FDA has completed their review of the rationale document titled, “Polymyxin Breakpoints for Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter spp. – CLSI Rationale Document MR01-Ed2, April 2020,” submitted by the Clinical and Laboratory Standards Institute (CLSI) to a public docket FDA-2017-N-5925-0012 in May 2020.

Of note, FDA does not currently recognize susceptibility interpretative criteria (STIC) for Enterobacterales, Pseudomonas aeruginosa, or Acinetobacter spp. for colistimethate for injection (henceforth referred to as colistin) or polymyxin B.

CLSI’s rationale document includes the following revisions to the breakpoints (BPs) for colistin and polymyxin B that were included in M100 standard (30th edition, January 2020):

• For Pseudomonas aeruginosa, the intermediate and resistant BPs are revised to ≤ 2 mg/L and ≥ 4 mg/L, respectively. No susceptible BP is proposed.
• For Acinetobacter spp., the intermediate and resistant BPs are revised to ≤ 2 mg/L and ≥ 4 mg/L, respectively. No susceptible BP is proposed.
• For Enterobacterales, an intermediate and resistant BPs of ≤ 2 mg/L and ≥ 4 mg/L, respectively, are added. No susceptible BPs are proposed.

FDA reviewed the submitted pharmacokinetic, pharmacodynamic, microbiology, and clinical information and determined that no change is recommended regarding the STIC for colistin or polymyxin B. This is based on the following:

• The pharmacokinetic and pharmacodynamic (PK/PD) data presented are limited and inconclusive to support the proposed BPs for either colistin or polymyxin B. For colistin, the target attainment analyses did not utilize traditional target attainment estimation with a population PK model and Monte Carlo simulation. Rather, individual patients’ average steady-state concentrations (Css,avg) were calculated with the recommended doses in various renal impairment categories.
• For polymyxin B, the murine infection model data used to support the proposed BPs for Enterobacterales included only three Klebsiella pneumoniae strains with a wide range in the ratio of the area under the unbound concentration–time curve to the MIC (fAUC/MIC) targets for stasis and 1-log kill. Target attainment analyses for polymyxin B are not included in the rationale. The animal infection model data for colistin against P. aeruginosa and A. baumannii cannot be accurately extrapolated to polymyxin B given the differences in bacterial strains and the limited nature of the data provided.
• The clinical data provided in the rationale do not allow determining colistin or polymyxin B breakpoints for Enterobacterales, A. baumannii, or P. aeruginosa because no assessment of clinical outcomes in patients treated with polymyxins based on MIC can be made. Overall, the provided data question whether an MIC ≤2 mg/L can be considered an intermediate breakpoint given the low likelihood of treatment success and do not allow defining susceptible and resistant categories for either colistin or polymyxin for any of the listed pathogens. Please see the Appendix for additional details.

Therefore, based on the information discussed above, the Agency does not recognize the CLSI breakpoints for Enterobacterales, Pseudomonas aeruginosa, or Acinetobacter spp. for colistin or polymyxin B.

APPENDIX

The rationale included four studies evaluating polymyxins in the treatment of infections caused by carbapenem-resistant gram-negative bacteria. ^1,2 Colistin breakpoints were defined in two out of the four cited publications. The study by Paul M et al. (2018) was the largest of the studies included in the rationale.¹ In this study a total of 406 patients with bacteremia, ventilator-associated pneumonia, hospital-acquired pneumonia, or urosepsis caused by carbapenem-resistant colistin-susceptible gram-negative bacteria were randomly assigned (1:1) to colistin or colistin plus meropenem. Susceptibility to colistin was defined as MIC ≤2 mg/L for A. baumannii and Enterobacteriaceae and ≤4 mg/L for P. aeruginosa. Both treatment arms performed equally poorly with the clinical failure rates at 14 days post-randomization of 79% (156/198) in the colistin versus 73% (152/208) in the colistin-meropenem arm. The 28-day mortality rate was 43% (86/198) in the colistin and 45% (94/208) in the colistin plus meropenem arm.

The other 3 studies enrolled patients with Enterobacterales infections. One of these studies defined colistin resistance as MIC >2 µg/ml.² In this study, colistin (n=20) in combination with either meropenem or tigecycline was compared with plazomicin (n=17). The supplemental materials for the publication noted that 40.9% of the isolates from patients in the colistin group were colistin resistant and that “poor outcomes for colistin-treated patients infected with colistin-susceptible or colistin-resistant pathogens underscore the limitations of this last-resort antibiotic.” Otherwise, the analyses of clinical outcomes by colistin MIC were not reported. The remaining two studies used colistin primarily in combination with other antibacterials and did not report efficacy results by colistin MIC. ^3,4 For polymyxin B, the provided clinical data are limited to one patient with K. pneumoniae infection with no MIC reported.⁴

---

¹ Paul, M., et al., Colistin alone versus colistin plus meropenem for treatment of severe infections caused by carbapenem-resistant Gram-negative bacteria: an open-label, randomised controlled trial. Lancet Infect Dis, 2018. 18(4): p. 391-400.
² McKinnell, J.A., et al., Plazomicin for Infections Caused by Carbapenem-Resistant Enterobacteriaceae. N Engl J Med, 2019. 380(8): p. 791-793.
³ van Duin, D., et al., Colistin Versus Ceftazidime-Avibactam in the Treatment of Infections Due to Carbapenem-Resistant Enterobacteriaceae. Clin Infect Dis, 2018. 66(2): p. 163-171.
⁴ Wunderink, R.G., et al., Effect and Safety of Meropenem-Vaborbactam versus Best-Available Therapy in Patients with Carbapenem-Resistant Enterobacteriaceae Infections: The TANGO II Randomized Clinical Trial. Infect Dis Ther, 2018. 7(4): p. 439-455. Supplementary material.