Monotherapy or Combination Therapy?

The Pseudomonas aeruginosa Conundrum

Kristi A. Traugott, Pharm.D.; Kelly Echevarria, Pharm.D.; Pamela Maxwell, Pharm.D.; Kay Green, B.S.; James S. Lewis, II, Pharm.D.


Pharmacotherapy. 2011;31(6):598-608. 

In This Article

Clinical Data on Monotherapy versus Combination Therapy


The study most often quoted to support combination therapy's superiority to monotherapy in patients with P. aeruginosa bacteremia was published in 1989.[23] This prospective, multicenter study was conducted in a mixed population of 200 patients, one third of whom were in the ICU at onset of bacteremia and of whom approximately 60% were considered immunosuppressed. Ninety percent of patients received an aminoglycoside (mainly tobramycin or gentamicin) and 75% received a β-lactam agent (mainly piperacillin or ticarcillin), with 70% receiving a combination of the two. These investigators demonstrated a clear mortality benefit for patients receiving combination therapy, with overall mortality of 47% in the monotherapy group versus 27% with combination therapy (p=0.023). Trends were in favor of combination therapy in every subgroup, although a significant mortality benefit was shown only in patients with nosocomial infections or pneumonia or in critically ill patients. In addition, multivariate analysis found combination therapy to be significantly associated with survival (p=0.004). However, in contrast to a previously mentioned study,[13] in vitro synergy testing showed no significant correlation with patient outcomes. One major criticism of this study is that 86% of patients (37/43) receiving monotherapy were receiving only an aminoglycoside,[23] a treatment option that is no longer considered optimal because of the association with increased mortality.[36,49–51] In fact, the mortality rate of 47% in this group is similar to mortality rates in studies where patients were treated with drugs not active in vitro against P. aeruginosa.[12,43] Most of the studies demonstrating the inferiority of aminoglycosides used traditional dosing with multiple daily doses of aminoglycosides and no mention of pharmacokinetic monitoring. It is unclear if appropriate kinetically adjusted traditional dosing or extended-interval dosing might have changed the outcomes.

In contrast, another group reviewed 410 episodes of P. aeruginosa bacteremia in patients with cancer and failed to demonstrate a significant difference in cure rates with combination therapy (antipseudomonal penicillin plus an aminoglycoside) versus monotherapy (antipseudomonal penicillin, 75% vs 79%, p=0.64).[36] Many different regimens were used with the most common antipseudomonal penicillins including older extended-spectrum penicillins (carbenicillin and ticarcillin) and with gentamicin as the most common aminoglycoside. The authors did demonstrate significantly higher cure rates in patients receiving appropriate empiric antibiotics (67% vs 14%, p<0.001) and in patients receiving antipseudomonal penicillins alone versus aminoglycosides alone (71% vs 29%, p=0.00001). Even in this immunosuppressed population, the addition of an aminoglycoside to an antipseudomonal penicillin did not appear to demonstrate benefit, regardless of the neutrophil count, but did clearly demonstrate the inferiority of amino-glycoside monotherapy compared with other antipseudomonal agents for serious P. aeruginosa infections.

A retrospective cohort study was performed in 115 patients with sepsis and P. aeruginosa bacteremia.[6] Approximately one third of patients were in the ICU and more than one half were immunocompromised. Initial evaluation of the data failed to find a relationship between empiric therapies and risk of death before receipt of susceptibility results. However, after controlling for independent prognostic factors and definitive treatment, patients in the adequate empiric monotherapy group and the inadequate treatment group were both found to have significantly higher 30-day mortality rates compared with those in patients in the adequate empiric combination therapy group. The increase in mortality was evident early on in patients receiving inappropriate therapy; however, the Kaplan-Meier curve lines did not begin to diverge for monotherapy versus combination therapy until approximately day 23. When examining only definitive therapy, adequate therapies were associated with better outcomes than were inadequate therapies regardless of monotherapy or combination therapy, but no evidence for the superiority of either monotherapy or combination therapy was found. Contrary to the first-mentioned study,[23] this study did not accept aminoglycoside monotherapy as appropriate therapy.

More recently, a retrospective cohort of 305 patients with P. aeruginosa bacteremia was evaluated with the intent to investigate the relationship between combination therapy and appropriate treatment.[2] Over half of the patients were in the ICU, and a large portion were immunocompromised. This study is distinctive in that during the study period, antimicrobial guidelines were in place recommending empiric combination therapy (antipseudomonal β-lactam or carbapenem plus an aminoglycoside or fluoroquinolone) in patients with gram-negative bacteremia at increased risk for resistant organisms. Inappropriate empiric therapy occurred in 24.6% of patients, with antimicrobial resistance playing a role in 88% of these inappropriate therapies. Patients receiving inappropriate therapy had a significantly higher hospital mortality rate than those receiving appropriate antibiotics (30.7% vs 17.8%, p=0.018). Furthermore, the study revealed that inappropriate initial therapy was significantly more likely to occur among patients receiving monotherapy versus those receiving combination therapy (34.5% vs 20.6%, p=0.011). However, when examining only patients who received appropriate empiric therapy, no significant difference in hospital mortality was shown regardless of whether patients received a single β-lactam, a single aminoglycoside, the combination of a β-lactam and an aminoglycoside, or ciprofloxacin alone. Results were similar for patients receiving appropriate definitive therapy. This study provides strong support that much of the benefit realized from combination therapy is due to improved initial therapy, especially in patients at high risk for death.

Other studies have demonstrated the benefit of appropriate empiric therapy in P. aeruginosa bacteremia.[42,43,49,51] However, similar to the aforementioned studies, none were able to replicate the mortality benefits of combination therapy versus appropriate monotherapy demonstrated in one study.[23]

Ventilator-associated Pneumonia

In contrast to the large body of literature examining combination therapy in P. aeruginosa bacteremia, a paucity of literature exists examining the same issue in patients with VAP. The largest pseudomonal VAP trial, to our knowledge, to examine monotherapy versus combination therapy was a retrospective, multicenter, observational, cohort study conducted in five ICUs with analysis of 183 cases of P. aeruginosa VAP.[37] Similar to the bacteremia studies, rates of appropriate empiric therapy were significantly higher in patients receiving combination therapy versus those receiving monotherapy (90.5% vs 56.7%, p<0.0001). Mortality rates were significantly higher in those receiving inappropriate empiric therapies. Although the trend in patients receiving empiric combination therapy was toward lower mortality rates versus those receiving empiric monotherapy, the results did not reach statistical significance (50.7% vs 37%, p=0.09), possibly due to the limited sample size. Similarly, excluding patients who received inappropriate empiric therapy, no significant mortality difference existed regardless of whether patients received definitive combination therapy or monotherapy. In addition, a recent observational study performed in 84 trauma patients in the ICU who had P. aeruginosa VAP supported the use of monotherapy for these infections.[52] Although this was an observational study with no comparator group, monotherapy achieved a microbiologic cure rate of 94.1% with no documented recurrences.

In a randomized trial comparing combination therapy with monotherapy for the empiric treatment of suspected late-onset VAP in 740 critically ill patients, monotherapy demonstrated similar outcomes compared with combination therapy.[38] However, again, the rate of appropriate empiric therapy was significantly greater in the combination group compared with that in the monotherapy group (93.1% vs 85.1%, p=0.01). Those who received inappropriate empiric therapy were more likely to be infected with MDR organisms, including P. aeruginosa, Acinetobacter species, Stenotrophomonas maltophilia, or other resistant gram-negative rods. In the 56 patients in the MDR subgroup analysis, trends toward greater eradication rates, shorter mechanical ventilator duration and ICU stays, and lower ICU mortalities were noted. However, none of these trends in the subgroup reached statistical significance, probably due to the extremely small numbers of patients in the subgroup analysis. Although this study was not focused on pseudomonal VAP, it suggests empiric combination therapy may be beneficial in critically ill patients with difficult-to-treat organisms, including P.aeruginosa. Other studies have analyzed combination therapy in pseudomonal VAP, but the extremely small numbers from their subgroup analyses make these trials difficult to interpret.[53,54]


Due to the conflicting data about monotherapy versus combination therapy, several meta-analyses have been performed. One meta-analysis examined use of a β-lactam versus lactam plus aminoglycoside in immunocompetent patients with sepsis.[47] Although not specifically examining bacteremia, the investigators showed no advantage of combination therapy in this population for all-cause mortality or treatment failures in the subgroup of patients with P.aeruginosa infections (Table 2 and Table 3).

A second meta-analysis, by the same group of researchers, examined the use of a β-lactam versus a β-lactam plus aminoglycoside, but this time only febrile neutropenic patients were included.[48] Again, subgroup analysis of P.aeruginosa infections found no significant difference between monotherapy and combination therapy in all-cause mortality or treatment failure (Table 2 and Table 3).

Another group of authors analyzed the relationship between combination therapy and reduced mortality rates in patients with gramnegative bacteremia.[55] In contrast to the other meta-analyses, subgroup analysis of P. aeruginosa bacteremia showed a significantly reduced mortality with combination therapy (p=0.007). Unfortunately, the poor quality and heterogeneity of the studies included in these meta-analyses may have impacted the results, making their application to patient management problematic.