Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
ORIGINAL ARTICLE
ABSTRACT
Objectives The development of resistance to multiple
antimicrobial agents has limited treatment options for
gonorrhoea. The potential emergence of cephalosporin
resistance in Neisseria gonorrhoeae and cephalosporin
allergy in some patients make it necessary to evaluate
the effectiveness of other available antimicrobial agents.
Gentamicin is widely available in the USA and is used for
gonorrhoea treatment in several countries. We conducted
a systematic review of the medical literature to assess
the effectiveness of gentamicin for treatment of
uncomplicated urogenital gonococcal infections.
Methods Two reviewers assessed relevant articles and
independently selected studies that met prespecied
selection criteria (including systematic enrolment and
assignment to treatment and culture-conrmed diagnosis
and outcome). Summary measures for selected studies
were pooled using inverse variance-weighted averages
with xed effects. Heterogeneity was assessed using I2,
which estimates proportion (0100%) of variability
attributable to heterogeneity between studies. Pooled
percentage with negative follow-up culture was
compared with Centers for Disease Control and
Prevention (CDC) criteria for selection of recommended
therapy (lower 95% CI of efcacy 95%).
Results Twenty-nine potentially relevant studies were
identied; three met inclusion criteria. Two studies used
240 mg intramuscular gentamicin and one used 280 mg.
Percentages with negative culture after single-dose
treatment were 90.7% (n=86), 91.4% (n=220) and
95.0% (n=40). Pooled percentage with negative culture
after single-dose treatment was 91.5% (95% CI 88.1% to
94.0%, I2=0%).
Conclusions Gentamicin does not meet current CDC
criteria for recommended treatment of gonorrhoea.
However, if cephalosporin resistance emerges,
gentamicin may be a useful alternative agent. Evaluation
of additional regimens, including combination therapy, is
warranted.
BACKGROUND
Neisseria gonorrhoeae has demonstrated the ability to
develop resistance to several antibiotics. Penicillin,
tetracycline and uoroquinolones are no longer
considered sufciently effective for the treatment of
gonococcal infections after the emergence of resistance to these antibiotics. Spectinomycin is efcacious but is no longer available in the USA. The
Centers for Disease Control and Prevention (CDC)
currently recommends dual therapy for gonorrhoea
with a cephalosporin (preferably ceftriaxone) and
azithromycin or doxycycline to ensure treatment of
co-occurring pathogens and reecting concern about
Sex Transm Infect 2012;88:589594. doi:10.1136/sextrans-2012-050604
METHODS
We developed and followed a review protocol (see
supplementary text, available online only). We
searched electronic databases (PubMed, EMBASE, the
Cochrane Central Register of Controlled Trials, Web
of Science), registers of ongoing trials (Clinicaltrials.
gov,9 CenterWatch,10 Current Controlled Trials11 and
the WHO International Clinical Trials Registry
Platform),12 and reference lists of found articles for
relevant studies. The PubMed search strategy was
entered as (gentamicin OR gentamicin(MeSH
Terms)) AND (gonorrhoea OR gonococcal OR N
gonorrhoeae OR sexually transmitted diseases OR
sexually transmitted infections OR urethritis
OR cervicitis) OR (gonorrhoea OR N gonorrhoeae OR
sexually transmitted diseases OR sexually transmitted diseases, bacterial(MeSH Terms)). Similar
searches were used for the other databases. Databases
were last searched on 17 June 2012. We did not apply
date or language restrictions. Abstracts were reviewed,
and full text articles were retrieved for studies that
appeared to be trials of gentamicin in patients with
gonococcal infection.
Two reviewers (DD and RDK) assessed relevant
articles and independently selected studies that
met prespecied selection criteria. We included
studies of adolescents and/or adults with uncomplicated urogenital gonococcal infection treated
with single-dose intramuscular (IM) or intravenous
gentamicin (any dose). Patients were considered to
have complicated gonococcal infection if they were
described as having a diagnosis of pelvic inammatory disease or disseminated gonococcal infection.
We required included studies to have used urethral
589
Clinical
or cervical culture for diagnosis of N gonorrhoeae at the time of
treatment and for follow-up cultures of N gonorrhoeae to have
been obtained from the same anatomical site as the initial
infection between 2 and 21 days post-treatment. To minimise
selection bias, we included controlled trials in which assignment to the treatment group was random or systematic and
not based on the severity of symptoms or patient characteristics. In addition, because we believed controlled trials would be
rare, we included single-arm case series, provided that enrolment in the study was systematic (eg, enrolment offered to
consecutive patients) and that any exclusions were prespecied.
Initial agreement and kappa for assessment of studies to be
included was calculated. Discrepancies in judgements regarding
inclusion were adjudicated by discussion. For included studies,
data including numbers of culture-positive patients treated,
numbers with negative test of cure and dose of gentamicin
used were extracted and recorded independently by two
reviewers (DD and RDK) using a data extraction form. Data
were compared, and differences were resolved through
discussion.
Treatment success was dened as the percentage of negative
cultures obtained from initially culture-positive anatomical
sites within 221 days post-treatment. In the primary analysis
we included all participants who received treatment and did
not exclude participants based on perceived re-exposure.
Summary measures of treatment success for all selected studies
were pooled using inverse variance-weighted averages (in which
the weight of each study is the reciprocal of the variance in the
studys estimate of treatment effect) with xed effects. Fixed
effects were assumed because we hypothesised there would be
minimal variability in effect of an antibiotic on different urogenital infections caused by N gonorrhoeae. Heterogeneity was
assessed using I2, which estimates proportion (0100%) of variability attributable to heterogeneity between studies. Pooled
percentage with negative follow-up culture was compared with
CDC criteria for selection of recommended therapy (95% lower
95% CI of efcacy at least 95%) and for alternative treatments
(efcacy of at least 95% with a lower 95% CI of at least
90%).13 Prespecied sensitivity analyses were performed to
determine impact of the following on pooled treatment effect
and precision of the estimate of treatment effect: excluding
trials that were not randomised, controlled trials; excluding
results for patients who reported sexual activity between gentamicin treatment and test of cure; and excluding results for
patients given doses other than 240 mg gentamicin IM.
Sensitivity analyses were also planned to determine impacts of
excluding women, of excluding results for patients given antibiotics in addition to gentamicin between 7 days before treatment and culture test of cure, of excluding patients known to
have symptoms of gonorrhoea for more than 7 days before
treatment, of excluding HIV-negative patients, and of excluding
HIV-positive patients.
RESULTS
We identied 913 records through database searches and reference lists of found articles. We retrieved the full text for the 29
studies among those that appeared to be trials of gentamicin
treatment for gonorrhoea. After independent review and discussion, we agreed that 26 studies (table 1) did not meet prespecied inclusion criteria because they were not treatment trials of
single-dose gentamicin, because we could not conrm that
study enrolment or assignment to treatment was systematic, or
because culture was not used for diagnosis and for test of cure
(gure 1). Agreement for initial judgement on meeting
590
DISCUSSION
A systematic review of the medical literature for trials of singledose gentamicin treatment of culture-conrmed uncomplicated
urogenital gonorrhoea found treatment efcacy of greater than
90% for all studies and a pooled efcacy of 91.5% with a lower
95% condence limit for efcacy of 88.1%. Only one study (a
randomised, controlled trial) found a treatment efcacy of 95%
(with lower 95% condence limit of 88%).14 Based on our
primary analysis, gentamicin does not meet current CDC
Sex Transm Infect 2012;88:589594. doi:10.1136/sextrans-2012-050604
Clinical
Table 1 Studies considered but excluded from the systematic review
Author
Bartunek et al
19
Year
1974
Bowie et al20
1974
Brown et al8
2007
Bukharovich and
Revunov23
1991
Danda et al w5
1976
Felarca et al29
1971
Gomez-Medinaw6
1969
Hantschke et al21
1973
Hantschke et al17
1973
Hantschke and
Mauss22
1970
Iskandar et al w7
1978
Lawrence et al w8
1974
Mirarchi et al w9
1971
Morrison and
Reevesw10
1973
Pandhi et al w11
1989
Pareek and
Chowdhuryw12
1981
1969
Price et al24
2003
Rajanw14
Reevesw15
Serriw16
Sibouletw17
1983
1974
1969
1972
Strauss and
Hantschkew18
Tan et al18
1973
Yoon et al w19
1988
Zabaneh et al w20
1980
1980
58/62 Patients (93.5%) with gonorrhoea were cured (criteria not specified) with
5 mg/kg gentamicin IM
83/86 Women (96.5%) and 10/10 men (100.0%) with penicillinase-producing N
gonorrhoeae had negative cultures at 3 days, 1 week and 2 weeks after treatment
with gentamicin 280 mg IM1
78/125 Male patients (62.4%) with uncomplicated gonococcal infection treated
with gentamicin 240 mg IM1 had negative Gram stain and culture results at 3
5 days
15/15 Men (100.0%) with gonococcal urethritis had disappearance of urethral
discharge and a negative culture 3 days after treatment with gentamicin 280 mg
IM1
IM, intramuscular.
591
Clinical
Figure 1 Flow of studies included in
the systematic review.
Table 2
alternative (95% efcacy with lower 95% CI of 90%) treatment, it could potentially prove to be a treatment option if
cephalosporin resistance emerges or for patients with severe
cephalosporin allergy. In an era of dwindling treatment options,
the current CDC criteria for recommended and alternative treatment may need to be reconsidered. It is also worth noting that
the CDC criteria for recommended treatment of gonorrhoea
have been applied to pooled estimates30 including trials that
excluded re-exposed patients in determining efcacy,31w2 and
that these pooled estimates may have been lower if all studies
were analysed according to assigned treatment (intent-to-treat)
without the exclusion of patients who were considered
re-exposed. As an example, the authors of one study that we
included retrospectively excluded patients deemed re-infected.16
Our analysis, based on an intent-to-treat approach, found a
lower efcacy (91.4%) than the original study authors (98.0%),
because we included all patients who were enrolled and received
the study drug. Although it may seem logical to exclude patients
who were re-exposed to N gonorrhoeae after treatment, retrospectively excluding these patients may introduce bias into the
results, particularly when patients who had positive follow-up
cultures were retrospectively evaluated for re-exposure while
patients with negative cultures were not.
Our ndings are subject to several limitations. First, we
found a small number of studies that met selection criteria
Author
Year
Study design
Population
Daly et al15
1997
Single-arm case
series
240 mg IM
Hira et al16
1985
Lule et al14
1994
240 mg IM
280 mg IM
Weight in pooled
analysis
Efficacy of gentamicin
(95% CI)
Culture on
post-treatment
day 7
Culture on
post-treatment
day 14
27.4%
65.4%
Culture on
post-treatment
day 810
7.2%
Outcome
100%
90.7% (88.1%94.0%)
IM, intramuscular.
592
Clinical
established to exclude studies with a higher risk of bias. We
excluded studies from our evaluation of efcacy for two main
reasons. Either they could not be compared to included studies
because they did not measure outcomes by the same method
(culture) or did not use single-dose gentamicin treatment, and/
or they did not report systematic assignment to treatment.
That is, it is possible that these studies selected patients judged
by investigators to be more likely to respond to treatment
rather than treating all patients with gentamicin. Of the three
studies we included, only one used randomised assignment to
treatment. However, the other two studies did use systematic
assignment to treatment, thus minimising a potential source of
bias. In addition, treatment was found to be as effective in the
single randomised, controlled trial as in the two trials without
control groups.
Second, given the limited patient population in the three
included studies, we cannot be sure that the results apply to
all populations. While none of the information from excluded
trials suggested women would have a lower rate of successful
treatment, none of the included trials included women, so it is
possible the ndings do not apply to women or girls. While we
were unable to analyse the effectiveness of gentamicin for
HIV-positive patients with gonorrhoea, the study that reported
44% of enrolees were HIV positive also reported in the text that
the cure rate for gonorrhoea was not affected by concurrent HIV
infection.14 Therefore, we did not nd evidence that gentamicin
was less effective among HIV-positive patients. All patients
were from two countries (Malawi and Zambia) in sub-Saharan
Africa, and all data included were collected before 1997, and so it
would be important to monitor current in-vitro susceptibility
and clinical effectiveness in these and other settings where gentamicin is used for the treatment of gonorrhoea. As an example,
a recent evaluation of gentamicin susceptibility among N gonorrhoeae isolates in Europe demonstrated that 95% of isolates had
MIC of 8 g/ml or less.w3 Interpreting in-vitro susceptibility
data will be challenging, however, as we lack MIC thresholds for
resistance to gentamicin.
Third, due to the available literature and our study design,
we cannot infer efcacy for non-genital infections. However,
spectinomycin has poor efcacy for the treatment of pharyngeal gonorrhoea,w4 so it is expected that single-dose gentamicin
would also have limited efcacy for pharyngeal gonorrhoea. We
did not nd data regarding the efcacy of gentamicin for rectal
gonorrhoea infections.
Finally, we assumed xed effects rather than random effects
in pooling summary measures of treatment efcacy, which
could potentially underestimate heterogeneity between studies.
Our nding of I2 of 0% for our primary analysis (indicating 0%
of variability was due to heterogeneity between studies) provides reassurance that an assumption of xed effects was
reasonable.
The pooled efcacy of single-dose gentamicin does not meet
CDC criteria for a recommended or alternative therapy for gonorrhoea. However, we are facing the potential emergence of
multidrug-resistant gonorrhoea at a time of dwindling treatment options. Gentamicin may be a treatment option for
patients with severe cephalosporin hypersensitivity or when
gonococcal cephalosporin resistance emerges. A clinical trial
sponsored by the National Institute of Allergy and Infectious
Diseases is currently evaluating the efcacy of a combination of
gentamicin and azithromycin for the treatment of urogenital
gonorrhoea. Evaluation of additional treatment options is
warranted.
Key messages
The emergence of cephalosporin-resistant gonorrhoea could
leave us with no treatment options that meet the criteria for
recommended gonorrhoea treatment.
A systematic review of gentamicin for gonorrhoea treatment
found gentamicin does not meet current CDC criteria for
recommended gonorrhoea treatment.
Gentamicin may be a useful alternative if cephalosporin
resistance emerges.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
593
Clinical
23.
24.
25.
26.
594
27.
28.
29.
30.
31.
Barclay ML, Begg EJ. Aminoglycoside toxicity and relation to dose regimen.
Adverse Drug React Toxicol Rev 1994;13:20734.
Jackson GG, Arcieri G. Ototoxicity of gentamicin in man: a survey and controlled analysis
of clinical experience in the United States. J Infect Dis 1971;124:S1307.
Felarca AB, Laqui EM, Ibarra LM. Gentamicin in gonococcal urethritis of Filipino
males: dosage and response. J Infect Dis 1971;124:S28792.
Moran JS, Levine WC. Drugs of choice for the treatment of uncomplicated
gonococcal infections. Clin Infect Dis 1995;20(Suppl 1):S4765.
Cavenee MR, Farris JR, Spalding TR, et al. Treatment of gonorrhea in pregnancy.
Obstet Gynecol 1993;81:338.
Copyright of Sexually Transmitted Infections is the property of BMJ Publishing Group and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.