TABLE 1.

Overview of changes to moxifloxacin breakpoints

Medium		Recommended moxifloxacin breakpoint (in mg·L −1 ); daily dose (in mg) ¶
		CLSI	WHO
		Since 2011 [37, 43, 44]	2008–2014 [45, 46]	2014–2018 [47]	Since 2018/19 [10, 48]
7H10	CC	0.5+; N/A§	−	0.5; N/Af 2; 400##	0.5¶¶; 400 (standard dose) in longer regimen or 400–800 (high dose) in shorter regimen++
	CBWHO	−	−	−	2; 400–800 (high dose) in longer regimen++
MGIT	CC	0.25+; N/A§	0.25; 400	0.5; N/Af 2; 400##	0.25¶¶; 400 (standard dose) in longer regimen or 400–800 (high dose) in shorter regimen++
	CBWHO	−	−	−	1; 400–800 (high dose) in longer regimen++

Based on World Health Organization (WHO) surveillance data^#, approximately 90% of moxifloxacin-resistant isolates could have been misclassified as susceptible using the BACTEC Mycobacterial Growth Indicator Tube (MGIT) because the WHO critical concentration (CC) of 2 mg·L⁻¹ was eight times higher than the epidemiological cut-off value (ECOFF) between 2014 and 2018. In practice, however, the rate of misclassification was far lower. First, many countries did not use moxifloxacin at all during this period and, consequently, did not use this CC. Second, even countries that prescribed moxifloxacin avoided or minimised the misclassifications because they completely or primarily relied on genotypic antimicrobial susceptibility testing, continued using the MGIT CC of 0.25 mg·L⁻¹ in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines, used a CC for another fluoroquinolone as surrogate for moxifloxacin resistance, or relied on 0.5 mg·L⁻¹ as the breakpoint for the standard dose of moxifloxacin in combination with 2 mg·L⁻¹ as the breakpoint for the high dose of moxifloxacin. The theoretical rate of false-susceptible results was lower using Middlebrook 7H10 medium because the WHO CC of 2 mg·L⁻¹ was four rather than eight times higher than the ECOFF. The clinical breakpoints introduced by WHO (CB^WHO) are not recognised by CLSI. ^#: comparing MGIT results for 2 mg·L⁻¹ moxifloxacin with 2 mg·L⁻¹ ofloxacin in the WHO surveillance study, which is equivalent to testing the currently recognised levofloxacin CC of 1 mg·L⁻¹ (i.e. 1.5 mg·L⁻¹ tested in that study was also too high) [10, 20]. ^¶: changes to WHO breakpoint/dose combinations relative to the previous guidelines are highlighted in bold. Breakpoints that correspond to ECOFFs are underlined [10]. ⁺: can be tested as surrogate for other fluoroquinolones [37, 43–45]. ^§: not applicable (N/A) as CLSI does not define doses for treatment. ^f: 0.5 mg·L⁻¹ moxifloxacin in 7H10 and MGIT were recommended as surrogates for resistance to ofloxacin and levofloxacin. Because the ECOFF for moxifloxacin is 0.25 mg·L⁻¹ in MGIT, this meant that some strains resistant to ofloxacin and levofloxacin were misclassified as susceptible [10]. In effect, the surrogate breakpoints at 0.5 mg·L⁻¹ and moxifloxacin CCs at 2 mg·L⁻¹ were set inconsistently for both media because the 7H10 data was extrapolated to MGIT, despite the systematic differences between both media [47]. ^##: the WHO-endorsed dosage for individualised multidrug-resistant/rifampicin-resistant tuberculosis (MDR/RR-TB) regimens was 400 mg [47]. However, operational research using a higher dosage of moxifloxacin (800 mg) in a standardised short-course MDR/RR-TB regimen was in progress, although not WHO-endorsed at the time [49]. ^¶¶: not recommended as surrogate for other fluoroquinolones [10]. ⁺⁺: levofloxacin is the preferred fluoroquinolone for the shorter all-oral bedaquiline-containing MDR/RR-TB regimen recommended by WHO in 2020, but high-dose moxifloxacin can be used instead. However, any moxifloxacin resistance, irrespective of the level, is an exclusion criterion for the shorter all-oral regimen (i.e. the CC is the relevant breakpoint) [48]. This exclusion criterion for moxifloxacin had also applied to the shorter amikacin-containing MDR/RR-TB regimen that was recommended by WHO between 2018 and 2020 [50]. High-dose moxifloxacin can only be used to treat low-level resistant strains as part of the longer MDR/RR-TB regimen, for which the CB^WHO is valid [48].