Antiviral Medications in Seasonal and Pandemic Influenza
Abstract
Background
Amantadine, oseltamivir, and zanamivir are currently available in Germany for the prevention and treatment of influenza. We review their efficacy and side-effect profiles.
Methods
This review is based on pertinent randomized and controlled trials (RCTs) and systematic reviews retrieved by a systematic literature search, and on other relevant literature.
Results
The efficacy of antiviral drugs for the prevention of symptomatic influenza ranges from 60% to 90% (number needed to treat [NNT], 8–89) depending on the population and type of drug in question. Antiviral drugs shorten the duration of illness by 0.5–1.5 days when given within 48 hours of the onset of symptoms. Neuraminidase inhibitors do not significantly lower the incidence of bronchitis in adults, or of otitis media in children; they do have a positive effect against reported, but not necessarily diagnostically confirmed pneumonia in adults (NNT, 89 [50–232]). The RCTs yielded no information about possible effects on severe cases of influenza, or on mortality, as they included only mildly or moderately ill patients, but observational studies have yielded some evidence of benefit. The most common side effects of oseltamivir (>10%) are headache, nausea, and vomiting; of zanamivir (>1%), a skin rash; and of amantadine (>1%), loss of appetite, nausea, and central nervous effects.
Conclusion
The benefits of antiviral drugs, particularly neuraminidase inhibitors, outweigh their risks. In deciding whether to use them, physicians should consider the properties of the currently circulating viruses and the patient’s individual risk constellation, as directed in clinical treatment recommendations.
In Germany, the drugs available for influenza prophylaxis and therapy are amantadine and the neuraminidase inhibitors (NIs) oseltamivir and zanamivir. Unlike neuraminidase inhibitors, amantadine, as an M2 membrane channel blocker, is only effective against influenza A viruses. The use of amantadine is no longer recommended, principally due to rapid development of resistance during its use and high resistance rates in circulating influenza viruses, as well as poor tolerability (1, 2).
Most randomized controlled trials (RCTs)—the gold standard for proving efficacy—for amantadine, zanamivir, and oseltamivir are more than 15 years old. Since they were conducted, these trials have been summarized in many systematic reviews and meta-analyses. This review came about as part of Germany’s national pandemic plan and was performed by a working group of the Robert Koch Institute (RKI) Expert Advisory Board on Influenza. It brings together the most important information from the chapter Medicines relevant in a pandemic of the scientific part of the German Influenza Pandemic Preparedness Plan. This provides comprehensive details on the amount of underlying data and the epidemiology of influenza (3). Regarding questions on political implications and data transparency, see earlier articles in Deutches Ärzteblatt (4, 5).
Public health institutions and specialist societies have published treatment recommendations on the use of antiviral drugs for influenza (box). The biased nature of some public debate on the issue makes an objective representation of the available evidence on antiviral drugs particularly important.
Below, we provide a brief description of the evidence on the efficacy and safety of the antiviral drugs available for influenza in Germany. This takes the form of answers to questions that are of particular clinical importance for the potential use of these drugs.
Methods
A systematic search was performed to locate meta-analyses, systematic reviews, and randomized controlled trials written in German or English and published no later than December 2, 2015. The search included the databases Cochrane Library, PubMed, and Scopus (see eBox 2, eTable 1, and the eFigure for details on the search of the literature).
Table 1
| Intervention | Cumulative no. of study participants | Relative risk [95% confidence interval]*1 | Number needed to treat (NNT) |
| Long-term prophylaxis: adults | 1039 to 2479 1107 to 5275 | Oseltamivir: 0.24 to 0.45 [0.09; 0.83] Zanamivir: 0.32 to 0.39 [0.17; 0.70] | OSV: 18 to 35 ZNV: 14 to 55 |
| Post-exposure prophylaxis (PEP): adults | 405 to 1747 1525 to 2416 | Oseltamivir: 0.0 to 0.20 [0.0; 0.44] Zanamivir: 0.21 to 0.33 [0.13; 0.36] | OSV: 8 to 15 ZNV: 9 to 22 |
| Prophylaxis (PEP and long-term combined): adults | 4645 | Amantadin: 0.39 [0.24; 0.65] | AMT: 16 |
| Post-exposure prophylaxis: children | 215 863 | Oseltamivir: 0.36 [0.15; 0.84] Oseltamivir und Zanamivir: absolute Risikoreduktion: 8% [5; 12] | OSV: 25*2 OSV + ZNV: 12 |
| Long-term prophylaxis: children | 773 | Amantadin: 0.11 [0.04; 0.30] | AMT: 12 |
| Long-term prophylaxis: elderly high-risk patients*3 | 548 1896 | Oseltamivir: 0.08 [0.01; 0.63] Zanamivir: 0.20 [0.02; 1.72] | OSV: 16 ZNV: 238*3 |
| Long-term prophylaxis: high-risk adolescents/ adults (68% vaccinated) | 3363 | Zanamivir: 0.17 [0.07; 0.44] | ZNV: 89 |
*1Relative risk for onset of symptomatic influenza disease. Duration of treatment is between 10 days and 6 weeks but is not stated for all studies of amantadine.
*2The 95% confidence interval has no upper bound. as a result of very similar risk rates or small total case numbers in both treatment groups. Substantial uncertainty is therefore associated with NNT or NNH.
*3High-risk patients: residents of chronic-care facilities. patients aged 65 years or older. and/or those with chronic cardiovascular. respiratory. or renal disease.
AMT. amantadine; OSV. oseltamivir; ZNV. zanamivir
PRISMA flowchart
PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses
RCT: Randomized controlled trial
In addition, randomized controlled trials, observational studies, and published data from the summaries of product characteristics of the drugs concerned were included on a selective basis.
Results
The findings presented here are based on one review of systematic reviews (11), 13 systematic reviews/meta-analyses of RCTs, and 6 additional RCTs.
Influenza prophylaxis
Essentially, influenza prophylaxis can be administered in the form of either oseltamivir and zanamivir, or amantadine (influenza A only).
How effective are antiviral drugs for prophylactic use?
For healthy adults, it has been shown that fewer symptomatic cases of influenza occur if antiviral drugs are administered as either post-exposure or long-term prophylaxis. For amantadine there is no clear division between types of prophylaxis in the studies examined (table 1) (12– 18). The number needed to treat (NNT) is between 8 and 89; neuraminidase inhibitors are more effective for post-exposure prophylaxis (NNT: 8 to 25). For children and high-risk patients, some of the evidence is patchy (12– 14, 17, 18); the NNT is between 12 and 89, depending on the drug and the population. Oseltamivir has been proven to be effective for long-term prophylaxis in elderly high-risk patients (NNT: 25) (12– 14).
Only one RCT addressed prophylactic use in immunosuppressed patients; this included 475 patients. It showed no significant difference between oseltamivir, at its standard dose for 12 weeks’ prophylaxis, and placebo in terms of symptomatic cases of confirmed influenza (19).
Does the timing of administration play a role in efficacy?
Only one RCT of zanamivir reported that post-exposure prophylaxis usually failed in the first few days of use. This may indicate that the time between exposure and the beginning of prophylaxis was too long in these cases (20).
Influenza therapy
Intention-to-treat (ITT) analysis includes data on all study participants with clinically diagnosed influenza, i.e. including those with influenza-like illness (ILI) not caused by influenza viruses. In contrast, intention-to-treat infected (ITTi) analysis includes only data on those in whom the diagnosis of influenza was confirmed, usually using subsequent laboratory diagnostics: antigen or antibody testing. As the mechanism of action of antiviral drugs is specific to influenza, the ITTi analysis reflects the efficacy of antiviral drugs against influenza viruses, whereas ITT analysis better reflects clinical practice. Because the analysis population is so important, this review describes the findings for both the ITT and the ITTi populations.
Do antiviral drugs affect disease duration?
With a disease duration for placebo of 6.6 to 7 days and start of therapy up to 48 hours after symptom onset, antiviral drugs reduce the time until the patient is symptom-free by a mean of approximately 0.5 to 1.5 days (etable 2) (15– 18, 21). A recent meta-analysis of patient data from trials of oseltamivir also shows that oseltamivir decreases the time until patients are symptom-free by approximately one day (21%) (22). For children and adults with no comorbidities this effect is almost always statistically significant. The fact that the 2014 Cochrane Review by Jefferson et al. found no significant difference for zanamivir in the treatment of children may be the result of the methods used: both the individual studies included in the meta-analysis did report statistically significant differences (15). In the elderly and high-risk patients (those with chronic respiratory diseases or cardiovascular diseases), there was usually no statistically significant difference between them and placebo, although in most cases numerical benefits were found for neuraminidase inhibitors.
eTable 2
| No. of RCTs (litera- ture reference) | Numeber of participants | Drug | Difference in time until patients symptom-free. days (median. [95% CI]) |
| Adults with no comorbidities | |||
| 4 (21) | 1410 | Oseltamivir | –0.55 [–1.04; –0.14] |
| 6 ITTi (21) | 1182 | Oseltamivir | –0.92 [–1.55; –0.29] |
| 8 (15) | 3954 | Oseltamivir | –0.69 [–1.04; –0.35] |
| 6 (21) | 1701 | Zanamivir | –0.57 [–1; –0.1] |
| 13 (15) | 5411 | Zanamivir | –0.60 [–0.81; –0.39] |
| 10 (16) | 542 | Amantadine | –0.99 [–1.3; –0.7]*1 |
| Children with no comorbidities | |||
| 2 (21) | 1029 | Oseltamivir | –0.88 [–1.40; –0.35] |
| 2 ITTi (21) | 631 | Oseltamivir | –1.2 [–1.82; –0.58] |
| 1 (15) | 669 | Oseltamivir | –1.22 [–1.86; –0.49]*2 |
| 2 (21) | 737 | Zanamivir | –0.94 [–1.43; –0.46] |
| 2 (18) | 451/266 | Zanamivir | –0.5*3 |
| 2 (15) | 723 | Zanamivir | –1.08 [–2.32; 0.15] |
| 2 (17) | 104 | Amantadine | Relative risk of fever on day 3: 0.39 [0.20; 0.79] |
| Elderly patients | |||
| 1 (21) | 736 | Oseltamivir | –0.42 [– 1.87; 1.04] |
| 1 ITTi (21) | 477 | Oseltamivir | –1.00 [–2.83; 0.83] |
| 5 (21) | 475 | Zanamivir | –1.13 [–2.90; 0.63] |
| 5 ITTi (21) | 323 | Zanamivir | –1.85 [–4.77; 1.07] |
| High-risk patients | |||
| 2 (21) | 1472 | Oseltamivir | –0.74 [–1.50; 0.02] |
| 3 ITTi (21) | 907 | Oseltamivir | –0.59 [–1.51; 0.34] |
| 2 (15)*4 | 660 | Oseltamivir | 0.21 [–0.46; 0.89] |
| 7 (21) | 1252 | Zanamivir | –0.98 [– 1.84; –0.11] |
| 6 ITTi (21) | 730 | Zanamivir | –1.83 [–2.81; 0.86] |
(Analysis population: ITT unless otherwise stated)
*1Number of days with fever (37 °C or more); *2Including studies in children with asthma: –0.33 [-1.39; 0.72];
*3Data from 2 RCTs not pooled; median identical for both trials; *4Data from 2 RCTs not pooled; median identical for both trials
ITT: Intention to treat; ITTi: Intention to treat infected; CI: Confidence interval; RCT: Randomized controlled trial
Patients treated with neuraminidase inhibitors return to normal activities a mean of 0.5 to 1 day sooner. In most studies the findings of ITTi analysis for antiviral drugs were more favorable than those of ITT analysis. It remains unclear why this is not the case in the study by Burch et al. (21).
Do antiviral drugs affect disease severity?
There is no RCT data on mortality or need for intensive care. A meta-analysis with datasets from approximately 30 000 patients hospitalized during the 2009/2010 pandemic found that the use of neuraminidase inhibitors, usually oseltamivir, in adults was associated with a reduced mortality risk, regardless of when they were used (odds ratio [OR]: 0.81; 95% confidence interval: [0.70; 0.93]) (23). For methodological reasons, neither the absolute risk reduction nor the NNT can be determined for the relevant adjusted analysis. Subgroup analyses for adults receiving intensive care, pregnant women, and patients who received neuraminidase inhibitors early (no more than 2 days after symptom onset) show more substantial risk reductions. No effect on mortality is found if therapy is begun later, or in children.
A meta-analysis of 3 observational studies, selected on the basis of very stringent criteria, on the effect of oseltamivir on mortality in seasonal or avian (H5N1) influenza indicates that neuraminidase inhibitors are effective (24).
ITT evaluation revealed no significant effect on rate or duration of hospitalization for either neuraminidase inhibitors or amantadine.
Do antiviral drugs affect influenza-related complications?
While point estimates in most evaluations do indicate that neuraminidase inhibitors effectively reduce influenza-related complications, statistically significant differences for individual neuraminidase inhibitors were not always seen (Tables 2 and and3)3) (15, 21, 25). Only one meta-analysis for oseltamivir in adults with no comorbidities found a significant reduction in reported cases of pneumonia (NNT: 89 [50; 232]) (15). Reduced use of antibiotics (as reported in some double-blind trials) can at best be interpreted as a surrogate parameter for fewer bacterial complications.
Table 2
| Outcome (literature reference) | Number of RCTs | Number of participants | Effect estimate | Value [95% CI] | Number needed to treat (NNT) |
| Adults with no comorbidities | |||||
| All complications (21) | 1 | 419 | OR | 0.61 [0.3; 1.1] | 21*2 |
| ITTi (21) | 3 | 564 | OR | 0.75 [0.41; 1.37] | 46*2 |
| Pneumonia (15) | 8 | 4752 | RR | 0.55 [0.33; 0.90] | 83 |
| Pneumonia (21) | 2 | 784 | OR | 0.33 [0.03; 3.16] | 195*2 |
| Pneumonia ITTi (21) | 3 | 610 | OR | 0.33 [0.05; 2.14] | 103*2 |
| Bronchitis (15) | 8 | 4752 | RR | 0.75 [0.56; 1.01] | 31 |
| Bronchitis (21) | 1 | 476 | RR | 1.38 [0.43; 4.4] | NNH: 129*2 |
| Bronchitis ITTi (21) | 2 | 572 | OR | 0.94 [0.39; 2.24] | 405*2 |
| Sinusitis (15) | 8 | 4752 | RR | 1.03 [0.76; 1.40] | 194*2 |
| Otitis media (15) | 6 | 4368 | RR | 1.11 [0.57; 2.15] | 769*2 |
| Antibiotic use (21) | 2 | 1652 | OR | 0.37 [0.29; 0.48] | 7 |
| Antibiotic use ITTi (21) | 3 | 845 | OR | 0.52 [0.27; 1.00] | 35 |
| Children with no comorbidities | |||||
| Otitis media (15) | 3*1 | 1359 | RR | 0.80 [0.62; 1.02] | 136*2 |
| Pneumonia (21) | 2 | 1029 | OR | 0.58 [0.26; 1.28] | 27*2 |
| Pneumonia (15) | 3*1 | 1359 | RR | 1.06 [0.62; 1.83] | 74*2 |
| Bronchitis (21) | 1 | 334 | OR | 4.94 [0.57; 42.74] | NNH: 573*2 |
| Bronchitis (15) | 3*1 | 1359 | RR | 0.65 [0.27; 1.55] | 87*2 |
| Sinusitis (15) | 3*1 | 1359 | RR | 1.00 [0.58; 1.72] | NNH: 3694*2 |
| Antibiotic use (21) | 1 | 695 | OR | 0.96 [0.46; 1.99] | 291*2 |
| Exacerbation of asthma (25) | 1 | 177 | RD | -0.05 [-0.15; 0.05] | 6*2 |
| High-risk patients | |||||
| Pneumonia (21) | 1 | 334 | OR | 0.48 [0.04; 5.34] | 157*2 |
| Pneumonia ITTi (21) | 2 | 728 | OR | 0.99 [0.36; 2.27] | 1489*2 |
| Bronchitis (21) | 1 | 334 | OR | 4.94 [0.57; 42.7] | NNH: 43*2 |
| Antibiotic use (21) | 1 | 334 | OR | 0.96 [0.46; 1.99] | 116*2 |
| Antibiotic use ITTi (21) | 4 | 878 | OR | 0.57 [0.33; 0.98] | 15 |
(Analysis population: ITT unless otherwise stated)
*1Meta-analysis of one study in healthy children (n = 695) and two studies in children with chronic asthma (n = 664). The combined meta-analysis does not substan?tially alter the findings of the studies in the various populations.
*2The 95% confidence interval has no upper bound. as a result of very similar risk rates or small total case numbers in both treatment groups. Substantial uncertainty is therefore associated with NNT or NNH.
95% CI. 95% confidence interval; ITTi. intention to treat infected; NNH. number needed to harm; OR. odds ratio; RCT. randomized controlled trial; RR. relative risk;
RD. risk difference
Table 3
| Outcome (literature reference) | Number of RCTs | Number of participants | Effect estimate | Value [95% CI] | Number needed to treat (NNT) |
| Adults with no comorbidities | |||||
| Pneumonia (21) | 1 | 588 | OR | 1.36 [0.63; 2.9] | NNH: 72* |
| Pneumonia (15) | 11 | 5876 | RR | 0.90 [0.58; 1.40] | 246* |
| Bronchitis (21) | 2 | 1054 | OR | 1.08 [0.54; 2.17] | NNH: 402* |
| Bronchitis (15) | 12 | 6072 | RR | 0.75 [0.61; 0.91] | 47 |
| Sinusitis (15) | 12 | 6072 | RR | 1.12 [0.84; 1.48] | 196* |
| Otitis media (15) | 10 | 5494 | RR | 0.81 [0.54; 1.20] | 186* |
| Antibiotic use (21) | 1 | 276 | OR | 0.68 [0.31; 1.51] | 29* |
| Exacerbation of asthma (15) | 9 | 5269 | OR | 0.54 [0.34; 0.86] | 88 |
| Children with no comorbidities | |||||
| All complications (21) | 2 | 732 | OR | 0.88 [0.62; 1.24] | NNH: 740* |
| Otitis media (21) | 1 | 266 | OR | 0.63 [0.16; 2.4] | 63* |
| Otitis media (15) | 2 | 737 | RR | 1.00 [0.59; 1.72] | 115* |
| Pneumonia (21) | 1 | 266 | OR | 0.51 [0.07; 3.65] | 93* |
| Pneumonia (15) | 2 | 737 | RR | 0.53 [0.12; 2.38] | 229* |
| Bronchitis (21) | 2 | 732 | OR | 1.05 [0.28; 3.89] | 278* |
| Bronchitis (15) | 2 | 737 | RR | 0.86 [0.26; 2.80] | NNH: 6128* |
| Sinusitis (15) | 2 | 737 | RR | 0.87 [0.12; 6.45] | NNH: 6128* |
| Antibiotic use (21) | 1 | 471 | OR | 0.05 [0.01; 0.23] | 8 |
| Elderly patients | |||||
| All complications (21) | 1 | 358 | OR | 0.84 [0.54; 1.32] | 28* |
| Pneumonia (21) | 1 | 358 | OR | 0.87 [0.17; 4.38] | 444* |
| Bronchitis (21) | 1 | 358 | OR | 0.46 [0.2; 1.02] | 19* |
| Antibiotic use (21) | 1 | 358 | OR | 0.73 [0.43; 1.24] | 21* |
| High-risk patients | |||||
| All complications (21) | 4 | 575 | OR | 0.73 [0.51; 1.04] | 14* |
| Pneumonia (21) | 2 | 881 | OR | 0.57 [0.15; 2.23] | 200* |
| Bronchitis (21) | 3 | 1210 | OR | 0.41 [0.24; 0.70] | 20 |
| Antibiotic use (21) | 4 | 575 | OR | 0.71 [0.47; 1.07] | 18* |
| Antibiotic use ITTi (21) | 4 | 565 | OR | 0.66 [0.42; 1.03] | 20* |
(Analysis population: ITT unless otherwise stated)
*The 95% confidence interval has no upper bound. as a result of very similar risk rates or small total case numbers in both treatment groups. Substantial uncertainty is therefore associated with NNT or NNH.
95% CI. 95% confidence interval; ITT. intention to treat; ITTi. intention to treat infected; NNH. number needed to harm; OR. odds ratio; RCT. randomized controlled trial; RR. relative risk
An additional meta-analysis of 11 RCTs of oseltamivir treatment in a total of 3908 patients, which also included data from unpublished trials (26), showed an absolute risk reduction for respiratory diseases (mostly bronchitis) requiring antibiotic treatment of 3.42% (relative risk reduction: 37%; NNT: 20). However, bronchitis is not a serious complication and does not usually require antibiotic treatment.
No data has been published on the effect of amantadine on complications.
Two large meta-analyses on hospitalized patients from the A(H1N1)pdm09 pandemic yielded no clear findings regarding the use of neuraminidase inhibitors and pneumonia (27, 28). Overall, the data from the available studies do not yet definitively show that antiviral drugs have an effect on relevant influenza-related complications.
Do antiviral drugs affect viral shedding?
Reduced viral shedding after 24 or 48 hours was found for study participants with no comorbidities receiving neuraminidase inhibitors (29, 30). As there is no evidence as yet that shortened or reduced viral shedding also reduces transmission to contact persons, the clinical relevance remains questionable.
Does the timing of administration play a role in efficacy?
One RCT showed that oseltamivir shortens the duration of influenza symptoms significantly, by 43 hours, if taken within 24 hours of symptom onset (31). Multiple observational studies yield findings in line with this (19, 24, 32, 33). There is no similar data for zanamivir or amantadine. However, shorter or milder course of disease was also confirmed when oseltamivir therapy was begun more than 48 hours after symptom onset (29, 33, 34). For post-exposure prophylaxis, systematic evaluations of efficacy depending on the timing of start of administration in relation to contact with ill individuals are lacking.
What are the side effects of antiviral drugs?
In terms of types of side effect, there is no difference between prophylactic and therapeutic use. For neuraminidase inhibitors this applies also to the frequencies of side effects. (etable 3). Regarding the safety of long-term neuraminidase inhibitor use, data is available for a period of up to 16 weeks (35). For amantadine, it seems that therapy in adults, which is shorter than prophylactic use, is better tolerated.
eTable 3
| Adverse event | Drug (literature reference) | No. of RCTs | No. of partici- pants | Effect estimate | Value [95% CI] | Number needed to harm (NNH) |
| Adults with no comorbidities | ||||||
| All events | Oseltamivir (21) | 4 | 1623 | OR | 0.81 [0.59; 1.12] | NNT: 18*2 |
| Zanamivir (21) | 2 | 1054 | OR | 1.03 [0.79; 1.34] | 177*2 | |
| Drug-related | Oseltamivir (21) | 2 | 509 | OR | 1.45 [0.83; 2.53] | 24*2 |
| Zanamivir (21) | 4 | 1406 | OR | 1.11 [0.76; 1.62] | 173*2 | |
| Serious | Oseltamivir (21) | 3 | 985 | OR | 0.32 [0.03; 1.17] | NNT: 251*2 |
| Oseltamivir (15) | 7 | 4394 | RR | 0.96 [0.51; 1.80] | NNT: 630*2 | |
| Zanamivir (21) | 3 | 1130 | 1.14 [0.28; 7.35] | 537*2 | ||
| Zanamivir (15) | 10 | 4388 | RR | 0.98 [0.49; 1.50] | NNT: 4 048*2 | |
| Treatment discontinued*1 | Oseltamivir (15) | 8 | 4452 | RR | −0.17 [−0.53; 0.87] | 105*2 |
| Zanamivir (15) | 13 | 6116 | RR | 0.96 [0.66; 1.39] | 581*2 | |
| Nausea*1 | Oseltamivir (15) | 8 | 4452 | RR | 1.57 [1.14; 2.15] | 21 |
| Vomiting*1 | Oseltamivir (15) | 9 | 4452 | RR | 2.43 [1.75; 3.38] | 17 |
| Nausea/vomiting | Zanamivir (15) | 15 | 6553 | RR | 0.60 [0.39; 0.94] | 1 967*2 |
| Diarrhea*1 | Oseltamivir (15) | 8 | 4452 | RR | 0.67 [0.46; 0.98] | NNT: 80*2 |
| Zanamivir (15) | 15 | 6553 | RR | 0.87 [0.66; 1.14] | NNT: 232*2 | |
| Gastrointestinal | Amantadine (16) | 3 | 494 | OR | 1.34 [0.32; 5.61] | 275*2 |
| Increased CNS activity | Amantadine (16) | 2 | 465 | OR | 0.77 [0.23; 2.53] | NNT: 66*2 |
| Reduced CNS activity | Amantadine (16) | 3 | 491 | OR | 0.65 [0.31; 1.38] | NNT: 12 |
| Skin alterations | Amantadine (16) | 2 | 465 | OR | 1.4 [0.14; 13.78] | 332*2 |
| Children with no comorbidities | ||||||
| All events | Oseltamivir (21) | 1 | 334 | OR | 0.91 [0.59; 1.40] | NNT: 42*2 |
| Zanamivir (21) | 2 | 737 | OR | 0.88 [0.62; 1.24] | NNT: 39*2 | |
| Drug-related | Zanamivir (21) | 2 | 737 | OR | 1.32 [0.59; 2.92] | 66*2 |
| Serious | Oseltamivir (21) | 1 | 695 | OR | 1.54 [0.25; 9.24] | 331*2 |
| Oseltamivir (15) | 2 | 1029 | RR | 1.97 [0.59; 6.56] | 127*2 | |
| Zanamivir (21) | 2 | 737 | OR | 2.29 [−0.24; 22.09] | 200*2 | |
| Oseltamivir/Zanamivir (18) | 4 | 2172 | RD | 0.00 [0.00; 0.01] | 165*2 | |
| Treatment discontinued*1 | Oseltamivir/Zanamivir (18) | 3 | 1143 | RD | 0.01 [−0.02; 0.03] | 131*2 |
| Oseltamivir (15) | 2 | 1029 | RR | 0.99 [0.33; 3.01] | NNT: 295 | |
| Zanamivir (18) | 2 | 737 | RR | 1.18 [0.16; 8.88] | NNT: 214*2 | |
| Nausea*1 | Oseltamivir/Zanamivir (18) | 4 | 1766 | RD | −0.01 [−0.03; 0.00] | 76*2 |
| Oseltamivir (15) | 3 | 1358 | RR | 0.87 [0.50; 1.51] | 180*2 | |
| Vomiting*1 | Oseltamivir (18) | 3 | 1435 | RD | 0.06 [0.03; 0.10] | 15 |
| Oseltamivir (15) | 3 | 1358 | RR | 1.70 [1.23; 2.35] | 19 | |
| Zanamivir (18) | 2 | 737 | RD | 0.00 [−0.02; 0.02] | NNT: 150*2 | |
| Nausea/vomiting*1 | Zanamivir (15) | 2 | 737 | RR | 0.54 [0.24; 1.22] | NNT: 45*2 |
| Amantadine (17) | 2 | 599 | RR | 0.54 [0.15; 2.0] | NNT: 27*2 | |
| Diarrhea*1 | Oseltamivir/Zanamivir (18) | 5 | 2172 | RD | −0.01 [−0.03; 0.00] | NNT: 36*2 |
| Oseltamivir (15) | 3 | 1358 | RR | 0.87 [0.58; 1.28] | NNT: 103*2 | |
| Zanamivir (15) | 2 | 737 | RR | 0.52 [0.15; 1.75] | NNT: 93*2 | |
| Amantadine (17) | 2 | 599 | RR | 0.81 [0.43; 1.47] | NNT: 85*2 | |
| Exanthema | Amantadine (17) | 2 | 599 | RR | 0.81 [0.43; 1.53] | NNT: 175*2 |
| Malaise | Amantadine (17) | 2 | 599 | RR | 0.89 [0.41; 1.96] | NNT: 10 |
| Muscle/limb pain | Amantadine (17) | 2 | 599 | RR | 0.85 [0.46; 1.59] | NNT: 91*2 |
| Headache | Amantadine (17) | 2 | 599 | RR | 0.73 [0.52; 1.03] | NNT: 18*2 |
| Dyspnea | Amantadine (17) | 1 | 159 | RR | 0.37 [0.02; 9.02) | NNT: 84*2 |
| Dizziness | Amantadine (17) | 2 | 599 | RR | 6.63 [0.32; 137.33] | 132*2 |
| Insomnia | Amantadine (17) | 2 | 599 | RR | 0.46 [0.12; 1.74] | 80*2 |
| Arrhythmia | Amantadine (17) | 2 | 599 | RR | 0.0 [0.0; 0.0] | N/A |
| High-risk patients | ||||||
| All events | Oseltamivir (21) | 2 | 452 | RR | 0.96 [0.63; 1.46] | NNT: 457*2 |
| Zanamivir (21) | 4 | 1286 | RR | 1.24 [0.96; 1.60] | NNT: 39*2 | |
| Drug-related | Zanamivir (21) | 1 | 5246 | RR | 1.01 [0.55; 1.85] | 1 439*2 |
| Serious | Zanamivir (21) | 3 | 1210 | RR | 0.72 [0.32; 1.62] | 321*2 |
*1Studies of prophylaxis and therapy
*2The 95% confidence interval has no upper bound. as a result of very similar risk rates or small total case numbers in both treatment groups. Substantial uncertainty is therefore associated with NNT or NNH.
95% CI: 95% confidence interval; N/A: Not available or not applicable; NNT: Number needed to treat; NNH: Number needed to harm; OR: Odds ratio; RCT: Randomized controlled trial; RD: Risk difference; RR: Relative risk
The most frequent and also the most serious adverse effects, such as neuropsychiatric effects, have been reported for amantadine. Gastrointestinal side-effects are characteristic of oseltamivir. Nausea and vomiting are significantly more common than with placebo use (NNT: 21 versus 17). In addition, neuropsychiatric effects have been the subject of critical discussion. These were reported in adolescents and young adults in particular, most commonly in Japan. Adverse effects of zanamivir are particularly likely to affect the respiratory tract; this is partly due to the route of administration and must be taken into account for patients with existing or concomitant respiratory diseases.
The current summaries of product characteristics of individual antiviral drugs must also be consulted, as these include all data from controlled clinical trials and post-authorization market surveillance (36– 38).
Discussion
For influenza prophylaxis, the efficacy of both post-exposure and long-term prophylaxis is rated as good. The protective effect of antiviral drugs lasts only while they are being used.
When used therapeutically during seasonal influenza, antiviral drugs reduce disease duration by 0.5 to 1 days if treatment begins no more than 48 hours after symptom onset. The efficacy of oseltamivir, zanamivir, and amantadine varies between individual patient groups; overall, it should be seen as moderate.
The efficacy of antiviral drugs for influenza-related complications and mortality is of particular clinical relevance. The 2014 Cochrane Review (15), which includes the greatest amount of trial data, found a reduction in cases of pneumonia for oseltamivir in adults with no comorbidities. This finding is not undisputed, however, as in some trials complications of influenza were reported retrospectively by trial participants themselves instead of being diagnosed by investigators. The diagnosis of pneumonia was often not confirmed.
The RCTs were conducted during seasonal influenza outbreaks. The course of disease was mostly mild and uncomplicated, and there were only isolated deaths. The trials were not primarily designed to show an effect on serious disease courses. It was therefore also impossible to draw any conclusions concerning mortality from a pooled analysis of RCTs, as a result of small sample sizes (15). In the event of a future serious influenza pandemic with high disease rates and highly pathogenic, NI-sensitive viruses, effects may be stronger in terms of both NNT and complications. However, the investigated strains of influenza from the 2008/2009 season in Germany and the USA showed primary oseltamivir resistance in 99% and 98% of cases respectively (39), whereas almost 100% of the pandemic H1N1 viruses in the following years were sensitive to oseltamivir. Viral resistance must therefore be continuously monitored.
Overall, although the available data does give some indications that antiviral drugs have a positive effect on clinically relevant complications of influenza, it does not provide conclusive evidence of this.
However, the findings of the observational study by Muthuri et al. during the 2009/2010 pandemic are noteworthy in this regard (23). This meta-analysis includes datasets from 30 000 hospitalized patients in 78 individual trials and shows that the use of neuraminidase inhibitors, usually oseltamivir, was associated with a statistically significant relevant risk reduction for mortality in adults. The finding for adults receiving intensive care, a probably more homogeneous subgroup, was consistent with the finding for the general population. No effect on mortality was found in children. This study’s strengths are its large sample size and the fact that its evaluation takes into account that the probability of administering antiviral medication therapy depends on various factors, including disease severity (propensity score). The study methods are considered valid. However, a fundamental weakness of observational studies when compared to RCTs is that it is impossible to rule out biased findings as result of influencing factors that are unknown and therefore not taken into account in the analysis.
Potential limitations of this systematic literature review are that most of the RCTs were included in all the systematic reviews considered for this article, so the findings of the analyses seem very homogeneous overall. In addition, most studies and some meta-analyses were funded by pharmaceutical companies.
After more than 10 years of use, antiviral drugs’ side effect profile is well characterized. Overall, the side effects of NIs are less severe than those of amantadine. Serious side effects have been reported in less than 1 in 100 treated patients. The authors therefore rate the side effect profile of neuraminidase inhibitors overall as giving little cause for concern.
Unlike amantadine, development of resistance has been infrequent for the neuraminidaseinhibitors to date. However, some years ago an influenza A (H1N1) virus with primary oseltamivir resistance circulated (39). Of course, it is possible that resistance may develop and render one or both neuraminidase inhibitors ineffective.
It became evident that for some at-risk populations evidence of efficacy is patchy or even absent, e.g. for pregnant women, neonates, immunosuppressed patients, and patients with cystic fibrosis (40). The authors believe that more research is needed in these patients, and also on the use of antiviral drugs in infections caused by highly pathogenic influenza viruses.
This systematic review does not aim to provide clinical recommendations on the use of antiviral drugs.
In line with the evidence we have presented, and regardless of the fact that for some issues the available data is weak or patchy, specialist societies and national and international public health institutions have for some years provided recommendations on the use of antiviral drugs in seasonal and pandemic influenza that are broadly in line with each other (box).
Summary
The benefits of antiviral drugs are rated to outweigh the risks, particularly for neuraminidase inhibitors. To date there are no alternative causal treatment options for influenza viruses. When prescribing antiviral drugs the characteristics of circulating viruses and the individual risk profile of the patient should be borne in mind, in line with the recommendations of specialist societies. Neuraminidase inhibitors are generally preferable, due to their safety profile. Even in a best case scenario, in the event of a pandemic approximately 6 months elapse between the appearance of a pandemic influenza virus and the first use of a corresponding vaccine; in addition, it is unclear how much protection a vaccine will afford. Antiviral drugs may therefore be important, particularly in reducing mortality and morbidity in the population during the first wave of a pandemic.
eTable 1
| ID | Search terms: PubMed |
| #1 | (oseltamivir* OR zanamivir* OR laninamivir* OR tamiflu* OR relenza* OR peramivir* OR gs4071* OR Neuraminidase* OR "neuraminidase inhibitors" OR Amantadin* OR Rimantadin* OR adamantan* OR antiviral*[Title/Abstract]) |
| #2 | (influenza* OR flu OR "Influenza A Virus" OR "H1N1 Subtype*" OR "Influenza A Virus" OR "H1N1 Subtype*" OR "swine flu" OR "swine influenza" OR "H1N1 pandemic influenza" OR "H1N1v" OR "pandemic influenza 2009" OR "novel influenza" OR H1N1pdm OR "swine-origin influenza" OR "swine-origin type A" OR nH1N1 OR pH1N1 OR H1N1 OR AH1N1 OR A/H1N1 OR "A (H1N1)" OR "A (H1N1) pdm09" OR *H1N1* OR "H1N1 virus*" OR "H3N2 virus*" OR "H5N1 virus*" OR "H7N9 virus*" OR *H3N2* OR AH3N2 OR A/H3N2 OR "A (H3N2)" OR *H5N1* OR AH5N1 OR A/H5N1 OR "A (H5N1)" OR *H7N9* OR AH7N9 OR A/H7N9 OR "A (H7N9)" OR "Influenzavirus A" OR "Influenzavirus B" OR "Influenzavirus C" OR "Influenza A Virus" OR "Influenza B Virus" OR "Influenza C Virus"[Title/Abstract]) |
| #3 | #1 AND #2 |
| #4 | #3 AND Filter "Meta-Analysis*” OR "Systematic Review*” |
| #5 | #3 AND ("systematic review*" OR "meta-analysis*") |
| #6 | #3 Filter Clinical Trial[ptyp] OR Clinical Trial. Phase I[ptyp] OR Clinical Trial. Phase II[ptyp] OR Clinical Trial. Phase III[ptyp] OR Clinical Trial. Phase IV[ptyp] OR Controlled Clinical Trial[ptyp] OR Randomized Controlled Trial[ptyp] |
| #7 | #3 AND ("Clinical Trial" OR "Controlled Clinical Trial" OR "Randomized Controlled Trial") |
| #8 | #3 AND ((randomized controlled trial*) OR (randomised controlled trial*) OR (RCT*) OR (randomi* trial*) OR ("randomized controlled trial") OR (randomi* study*)) |
| #9 | #3 AND Filter Randomized Controlled Trial |
| ID | Search terms: Cochrane Library |
| #1 | (oseltamivir* OR zanamivir* OR laninamivir* OR tamiflu* OR relenza* OR peramivir* OR gs4071* OR Neuraminidase* OR "neuraminidase inhibitors" OR amantadin* OR rimantadin* OR adamantan* OR antiviral*):ti.ab.kw |
| #2 | (influenza* OR Influenza* OR flu OR "Influenza A Virus" OR "H1N1 Subtype" OR "swine flu" OR "swine influenza" OR "H1N1 pandemic influenza" OR H1N1v OR "pandemic influenza 2009" OR "novel influenza" OR H1N1pdm OR "swine-origin influenza" OR "swine-origin type A" OR nH1N1 OR pH1N1 OR H1N1 OR AH1N1 OR A/H1N1 OR "A (H1N1)" OR "A (H1N1) pdm09" OR *H1N1* OR "H1N1 virus*" OR "H3N2 virus*" OR "H5N1 virus*" OR "H7N9 virus" OR *H3N2* OR AH3N2 OR A/H3N2 OR "A (H3N2)" OR *H5N1* OR AH5N1 OR A/H5N1 OR "A (H5N1)" OR *H7N9* OR AH7N9 OR A/H7N9 OR "A (H7N9)" OR "Influenzavirus A" OR "Influenzavirus B" OR "Influenzavirus C" OR "Influenza A Virus" OR "Influenza B Virus" OR "Influenza C Virus"):ti.ab.kw |
| #3 | #1 AND #2 |
| #4 | #3 in Cochrane Reviews AND Other Reviews |
| #5 | #3 AND ("systematic review*" OR "meta-analysis*") |
| #6 | #3 in Trials |
| #7 | #3 AND ("Clinical Trial" OR "Controlled Clinical Trial" OR "Randomized Controlled Trial") |
| #8 | #3 AND ((randomized controlled trial*) OR (randomised controlled trial*) OR (RCT*) OR (randomi* NEAR trial*) OR ("randomized controlled trial") OR (randomi* NEAR study*)) |
| ID | Search terms: Scopus |
| #1 | (TITLE-ABS-KEY(oseltamivir* OR zanamivir* OR laninamivir* OR tamiflu* or relenza* or peramivir* or gs4071* or Neuraminidase* or "neuraminidase inhibitors” or amantadine* or rimantadin* or adamantan* or antiviral*)) |
| #2 | (TITLE-ABS-KEY(influenza* OR flu OR "Influenza A Virus” OR "H1N1 Subtype” OR "swine flu” OR "swine influenza” OR "H1N1 pandemic influenza” OR H1N1v OR "pandemic influenza 2009” OR "novel influenza” OR H1N1pdm OR "swine-origin influenza” OR "swine-origin type A” OR nH1N1 OR pH1N1 OR H1N1 OR AH1N1 OR A/H1N1 OR "A (H1N1)” OR "A (H1N1) pdm09” OR *H1N1* OR "H1N1 virus*” OR "H3N2 virus*” OR "H5N1 virus*” OR "H7N9 virus” OR *H3N2* OR AH3N2 OR A/H3N2 OR "A (H3N2)” OR *H5N1* OR AH5N1 OR A/H5N1 OR "A (H5N1)” OR *H7N9* OR AH7N9 OR A/H7N9 OR "A (H7N9)” OR "Influenzavirus A” OR "Influenzavirus B” OR "Influenzavirus C” OR "Influenza A Virus” OR "Influenza B Virus” OR "Influenza C Virus”)) |
| #3 | #1 AND #2 |
| #4 | #3 Filter german language |
| #5 | #4 AND Filter Review |
| #6 | #4 AND ("systematic review*" OR "meta-analysis*") |
| #7 | #4 AND ("Clinical Trial" OR "Controlled Clinical Trial" OR "Randomized Controlled Trial") |
| #8 | #4 AND ((randomized controlled trial*) OR (randomised controlled trial*) OR (RCT*) OR (randomi* PRE/10 trial*) OR ("randomized controlled trial") OR (randomi* PRE/10 study*)) |
Acknowledgments
Acknowledgement
We would like to thank the members of the Expert Advisory Board on Influenza of the Robert Koch Institute (RKI) for their valuable contributions to the content of this article: Dr. Bernhard Bornhofen, PD Dr. Roswitha Bruns, Prof. Dr. Petra Gastmeier, Prof. Dr. Timm Harder, Prof. Dr. Ulrich Hartenauer, Prof. Dr. Eberhard Hildt, Prof. Dr. Hanna Kaduszkiewicz, Peter Lang, Prof. Dr. Thomas Mertens, Prof. Dr. Georg Peters, and Prof. Dr. Horst Schroten.
We would also like to thank the following guests of the Expert Advisory Board on Influenza for their advice during sessions: the Consortium of the Upper Federal State Authorities (AOLG, Arbeitsgemeinschaft der oberen Landesbehörden), the Working Group on Protection Against Infection of the AOLG, the German Federal Office of Civil Protection and Disaster Assistance (BBK, Bundesamt für Bevölkerungsschutz und Katastrophenhilfe), the German Medical Association (Bundesärztekammer), the German Federal Ministry of Labour and Social Affairs, the German Federal Ministry of Health, the German Federal Ministry of Defence, the German Federal Ministry of the Interior, the Federal Union of German Associations of Pharmacists (ABDA, Bundesvereinigung Deutscher Apothekerverbände), the German Hospital Federation (Deutsche Krankenhausgesellschaft), the Federal Joint Committee (G-BA, Gemeinsamer Bundesausschuss), the National Association of Statutory Health Insurance Physicians (KBV, Kassenärztliche Bundesvereinigung), the Umbrella Organization of Health Insurers (Spitzenverband Bund der Krankenkassen), and the German Standing Vaccination Committee (STIKO, Ständige Impfkommission) of the Robert Koch Institute.
In addition, our thanks go to Prof. Dr. med. Walter Haas for his helpful specialist comments and to Yvonne Bichel for performing the literature search.
Footnotes
Conflict of interest statement
The authors declare that no conflict of interest exists.
Translated from the original German by Caroline Shimakawa-Devitt, M.A.
Facebook
Twitter
Google+