Striving to be the best.

Moxifloxacin is a recent addition to the fluoroquinolone class, differing from ciprofloxacin and other older agents in having much better in vitro activity against Grampositive aerobes while retaining potent activity against Gram-negative aerobes. It is also active against the pathogens of human and animal bite wounds and those species of atypical mycobacteria associated with dermatologic infections. Its activity against anaerobes is quite variable. Moxifloxacin penetrates well into inflammatory blister fluid and muscle and subcutaneous adipose tissues. Moxifloxacin should thus be a reasonable option for the treatment of skin and skin structure infections (SSSIs). In 3 randomized controlled trials (RCTs), oral moxifloxacin was as effective as cephalexin in the treatment of uncomplicated SSSIs in adults while in 2 RCTs, intravenous/oral moxifloxacin was as effective as intravenous/oral β-lactam/ β-lactamase inhibitor therapy in the treatment of complicated SSSIs in adults. Moxifloxacin does not inhibit cytochrome P450 enzymes and thus interact with warfarin or methylxanthines. However, multivalent cations can reduce its oral bioavailability substantially. Dosage adjustment is not required in the presence of renal or hepatic impairment. The clinical relevance of its electrophysiologic effects (QTc prolongation) remains unresolved.


Introduction
Moxifloxacin is one of most-recently marketed of the fluoroquinolone (hereafter called quinolone) antimicrobials (Avelox ® , Bayer, Leverkusen, Germany). Of this group (gatifloxacin, gemifloxacin, moxifloxacin), only moxifloxacin has regulatory approval for use in skin and skin structure infections (SSSIs) (Anonymous 2005). Until recently, this approval was limited to uncomplicated SSSIs in adults caused by Staphylococcus aureus or Streptococcus pyogenes (Anonymous 2005). In June of 2005, it also received approval for use in complicated SSSIs caused by methicillinsusceptible S. aureus, Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae (Anonymous 2005). Complicated SSSIs are defined as those which involve deeper layers such as fascia or muscle, require significant surgical interventions, or arise in the presence of significant comorbidities such as diabetes or HIV infection (Raghavan and Linden 2004). The purpose of this paper is to review the in vitro antimicrobial activity, pharmacokinetics, clinical and bacteriological efficacy, safety, tolerability, and drug-drug interaction potential of moxifloxacin, concentrating on its potential role in the treatment of complicated (and uncomplicated) SSSIs. quinolones by either changes in the target site (DNA gyrase and/or topoisomerase IV) or changes that alter intracellular drug accumulation (ie, enhanced efflux).
Moxifloxacin has significant activity against many strains of atypical mycobacteria, microorganisms of significance in dermatological infections. For example, the following minimum inhibitory concentrations of 90% of isolates (MIC 90 (Braback et al 2002), 1.0 (53 isolates) (Aubry et al 2000); b) 16.0 (69 isolates) (Yang et al 2003), 1.0 (7 Table 1 In vitro activity of moxifloxacin against potential Gram-positive and Gram-negative aerobic pathogens in SSSIs Organism
In another study with potential application to foreign device-related infections, an in vitro study was performed evaluating the bactericidal activity of moxifloxacin monotherapy and combination moxifloxacin plus vancomycin/teicoplanin therapy in 9 S. aureus isolates (1 being methicillin-and quinolone-sensitive and 8 being resistant to both). These isolates came from patients unresponsive to glycopeptide therapy or relapsing after completion of glycopeptide therapy despite device removal. Moxifloxacin MICs were 0.125 mg/L (in the 1 methicillinsensitive isolate) and 2.0 mg/L (in all of the methicillinresistant isolates). All isolates were tolerant to vancomycin, teicoplanin, and moxifloxacin at medium concentrations 2 X MIC. However, moxifloxacin plus glycopeptide combination therapy was bactericidal at the glycopeptide MIC and 0.5 X MIC of moxifloxacin (Tarasi et al 2003). Thus, combination therapy may be worthwhile considering in difficult-to-treat patients.
Characteristic of quinolones, moxifloxacin has concentration-dependent bactericidal activity (Dalhoff et al 1996). This has been shown with methicillin-sensitive and -resistant S. aureus and coagulase-negative staphylococci (Lister 2001, Speciale et al 2002Noviello et al 2003;Bogdanovich et al 2005), S. pyogenes (Esposito et al 2000; 5-4 (1, 5, 7, 19, 22, 26, 27) Odenholt et al 2002; Speciale et al 2002;Noviello et al 2003), Enterococcus faecalis (Speciale et al 2002), and a variety of anaerobes (Credito et al 2003). Bactericidal activity using serum bactericidal titer technology has been demonstrated for S. aureus, S. epidermidis, S. pyogenes, E. coli, and Klebsiella pneumoniae (Lemmen et al 2003;Dan et al 2004). Also, bactericidal activity using synovial fluid bactericidal titer technology has been demonstrated for S. aureus, S. pyogenes, and K. pneumoniae (Dan et al 2004). Activity against S. pyogenes is independent of the presence/ absence of macrolide resistance determinants (Critchley et al 2002). Two recent studies have evaluated the activity of moxifloxacin in in vitro pharmacokinetic/pharmacodynamic models (IVP/PM) of mixed aerobic-anaerobic infection. Hermsen and colleagues (2005) demonstrated bactericidal activity against Escherichia coli and Bacteroides fragilis in a study using a limited number of isolates of both microorganisms. Using different strains of the same microorganisms, Schaumann and colleagues (2005) demonstrated that the presence of B. fragilis did not affect the activity of moxifloxacin against E. coli but the presence of E. coli reduced the activity against B. fragilis significantly.
Another recent study evaluated the pharmacodynamics of moxifloxacin against anaerobes in an IVP/PM. B. fragilis, Clostridium perfringens, and Gram-positive anaerobic cocci (GPAC) were studied. Dose-ranging using area under the concentration-time curve/minimum inhibitory concentration (AUC/MIC) ratios of 6.7 to 890 (for the reference aerobe E. coli) and 9 to 216 (for the anaerobes) revealed differing AUC/MIC antibacterial effect patterns for anaerobes versus the aerobe. For E. coli, the AUC/MIC ratios for 50%/90% effect were 34/59 while the corresponding values for B. fragilis, C. perfringens, and GPAC were 11/26, 9/16, and 7/17, respectively. At similar AUC/MIC ratio values, the activity against anaerobes was less than that against the aerobe. Using Monte Carlo simulations of human pharmacokinetic data and a target anaerobic AUC/MIC ratio value of 7.5, over 90% achievement of this target was possible at MICs < 2 mg/L. The clinical implications of this "pharmacokinetic breakpoint" of 2 mg/L for anaerobes deserves further study (Noel et al 2005).
Despite 10-to 20-fold accumulation of drug intracellularly (vida infra), the bactericidal effect of moxifloxacin intracellularly develops slowly, with only a 1.51 log reduction in post-phagocytosis S. aureus inoculum within 24 h (Seral et al 2003). This may be explainable, in part, by the deleterious effect of acid pH on activity. For example, at pH 5 (mimicking phagolysosome pH), moxifloxacin activity falls at least 4-fold compared with physiologic pH (Seral et al 2003). However, over a pH range of 6 to 8, moxifloxacin activity is unaffected for S. aureus and S. pyogenes (Dalhoff et al 2005).
The presence of 10%, 30%, and 50% albumin and human serum had no significant effect on the in vitro activity of moxifloxacin (Woodcock et al 1997;Rubinstein et al 2000). The presence of dead bacteria, sterile pus under aerobic conditions, sterile pus under anaerobic conditions, and anaerobic conditions alone had no significant effect on the in vitro activity of moxifloxacin against methicillin-sensitive and -resistant staphylococci, S. pyogenes, and E. coli Rubinstein et al 2000;Wright et al 2002;Noel et al 2005). Anaerobic conditions also do not alter the AUC/MIC ratios for 50% and 90% of effect for E. coli and moxifloxacin (Noel et al 2005). Mixed effects have been noted when testing for an inoculum effect Woodcock et al 1997).
Moxifloxacin has been evaluated in two mouse models relevant to skin and skin structure infections (systemic infection induced by intraperitoneal inoculation and cellulitis induced by subcutaneous (SC) inoculation). In the systemic infection model using 2 strains of methicillin-sensitive S. aureus (both having a moxifloxacin MIC of 0.06 mg/L), the ED 50 and ED 90 values for SC dosed drug were 0.6 mg/kg, 1.0 mg/kg and 1.7 mg/kg, 2.3 mg/kg, respectively, while corresponding ED values for orally dosed drug were 3.2 mg/ kg (for both) and 9.6 mg/kg (for both) (Patel et al 2004). The term ED 50/90 refers to the doses effective in producing survival rates of 50% and 90%, respectively. In the mouse cellulitis model, one methicillin-sensitive strain (MIC 0.03mg/L) and two methicillin-resistant strains (MICs 1 mg/ L and 4 mg/L) of S. aureus were used. For the methicillinsensitive strain, 2.5 mg/kg and 5 mg/kg SC doses of moxifloxacin were administered 1 h, 4 h, 24 h, and 27 h postinoculation while for the methicillin-resistant strains, 25 mg/ kg and 50 mg/kg SC doses were administered 1 h, 3 h, and 5 h post-inoculation. With the methicillin-sensitive strain, moxifloxacin proved to be bacteriostatic (even 5 mg/kg doses led to a <2 log colony forming unit (CFU) reduction/ lesion). With one methicillin-resistant strain, there was a marginal fall in CFU/lesion while with the other, it had no effect (CFU increased) (Patel et al 2004).
Two recent studies have evaluated the activity of moxifloxacin in animal models of osteomyelitis (bone infection). Bone infection can occur as a result of hematogenous seeding or contiguous spread from an area of soft tissue infection. One study compared moxifloxacin and vancomycin in the treatment of bone infections due to methicillin-sensitive S. aureus in rats. Moxifloxacin (10 mg/ kg) and vancomycin (15 mg/kg) were both dosed twice daily by the intraperitoneal route. Moxifloxacin was significantly more effective in reducing bacterial colony counts in femoral bone than were placebo and vancomycin (both p<0.001; placebo vs vancomycin, p=0.53). However, no treatment sterilized the infected bone, probably due to the presence of foreign material (ie, the hollow needle maintained in the cavity) (Kalteis et al 2006a). The second study extended the findings of the first to include bacterial counts in infected capsular soft tissue and biofilm adherent to the implanted needle. Moxifloxacin was significantly superior to vancomycin and placebo in reducing bacterial counts in capsular soft tissue (both p<0.001; placebo vs vancomycin, p=0.09) and biofilm (p=0.009 and p<0.001, respectively; placebo vs vancomycin, p=0.20) (Kalteis et al 2006b). Further investigation of moxifloxacin in osteomyelitis appears warranted.
Another atypical mycobacterial infection with dermatological manifestations is leprosy (due to M. leprae). The classic test for evaluating the activity of antimicrobials against this microorganism is the mouse footpad model wherein the microorganism is injected into the footpad. In this model, where all drugs were administered orally, moxifloxacin proved superior to the quinolone used most frequently to date in combination regimens for leprosy, ofloxacin (percentages killed of 92.1% vs 60.2%, respectively; p<0.05). In addition, the combination of minocycline plus moxifloxacin was superior to that of minocycline plus ofloxacin (percentages killed of 93.7% vs 74.9%, respectively; p<0.05). Moxifloxacin was equipotent with rifampin, one of the key anti-leprosy drugs (percentage killed by both drugs was 92.1%, p=NS). The most potent combination tested was rifapentine (a rifamycin) plus minocycline plus moxifloxacin which produced a 99.9% killing rate (Ji and Grosset 2000). Moxifloxacin is an important agent to evaluate in vivo in human leprosy.
Performing large-scale surveillance of antimicrobial susceptibility can assist in deciding upon the selection of empiric antimicrobial therapy. In one such study, performed between 1994 and 2001, 12 US hospitals collected 4434 Bacteroides isolates for susceptibility testing. The overall moxifloxacin resistance rate rose from 30% to 43% (p<0.001) from 1998 to 2001 (2001 MIC 90 = 16 mg/L). Increased moxifloxacin resistance over time was noted in all species, all infection sites, and in 11 of 12 hospitals. For blood isolates, the resistance rate rose from 38% in 1998 to 52% in 2001. Resistance rates were especially high in the species ovatus, uniformis, and vulgatus. The largest increase in resistance occurred with the species distasonis (22% in 1999 to 37% in 2001). Of all infection sites, decubitus ulcer isolates had the highest resistance rates. On multivariate analysis, the association between the year of microorganism isolation and increased resistance rates remained significant after adjusting for hospital, species, and infection site (adjusted odds ratio=1.33, p<0.001) (Golan et al 2003).
In the LIBRA surveillance study in the US in 1999, 324 laboratories participated in an examination of resistance rates in S. pyogenes isolates. Isolates came from the respiratory tract (n=2039), skin and skin structure (n=405) and blood (n=148). The modal moxifloxacin MIC was 0.12 mg/L and the MIC 90 was 0.25 mg/L. One hundred percent of isolates were susceptible (based on a resistance breakpoint of ≥4 mg/ L) (Critchley et al 2002).

Pharmacokinetics
Moxifloxacin is well-absorbed after oral administration, with absolute oral bioavailabilities of 92% (mean) (range 77%-120%) and 86.2 ± 1.11% (mean ± standard deviation [SD]) being noted in 2 healthy-volunteer studies (Ballow et al 1999;Stass and Kubitza 1999). Oral administration in the fed state (high fat breakfast) had nonsignificant effects on moxifloxacin peak plasma concentration (C max ) and area under the plasma concentration-versus-time curve (AUC) but did prolong the median time to C max (T max ) from 1 to 2.5 hours (not clinically significant) . Milk (240-300 mL) and yogurt (250-300 mL) coadministration did not significantly alter moxifloxacin bioavailability (Guay 2005) nor did enteral feeding coadministration (Burkhardt et al 2005). Minor accumulation occurs during multiple dosing and its pharmacokinetics are linear and doseproportional (Stass et al 1998(Stass et al , 2001Sullivan et al 1999).
In yet another healthy volunteer study, the penetration of moxifloxacin into the interstitial fluid compartments of muscle and SC adipose fat tissue and inflammatory blister fluid were evaluated after single 400 mg oral and intravenous doses. Mean ± SD penetration into muscle interstitial fluid was 55% ± 12% (using AUC muscle /AUC total plasma data) and 86% ± 17% (using AUC muscle /AUC free plasma data data). Corresponding SC adipose fat interstitial fluid penetration was 38% ± 9% and 81% ± 19%. These results suggest a virtually total equilibration between these two body compartments and free (unbound) drug in plasma. The concentration over time profiles in these two compartments were virtually superimposable. Mean ± SD penetration into inflammatory blister fluid (BF) was 64% ± 21% (using AUC BF /AUC total plasma data) (Muller et al 1999).
The penetration of moxifloxacin into healthy and inflamed SC adipose fat tissues was assessed in 11 patients with skin and skin structure infections (n=5 had diabetes and peripheral occlusive arterial disease (POAD), n=6 did not). A single 400 mg IV dose of moxifloxacin was administered over one hour. In the aggregate analysis (n=11), drug concentrations in healthy and inflamed tissues were consistently below total plasma concentrations and approximately equal to free plasma concentrations. Moxifloxacin total plasma concentration over time profiles were similar in patients with and without diabetes/POAD. Penetration into tissue did not statistically differ in the two groups. In patients with diabetes/POAD, the concentrations in healthy tissue exceeded those in inflamed tissue. Only the healthy tissue concentrations were similar to the free concentrations in plasma (the inflamed tissue concentrations were about 1/2 of the free concentrations in plasma). In the patients without diabetes, exactly the opposite findings were noted. Thus, the mean ± SD AUC 0-8h for inflamed tissue/ AUC 0-∞ for free drug in plasma was 0.5 ± 0.4 in patients with diabetes/POAD and 1.2 ± 0.8 in the other group. Mean ± SD inflamed tissue C max values were 0.8 ± 0.5 mg/L in patients with diabetes/POAD and 2.3 ± 1.2 mg/L in the other group. The reduced concentrations in inflamed tissues in patients with versus without diabetes/POAD was perhaps due to the vascular compromise associated with POAD. None of these differences between patients with/without diabetes/POAD were significant due to small sample sizes and large interpatient variability. In all cases, tissue and plasma drug concentrations exceeded the MICs of most dermatologic pathogens, suggesting a role for moxifloxacin in the treatment of skin and skin structure infections (Joukhadar et al 2003).
Thirty patients undergoing total knee arthroplasty participated in a study investigating the penetration of moxifloxacin into bone. Ten patients received oral moxifloxacin 2 h (range 1.5-2.5 h) preoperatively and mean plasma, cancellous bone, and cortical bone drug concentrations were 3.45 mg/L, 1.89 mg/kg, and 1.43 mg/ kg, respectively. In the ten patients receiving moxifloxacin 4 h (range 3.5-4.5 h) preoperatively, corresponding mean drug concentrations were 3.73 mg/L, 1.81 mg/kg, and 1.56 mg/kg. In the ten patients receiving the drug 14 h and 2 h (range 1.5-2.5 h) preoperatively, corresponding drug concentrations were 6.26 mg/L, 2.97 mg/kg, and 2.54 mg/ kg (Malincarne et al 2006).
As with other quinolones, moxifloxacin penetrates well into polymorphonuclear leukocytes (mean ± SD internal concentration = 54 ± 0.5 mg/L with an external medium concentration of 4.5 mg/L) (Mandell and Coleman 2001). Moxifloxacin also penetrates well into macrophages (with an external concentration of 4 mg/L, the mean internalexternal concentration ratio = 13.4 (in cells infected with S. aureus) and 11.4 (in non-infected cells), p=0.02) (Seral et al 2003).
When pharmacokinetic parameters are weightnormalized, there are no significant age-or gender-based effects on moxifloxacin pharmacokinetics (Sullivan et al 2001;Pea et al 2006). Thirty-two individuals participated in a single 400 mg oral dose study of the effects of renal impairment on the pharmacokinetics of moxifloxacin and its M1 and M2 metabolites. With moxifloxacin, there was no significant relationship of AUC with creatinine clearance (CrCl) (p=0.59). However, C max did fall as CrCl fell (p=0.03). Significant correlations existed between moxifloxcin C max (r=0.438, p=0.012), fractional renal excretion (F e ) (r=0.613, p=0.0002) and CL R (r=0.6405, p=0.0001) and CrCl. The M1 metabolite was little affected by changes in renal function although significant correlations existed between M1 metabolite F e (r=0.620, p=0.0002) and CL R (r=0.703, p<0.0001) and CrCl. For the M2 metabolite, AUC rose as CrCl fell (p=0.0167) while C max did not change (p=0.3042). Significant correlations existed between M2 metabolite AUC (r=-0.565, p=0.0011), CL R (r=0.736, p<0.0001) and F e (r=0.514, p=0.0026) and CrCl .
Single and multiple dose pharmacokinetics of moxifloxacin have also been evaluated in end-stage renal disease patients undergoing hemodialysis (HD) and continuous ambulatory peritoneal dialysis (CAPD). Following a single oral 400 mg dose, the AUC's of parent compound in these patients were statistically similar to those found generally in healthy volunteers. However, C max values were reduced by about 45% (HD) and 33% (CAPD) compared with historical healthy control subject data. The systemic exposure to the M1 metabolite rose by 1.4-to 1.5fold (based on AUC) while that to the M2 metabolite rose 7.5-fold (based on AUC) and 2.5-to 3-fold (based on C max ) compared with historical healthy control subject data. The possible consequences of increased metabolite exposure are not known. Multiple once-daily oral dosing of 400 mg for 7 days in these patients produced systemic exposure (AUC) similar to that seen in healthy volunteers. Steady-state C max values in HD patients were approximately 22% lower than those in healthy volunteers while they were comparable for CAPD patients. Dialysis removal of moxifloxacin was low (9% by HD, 3% by CAPD) (Anonymous 2005). Continuous renal replacement therapy (continuous venovenous hemodiafiltration) does not affect

Moxifloxacin in skin infections
moxifloxacin disposition to a clinically-meaningful extent (Fuhrmann et al 2004).

Clinical efficacy
This section will deal only with the clinical efficacy of moxifloxacin in skin and skin structure infections (SSSIs). All have been active-controlled in design, with cephalexin as the control in the three uncomplicated SSSI studies, unspecified β-lactam/β-lactamase-inhibitor combinations in the 2 unpublished complicated SSSI studies, and piperacillin-tazobactam followed by amoxicillin-clavulanate as the control in the one published complicated SSSI study. As may be seen, there were no significant or substantial inter-group differences in clinical or bacteriological efficacy or tolerability. These results should be interpreted while keeping in mind potential study limitations. In general, these studies did not provide evidence of power analyses conducted to establish proper group sizes. Thus, they were underpowered to avoid type II statistical errors (ie, false negative intergroup differences), especially with respect to subgroup analyses based on comorbidities or individual pathogens. Three of five studies remain unpublished, available only as abstracts or in the product information. The limitations of this should be obvious to the reader. The extensive number and breadth of exclusion criteria generated study populations much different from the "real world" patients who would require treatment. Hence, the generalizability of study results can be questioned. Clinical trials of SSSIs have long been difficult to design in a way that renders them rigorous and reproducible. Although the publication of SSSI study design criteria by the Infectious Diseases Society of America (IDSA) have helped in this regard (Calandra et al 1992), deficient study designs are, unfortunately, still far too frequently seen. The ability of investigators to add antianaerobic coverage (eg, metronidazole) for presumed mixed SSSIs makes it impossible to judge whether or not the anaerobic coverage of moxifloxacin is clinically-relevant. Lastly, the definitions of clinical endpoints can vary between SSSI studies, making it difficult to obtain a global perspective of drug efficacy.
A recent case report detailed a case of a patient with an infected cat-bite wound failing therapy with IV cefuroxime, ciprofloxacin, and metronidazole which responded to the addition of IV moxifloxacin (Draenert et al 2005). Another case report documented successful therapy of a M. thermoresistibile infection following knee replacement surgery with long-term therapy using moxifloxacin and doxycycline (Labombardi et al 2005).
A recent review of potential antitubercular drugs included encouraging in vitro, in vivo, (animal) and pilot human clinical data regarding moxifloxacin use in infections due to M. tuberculosis (Duncan and Barry 2004). These data, in conjunction with in vitro susceptibility data reviewed earlier in this paper, suggest that moxifloxacin may play an important role in the treatment of SSSIs caused by atypical mycobacteria. Clinical trials in this area are warranted.

Safety
As with other quinolones, moxifloxacin is generally welltolerated, with the majority of adverse events referable to the gastrointestinal tract (nausea, diarrhea, vomiting, constipation, abdominal pain, vaginitis, taste alteration) or the central nervous system (headache, dizziness) (Anonymous 2005). There were only three adverse events, judged by investigators to be at least possibly drug-related, occurring in at least 2% of over 8600 moxifloxacin-treated patients in the drug development process: nausea (6%), diarrhea (5%), and dizziness (2%) (Anonymous 2005). Premature study discontinuation rates for drug-related adverse events were 2.9% (oral) and 4.6% (IV → oral) (Anonymous 2005). Moxifloxacin is not associated with the phototoxicity characteristic of lomefloxacin and sparfloxacin (Man et al 1999) and has only been associated with the "temafloxacin hypersensitivity syndrome" in 1 patient (Nori et al 2004). This patient exhibited toxic epidermal necrolysis and fulminant hepatic failure leading to death. Moxifloxacin is not associated with the hypo or hyperglycemia seen with gatifloxacin therapy (Park-Wyllie et al 2006).
In 2004, a cumulative safety review of oral and intravenous moxifloxacin from the drug development program and postmarketing surveillance studies was published. For the oral agent, the database comprised 30 phase II/III active comparator studies (moxifloxacin n=7368, comparators n=5687), one phase IV study (n=18374), and 4 postmarketing observation studies (n=27756). For the intravenous agent, the database comprised two phase III active comparator studies (moxifloxacin n=550, comparators n=579). With both formulations, the commonest adverse events involved the gastrointestinal tract (for the oral agent, nausea in 7.1% and  Pip-Tazo 3.375 g IV infection type in the 2 groups except abscess qd → Amox-Clav (79% Moxi vs. 93% Pip-Tazo → Amox-Clav, PO 800 mg bid p=0.04). Univariate followed by multivariate (total of 7-14 days) regression analysis identified the no. of surgeries (ie, ≥2) as being an independent risk factor for minimum duration failure of abscess cure. After adjusting for risk of IV therapy was factors for failure of abscess cure, the difference 3 days.
between groups disappeared (odds ratio, 1.05; p=0.12). There were no significant differences between the groups in bacteriologic eradication rates by organism or for monomicrobial or polymicrobial infections. AE rates were comparable in the 2 groups (drug-related AEs in 31% and 30% of Moxi and Pip-Tazo → Amox-Clav patients, respectively; corresponding rates of diarrhea were 5% and 8%, and nausea were 4% and 2%; premature D/C for drug-related AEs in 14 and 17 patients). Drug-related serious AEs with Moxi included weakness, worsening of drug reaction rash, cellulitis exacerbation, PMC, and clinical failure and with Pip-Tazo → Amox-Clav included cardiopulmonary arrest, worsening CHF, allergic reaction, asthenia, worsened skin eruption, allergic reaction, persistent right leg abscess, bloody diarrhea, osteomyelitis, and clinical failure.
In the same year, a similar retrospective review of the phase II/III clinical trial database (32 trials, n=14731 patients (moxifloxacin n=8474, active comparators n=6257)) and postmarketing surveillance studies (n=46130) were reviewed with a focus on adverse events involving glucose metabolism. In the clinical trials, there were 0 and 3 drugrelated hypoglycemic adverse events with moxifloxacin and active comparator patients, respectively (latter were 2 cases with levofloxacin and 1 case with trovafloxacin). Drugrelated hypoglycemic events occurred in 7 moxifloxacin and 1 active comparator patients. No evidence was found supporting the presence of drug-drug interactions with oral antidiabetic drugs. In the postmarketing studies, no cases of drug-related hypoglycemia were noted. This same paper described results of in vivo (rat) studies utilizing single 30 mg/kg and 100 mg/kg oral doses of moxifloxacin, gatifloxacin, and levofloxacin (positive control was glibenclamide 10 mg/kg). Blood was sampled predosing and at 0.5 h and 1.5 h postdosing. In fasted rates, glibenclamide had the expected effect (blood glucose fell a mean of 24% at 1.5 h postdose which was preceded by a rise in serum insulin). Blood glucose was not significantly affected by either moxifloxacin or levofloxacin but fell means of 8% and 18% with 30 mg/kg and 100 mg/kg doses of gatifloxacin, respectively. No significant effects on serum insulin were noted with any quinolone. In fed rats, glibenclamide increased serum insulin and reduced blood glucose approximately 40% at 1.5 h postdosing. Again, moxifloxacin and levofloxacin had no significant effect on blood glucose or serum insulin. Gatifloxacin 30 mg/kg and 100 mg/kg reduced blood glucose by means of 17% and 26%, respectively. This was preceded by a reduction in serum insulin (only significant at 0.5 h with 100 mg/kg dose). Thus, the gatifloxacin effect was felt to be insulin-independent (Gavin et al 2004).
Results of the last of the large retrospective database reviews for safety purposes were published in 2005 and focused on the safety of oral moxifloxacin in the clinical trial database as a function of aging. The database involved 27 phase II/III clinical trials analyzed by age grouping (<65, 65-74, ≥75 years old) and treatment (moxifloxacin and active comparators). Valid data were available from 12231 subjects (moxifloxacin n=6270, active comparators n=5961; comparators were primarily clarithromycin and cefuroxime axetil). The subject distribution by age was as follows: n=9671 were <65, n=1631 were 65-74, and n=924 were ≥75 years old. The distribution of drug-related adverse events comparing moxifloxacin with active comparators was not affected by increasing age (p=0.43) nor were premature discontinuations due to adverse events (p=0.552). The number of deaths were also similar (moxifloxacin n=17, active comparator n=19). No arrhythmias due to QT c interval prolongation were noted in the entire population. The differences in adverse event profiles across the 3 age groups were not significant for moxifloxacin (p=0.599) with 1 exception. For drug-related adverse events with moxifloxacin, the profiles did differ across the 3 age groups (p=0.001) but, unexpectedly most rates increased as age fell (eg, nausea, vomiting, dyspepsia, and vaginal moniliasis rates rose as age fell but insomnia rates rose as age rose). There were no clinically-significant differences between the treatments in the prevalence of serious drug-related events (≤1%) and these were not related to age. In 787 patients who had pre and 3 day-post 12-lead electrocardiograms (n=651 were <65, n=136 were ≥65 years old), the mean ± SD QT C interval prolongation was 7 ± 25msec in the <65 group and 2 ± 28 msec in the ≥65 group (p=0.055, trend) (Andriole et al 2005).

Pharmacokinetic drug-drug interactions
Based on the significant role of metabolism in the elimination of moxifloxacin, a study was conducted to evaluate the effect of concurrent use of a potent cytochrome P450 isozyme 3A4 inhibitor (itraconazole) on moxifloxacin pharmacokinetics. Nine days of itraconazole 200 mg orally daily had no significant effect on the single-dose kinetics of oral moxifloxacin 400 mg (administered on day 7 of the 9-day itraconazole regimen). Moxifloxacin also had no significant effect on itraconazole kinetics (Stass et al 2004). Moxifloxacin 400 mg orally daily had no significant effect on theophylline total body clearance and terminal disposition half-life in 2 studies (Guay 2005). Moxifloxacin 400 mg orally once daily had no significant effect on the pharmacokinetics or pharmacodynamics of warfarin in healthy volunteers (Guay 2005). However, case reports have documented moxifloxacin-associated increases in INR (international normalized ratio of prothrombin time) in warfarin recipients (Elbe and Chang 2005;Guay 2005). Pending additional information, patients receiving long-term warfarin therapy in whom moxifloxacin is to be used should be monitored for changes in INR. Moxifloxacin does not significantly affect low-dose oral contraceptive or digoxin pharmacokinetics (Guay 2005). Unlike the case with ciprofloxacin, norfloxacin, levofloxacin, and gatifloxacin, probenecid does not significantly alter the systemic pharmacokinetics of moxifloxacin (Guay 2005).
Like all quinolones, moxifloxacin is susceptible to interactions with multivalent cations, wherein the reductions in bioavailability may be clinically-relevant and lead to a loss of therapeutic benefit. For example, calcium carbonate or calcium lactate-gluconate given immediately before and 12 h and 24 h after single-dose moxifloxacin resulted in a mean 15% reduction in C max (p≤0.05) and a mean 2% reduction in AUC (p=NS) (Guay 2005). Simultaneous administration of single-dose ferrous sulfate lead to mean reductions of 59 and 39% in single-dose moxifloxacin C max and AUC, respectively (both, p≤0.05) (Guay 2005). Singledose magnesium-aluminum antacid administered simultaneously with single-dose moxifloxacin resulted in mean 61% and 59% reductions in C max and AUC, respectively (both, p≤0.05). When the antacid administration was delayed until 2 h after moxifloxacin administration, the mean 7% reduction in C max was non-significant but the mean 26% reduction in AUC was significant (p≤0.05). Even administering the antacid 4 h before moxifloxacin administration still resulted in a significant reduction in quinolone AUC (mean 23%, p≤0.05) although the mean 1% fall in C max was not (Guay 2005). Sucralfate administered simultaneously with and then 5 h, 10 h, 15 h, and 24 h following single-dose moxifloxacin administration resulted in mean 71% and 60% reductions in quinolone C max and AUC, respectively (both, p≤0.05) (Guay 2005). Ranitidine, a histamine-2 receptor antagonist, does not interact with moxifloxacin (2 studies) and can be recommended as a noninteracting antiulcer therapy alternative to sucralfate and antacids (Guay 2005).

Pharmacodynamic drug-drug interactions
The only potentially clinically-relevant pharmacodynamic drug-drug interaction with moxifloxacin involves additive electrophysiological effects with other drugs that can also prolong the QT c interval on the electrocardiogram (Table  5) (Guay 2005). Drugs which prolong cardiac repolarization may, on rare occasions, be associated with the development of polymorphous ventricular tachycardia ("torsades de pointes") which, in turn, may degenerate into ventricular fibrillation. Quinolones cause a drug-specific, dosedependent prolongation of QT c interval by inhibiting outward potassium currents in myocytes. In numerous in vitro models examining the mechanism(s) underlying this arrhythmogenic effect, sparfloxacin has been more potent than moxifloxacin (= grepafloxacin) which, in turn, has been more potent than the other tested quinolones (Guay 2005). In in vivo (animal) models, sparfloxacin again was most potent followed by the trio of moxifloxacin-gatifloxacingrepafloxacin (latter agent has been withdrawn from the market) (Guay 2005). Concentration-response relationships have been demonstrated in both in vitro and in vivo preclinical models (Chen et al 2005).
In one healthy volunteer study, single 400 mg and 800 mg oral doses of moxifloxacin caused mean ± SD 4.0% ± 5.1% and 4.5% ± 3.8% QT c interval prolongation at rest compared with baseline, respectively (both, p≤0.05). Significant QT c interval prolongation occurred at all heart rates and across the entire RR interval range (400-1000 msec). There were no gender-dependent differences in results. The correlation of moxifloxacin plasma concentrations and QT interval was significant but weak (r=0.35) (Guay 2005).
Only two comparative studies of the effect of quinolones on QT c intervals have been published. In the first, single oral doses of levofloxacin (1000 mg), ciprofloxacin (1500 mg) and moxifloxacin (800 mg) were evaluated in a placebo-controlled, crossover trial in healthy volunteers. Mean QT and QT c interval prolongation was significantly greater for moxifloxacin compared with placebo for all end points, but it was generally not for the other two quinolones. The proportion of subjects with QT c interval prolongation of 30 msec. or higher was greater with moxifloxacin (72%-81%) compared with levofloxacin (33%-38%) and ciprofloxacin (34%-40%) (Guay 2005). In the second, 13 healthy volunteers received three 7-day regimens in random order: ciprofloxacin 500 mg twice daily, levofloxacin 500 mg once daily, and moxifloxacin 400 mg once daily. Washout periods between treatments were one week in length. Twelve-lead electrocardiograms were obtained before, 2 h after the first dose, and at the end of each regimen. At the end of 7 days, moxifloxacin prolonged QTc by a mean of 6 msec compared with baseline (p=0.022) and a mean of 11 msec from the 2 h baseline value (p=0.003). Ciprofloxacin and levofloxacin had no significant effect on QTc. No quinolone had a significant effect on QT dispersion, a parameter felt to be a more selective marker for risk of polymorphous ventricular tachycardia (Tsikouris et al 2006). Further studies of QT dispersion, especially in patients receiving quinolone therapy, with/without concurrent use of other drugs which also prolong QTc, are warranted.
The lack of multiple dose studies and well-done epidemiologic studies do not allow an assessment of the clinical relevance of the above-cited findings. In a recent randomized trial comparing the cardiac safety of IV/oral moxifloxacin and levofloxacin in elderly patients hospitalized for community-acquired pneumonia, the two agents had similar cardiac rhythm safety profiles (Morganroth et al 2005). However, it is prudent to use caution when using moxifloxacin (and other quinolones) in patients already receiving potentially arrhythmogenic drugs (Table 5) (Guay 2005). Similarly, caution is warranted when use is contemplated in patients with abnormal pretreatment QT c intervals, electrolyte abnormalities (especially hypokalemia, hypomagnesemia, rarely hypocalcemia), starvation/liquid protein fast diets, and prior/current history of coronary artery disease or arrhythmias.

Dosing
Moxifloxacin is available as 400 mg oral tablets and premixed moxifloxacin 400 mg in 250 mL 0.8% sodium chloride IV bags (Anonymous 2005). For SSSIs, the recommended daily dose is 400 mg and the recommended durations of therapy are 7 days (uncomplicated infections) and 7-21 days (complicated infections). Oral doses of moxifloxacin should be administered at least 4 h before or 8 h after magnesium/aluminum antacids, sucralfate, metal cations (eg, iron), multivitamin + mineral supplements (especially with zinc), and didenosine chewable/buffered tablets or pediatric powder for solution (Anonymous 2005). No dosage adjustment is needed with any level of renal impairment or mild or moderate (Child-Pugh Class A or B) hepatic impairment (Anonymous 2005). In severe (Child-Pugh Class C) hepatic impairment, there are no data to guide the clinician (Anonymous 2005).

Conclusions
Optimal therapy for SSSIs requires careful patient evaluation, including obtaining a thorough medical and social history, performing a meticulous examination of the lesion(s), and examination of Gram-stained smears of lesion exudates. Early empirical antimicrobial therapy is the next step after obtaining culture specimens, with subsequent tailoring of therapy based on culture and sensitivity results. Local measures, including debridement, are important adjuncts to antimicrobial therapy. Selection of empiric therapy for SSSIs depends on a number of factors, including type of infection (uncomplicated vs complicated), location of acquisition (community versus hospital), and local sensitivity patterns for the commonest dermatologic pathogens.
Older quinolones have been demonstrated to be efficacious in the treatment of uncomplicated and complicated SSSIs over the past 20 years (Blondeau 2002;Sable and Murakawa 2003). In addition, more limited data support the efficacy of the newer quinolones in the treatment of uncomplicated SSSIs (gatifloxacin, levofloxacin 500 mg/ d, and moxifloxacin) (Leal del Leal de Rosal, Martinez, et al 1999;Parish et al 2000;Raghavan and Linden 2004). To these data may be added those documenting the efficacy of high-dose levofloxacin (750 mg/d) and moxifloxacin in the treatment of complicated SSSIs (Raghavan and Linden 2004;Anonymous 2005).
What have other reviews recently said regarding the role of quinolones in the treatment of SSSIs? In a recent review of the diagnosis and treatment of facial bite wounds, moxifloxacin was felt to be a reasonable monotherapy for infected human facial bite wounds, especially in patients who are penicillin-allergic. The suggested duration of a prophylactic regimen ranged from 3-5 days (10-14 days if bone was involved). For treatment, if bones or joints were involved, a duration of 21+ days was recommended (Stefanopoulos and Tarantzopoulou 2005).
In a review of the role of quinolones in the treatment of SSSIs, Blondeau felt that both older and newer quinolones had roles to play. The older ciprofloxacin-type quinolones were recommended only in SSSIs where Gram-negative aerobic bacilli were known pathogens, adding clindamycin or metronidazole in suspected mixed infections. The newer quinolones (including moxifloxacin) were recommended in SSSIs where non-pseudomonal Gram-negative aerobic bacilli were known pathogens, where there was a mixed infection with anaerobes within the quinolone's spectrum of activity, where non-methicillin-resistant S. aureus grampositive aerobes were involved, and where Gram-positive and Gram-negative aerobic pathogens were coexisting (Blondeau 2002).
The first organization-based guidelines for the diagnosis and management of SSSIs were published by IDSA in November 2005 (Stevens et al 2005). Quinolones are mentioned in the context of several general and organismspecific infections. Quinolones are one of many empirical choices in human and animal bite wounds, especially in patients with severe penicillin allergies. Ciprofloxacin, gatifloxacin, and moxifloxacin with/without clindamycin or metronidazole are those specifically mentioned. Again, quinolones are one of many empirical choices in SSSIs in the immunocompromised host. A quinolone (probably ciprofloxacin) plus an extended-spectrum β-lactam is the specific quinolone mentioned. Quinolones are one of several choices for the empiric management of each of three defined types of surgical wound infections. In community-acquired mixed necrotizing fasciitis, one option is a combination of ciprofloxacin plus a β-lactam-β-lactamase inhibitor. In SSSIs due to community-acquired MRSA, the newer quinolone group (including moxifloxacin) is one option (author's comment: based on 20 years of experience with quinolones in infections due to MRSA, the likelihood of the development of resistance to quinolones during such use is quite high. caution is warranted). Lastly, in infected wounds after intestinal or genital tract surgeries, a combination of a quinolone plus one of clindamycin/ metronidazole/chloramphenicol/β-lactam-β-lactamase inhibitor is one option. Quinolones are mentioned in the context of the patient with severe penicllin allergies. In cutaneous anthrax, ciprofloxacin is the drug-of-choice (newer quinolones are also considered likely effective). In erysipeloid, quinolones are options only in patients with severe penicillin allergies. Lastly, among many options, levofloxacin and ciprofloxacin are considered options for tularemia but only in mild to moderate illness.
Evolving resistance trends, especially amongst staphylococci and enterococci, and sporadic intolerance to conventional β-lactam therapy have led to the need for novel options with proven efficacy and tolerability. Moxifloxacin is comparable in both efficacy and tolerability with conventional agents for SSSIs and is also effective against many organisms resistant to these agents (exception: methicillin-resistant staphylococci). It also possesses activity against atypical mycobacteria associated with SSSIs (moxifloxacin may develop into a first-line therapy for these types of SSSIs but much work remains to be done before this can be recommended). Despite its proven activity against dermatologic pathogens, one could argue that its use be carefully considered and, in general, restricted to serious, life-threatening infections caused by resistant pathogens or those patients with serious intolerances to conventional therapy. In this way, the development and progression of resistance might be blunted and its "lifespan" as an antimicrobial for dermatologic infections prolonged.