Microbiological Analysis of a Prospective, Randomized, Double-Blind Trial Comparing Moxifloxacin and Clindamycin in the Treatment of Odontogenic Infiltrates and Abscesses

Abstract

The objective of this study was to identify the oral pathogens found in odontogenic infections, to determine their susceptibilities to amoxicillin-clavulanic acid (AMC), clindamycin (CLI), doxycycline (DOX), levofloxacin (LVX), moxifloxacin (MXF), and penicillin (PEN), and to search for associations between specific pathogens and types of infection. Swabs from patients enrolled in a randomized, double-blind phase II trial comparing MXF with CLI for the treatment of odontogenic abscesses or inflammatory infiltrates were cultured on media for aerobes and anaerobes. All bacterial isolates were identified at the species level. Overall, 205 isolates were cultured from 71 patients: 77 viridans group streptococci, 56 Prevotella spp., 19 Neisseria spp., 17 Streptococcus anginosus group isolates and hemolytic streptococci, 15 other anaerobes, and 21 other bacteria. Ninety-eight percent of pathogens were susceptible to MXF, 96% to AMC, 85% to LVX, 67% to PEN, 60% to CLI, and 50% to DOX. S. anginosus group and hemolytic streptococci were found significantly more frequently (P = 0.04) in patients with abscesses (12/95) than in patients with infiltrates (5/110). In four patients with infiltrates who failed to respond to CLI therapy, three isolates of the Streptococcus mitis group and four Neisseria spp. resistant to CLI were found. In this study, S. anginosus group and hemolytic streptococci were clearly associated with odontogenic abscesses. Our analysis suggests that viridans group streptococci and Neisseria spp. play a decisive role in the etiology of odontogenic infiltrates. The high in vitro activity of MXF against odontogenic bacteria corresponds well to its clinical results in the treatment of odontogenic abscesses and infiltrates.

INTRODUCTION

Odontogenic infections of bacterial or inflammatory origin penetrate primarily into the soft and bony oromaxillofacial tissues to produce submucosal infiltrates and abscesses. Often taking a mild course, these infections may also produce life-threatening complications, depending on a patient's immunocompetence and the site of the inflammatory process. These infections are typically polymicrobial, and anaerobic bacteria are thought to play a central etiologic role (1, 15, 18, 21). Antibiotics are an important component in the treatment of odontogenic infections. The drugs most frequently recommended (2, 14) are penicillin (PEN), amoxicillin-clavulanic acid (AMC), and clindamycin (CLI)—despite reports of substantial resistance to CLI among oral pathogens (16, 20).

In an earlier study of the bacterial distribution and in vitro susceptibilities of isolates from 37 patients with odontogenic abscesses, we found viridans group streptococci and Prevotella species to be the most prevalent pathogens. The overall rates of susceptibility to AMC, CLI, doxycycline (DOX), levofloxacin (LVX), moxifloxacin (MXF), and PEN were 100%, 75%, 76%, 98%, 98%, and 69%, respectively (20). These in vitro data justified clinical trials to assess whether MXF may be a rational choice for the treatment of odontogenic infections. In addition, by use of an animal model with Wistar rats, good penetration of MXF into tissues has been demonstrated, with a calculated musculus masseter AUC (area under the curve)/plasma AUC ratio of 2.64 and a calculated mandible AUC/plasma AUC ratio of 1.13 (4).

In view of the promising in vitro and pharmacokinetic data for MXF, and considering that CLI is widely used in Germany for empirical treatment of odontogenic infections, a prospective, randomized double-blind multicenter phase II trial was performed comparing the efficacy and tolerability of MXF with those of CLI in the treatment of inflammatory infiltrates and odontogenic abscesses. The detailed clinical results of this study are described elsewhere (3).

The microbiological part of this study aimed to identify the oral pathogens found in odontogenic abscesses and gingival infiltrates and to determine their susceptibilities to a number of relevant antibiotics but also to check for possible associations between specific pathogens and types of infection and to examine whether the in vitro susceptibilities of the pathogens correspond to the clinical results observed in this patient population.

(The results of this study were presented in part at the 21st European Congress of Clinical Microbiology and Infectious Diseases [ECCMID], Milan, Italy [19]).

MATERIALS AND METHODS

Outpatients with either inflammatory infiltrates or odontogenic abscesses at the Department of Restorative and Preventive Dentistry, University Hospital Hamburg-Eppendorf, Hamburg, Germany, and the Department of Maxillofacial Surgery, Emergency Hospital Berlin, Berlin, Germany, were enrolled in this study. Infiltrate and abscess diagnoses and treatment modalities were based on the International Statistical Classification of Diseases and Related Health Problems, 10th revision, using codes associated with odontogenic infections (K04 to K05) (21a). All patients received oral treatment with either 400 mg MXF once daily or 300 mg CLI four times daily for 5 days in a randomized double-blind, double-dummy manner, either as the main component of medical therapy for inflammatory infiltrates or as an adjuvant therapy to the surgical treatment of odontogenic abscesses. The primary efficacy endpoint was the percentage of reduction in patients' perceived pain from day 1 to days 2 and 3 on a visual analogue scale (VAS). The primary efficacy analysis included 21 MXF-treated patients and 19 CLI-treated patients with inflammatory infiltrates and 15 MXF and 16 CLI patients with odontogenic abscesses (intention-to-treat population). The microbiological analysis was a secondary endpoint of the study.

Swabs of tissues with odontogenic infections were collected from all patients at the initial visit. For a few patients, a second swab was taken at a later time. None of the patients had antimicrobial treatment prior to the initial collection of specimens. Swabs were obtained by surgical incision following disinfection of the mucosa with a povidone-iodine solution, placed in Amies charcoal medium (BBL CultureSwab; Becton Dickinson GmbH, Heidelberg, Germany), and cultured within 24 h of collection.

Columbia agar plus 5% sheep blood and Schaedler agar were used for the cultivation of aerobes/facultative anaerobes and anaerobes, respectively. All bacterial isolates were identified at the species level using the Vitek 2 system (bioMérieux) according to the manufacturer's instructions (9). Streptococci were categorized according to the work of Facklam (8). The MICs of PEN, AMC, CLI, DOX, LVX, and MXF were determined by Etest (AB Biodisk, Solna, Sweden) according to the manufacturer's instructions. Acceptable quality control limits of MICs for Staphylococcus aureus ATCC 29213, Streptococcus pneumoniae ATCC 49619, and Escherichia coli ATCC 25922 were achieved according to current EUCAST (European Committee on Antimicrobial Susceptibility Testing) recommendations (7). MIC values were interpreted according to recent EUCAST-recommended breakpoints (6).

RESULTS

Eighty swabs of tissues with odontogenic infections were obtained from 71 patients. Male/female ratios for patients with infiltrates treated with MXF or CLI were 12/9 or 13/6, respectively, while those for patients with abscesses treated with MXF or CLI were 9/6 or 11/5, respectively. Seventy-one swabs were taken at the initial visit, and nine swabs were taken 1, 2, or 3 days later.

In total, 205 bacteria (134 aerobes/facultative anaerobes and 71 anaerobes) were cultured from the 80 swabs. In 89% of cases, the odontogenic infections had polymicrobial etiology. There was an average of 2.6 isolates per specimen. The most prevalent bacteria (Table 1) were Prevotella species (n = 56), the Streptococcus mitis group (n = 53), other viridans group streptococci (n = 24), and Neisseria spp. (n = 19).

Table 1.

Distribution of the most common bacteria/bacterial groups in abscesses and infiltrates

Bacterium or bacterial group	No. of isolates from:		Pa
	Abscess patients (n = 95)	Infiltrate patients (n = 110)
Streptococcus mitis group	23	30	0.64 (NS)
Other viridans streptococci	8	16	0.20 (NS)
Streptococcus anginosus group and hemolytic streptococci	12	5	0.04
Neisseria spp.	6	13	0.23 (NS)
Prevotella oralis	6	9	0.79 (NS)
Prevotella intermedia	6	7	1.00 (NS)
Other Prevotella spp.	16	12	0.23 (NS)

^aBy Fisher's exact test. NS, not significant.

In vitro susceptibility testing demonstrated that 98% of all odontogenic bacteria (Table 2) were susceptible to MXF. Overall susceptibility rates for AMC, LVX, PEN, CLI, and DOX were 96%, 85%, 67%, 60%, and 50%, respectively. DOX showed its best activity against “other anaerobes” (80% susceptible) and the least activity against the S. mitis group, with only 25% of isolates susceptible. CLI showed its best activity against “other anaerobes” (87% susceptible) but lacked activity against various species of Neisseria. For PEN, susceptibility rates ranging from 100% for Streptococcus anginosus group/hemolytic streptococci to 0% for various species of Neisseria were observed. Rates of susceptibility to LVX ranged from 100% for Neisseria spp. and Prevotella intermedia down to 73% for Prevotella oralis and other anaerobes. Anaerobes and Neisseria spp. had 100% susceptibility to AMC, whereas 83% of isolates of other viridans group streptococci were susceptible. For MXF, susceptibility rates of 100% were found for the most prevalent bacterial groups apart from Neisseria spp., “other Prevotella spp.,” and “other anaerobes,” with 90%, 96%, and 87% of isolates susceptible, respectively.

Table 2.

In vitro susceptibilities of 205 bacterial isolates obtained from 71 patients with odontogenic abscesses or infiltrates

Bacterium or bacterial group and antibiotic	MIC (μg/ml)a			% Susceptibleb
	Range	50%	90%
Streptococcus mitis group (n = 53)c
PEN	0.008–0.5	0.064	0.25	91
AMC	0.016–1	0.064	0.25	94
CLI	0.064–>256	0.25	>256	60
DOX	0.125–>128	8	16	25
LVX	0.25–4	1	2	75
MXF	0.032–0.5	0.125	0.25	100
Other viridans group streptococci (n = 24)d
PEN	0.016–0.5	0.064	0.5	83
AMC	0.016–0.5	0.064	0.5	83
CLI	0.064–>256	0.125	256	67
DOX	0.032–16	0.25	8	58
LVX	0.25–2	1	2	83
MXF	0.032–0.5	0.125	0.25	100
Streptococcus anginosus group and hemolytic streptococci (n = 17)e
PEN	0.016–0.25	0.064	0.125	100
AMC	0.016–0.25	0.064	0.25	100
CLI	0.064–>256	0.25	>256	71
DOX	0.064–>256	0.5	16	71
LVX	0.25–2	0.5	2	88
MXF	0.032–0.5	0.125	0.5	100
Neisseria spp. (n = 19)f
PEN	0.5–8	2	4	0
AMC	0.5–2	1	2	100
CLI	16–>256	128	>256	0
DOX	0.5–>256	2	>256	37
LVX	0.008–1	0.032	0.5	100
MXF	0.016–1	0.064	1	90
Prevotella oralis (n = 15)
PEN	0.008–>32	4	>32	40
AMC	0.016–1	0.064	0.25	100
CLI	0.016–>256	0.064	>256	73
DOX	0.032–8	0.5	4	53
LVX	0.25–2	0.5	2	73
MXF	0.125–0.5	0.25	0.5	100
Prevotella intermedia (n = 13)
PEN	0.008–>32	0.016	>32	77
AMC	0.016–0.064	0.016	0.064	100
CLI	0.016–>256	0.016	>256	85
DOX	0.032–4	0.5	4	69
LVX	0.064–0.5	0.25	0.5	100
MXF	0.016–0.5	0.125	0.5	100
Other Prevotella spp. (n = 28)g
PEN	0.016–>32	0.064	>32	64
AMC	0.016–0.25	0.032	0.125	100
CLI	0.016–>256	0.016	>256	82
DOX	0.032–16	0.25	8	61
LVX	0.125–4	1	1	93
MXF	0.032–4	0.25	0.5	96
Other anaerobes (n = 15)h
PEN	0.016–>32	0.125	>32	77
AMC	0.016–0.5	0.032	0.5	100
CLI	0.016–>256	0.032	>256	87
DOX	0,032–16	0.25	4	80
LVX	0.25–>32	0.5	16	73
MXF	0.064–4	0.125	1	87
Other bacteria (n = 21)i
PEN	0.004–32	1	>32	38
AMC	0.016–64	0.25	1	95
CLI	0.016–>256	2	>256	19
DOX	0.064–>32	1	8	48
LVX	0.006–2	0.25	2	91
MXF	0.016–0.5	0.125	0.25	100
All bacterial isolates (n = 205)
PEN	0.004–>32	0.125	8	67
AMC	0.016–64	0.064	0.5	96
CLI	0.016–>256	0.25	>256	60
DOX	0.032–>256	1	16	50
LVX	0.006–32	0.5	2	85
MXF	0.016–4	0.125	0.5	98

^a50% and 90%, MICs at which 50 and 90% of isolates are inhibited, respectively.

^bMICs were interpreted according to the breakpoints recommended by EUCAST in 2011 (6).

^cStreptococcus mitis (n = 44) and Streptococcus oralis (n = 9).

^dStreptococcus sanguinis (n = 8), S. salivarius (n = 6), S. parasanguinis (n = 3), S. pluranimalium (n = 3), S. acidominimus (n = 1), S. ovis (n = 1), S. sobrinus (n = 1), and S. thoraltensis (n = 1).

^eStreptococcus constellatus (n = 10), S. anginosus (n = 2), S. intermedius (n = 1), group C hemolytic streptococci (n = 3), and group G hemolytic streptococci (n = 1).

^fNeisseria spp. (n = 8), Neisseria cinerea (n = 5), N. sicca (n = 5), and N. elongata (n = 1).

^gPrevotella denticola (n = 9), P. buccae (n = 8), P. melaninogenica (n = 3), P. oris (n = 3), P. loescheii (n = 2), and Prevotella spp. (n = 3).

^hFusobacterium nucleatum (n = 6), Fusobacterium spp. (n = 3), Bacteroides merdae (n = 1), Bacteroides ovatus (n = 1), Bifidobacterium adolescentis (n = 1), Bifidobacterium sp. (n = 1), Porphyromonas endodontalis (n = 1), and Propionibacterium sp. (n = 1).

ⁱStaphylococcus aureus (n = 3), Moraxella catarrhalis (n = 3), Rothia mucilaginosa (n = 3), Gemella morbillorum (n = 2), Haemophilus influenzae (n = 2), Capnocytophaga spp. (n = 2), Enterococcus faecalis (n = 1), Escherichia coli (n = 1), Klebsiella pneumoniae (n = 1), Haemophilus parainfluenzae (n = 1), Staphylococcus epidermidis (n = 1), and Rothia dentocariosa (n = 1).

An itemized breakdown of the resistance rates of bacteria associated with various odontogenic infections demonstrated comparable overall susceptibility rates of isolates from abscesses versus isolates from infiltrates for MXF (98% versus 98%), LVX (83% versus 86%), PEN (66% versus 67%), AMC (97% versus 96%), CLI (59% versus 60%), and DOX (51% versus 49%), respectively.

Analysis of the distribution of the most common bacteria or bacterial groups showed that the Streptococcus mitis group, other viridans group streptococci, and Neisseria spp. tend to occur more frequently in infiltrates than in abscesses (Table 1), though without attaining the level of significance. In contrast, the S. anginosus group (S. anginosus, Streptococcus constellatus, and Streptococcus intermedius) and hemolytic streptococci were detected significantly more frequently (P = 0.04) in patients with abscesses (12 of 95) than in patients with infiltrates (5 of 110). All other bacteria or bacterial groups were found at nearly equal frequencies in the two patient groups.

Overall, only eight patients with odontogenic infections failed to recover following surgical intervention and primary antibiotic therapy. Bacteria isolated from patients with treatment failure and found to be resistant against the antibiotic used for treatment are listed in Table 3. From the abscess group a wide range of bacteria was isolated, including the S. anginosus group (n = 1), viridans group streptococci (n = 2), Enterococcus faecalis (n = 1), Neisseria spp. (n = 2), and anaerobes (n = 4). In contrast, from patients with treatment failure in the infiltrate group, only viridans group streptococci (n = 3) and Neisseria spp. (n = 4) were isolated.

Table 3.

Bacterial isolates from patients with treatment failure that were resistant against the antibiotic administered

Patient group and bacterium or bacterial group	Antibiotic
Abscess patients
Streptococcus anginosus group	CLI
Streptococcus sanguinis	CLI
Streptococcus salivarius	CLI
Enterococcus faecalis	CLI
Neisseria sicca	CLI
Neisseria sp.	CLI
Prevotella oralis	CLI
Prevotella intermedia	CLI
Peptostreptococcus sp.	CLI
Propionibacterium sp.	MXF
Infiltrate patients
Streptococcus mitis group (3 isolates)	CLI
Neisseria sicca	CLI
Neisseria cinerea	CLI
Neisseria elongata	CLI
Neisseria sp.	CLI

DISCUSSION

To our knowledge, the study from which the present data are taken was the first double-blind clinical trial of MXF for the treatment of odontogenic abscesses and gingival infiltrates. The study design allowed the analysis of the microbiota of odontogenic abscesses compared to those of gingival infiltrates, as well as a comparison of the in vitro susceptibility rates of the pathogens.

Statistical analysis by Fisher's exact test revealed that the S. anginosus group and hemolytic streptococci could be detected significantly more frequently in patients with abscesses than in patients with infiltrates. On the other hand, the S. mitis group, other viridans group streptococci, and Neisseria spp. tended to occur more frequently in infiltrates than in abscesses, although the difference did not reach statistical significance (Table 1). However, the detection of CLI-resistant S. mitis group isolates (S. mitis and Streptococcus oralis) and Neisseria spp. (Neisseria cinerea, N. elongata, N. sicca, and a Neisseria sp.) in patients with inflammatory infiltrates who failed to improve following CLI therapy clearly indicates that the S. mitis group and Neisseria spp. play decisive roles in the pathogenesis of odontogenic infiltrates (Table 3).

Our microbiological findings are in general agreement with the results of other studies, mostly of odontogenic abscesses, that also found strict anaerobes, such as Prevotella spp. and Fusobacterium spp., and facultative anaerobes, such as viridans group streptococci, to be the predominant pathogens (1, 15, 21). Our average number of isolates per specimen, 2.6, is well in line with the range of 1 to 24 reported in the literature (10, 18, 21). To our knowledge, there are as yet no published data on the bacterial spectrum in inflammatory gingival infiltrates.

Newer molecularly based methods, such as a reverse-capture checkerboard assay, were able to detect as yet uncultivated phylotypes in odontogenic abscesses, suggesting that previously unrecognized taxa are involved in the pathogenesis of this disease (18). Therefore, antibiotic therapy for odontogenic infections should be active against a wide range of bacterial pathogens, including viridans group streptococci and Prevotella spp., which accounted in our study for 77/205 and 56/205 isolates, respectively (Table 2). For this reason, antibiotics with broad activity against aerobes and anaerobes, such as PEN, CLI, DOX, and AMC, are mainly used or recommended for the therapy of odontogenic infections (2, 11, 14, 17).

With all antibiotics tested, susceptibility rates were similar for bacteria from abscesses and bacteria from infiltrates in our population.

As in our earlier reported clinical results (3), MXF was significantly more effective than CLI at reducing pain in patients with inflammatory infiltrates up to days 2 and 3 and showed significantly higher global clinical efficacy than CLI. The better outcome of MXF-treated patients with infiltrates may be explained in part by the higher impact of antibiotic therapy in the medical treatment of infiltrates, in contrast to abscesses, where the antimicrobial is adjuvant to incision and drainage (3). These clinical observations correlate well with our in vitro findings demonstrating that MXF is the most active antibiotic, with 98% of all isolates susceptible, whereas 60% of all isolates are susceptible to CLI.

In agreement with our results, Warnke et al. (21) observed MXF susceptibility rates of 99% versus 96% for aerobes/facultative anaerobes versus anaerobes from odontogenic abscesses. In a clinical trial, MXF, given prophylactically, was shown to be highly effective at reducing the prevalence and duration of bacteremia following dental extraction, while CLI was not effective (5).

However, susceptibility rates and MICs alone are insufficient for the prediction of in vivo potency. Concentrations achievable in serum and tissues, as well as pharmacokinetic characteristics, must be considered. MXF provides high bioavailability, a long half-life, and good penetration into tissue, achieving odontogenic site concentrations exceeding the concentrations in serum, as shown previously (4). When pharmacokinetic and pharmacodynamic values are combined, the ratio of the maximum concentration of the drug in serum (C_max) to the MIC predicts clinical cure for fluoroquinolones. Clinical data and animal models indicate that a C_max/MIC ratio of ≥8 is predictive of clinical cure for fluoroquinolone therapy (12, 13). In our study, a C_max/MIC ratio of 9 (4.5/0.5) was found for MXF, which is predictive of clinical cure (a C_max of 4.5 mg/liter following oral administration of 400 mg of MXF was taken as the basis [13]).

In conclusion, we found that MXF has promising in vitro activity against odontogenic pathogens, which justifies its use for the treatment of odontogenic abscesses and inflammatory infiltrates. Our findings clearly indicate that the S. anginosus group and hemolytic streptococci are associated with odontogenic abscesses and that the S. mitis group and Neisseria spp. are involved in the pathogenesis of odontogenic infiltrates.