General Methodological Considerations

Patient Analysis Populations²⁵

The randomized or intent-to-treat (ITT) population included screened subjects who were randomly assigned to study drug.

The modified intent-to-treat (mITT) population included ITT subjects who received at least one dose of study drug.

The clinical modified intent-to-treat (c-mITT) population included all subjects who received at least one dose of study drug and who had clinical evidence of a cIAI, as defined in the inclusion criteria.

The microbiological modified intent-to-treat (m-mITT) population included c-mITT subjects for whom one or more potentially causative baseline isolates were identified.

The clinically evaluable (CE) population included c-mITT subjects who met the following criteria: met inclusion and exclusion criteria as stated in the protocol under which the subject was enrolled (any exemptions from these criteria were decided while the data were blinded and included prospective exemptions, considered by the medical monitor on a case-by-case basis, and prospective or retrospective exemptions for creatinine clearance [CL_CR] [based on comparator labeling] for subjects enrolled under any version of the protocol, if the CL_CR was not lower than 41 mL/min [–15%]); had a test-of-cure assessment of cure or failure, but not indeterminate; met the criteria for either a clinical cure or a clinical failure; treatment regimen remained blinded throughout the study duration; received no more than one dose of a potentially effective concomitant antibacterial treatment after the baseline culture was obtained and before the first dose of test article.

The microbiologically evaluable (ME) population included CE subjects who met the following criteria: had a culture taken from the infection site at baseline that identified one or more isolate(s), and at least one isolate was susceptible to both study drugs (ie, tigecycline and imipenem/cilastatin); had clinical outcome and microbiologic information available to allow classification of a microbiologic response of eradication, persistence, or superinfection at the test-of-cure assessment.

Outcome Parameters²⁵

The data for evaluable patients were analyzed in terms of the patient’s clinical, microbiological, and pathogen response at the test-of-cure assessment. Clinical efficacy results were communicated as either cure or failure. Microbiological and pathogen responses were reported in terms of eradication, persistence, superinfection, or indeterminate response.

Statistical Analysis²⁵

The use of a noninferiority trial design is standard in phase 3 registration studies of antimicrobial agents. For the cSSSI and cIAI indications, a noninferiority margin of –15% was selected as a balance between the severity of the diseases under study and practical considerations regarding study enrollment. In fact, in each study, the lower bound of the
2-sided 95% confidence interval in both the evaluable and relevant ITT populations was above or near –10%. Furthermore, the 2 studies within each indication were virtually identical to allow for pooling. The final pooled data resulted in a lower bound well above –10% for each indication.

For all of the subpopulation analyses (eg, subgroup analyses, monomicrobial vs polymicrobial infection), an adjusted difference between treatment groups with its 95% confidence interval was calculated from a generalized linear model with a binomial probability function and an identity link (Proc GENMOD). For end points involving comparisons with small sample sizes, the method of Wilson corrected for continuity was used.

Duration of hospitalization, time to defervescence, duration of antibiotic therapy, duration of ICU stay, duration of home care, number of emergency department visits, and number of major procedures required in the operating room (eg, amputations) were analyzed by using the log-rank test. The median was used to estimate central tendency for each treatment group. The Hodges-Lehmann estimate and the 95% confidence intervals were given for the difference between medians, assuming that group distributions differ only by location.

Demographic and baseline characteristics of study subjects were compared between treatment groups. Characteristics that were continuous variables were compared by using a one-way analysis of variance with treatment as a factor. For categorical variables, the Fisher exact test was used. The difference between treatment groups in the percentage of discontinuations from study drug, both overall and for each primary reason, was evaluated by using a 2-sided Fisher exact test.

Complicated Intra-abdominal Infections

Protocols 301 and 306 compared the safety and efficacy of tigecycline with imipenem/cilastatin in hospitalized subjects with cIAI. Subjects were stratified at randomization according to Acute Physiologic and Chronic Health Evaluation (APACHE) II scores (≤15, or >15 but <31). Subjects were randomly assigned (1:1) to receive either tigecycline (an initial 100-mg dose, followed by 50 mg twice a day) or imipenem/cilastatin (200 to 500 mg, dose based on body weight and calculated creatinine clearance) every 6 hours via IV administration for up to 14 days. Subject participation in the study was approximately 4 to 8 weeks. Subjects were treated intravenously for up to 14 days (minimum of 5 days) of study drug administration, and, unless the subject was a clinical failure, a test-of-cure assessment was performed 12 to 44 days after the final dose of study drug. The coprimary efficacy analyses were the clinical response rates for the ME and m-mITT subjects at the test-of-cure assessment.²⁵

Inclusion/Exclusion Criteria, General

Subjects included hospitalized male or female subjects, at least 18 years of age, with cIAI who were candidates for or had received a laparotomy, laparoscopy, or percutaneous drainage of an intra-abdominal abscess.²⁵ Infections included intra-abdominal abscess, intra-abdominal abscess (including liver and spleen) that developed in a postoperative subject who received more than 48 hours but not more than 5 days of a nonstudy antibiotic and an intra-abdominal culture was obtained from the infected site, appendicitis complicated by perforation (grossly visible) and abscess or periappendicular abscess, perforated diverticulitis complicated by abscess formation or fecal contamination, complicated cholecystitis with evidence of perforation or empyema, perforation of the large or small intestine with abscess or fecal contamination, purulent peritonitis or peritonitis associated with fecal contamination, gastric or duodenal ulcer perforation with symptoms lasting at least 24 hours before operation, or traumatic bowel perforation with symptoms lasting at least 12 hours before operation.²⁵

Patients were excluded from participation if they had received more than one dose of an antibiotic (single broad-spectrum agent or one dose of each antibiotic in a combination regimen such as metronidazole, ampicillin, gentamicin) after the baseline intra-abdominal culture was obtained from the infected site. Subjects were also excluded if the intra-abdominal infection was known to be caused by one or more bacterial pathogens not susceptible to both of the study drugs (eg, P. aeruginosa, Proteus mirabilis) and in the investigator’s opinion required treatment with an additional antibacterial agent; anticipation of leaving the fascia or deep muscular layers open or expectation of planned abdominal reexploration either in or out of the operating room; suspected preoperatively to have had a diagnosis of spontaneous bacterial peritonitis, simple cholecystitis, gangrenous cholecystitis without rupture, simple appendicitis, acute suppurative cholangitis, pancreatic abscess, or infected necrotizing pancreatitis; APACHE II score >30; weight <40 kg; or administration of intraoperative antibacterial irrigants or peritoneal antibacterial agents (eg, irrigants, antibiotic-impregnated sponges).²⁵

Efficacy Results

Demographic/Baseline Characteristics

Study 301 was conducted in 17 countries, and study 306 was conducted in 27 countries.²⁵ Demographic and baseline characteristics (Table 1) were similar between the treatment groups in each study, with the exception of ethnic origin. In study 301, 39% of subjects were white and 61% were of other ethnic origin (primarily Hispanic). Most of the subjects (88%) in study 306 were white. Approximately 8% of the subjects were older than 75 years.²⁵

Table 1. Demographic and Baseline Characteristics Within the ME and m-mITT Populations²⁵

ME = microbiologically evaluable; m-mITT = microbiological modified intent-to-treat; APACHE = Acute Physiologic and Chronic Health Evaluation.

Infection type, etiology of disease, and comorbid conditions were comparable between treatment groups in each study. The most common clinical diagnosis in both studies was complicated appendicitis (52% in study 301 and 41% in study 306). Overall, severity of illness was moderate. Although APACHE II scores were generally less than 15, the subjects did not represent an easy-to-treat population, as evidenced by the substantial rate of fecal contamination (22.5% for all subjects), a high rate of bacteriological documentation (76.9%), and a high number of subjects with large (>100 mL, 31.9%) or multiple abscesses (10.7%). There was a 10% rate of prior antibiotic failure.²⁵

Table 2. Clinical Diagnosis of Infections Within the ME and m-mITT Populations²⁵

ME = microbiologically evaluable; m-mITT = microbiological modified intent-to-treat.

In study 301, a favorable clinical and microbiological response was observed in 199/247 ME subjects in the tigecycline group (80.6%) and 210/255 subjects in the imipenem/cilastatin group (82.4%). Similarly, a favorable clinical and microbiological response was observed in 227/309 m-mITT subjects in the tigecycline group (73.5%) and 244/312 subjects in the imipenem/cilastatin group (78.2%). Between-group differences in terms of clinical and microbiological response were not found to be statistically significant.²⁵

In study 306, a favorable clinical response was observed in 242/265 ME subjects in the tigecycline group (91%) and 232/258 subjects in the imipenem/cilastatin group (90%). A favorable clinical response was observed in 279/322
m-mITT subjects in the tigecycline group (87%) and 270/319 subjects in the imipenem/cilastatin group (85%). In this same study, a favorable microbiological response was observed in 242/265 ME subjects in the tigecycline group (91%) and 232/258 subjects in the imipenem/cilastatin group (90%). A favorable microbiological response was also observed in 279/322 m-mITT subjects in the tigecycline group (87%) and 270/319 subjects in the imipenem/cilastatin group (85%). Between-group differences in terms of clinical and microbiological response were not found to be statistically significant.²⁵

Results of the pooled data for studies 301 and 306 show a favorable clinical response was observed in 441/512 ME subjects in the tigecycline group (86.1%) and 442/513 subjects in the imipenem/cilastatin group (86.2%). Similarly, a favorable clinical response was observed in 506/631 m-mITT subjects in the tigecycline group (80.2%) and 514/631 subjects in the imipenem/cilastatin group (81.5%). Between-group differences in terms of clinical and microbiological response were not found to be statistically significant (see Figure 1).²⁵

Figure 1. Clinical Cure Rates in cIAI at Study End in the ME and m-mITT Populations²⁵

P < 0.0001 for noninferiority of tigecycline vs imipenem/cilastatin for each population comparison.
ME = microbiologically evaluable; m-mITT = microbiological modified intent-to-treat.

Figure 2 summarizes clinical responses within the pooled ME population by major clinical diagnosis: complicated appendicitis, complicated cholecystitis, perforation of the intestines, intra-abdominal abscess, complicated diverticulitis, gastric and duodenal perforations, peritonitis, and other. The results showed that treatment by clinical diagnosis interaction for clinical response was not significant.

Figure 2. Clinical Cure Rates by Clinical Diagnosis in cIAI at Study End in the ME Population²⁵

Pooled clinical cure rates by all baseline isolates at the test-of-cure assessment are shown in Figure 3 for the ME population.

Figure 3. Clinical Cure Rates by Pathogen in cIAI in the ME Population²⁵

*S. anginosus category includes S. anginosus, S. intermedius, and S. constellatus.

VSE = vancomycin-susceptible enterococci; MSSA = methicillin-susceptible Staphylococcus aureus.

At the ME subject level, 1 of 3 microbiological responses was possible: eradication, persistence, or superinfection.
Figure 4 presents the pooled responses in the ME population at the last day of therapy. In the ME population, eradication rates were not statistically different between the 2 treatment groups. The eradication rate was 88.1% for tigecycline-treated subjects and 91.0% for imipenem/cilastatin-treated subjects. The integrated analysis of microbiological response shows that tigecycline was statistically noninferior to imipenem/cilastatin in the ME population based on the adjusted lower bound of the 2-sided 95% confidence interval (95% CI, –6.9,1.0).²⁵

Figure 4. cIAI: Eradication Rates²⁵

P < 0.0001 for noninferiority of tigecycline vs imipenem/cilastatin for each population comparison.
ME = microbiologically evaluable; m-mITT = microbiological modified intent-to-treat.

Figure 5 summarizes pooled microbiological responses at the subject level for ME subjects who had monomicrobial or polymicrobial infections. At the last day of therapy assessment, the difference in eradication rates among ME subjects with monomicrobial infections was –3.2% (91.7% in the tigecycline group and 94.8% in the imipenem/cilastatin group). Among those with polymicrobial infections, the difference was –2.6% (86.1% in the tigecycline group and 88.7% in the imipenem/cilastatin group). The results of a generalized linear model showed that the type of infection and treatment interaction for microbiological response was not significant.²⁵

Figure 5. cIAI: Eradication Rates by Monomicrobial/Polymicrobial Infection²⁵

The clinical response of failure was analyzed using stepwise logistic regression. The same variables were analyzed as for cSSSI, with the following modifications and additions: clinical diagnosis (complicated appendicitis, complicated diverticulitis, complicated cholecystitis, intra-abdominal abscess, peritonitis, gastric and duodenal perforations, perforation of the intestines), APACHE II score (<15, ≥15 and <10, 10 to 15, and >15), percutaneous drainage, and bacteremia. Treatment group was not an independent predictor of failure. APACHE II score ≥10 was identified as significantly associated with failure in the ME population, supporting the finding of noninferiority in the coprimary analyses. Other statistically significant factors included geographic region (Western Europe, Canada, and the United States combined), ethnicity (other vs white), the presence of B. fragilis, underlying condition other than complicated appendicitis or complicated cholecystitis, and diabetes mellitus.²⁵

Subgroup analyses of clinical responses in the ME population were performed on pooled data to evaluate the consistency of tigecycline cure rates and the noninferiority of tigecycline to imipenem/cilastatin. No significant interactions of age, gender, ethnicity, clinical diagnosis, creatinine clearance, and bacteremia status with treatment were identified.²⁵

These secondary analyses were not powered sufficiently or designed to statistically show an effect; therefore, these results should be interpreted with caution.

The clinical cure rate for tigecycline-treated subjects with bacteremia at baseline in the cIAI studies was 78.8% (26 of 33).²⁵ The majority of isolated organisms were Bacteroides spp., followed by E. coli, S. aureus (methicillin-susceptible only), and Klebsiella spp.²⁵

Patients with P. aeruginosa isolated at baseline as the causative organism were excluded from the studies; however, if the organism was present as part of a polymicrobial infection, these patients were included. The clinical cure rate in the m-mITT group for these patients was 85.7% (6/7) in the tigecycline group versus 57.1% (4/7) in the imipenem/cilastatin group.²⁵

Conclusions

The results of these studies demonstrated that the clinical efficacy of 50 mg tigecycline, as a single agent infused every 12 hours (after an initial loading dose of 100 mg), is noninferior to the combination antibiotic regimen of imipenem/cilastatin (200 to 500 mg, adjusted according to body weight and creatinine clearance) administered every 6 hours in the treatment of subjects with cIAI.²⁵

Tigecycline met the statistical criteria of noninferiority to the comparative regimen for the primary end points within each of the coprimary populations. Tigecycline met the statistical criteria for noninferiority to comparative regimens in secondary end points, including bacteriological eradication rates in the ME and m-mITT populations. Bacterial susceptibility data indicated that tigecycline is efficacious against isolates commonly observed in subjects with cIAI. In addition, tigecycline was efficacious against isolates containing genes known to confer antibiotic resistance.²⁵

For these cIAI studies, consistent with the findings for clinical cure rates, few significant differences were found between treatment groups in the rate, duration, or intensity of use of various health care services during the inpatient phase or in outpatient health care resources.²⁵

Multiple subpopulation analyses of clinical responses were generally consistent with the findings from the coprimary populations. For each subpopulation analysis, no interaction was observed between treatment and the subpopulation tested.²⁵

Tigecycline is efficacious for the treatment of cIAI in patient populations broadly representative of those seen in clinical practice. These populations included subgroups that were challenging to treat by virtue of the comorbid conditions as well as characteristics of the infection. Efficacy was also demonstrated in relevant demographic and disease category subgroups.

Tigecycline is indicated for the treatment of adults with complicated intra-abdominal infections caused by Citrobacter freundii, Enterobacter cloacae, Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Enterococcus faecalis (vancomycin-susceptible isolates), Staphylococcus aureus (methicillin-susceptible and –resistant isolates), Streptococcus anginosus grp. (includes S. anginosus, S. intermedius, and S.constellatus), Bacteroides fragilis, Bacteroides thetaiotaomicron, Bacteroides uniformis, Bacteroides vulgatus, Clostridium perfringens, and Peptostreptococcus micros.

Results for the cIAI indication confirm the efficacy of tigecycline against infection caused by organisms with tetracycline resistance determinants.²⁵

Indications

TYGACIL^® (tigecycline) is indicated for the treatment of adults with:

• Complicated skin and skin structure infections caused by Escherichia coli, Enterococcus faecalis (vancomycin-susceptible isolates), Staphylococcus aureus (methicillin-susceptible and -resistant isolates), Streptococcus agalactiae, Streptococcus anginosus grp. (includes S. anginosus, S. intermedius, and S. constellatus), Streptococcus pyogenes, Enterobacter cloacae, Klebsiella pneumoniae, and Bacteroides fragilis
• Complicated intra-abdominal infections caused by Citrobacter freundii, Enterobacter cloacae, Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Enterococcus faecalis (vancomycin-susceptible isolates), Staphylococcus aureus (methicillin-susceptible and -resistant isolates), Streptococcus anginosus grp. (includes S. anginosus,
  S. intermedius, and S. constellatus), Bacteroides fragilis, Bacteroides thetaiotaomicron, Bacteroides uniformis, Bacteroides vulgatus, Clostridium perfringens, and Peptostreptococcus micros
• Community-acquired bacterial pneumonia caused by Streptococcus pneumoniae (penicillin-susceptible isolates), including cases with concurrent bacteremia, Haemophilus influenzae (beta-lactamase negative isolates), and Legionella pneumophila

Important Safety Information

• TYGACIL is contraindicated in patients with known hypersensitivity to tigecycline
• Anaphylaxis/anaphylactoid reactions have been reported with nearly all antibacterial agents, including tigecycline, and may be life-threatening. TYGACIL should be administered with caution in patients with known hypersensitivity to tetracycline-class antibiotics
• Isolated cases of significant hepatic dysfunction and hepatic failure have been reported in patients being treated with tigecycline. Some of these patients were receiving multiple concomitant medications. Patients who develop abnormal liver function tests during tigecycline therapy should be monitored for evidence of worsening hepatic function. Adverse events may occur after the drug has been discontinued
• The safety and efficacy of TYGACIL in patients with hospital-acquired pneumonia have not been established
• An increase in all-cause mortality has been observed across phase 3 and 4 clinical studies in TYGACIL-treated patients versus comparator-treated patients. The cause of this increase has not been established. This increase in all-cause mortality should be considered when selecting among treatment options
• TYGACIL may cause fetal harm when administered to a pregnant woman
• The use of TYGACIL during tooth development may cause permanent discoloration of the teeth. TYGACIL should not be used during tooth development unless other drugs are not likely to be effective or are contraindicated
• Acute pancreatitis, including fatal cases, has occurred in association with tigecycline treatment. Consideration should be given to the cessation of the treatment with tigecycline in cases suspected of having developed pancreatitis
• Clostridium difficile-associated diarrhea (CDAD) has been reported with use of nearly all antibacterial agents, including TYGACIL, and may range in severity from mild diarrhea to fatal colitis
• Monotherapy should be used with caution in patients with clinically apparent intestinal perforation
• TYGACIL is structurally similar to tetracycline-class antibiotics and may have similar adverse effects. Such effects may include: photosensitivity, pseudotumor cerebri, and anti-anabolic action (which has led to increased BUN, azotemia, acidosis, and hyperphosphatemia). As with tetracyclines, pancreatitis has been reported with the use of TYGACIL
• To reduce the development of drug-resistant bacteria and maintain the effectiveness of TYGACIL and other antibacterial drugs, TYGACIL should be used only to treat infections proven or strongly suspected to be caused by susceptible bacteria. As with other antibacterial drugs, use of TYGACIL may result in overgrowth of non-susceptible organisms, including fungi
• The most common adverse reactions (incidence >5%) are nausea, vomiting, diarrhea, abdominal pain, headache, and increased SGPT
• Prothrombin time or other suitable anticoagulant test should be monitored if TYGACIL is administered with warfarin
• Concurrent use of antibacterial drugs with oral contraceptives may render oral contraceptives less effective
• The safety and effectiveness of TYGACIL in patients below age 18 and lactating women have not been established

Please see full Prescribing Information.