of susceptibility to oseltamivir, but these viruses remained susceptible to zanamivir and adamantanes [2, 3]. However, little evidence is available concerning the clinical effectiveness of oseltamivir against oseltamivir-resistant virus. In Japan, oseltamivir has been commonly used to treat influenza A and B infections [4–9] and was widely used during the 2008–2009 season, because clinicians were not able to differentiate between H1N1 and H3N2 viral infections using commercial antigen detection kits.
In this report, based on patient medical records from the 2007–2008 and 2008–2009 influenza seasons, we compare the duration of fever (temperature, ⩾37.5°C) after the first dose and after the onset of symptoms in patients undergoing oseltamivir or zanamivir therapy. We also calculated the rate at which fever was prolonged to 48 h and 72 h after the start of treatment.
Study procedures. Family physicians, pediatricians, and general physicians at 15 clinics that belong to the Influenza Study Group of the Japan Physicians Association participated in the study. Patients were enrolled from 7 December 2007 through 27 March 2008 and from 6 December 2008 through 14 February 2009. Patients who reported to any of our 15 clinics with influenza-like illness who manifested such symptoms as a body temperature of ⩾37.5°C, upper respiratory tract symptoms, and systemic symptoms received a diagnosis of influenza A or B on the basis of the results of commercial antigen detection kits. Among the outpatients with influenza diagnosed by antigen detection kits and without severe underlying diseases, such as chronic obstructive pulmonary diseases or chronic heart diseases, those who received oseltamivir or zanamivir within 48 h after the onset of symptoms were registered in this study after providing informed consent.
Oseltamivir has been reported to be related to neuropsychiatric symptoms in young adults and has been prohibited, in most cases, for use by patients 10–19 years of age in Japan. Zanamivir is not recommended for patients with underlying respiratory disease or children <5 years old. Therefore, the decision whether to administer oseltamivir or zanamivir to patients with influenza was left to the discretion of the clinician, who followed the guidelines outlined herein and patient preference.
Specimens from throat swabs, nasal swabs, nasal aspirates, or blown nasal discharge were subjected to antigen detection and virus isolation. Of the commercially available antigen detection kits based on immunochromatography, Capilia FluA+B (Alfresa Pharma), QuickVue Rapid-SP influ (DS Pharma Biomedical), and Imuno Ace Flu (Touns) were mainly used.
The subtype of influenza A, H1N1 or H3N2, was determined by hemagglutinin inhibition assay with isolated virus using Madin-Darby canine kidney cells and also by the polymerase chain reaction method using subtype-specific primer sets for nasopharyngeal specimens, as previously reported [6]. A neuraminidase gene segment was amplified by reverse-transcriptase polymerase chain reaction, and the presence of H274Y mutation was determined by nucleotide sequencing for 93 patients with H1N1 virus (44 consecutive patients during the 2007–2008 season and 49 during the 2008–2009 season; 51 male and 42 female patients; mean age (±SD), 27.3±19.7 years).
Oseltamivir (75 mg for adults and for children who weighed ⩾37.5 kg and 2 mg/kg for children who weighed <37.5 kg) was taken orally twice per day for 5 days, and zanamivir (10 mg for adults and for children ⩾5 years old) was inhaled twice per day for 5 days. Antipyretics were not administered, but acetaminophen was used temporarily in the case of emergency.
Age, sex, vaccination status, results of the antigen detection test kit, and body temperature were recorded for all patients. The date and time of the onset of fever, date and time of administration of oseltamivir or zanamivir, and time of resolution of fever were recorded by the physician, patient, or an attending family member. The first time that a patient reported a fever (temperature, ⩾37.5°C) was defined as the time of onset. Patients were asked to measure body temperature at least 3 times per day (8 am, 2 pm, and 8 pm). The time at which a body temperature of <37.5°C was attained and maintained for >24 h was defined as the time at which the patient became afebrile. The highest body temperature during the disease was also recorded.
All data were collected using an internet-based protocol located in a secure room at the Gifu City Medical Association (Gifu, Japan) [10]. The time from the initial administration of oseltamivir or zanamivir to the resolution of fever and the duration of fever between the onset of symptoms and resolution were calculated automatically in the SQL database [5, 7]. The percentage of patients who were febrile at 48 h and 72 h after the first dose of oseltamivir or zanamivir was calculated [11].
Statistical analysis. Student's t test was used for between-group comparisons of the duration of fever. Fisher's exact test was also performed to compare between-group differences in the percentage of patients. To address factors that might influence the duration of fever after the onset, multiple regression analysis was performed for patients with H1N1 virus infection. The factors analyzed were patient age, sex, treatment (oseltamivir or zanamivir), vaccination status, antipyretics use, peak body temperature, time to administration of the first dose after the onset of fever, and the season. P <.05 was considered to be statistically significant.
N1 sequence analysis. Sequence analysis revealed that all 49 of the H1N1 virus isolates during the 2008–2009 season but none during the 2007–2008 season contained the H274Y mutation.
Patient characteristics. A total of 373 patients enrolled in the study during the 2007–2008 and 2008–2009 seasons. Finally, data for analysis were obtained from 291 patients, 68 with H1N1 virus infection during 2007–2008, 164 with H1N1 virus infection during 2008–2009, and 59 with H3N2 virus infection during 2008–2009. The demographic characteristics of the patients are summarized in Table 1. Because H3N2 virus was rare in the 2007–2008 season, it was not included in this study. Of 68 patients with H1N1 virus in the 2007–2008 season, 41 were treated with oseltamivir, and 27 were treated with zanamivir. During the 2008–2009 season, 77 patients with H1N1 virus infection were treated with oseltamivir and 87 were treated with zanamivir; and 31 patients with H3N2 virus were treated with oseltamivir and 28 were treated with zanamivir. No statistically significant differences other than mean age were detected between patients with H3N2 virus who received oseltamivir or zanamivir therapy or between patients with the H1N1 and H3N2 viruses who received zanamivir therapy during the 2008–2009 season (Table 1). Minor adverse reactions were observed for 2 patients who were treated with oseltamivir and for 3 patients who were treated with zanamivir. No severe adverse reactions were reported.
Baseline Demographic Characteristics of Patients with Influenza A Virus Subtype H1N1 during the 2007–2008 and 2008–2009 Seasons and for Influenza A Virus Subtype H3N2 in the 2008–2009 Season
Duration of fever after administration of the first dose of oseltamivir or zanamivir. The duration of fever after administration of the first dose of oseltamivir or zanamivir is given in Table 2. The duration after the start of oseltamivir therapy was significantly longer for patients with H1N1 virus infection during the 2008–2009 season than it was for those with infection during the 2007–2008 season (P <.001) and for patients with H3N2 virus during the 2008–2009 season (P <.01). No significant difference was found in the duration of fever after the start of zanamivir therapy among the 3 groups with H1N1 virus infection during the 2007–2008 season, H1N1 virus infection during the 2008–2009 season, or H3N2 virus infection during the 2008–2009 season.
Duration of Fever for Patients with Influenza Virus A Subtypes H1N1 and H3N2 Treated with Oseltamivir or Zanamivir Therapy
The duration of fever after the start of oseltamivir therapy for patients in the ⩽15-year-old and >15-year-old age groups was significantly longer for patients of both groups in 2008–2009 than in patients with H1N1 virus in 2007–2008 (Table 3). The duration of oseltamivir therapy in the 2008–2009 season was significantly longer than that of zanamivir therapy in each age group in the 2008–2009 season (P<.001 and P<.01, respectively).
Duration of Fever in Patients with Influenza A Virus Subtype H1N1 who received Oseltamivir and Zanamivir Therapy by Age
Duration of fever after onset of symptoms. The duration of fever after onset of symptoms was significantly longer for patients with H1N1 virus infection in the 2008–2009 season than for patients with H1N1 virus infection in the 2007–2008 season and for patients with H3N2 virus infection in the 2008–2009 season (Table 2). A significant difference was found between oseltamivir and zanamivir therapy for patients with H1N1 virus infection in the 2008–2009 season (P<.001). The duration of fever for patients treated with oseltamivir was significantly longer during the 2008–2009 season than it was during the 2007–2008 season for patients ⩽15 years old (P<.01) but was not statistically significant for patients >15 years old (Table 3). The duration of zanamivir therapy was significantly shorter than the duration of oseltamivir therapy in both age groups in the 2008–2009 season.
By multiple regression analysis, the time to the first administration of the drug, the drug administered (zanamivir or oseltamivir), the highest body temperature, and the season were found to be independent factors that affected the duration of fever. Vaccination status, antipyretics, sex, and age were not independent factors.
Percentage of patients febrile at 48 and 72 h after the first dose of oseltamivir or zanamivir. The percentages of febrile patients (temperature, ⩾37.5°C) at 48 or 72 h after the first dose of drug for H1N1 virus infection treatment are given in Table 4. The percentages of febrile patients at 48 and 72 h after oseltamivir therapy were significantly higher in the H1N1 virus infection group during 2008–2009 than in the H1N1 virus infection group during 2007–2008 or the H3N2 virus group during the 2008–2009 season in all age groups. The percentage of febrile patients at 48 and 72 h after oseltamivir therapy for the H1N1 virus infection group during the 2008–2009 season was also significantly higher than for the H1N1 virus group during 2007–2008 for children <10 years old. Table 4gives the age distribution of the patients. For patients aged 10–19 years, zanamivir was mainly used (62 of 66 patients), with oseltamivir administered to only 4 patients during these 2 seasons.
Proportion (%) of Febrile Patients with Influenza A Virus Subtypes H1N1 and H3N2 at 48 or 72 h after Initiation of Oseltamivir or Zanamivir Therapy
Influence of antipyretics or vaccination on the percentage of febrile patients. Antipyretics were temporarily used with oseltamivir in 15 (36.6%) of 41 patients, 16 (21.1%) of 76 patients, and 8 (27.6%) of 29 patients with H1N1 virus infection during 2007–2008, H1N1 virus in 2008–2009, and H3N2 virus, except for 3 patients in whom use of antipyretics was unknown. Antipyretics were also temporarily used with zanamivir in 8 (30.8%) of 26 patients, 24 (27.9%) of 86 patients, and 6 (22.2%) of 27 patients in each group, respectively, except for 3 unknown patients. Acetaminophen was prescribed in 62 (80.5%) of 77 patients who received an antipyretic. No significant differences were found in the percentage of patients who used antipyretics, either between oseltamivir and zanamivir or among the 3 groups: H1N1 virus in 2007–2008, H1N1 virus in 2008–2009, and H3N2 virus.
Table 5gives the percentage of H1N1 virus patients febrile at 48 or 72 h, with or without antipyretics, after starting oseltamivir or zanamivir therapy. No significant differences were found in the percentage of patients febrile at 48 or 72 h between patients taking and not taking antipyretics in the oseltamivir and zanamivir groups in the 2007–2008 or 2008–2009 seasons. The percentage of patients febrile at 48 and 72 h was significantly higher for the oseltamivir than for the zanamivir group for patients not taking antipyretics during the 2008–2009 season (P<.001for both).
Proportion (%) of Febrile Patients at 48 or 72 h with Influenza Virus A Subtype H1N1 in the 2008–2009 Season after Initiation of Antiviral Therapy with or without Antipyretics
The numbers of febrile patients with H1N1 virus infection at 48 h during the 2008–2009 season were 14 (43.8%) of 32 patients and 15 (34.1%) of 44 patients with or without vaccination, respectively, in the oseltamivir therapy group. The numbers were 6 (12.8%) of 47 and 5 (12.5%) of 40, respectively, for zanamivir therapy. No significant difference was found between vaccinated and unvaccinated patients.
Oseltamivir-resistant virus was reported to be rare (2%–3% of isolates) in Japan during the 2007–2008 season but common (98%–100% of isolates) during the 2008–2009 season [1]. In this study, the H274Y mutation was found in all H1N1 virus isolates obtained during the 2008–2009 season but in none of the isolates obtained during the 2007–2008 season. Therefore, we compared oseltamivir and zanamivir treatment for patients infected with H1N1 virus during the 2007–2008 and 2008–2009 seasons to investigate the influence of mutation on the effectiveness of neuraminidase inhibitors.
The duration of fever after the first dose of the drug and from the onset were analyzed to evaluate the clinical effectiveness of neuraminidase inhibitors [5, 7–9]. Limitations of the findings of our studies are that they were performed in general practices and not in the context of a rigorous clinical protocol. The body temperature of our outpatients was obtained from self-recorded reports by the patients or their family members. In our previous analysis, the duration of fever after the first dose of oseltamivir or zanamivir in patients with influenza A was ∼30 h [5, 7], and no significant difference was found in the duration of fever in patients with H1N1 and H3N2 virus infection [9]. Oseltamivir was less effective for influenza B than for influenza A. The data obtained using these self-recorded reports seem informative.
In this study, the durations of fever after the first dose of oseltamivir and from onset were significantly longer for patients with H1N1 virus infection in the 2008–2009 season than for patients with H1N1 virus in the 2007–2008 season or H3N2 virus in the 2008–2009 season. In contrast, no significant differences were found in duration among 3 zanamivir therapy groups. The H1N1 strains from 2007–2008 lacked the H247Y mutation, and those from 2008–2009 possessed it. The results of analysis of the percentage of febrile patients were concordant with the results of duration of fever. The reduction of the effectiveness of oseltamivir therapy for H1N1 virus infection during the 2008–2009 season was most likely attributable to the H274Y mutation.
Antipyretics are often used in the treatment of influenza patients, and their use may affect the clinical effectiveness of neuraminidase inhibitors. In this study, the percentage of febrile patients who were not taking antipyretics at 48 or 72 h was significantly higher for oseltamivir therapy than for zanamivir therapy in the 2008–2009 season. No influence of vaccination was shown. Thus, the reduction of effectiveness in therapy for patients with H1N1 virus infection during the 2008–2009 season was not attributable to antipyretic use or vaccination.
The re-isolation rates (proportion of patients with persistent virus infection) after zanamivir and oseltamivir therapy were 0% (0 of 8) and 11.4% (5 of 44), respectively, in 2007–2008 and 13.3% (4 of 30) and 27.6% (8 of 29), respectively, in the 2008–2009 season for all ages (data not shown). In another study, we reported that the mean (±SD) IC50of oseltamivir was significantly higher during the 2008–2009 season (319.3±185.4 nM) than it was in the 2007–2008 season (1.5±0.8 nM; P<.001) season [12]. The results of analysis of the duration of fever agreed with the results of virus re-isolation rates. Furthermore, plasma concentration of oseltamivir after administration has been estimated to be from 500- to 1000-fold that of IC50and was more than 300 nM [13–15], which was the mean IC50value in our other study [12]. This may explain why oseltamivir has demonstrated reduced but considerable effectiveness for outpatients with the H274Y mutation.
In Japan, the administration of zanamivir is not recommended for children <5 years old, and oseltamivir has been prohibited for use by patients 10–19 years old (Table 4). This makes controlled studies of anti-influenza drugs that include patients of all ages difficult in Japan for ethical reasons. Although there was no age-matched control (untreated) group in this study, comparisons were made between the H1N1 strains from 2007–2008 and 2008–2009; thus, the lack of traditional controls is mitigated. The mean (±SD) duration of fever from the onset of symptoms in patients with H1N1 virus who received oseltamivir in the 2008–2009 season (66.6±31.1 h) was significantly shorter than that observed for patients with influenza A who received no anti-influenza drug during the 2003–2004 and 2004–2005 seasons (82.4±36.0 h; P<.05) [5].
In this study, the reason for the age difference found between patients with H3N2 virus infection who received oseltamivir and zanamivir therapy and between patients with H1N1 virus infection and H3N2 virus infection who received zanamivir therapy in the 2008–2009 season is not clear. A reduction in the effectiveness of oseltamivir for treating H1N1 virus infection in the 2008–2009 season was seen not only among children <16 years old but also in adults ⩾16 years old; thus, the reduction in the effectiveness of oseltamivir for treating H1N1 virus infection during the 2008–2009 season may not be affected only by age.
We previously reported that there was no significant difference in IC50by age, but patients ⩽15 years old with oseltamivir-resistant virus infection had a higher viral re-isolation rate than did patients >15 years old (50% and 11.8%, respectively; P=.038) [12]. In this study, the duration of fever after the onset for oseltamivir therapy was not significantly different between the 2008–2009 and 2007–2008 seasons for adults >15 years old. Oseltamivir therapy may be to some extent effective, especially for adults. It is not clear whether age itself is an important factor. Factors such as a prior exposure to any type of influenza or an insufficient dosage of oseltamivir may be related to these results.
Of great concern is a possible relation between oseltamivir use and the emergence of resistant virus. Oseltamivir-resistant viruses have been detected in immunocompromised patients treated with oseltamivir [16, 17] and in children [18]. However, oseltamivir-resistant H1N1 virus with the H274Y mutation has emerged outside of Japan, where oseltamivir is commonly used. Resistant viruses were not detected in adult Japanese outpatients treated with oseltamivir in our previous study [6]. We cannot deny the possibility that mutations other than H274Y will emerge as a result of uncontrolled mass administration of neuraminidase inhibitors; however, the H274Y mutation of the H1N1 virus does not seem to have been caused by the general use of oseltamivir in Japan.
In conclusion, the clinical effectiveness of oseltamivir for treatment of H1N1 virus infection during the 2008–2009 influenza season was reduced, especially in children, in comparison with the 2007–2008 season, probably because of the H274Y mutation. Clinicians also should bear in mind that oseltamivir-resistant viruses have been reported to have significant pathogenicity and lethality for patients at high risk [19, 20]. Thus, oseltamivir should not be recommended for children or patients with high-risk underlying diseases who are infected with H274Y-mutated H1N1 virus. Zanamivir or therapy with a combination of antivirals has been recommended for the treatment of oseltamivir-resistant H1N1 virus infection in high-risk patients [21–23].
We thank Koichiro Kudo and Yasuyuki Kato, both of the International Medical Center of Japan.
Members of the Influenza Study Group of the Japan Physicians Association who participated in this study include the following: Hideo Kanazawa, Hideyuki Ikematsu, Ietaka Sato, Ken-ichi Doniwa, Kenichi Kawamura, Kunio Kondou, Naoki Kawai, Nobuo Hirotsu, Norio Iwaki, Osame Tanaka, Satoshi Yamauchi, Seizaburo Kashiwagi, Shinro Matsuura, Takashi Kawashima, Tetsunari Maeda, Tomoyuki Harada, and Tuyoshi Okayama.
Financial support. The Ministry of Health, Labour and Welfare of Japan (20A–028).
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy