High Frequency of Resistant Viruses Harboring Different Mutations in Amantadine-Treated Children with Influenza

Influenza A virus causes an unacceptable number of deaths worldwide, even in nonpandemic seasons, making its control a paramount medical goal. Amantadine has been used as an anti-influenza drug in numerous countries for >30 years [1]. Despite its ability to shorten the duration of illness and to reduce the severity of symptoms, this compound has adverse effects on the central nervous system and can give rise to resistant viruses.

Amantadine inhibits influenza A virus replication by interfering with M2 protein ion channel activity [2]. Viruses become resistant to this drug through a single amino acid substitution at position 26, 27, 30, 31, or 34 in the transmembrane region of the M2 protein (amino acid substitution at position 34 was demonstrated in drug-resistant avian influenza viruses but not in human isolates) [3]. Mutations at position 31 are most frequent in resistant viruses [4]. Approximately 30% of children treated with amantadine have been reported to shed resistant viruses [3]. Amantadine-resistant viruses are genetically stable [5] and can be transmitted from person to person [3, 4]. This property has not led to influenza epidemics due to resistant viruses [4, 6, 7], but it could pose serious problems in closed settings, such as nursing homes [3].

Traditionally, amantadine-resistant viruses have been detected by testing the drug sensitivity of isolates or by directly sequencing their M2 genes without molecular cloning [4, 10]. Therefore, unless the drug-resistant viruses represent the dominant population, they can be overlooked, leading to under-estimation of the prevalence of amantadine resistance among patients. In view of the international importance of amantadine in controlling influenza A virus infection, we sought to determine the prevalence of amantadine-resistant viruses in a pediatric patient population and to analyze the pattern of emergence of these mutant strains.

Patients, materials, and methods. Samples for influenza virus analysis were collected from 11 patients (patients 1–5, 7, 8, 11, 12, 14, and 15) hospitalized during the 1999–2000 influenza season and from 4 patients (patients 6, 9, 10 and 13) hospitalized during the 2000–2001 influenza season (table 1). Informed consent was obtained from the parents of all patients in this study, and the human experimentation guidelines of the US Department of Health and guidelines of Nippon Kokan Hospital were followed. Their ages ranged from 11 months to 14 years 4 months. Three of the patients (patients 1, 2, and 4) were infected with influenza virus during their hospitalization for other diseases (Kawasaki disease, pertussis, and pneumonia, respectively). Of these 3 patients, 1 (patient 4) had exposure to an amantadine-treated patient before the onset of her influenza symptoms. The others were healthy before hospitalization for high-grade fever (temperature ⩾38.0°C) and other influenza-like symptoms (e.g., headache, malaise, myalgia, sore throat, and cough). Treatment with amantadine (5 mg/kg of body weight/day; maximum dose, 100 mg/day) did not begin until influenza A infection was diagnosed with a Directigen FluA kit (Becton Dickinson). Common influenza complications, such as bronchitis, otitis media, or febrile convulsions, were seen in some patients, but there was no single complication common in all patients. All patients were released from the hospital, free of symptoms, within 13 days after amantadine treatment.

We collected nasal swabs, nasal aspirates, or throat swabs from all patients at least twice during the study and >4 times from 6 patients. Time points of sample collection are listed in table 1. These samples were used for direct RNA extraction or for isolation of influenza virus on Madin-Darby canine kidney (MDCK) cells or Caco-2 cells. Viral subtypes were identified by use of a hemagglutination inhibition test and by use of reference serum against A/New Caledonia/20/99 (H1N1) and A/Panama/2007/99 (H3N2) for viruses in the 2000–2001 season and A/Beijing/262/95 (H1N1) and A/Sydney/05/97 (H3N2) for viruses in the 1999–2000 season.

Viral RNA was extracted from the collected specimens (patients 5, 7, 8, 11, 12, 14, and 15) or from virus in cell culture fluids (patients 1–4, 6, 9, 10, and 13) by use of an RNA extraction kit (ISGEN-LS; Nippon Gene), according to the manufacturer's protocol.

Reverse-transcription reactions were performed with the BsmBI-U12 primer (5′CACACACGTCTCCGGGAGCAAAAGCAGG-3′), which has a BsmBI site and is complementary to the 3′ end of viral RNA of the M gene segment, and a reverse transcriptase (SUPERSCRIPT II or M-MLV Reverse Transcriptase; Invitrogen), according to the manufacturer's protocol. The resultant cDNA products were used in polymerase chain reaction (PCR) amplifications (Pwo DNA polymerase; Roche) of the M gene with the forward primer BsmBI-MF (5′-CACACACGTCTCCGGGAGCAAAAGCAGGTAGATATTGAAAGATGAGCCTTCTAACC-3′), which has a BsmBI site and corresponds to the region spanning positions 1–40 in the cDNA sense, and the reverse primer BsmBI-MR (5′-CACACACGTCTCCTATTAGTAGAAACAAGGTAGTTTTTTACTCCAGCTCTATGCTGAC-3′), which has a BsmBI site and is complementary to the region spanning positions 987–1027 in the cDNA sense. The thermocycler program was as follows: initial denaturation at 94°C for 2 min, followed by 30 cycles at 94°C for 30 s, 50°C for 30 s, and 72°C for 2 min, with a final extension at 72°C for 10 min. The resultant PCR products were purified by use of the QIAquick Gel Extraction Kit (Qiagen).

The PCR products were ligated into the pCR-Blunt II-TOPO vector (Invitrogen) and were transformed into TOP10 cells (Invitrogen). The positive clones were cultured in Luria broth containing 50 µg/mL kanamycin and were incubated overnight at 37°C in a shaking incubator. The bacterial culture was centrifuged at 2300 g for 10 min, and the cell pellet was then treated with the Wizard Plus SV Minipreps DNA Purification System (Promega), to extract plasmid DNA for sequencing. Sequence analysis was done on the Applied Biosystem 3700 Auto Sequencer by use of cycle sequencing dye terminator chemistry (Perkin Elmer). An M13 forward (−20) primer (5′-GTAAAACGACGGCCAG-3′) and an M13 reverse primer (5′-CAGGAAACAGCTATGAC-3′) were used to sequence the M gene.

Results. We tested a total of 53 samples from 15 patients who received amantadine. To determine the prevalence of amantadine resistance and the time of appearance of drug-resistant viruses, we first examined samples from patients 1–6, which represented at least 4 of the planned testing time points. Ten cDNA clones of the M gene were sequenced for each sample. Amino acid mutations known to confer amantadine resistance were detected in 5 of the 6 patients, either during or immediately after amantadine treatment (figure 1). Surprisingly, 4 patients carried ⩾2 different mutant clones.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1.

Prevalence of M2 mutant viruses during amantadine therapy in 6 patients with serial tests. Amantadine treatment was started on day 1 after clinical samples were collected. Ten cDNA clones of the M gene were sequenced for each sample; colored bars represent the percentages of wild-type virus and each amantadine-resistant mutant.

We noted that mutant viruses had replaced the wild-type population in patients 2, 4, and 6. However, in patients 1 and 3, the wild-type virus was either decreased or undetectable after amantadine treatment but reappeared 1 day later. Examination of 50 additional clones from the day-6 sample of patient 3 revealed a single wild-type clone, which indicated the presence of a minor population of virus with a wild-type M2 gene. Viruses with the S31N mutation, detected in 4 of the 6 patients, tended to be more prevalent than competing mutants, which suggests greater replicative competency.

To confirm the strikingly high prevalence of amantadine-resistant viruses, we studied an additional 9 patients, examining 3 clones of M gene cDNA from the day-1 sample of each subject (before amantadine treatment) and 10 clones from samples that were collected during or after amantadine therapy (table 1). None of the day-1 clones demonstrated drug-resistant mutations. cDNA clones with a known amantadine resistance-conferring mutation were identified in 7 (patients 7–10, 12, 14, and 15) of the 9 patients during or after amantadine therapy. Five patients in this group (patients 7, 8, 10, 12, and 15) had >1 mutant virus, which is consistent with findings in the previous series. In addition, mutations known to confer amantadine resistance, we found 4 single amino acid substitutions (A30G, I43V, A29V, and G34E) in the transmembrane region of M2 protein that have not been implicated in amantadine resistance in clinical samples. The biologic significance of these mutations is unclear.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1.

Patient information and summary of detection of multiple drug resistance-conferring mutations in the M2 gene.

Discussion. Previous reports indicate that amantadine-resistant viruses emerge in ∼30% of patients treated with this drug [3]. In our study, 12 (80%) of 15 patients whose influenza A viruses were initially sensitive to amantadine acquired mutations known to confer resistance to this drug. An even higher prevalence of amantadine resistance might have been found if we had been able to test patients at more frequent intervals. Indeed, 8 patients had only a single sample collected after the instigation of therapy. We attribute the discrepancy in the prevalence of amantadine resistance between our findings and earlier reports [4] to the greater likelihood of detecting mutant viruses by sequencing after molecular cloning versus direct sequencing of PCR products, because, in the latter analysis, the signals produced by mutant clones at a given nucleotide position are overlooked, unless they are dominant in the population.

Although others have detected resistant and sensitive viruses in the same clinical sample from an amantadine-treated patient [1], as well as different mutants in each of 2 patients [11], our study is the first to clearly demonstrate the coexistence of wild-type virus and several different mutants in a high proportion of drug-treated patients. The mutation S31N has been detected in the vast majority of previously studied patients [4]. In the present study, we show a high frequency of S31N and A30V mutations in H3 viruses, as well as V27A in H1 viruses. Because H3 and H1 viruses have phylogenetically different M2 proteins, mutations advantageous for conferring amantadine resistance may be influenced by structural differences in M2. The S31N substitution appeared to persist longest in our patients, which suggests greater replicative potential, although a clear trend could not be established.

It may be important that wild-type virus continued to be detected in serial samples from 2 of the 5 patients with drugresistant mutants. This suggests that wild-type viruses are only diminished in many patients with amantadine-resistant mutants and regain dominance when amantadine is withdrawn. The continued presence of wild-type viruses might explain why amantadine-resistant viruses have not been linked to a major influenza epidemic [4, 6, 7], despite their relatively high rate of emergence, genetic stability [5], and transmissibility [4, 5]. Indeed, there is experimental support for the hypothesis that wild-type viruses are more replication competent than resistant mutants in an environment without amantadine. For example, amantadine-resistant avian H7 viruses were attenuated in their growth, compared with wild-type viruses, and some amantadine-resistant viruses reverted to the wild-type in the absence of the drug [12], although resistant variants of H5N2 virus showed no such reversion after multiple passages in contact birds [5]. Possibly, the replicative competency of amantadineresistant viruses depends on the site of the M2 mutation, as well as the genetic background of the virus.

Four mutations that we detected in the M2 transmembrane region (A29V, A30G, G34E, and I43V) do not appear to have been found in clinical samples examined by others [4]. In experimental settings, the A30G mutation was identified in an amantadine-resistant avian H7N1 (Rostock) virus with an A30G/S31R double mutation in the M2 protein [12], whereas the G34E mutation was detected in avian H7N1 and H7N7 (Weybridge) viruses [12, 13]. The A29V and I43V mutations have not been reported to our knowledge. Residues at positions 27, 30, 31, 34, 37, 38, and 41 in the M2 protein constitute the face of the α-helix of the M2 transmembrane domain, which is critical for channel activity and amantadine sensitivity [14]. The 2 residues at positions 29 and 43 are not located on the face of the M2 α-helix, making it difficult to assess their contribution to amantadine resistance at these sites.

Amantadine-resistant viruses can cause serious problems in closed settings, such as nursing homes [8, 9]. Many patients with chronic central nervous system illness, including Parkinson disease or parkinsonism associated with cerebrovascular disease, are living in long-term care facilities and often receive preventive amantadine treatment. The mutation rate among such patients might be reduced by substituting the neuraminidase inhibitors zanamivir and oseltamivir, the former having been found to be effective against amantadine-resistant viruses in a nursing home [15]. To impede the induction of drug-resistant viruses still further, one could consider a combination of amantadine and neuraminidase inhibitors for treatment and prophylaxis.

• Received October 15, 2002.
• Accepted January 8, 2003.

• © 2003 by the Infectious Diseases Society of America