ABSTRACT
This study investigated the potential of the novel systemic pleuromutilin antibiotic BC-3781 to treat patients with an acute bacterial skin and skin structure infection (ABSSSI) caused by a Gram-positive pathogen. Patients were randomized to intravenous BC-3781 100 mg, BC-3781 150 mg, or vancomycin 1 g every 12 h. Response to treatment was assessed daily and at test of cure (TOC). The primary endpoint was the clinical success rate at TOC in the modified intent-to-treat (MITT) and clinically evaluable (CE) analysis populations. Baseline characteristics, including the frequency of methicillin-resistant Staphylococcus aureus (MRSA), were comparable between the different treatment groups. Of 210 patients randomized, 186 (88.6%) patients completed the study. Clinical success at TOC in the CE population occurred in 54 (90.0%) patients in the BC-3781 100-mg group, 48 (88.9%) in the BC-3781 150-mg group, and 47 (92.2%) in the vancomycin group. At day 3, the clinical response rate was similar across the three treatment groups. Six patients discontinued study medication following an adverse event. The incidence rate for drug-related adverse events was lower for patients receiving BC-3781 (34.3% and 39.4% in the 100-mg and 150-mg groups, respectively) than those receiving vancomycin (53.0%). When BC-3781 was used to treat ABSSSIs caused by a Gram-positive pathogen, including MRSA, clinical success rates were comparable to those of the comparator, vancomycin. BC-3781 was generally well tolerated. These results provide the first proof of concept for the systemic use of a pleuromutilin antibiotic for the treatment of ABSSSIs.
INTRODUCTION
Skin and skin structure infections (SSSIs) are predominantly caused by Staphylococcus aureus and Streptococcus pyogenes, although other pathogens may be involved (1). Effective antibiotic therapy is essential to control such infections (2, 3), although successful treatment is made more difficult by the continued emergence of bacterial resistance. There is an urgent need for new antibiotics active against resistant strains, preferably via a novel mechanism of action (1, 4–6).
BC-3781 is a novel semisynthetic systemic agent belonging to the pleuromutilin class of antibiotics. Pleuromutilins interfere with bacterial protein synthesis via a specific interaction with the 23S rRNA of the 50S bacterial ribosome subunit (7). In vitro, BC-3781 demonstrates clinically relevant activity against the most frequently identified Gram-positive skin pathogens, including methicillin-susceptible Staphylococcus aureus (MSSA) and methicillin-resistant S. aureus (MRSA), S. pyogenes, Streptococcus agalactiae, and vancomycin-resistant Enterococcus faecium (8, 9). Furthermore, BC-3781 exhibits significant in vitro activity against coagulase-negative Staphylococcus spp. and community respiratory pathogens, including Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, Legionella pneumophila, Chlamydophila pneumoniae, and Mycoplasma pneumoniae (9, 10).
The therapeutic administration of pleuromutilin antibiotics has so far been limited to veterinary medicine or to the topical route in humans, due to associated toxicities. Intravenous (i.v.) BC-3781 was well tolerated in phase I clinical studies in healthy subjects at exposures predicted to be therapeutically efficacious for the treatment of acute bacterial SSSIs (ABSSSIs) (11, 12). The present investigation was the first in which a pleuromutilin was administered systemically to patients for treatment of ABSSSIs.
(This work was presented, in part, at the 51st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL, 2011 [13] and at the 22nd European Congress of Clinical Microbiology and Infectious Diseases meeting, 2012, London, United Kingdom [14].)
MATERIALS AND METHODS
This double-blind, parallel-group, phase II study was conducted at 20 centers in the United States in accordance with the Declaration of Helsinki and with the International Conference on Harmonisation Good Clinical Practice. The study protocol and informed consent form were approved by the appropriate local institutional review board. All patients provided written informed consent.
Patients.The study population comprised male and female patients aged ≥18 years with an ABSSSI (wound infection, burn infection, cellulitis, erysipelas, or abscess with cellulitis) that was known or suspected to be caused by a Gram-positive pathogen requiring i.v. antibiotic therapy. Cellulitis was required to have a minimum dimension of ≥10 cm, with the diameter of any associated abscess being <75% of the maximum dimension of the cellulitis. Patients suffering from abscess with cellulitis could comprise a maximum of 30% of the enrolled study population. In addition, patients were required to have at least two signs of systemic inflammation (e.g., elevated temperature) or evidence of either a significant underlying systemic disease (e.g., diabetes mellitus) or a local medical condition (e.g., peripheral arterial disease). Exclusion criteria included those typical for studies of infections previously defined as “complicated SSSIs” (15). Receipt of >24 h of treatment with more than a single dose of a systemic antibiotic excluded a patient unless the patient was a proven failure (Gram-positive organism isolated and compliance with all other inclusion criteria).
Study design.Patients with ABSSSIs were randomized to receive BC-3781 100 mg, BC-3781 150 mg, or vancomycin 1 g (or vancomycin adjusted individually according to institutional guidelines) via i.v. infusion twice daily (once every 12 h) for between 5 and 14 days of treatment. Study medication was prepared by a nonblinded, qualified member of the site staff and administered by a different staff member who was blinded to the study medication. BC-3781 was diluted from a stock solution of 10 mg/ml according to the manufacturer's instructions. Vancomycin hydrochloride powder (Hospira Inc., Lake Forest, IL) was diluted according to the manufacturer's instructions. Patients with a confirmed Gram-negative pathogen were permitted to receive aztreonam i.v. every 8 h.
Hospitalized male and female patients aged 18 years or over with a clinical diagnosis of an ABSSSI that was known or suspected to be caused by a Gram-positive pathogen and required i.v. antibiotic therapy were enrolled in the study. Study assessments were performed at enrollment, during treatment, at the end-of-treatment (EOT) visit, at the test-of-cure (TOC) visit 7 to 14 days after treatment, and at long-term follow-up 30 days after treatment. Procedures included evaluation by the investigator of the infection site, adverse events (AEs), electrocardiograms (ECGs; including calculation of the QT interval duration by the correction of Fridericia [QTcF]), vital signs, and collection of laboratory safety samples (hematology, blood chemistry, and urinalysis). Electrocardiogram data were reviewed by the attending physician. Samples for microbiological evaluation were taken at enrollment and, if appropriate, at EOT and TOC for Gram's staining, culture, and determination of antibiotic susceptibility. All S. aureus isolates were tested for Panton-Valentine leukocidin (PVL) status and pulsed-field gel electrophoresis type.
At the TOC visits, treatment response was categorized as success, improved, failure, or not evaluable (for example, because of patient loss to follow-up). Success was defined as resolution of all signs and symptoms of the infection or improvement in these signs and symptoms such that no further antimicrobial therapy was necessary. Treatment was judged to have failed if the infection did not resolve, if a nonstudy antibiotic was administered, or if study drug treatment was prematurely discontinued due to an AE. The microbiological outcome at the TOC visit was categorized as eradication, presumed eradication, persistence, presumed persistence, superinfection, missing, and indeterminate.
Statistical methods.The primary outcome measure was the clinical success rate at the TOC visit. Clinical success in the clinically evaluable (CE) population and clinical success in the modified intent-to-treat (MITT) population were considered to be coprimary endpoints. The CE population included those patients who had a confirmed ABSSSI, met the inclusion/exclusion criteria, and fully complied with the protocol. The MITT population included those patients in the intent-to-treat (ITT) population (patients who received ≥1 dose of study drug) who had a Gram-positive pathogen isolated from a culture of blood or the ABSSSI site. Eligibility for inclusion in the CE analysis population and assignment as a clinical success or failure were determined by the sponsor's assessment of the blinded data. That is, for a patient to be a success, the patient had to have met all the protocol criteria, to be rated a success by the investigator, and to have completed a TOC visit. No patient response was changed from a failure to a success, but some patients for whom the investigator had judged the response a success were changed to a failure (e.g., for receipt of nonstudy antimicrobials). Secondary outcome measures included microbiological outcome and results of assessments of safety parameters (vital signs, ECG, clinical laboratory findings [hematology, blood chemistry, and urinalysis], physical examination, and AEs).
Lesion size and temperature were measured daily to EOT. This paper specifically presents clinical responders at day 3 measured using four different definitions: (i) absence of fever, (ii) no increase in the area of erythema plus absence of fever, (iii) no increase in the area of erythema, no increase in the area of swelling, and absence of fever, and (iv) a ≥20% reduction in the area of erythema from baseline at day 3. Patients were defined as nonresponders if they discontinued from study medication for clinical failure at or before day 3.
The planned sample size of 210 was chosen to provide information on the efficacy, safety, and tolerability of BC-3781. The study was not powered for inferential statistical analysis. Two-sided 95% confidence intervals (CIs) were calculated for the between-treatment group differences in the clinical success rate for each treatment group. The microbiological success rate was calculated as the proportion of patients with microbiological eradication or presumed eradication at the TOC visit.
RESULTS
Patient demographics and baseline characteristics.Of the 210 randomized patients, 207 received at least one dose of study medication. Patient disposition was similar among the treatment groups (Table 1). The three treatment groups were comparable with respect to age, gender, and race/ethnicity, but more patients with diabetes mellitus and more patients of higher weight were treated with BC-3781 150 mg (Table 2). One hundred eighty-six (88.6%) patients completed the study through TOC. Eighteen (8.6%) patients (4 [5.7%] patients on BC-3781 100 mg, 6 [8.3%] on BC-3781 150 mg, and 8 [11.8%] on vancomycin) did not complete the study; the incidence of withdrawal was not significantly different among the treatment groups. Mean overall adherence to the i.v. treatment regimen was 99.5% across the analysis populations.
Disposition of randomized patient population
Patient demographic characteristics (ITT population)
The most frequent infection types were cellulitis and abscess with cellulitis (Table 3). The mean duration of infection prior to enrollment was slightly longer in the BC-3781 150-mg group than in the other groups. This difference was driven by one patient in this group who had a much longer duration of infection than any other patient in the study (183 days). Signs of systemic inflammation were common, suggesting a substantial severity of illness at baseline; 54 (26.1%) patients had at least one comorbid condition. Fifty-eight (28.0%) patients had experienced a prior treatment failure. Sixty-one (29.5%) patients were naive to antibiotic therapy. Overall, there were no significant differences in the infection characteristics among the treatment groups at baseline.
Summary of baseline characteristics (ITT population)
Of 155 patients with a baseline pathogen, 151 (97.4%) had at least one Gram-positive pathogen. The remaining patients had infections with a Gram-negative organism or mixed infections with a Gram-positive organism/Gram-negative organism (1.9% and 0.6%, respectively). The most frequent Gram-positive pathogen was S. aureus (in 90.8% of patients in the MITT population; for 69.1% of these patients, the pathogen was MRSA). The distribution of pathogens was generally comparable across the treatment groups.
Clinical success at TOC.The clinical success rate in the CE population was comparable for all treatment groups: 54 (90.0%) patients in the BC-3781 100-mg group, 48 (88.9%) in the BC-3781 150-mg group, and 47 (92.2%) in the vancomycin group (Table 4). There were similar results in the other analysis populations. The clinical success rate was high for all important subgroups of patients (e.g., primary infection type, diabetes mellitus status) (Table 5). The clinical success rate by baseline pathogen, including MRSA, was comparable across the three treatment groups (Table 6). Clinical success rates for BC-3781 against PVL-positive MRSA and USA300 MRSA strains were similar to or numerically higher than the corresponding clinical success rates for vancomycin (Table 6).
Clinical success rate at TOC visit (CE and MITT populations)a
Clinical success rate at TOC visit by selected baseline characteristics (MITT population)
Clinical success rate at TOC visit by major baseline target pathogens (S. aureus) (MITT population)
Microbiological success at TOC.In the MITT population, microbiological success was achieved in 40/50 (80.0%) patients in the BC-3781 100-mg group, 43/51 (84.3%) patients in the BC-3781 150-mg group, and 42/51 (82.4%) patients in the vancomycin group. Development of decreased susceptibility to BC-3781 or vancomycin was not observed.
Other endpoints.The decrease in ABSSSI lesion size over the first 5 days of treatment, at EOT, and at TOC was similar for all three treatment groups (Fig. 1). Time to cessation of the spread of erythema was less than 3 days in most patients (Table 7). When the clinical response on day 3 was analyzed according to a range of criteria (absence of fever, cessation of spread of erythema plus absence of fever, cessation of spread of erythema and swelling, and a ≥20% reduction in erythema), the response was similar in all treatment groups (Table 7). The percent change of the area of erythema from baseline for each patient in the study is shown in Fig. 2 together with the median change. Comparison of the clinical response at TOC with the clinical response at day 3 showed that most of the patients that achieved success at TOC also achieved a response of success at the early time point (Table 8). Also, most patients who were not responders at the early time point became successes at TOC. A small number of patients (4.8%) who showed a response of success at the early time point became a failure at TOC. A small number of patients experienced a recurrence of infection at the site of the original infection (BC-3781 100 mg, one patient; BC-3781 150 mg, two patients; vancomycin, one patient).
Median percent change in area of erythema from baseline over time for the ITT population. The standard deviation of the median is calculated as (Q3 − Q1)/1.075, where Q1 is the first quartile and Q3 is the third quartile.
Clinical response at day 3 (ITT population)
Individual percent change of area of erythema from baseline (together with medians) on day 3 for the ITT population.
Concordance of clinical response at TOC visit to early clinical response on day 3
Safety and tolerability.The mean duration of exposure to study drug treatment was approximately 7 days for all groups, and almost 70% of patients completed therapy within that time. Table 9 shows the duration of treatment by treatment group.
Duration of treatment in ITT population
An AE was reported by 148 (71.5%) patients; 87 (42.0%) reported an AE considered drug related (BC-3781 100 mg, 34.3%; BC-3781 150 mg, 39.4%; vancomycin, 53.0%). Overall, the most frequently reported AEs in the BC-3781 treatment groups were headache, nausea, and diarrhea (Table 10). Most were mild or moderate in intensity. Incidences of nausea, headache, and pruritus were higher in patients who received vancomycin treatment than in those who received BC-3781 treatment. No cases of Clostridium difficile infection occurred. Local signs and symptoms at the site of infusion, such as pain, erythema, and phlebitis, were more frequent in BC-3781-treated patients than in vancomycin-treated patients; most of these AEs were considered mild in intensity (Table 10). Skin disorders were more frequent in vancomycin-treated patients than in BC-3781 100-mg- and 150-mg-treated patients (22.7%, 4.3%, and 4.2%, respectively). Changes in laboratory parameters, including isolated increases in alanine aminotransferase and aspartate aminotransferase levels, were comparable among groups.
Drug-related treatment-emergent AEs by preferred term reported by >2% of patients in ITT population
No change in ECG was considered to be of clinical significance, and no drug-related cardiovascular AE was reported. The ECG RR interval increased (i.e., the heart rate decreased) in all treatment groups during treatment. Both BC-3781 and vancomycin treatments were associated with a small increase in the QTcF interval. The mean maximum change with vancomycin was 7 ms, whereas with BC-3781 the change was slightly larger (BC-3781 100 mg, 11.2 ms; BC-3781 150 mg, 12.3 ms). Three subjects, one in each treatment group, had a QTcF interval of >450 ms. No QTcF value of >480 ms and no increase in QTcF interval of >60 ms from baseline were observed in any treatment group.
Four patients discontinued study medication following a drug-related AE: one patient (1.4%) in the BC-3781 100-mg group (events of hyperhidrosis, vomiting, and headache), two (2.8%) in the BC-3781 150-mg group (infusion site pain in one patient and dyspnea in the other), and one (1.5%) in the vancomycin group (drug eruption).
Five serious adverse events resulting in hospitalization occurred, but none of these was drug related: one abscess in the BC-3781 100-mg group, one respiratory failure and one case of cellulitis in the BC-3781 150-mg group, and one accidental overdose of an analgesic and one convulsion in the vancomycin group. All events resolved. No patient died.
DISCUSSION
In this first phase II study of a systemically available semisynthetic pleuromutilin, BC-3781 at both doses achieved high clinical and microbiological treatment success rates. Clinical response rates were consistent across multiple endpoints. The lack of a detectable dose-response effect on clinical outcomes for BC-3781 is of interest, but further analysis of pharmacokinetic and pharmacodynamic data has provided additional information suggesting a likely exposure-response relationship (16).
The comparator used in this study, vancomycin, achieved a clinical success rate at TOC in CE patients of 92.2%, similar to that reported in other recently published studies using only vancomycin for coverage of Gram-positive organisms (15, 17, 18). Although cross-study comparisons of outcomes are difficult because of differences in patient populations and other uncontrollable factors, the consistency of reported outcomes for vancomycin in the literature and the current study supports the sensitivity of this study to define the treatment effect of BC-3781.
The high clinical success rates achieved for each of the treatment groups in this study were paralleled by high microbiological eradication rates. The clinical success rates at TOC for both MSSA and MRSA were similar for the three treatment groups (Table 5).
A higher percentage of patients in the BC-3781 groups than in the vancomycin group had diabetes mellitus at baseline (20% and 26.6% versus 15.2%). A higher percentage of patients in the vancomycin group than in the BC-3781 groups had a baseline diagnosis of cellulitis. Diabetes mellitus has been documented to be a risk factor for poorer outcomes in ABSSSIs, favoring a greater clinical response in the vancomycin group (19). However, the higher proportion of abscesses in the BC-3781 groups may have favored the response rate for BC-3781, since the response of these types of ABSSSIs can be influenced by successful incision and drainage.
On the basis of recent Food and Drug Administration guidance (20), early clinical response was assessed using a composite endpoint of erythema and swelling no greater than those at study enrollment and absence of fever at day 3, which was proposed to be a primary endpoint for ABSSSIs. Also, in the present study, additional analyses of the early response were undertaken. The percentage of patients showing a reduction in lesion size of ≥20%, as measured by the area of erythema on study day 3 compared with that at enrollment, was determined (21). Also, at study day 3, the percentage of patients in whom there was no increase in the area of swelling compared to that at study enrollment and no fever was determined (Table 7). The addition of swelling to make a more complex composite endpoint did not alter the result or the conclusions. In all these analyses, BC-3781 at both doses showed clinical and microbiological success rates similar to those of vancomycin. At study day 3, BC-3781 showed a response rate of ≥70% by all methods of analysis used, which is consistent with its preclinical microbiological profile (8, 9) and confirmed by population pharmacokinetics (22) and dose selection rationale (16).
BC-3781 administered i.v. every 12 h over 5 to 14 days at a dose of either 100 mg or 150 mg was generally well tolerated. Some local signs and symptoms related to the site of infusion, such as local pain, erythema, and phlebitis, were reported more frequently in the BC-3781 treatment groups than in the vancomycin group. However, since the numbers reported were low in each group, it is difficult to make a conclusion about the significance of these reports. The types of AEs documented were typical for a patient population with ABSSSIs (15). The incidence of drug-related AEs was lower in patients treated with BC-3781 100 and 150 mg than in patients treated with vancomycin. Changes in laboratory parameters, including changes in alanine aminotransferase and aspartate aminotransferase, were comparable between treatment groups.
ECG analysis showed a decrease in heart rate in all treatment groups during treatment, likely reflecting an increased heart rate at study admission due to infection and subsequent successful treatment of the ABSSSI. This increase in the RR interval following treatment potentially confounded the assessment of QTcF. A small increase in QTcF was associated with each study treatment regimen, including vancomycin. No subject in any treatment group had an increase from baseline in QTcF of ≥60 ms, and only one subject in each group had a QTcF value of >450 ms. No subject had a QTcF value of >480 ms. None of these changes was considered to be of clinical concern, and no drug-related cardiovascular AE was reported.
Overall, BC-3781 was well tolerated and demonstrated clinical and microbiological success rates similar to those of the comparator, vancomycin. These results provide the first proof of concept for the systemic use of a pleuromutilin antibiotic for the treatment of ABSSSIs and support the continued clinical evaluation of BC-3781 for the treatment of ABSSSIs and other serious infections where the need for new treatment options has been widely emphasized.
ACKNOWLEDGMENTS
We thank the patients and staff who participated in the studies. Gareth Cuttle of Niche Science and Technology Ltd. provided support during the preparation of the manuscript.
These studies were funded by Nabriva Therapeutics AG.
W. T. Prince, Z. Ivezic-Schoenfeld, C. Lell, and F. Obermayr are employees of Nabriva AG. R. Novak is now an employee of Sanofi. K. J. Tack was a consultant to MedPace, the contract research organization that conducted the study on behalf of the sponsor. G. H. Talbot is a consultant to Nabriva and a member of the company's supervisory board. With the exception of K. J. Tack, each author holds an equity position in the company.
FOOTNOTES
- Received 17 October 2012.
- Returned for modification 16 December 2012.
- Accepted 12 February 2013.
- Accepted manuscript posted online 19 February 2013.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.