Among the 150 culture isolates, 11 isolates failed in the MGIT 960 PZA susceptibility test due to under inoculums size (3 isolates for both standard and reduced inoculums; 8 isolates for reduced inoculums only). The failure rate for standard and reduced inoculums was 1.8% (3/171) and 6.4% (11/171) respectively. Only 139 isolates were eligible for further study and analysis. Another 21 archived PZA resistant strains collected between 2003-2013 were also included for comparative study. However, one strain was later confirmed to be PZA susceptible that 140 PZA susceptible and 20 PZA resistant strains were used for the McNemar’s analysis, indicating there is no significant difference between standard and reduced inoculums for MGIT960 PZA susceptibility test (p = 0.6171; Odd Ratio = 1; 95% CI: 0.072 – 13.796).

The 20 PZA resistant archived strains collected between 2003-2013 were further characterized by MGIT960 PZA (standard and reduced inoculums) test, PZase activity and DNA sequencing to identify pncA mutations. The most common PZA resistant mutations at pncA were G162D and I90S. Novel mutations including insertions and deletions were also detected among PZA resistant strains. PZase activity were negative for all PZA resistant strains while 20 randomly selected PZA susceptible strains were positive to PZase activity with no mutations in the pncA gene. One strain (WC036) reported as PZA susceptible by TB Reference laboratory (PHLSB) was confirmed resistant to PZA by MGIT960 (both standard and reduced inoculum) with pncA mutations and negative for PZase activity. Another strain (WC205) reported as susceptible to all first line drugs but resistant to PZA was also confirmed susceptible to PZA by MGIT960 (both standard and reduced inoculum) with wild type pncA and positive for PZase activity.

In this study, testing 160 M. tuberculosis isolates indicated no significant difference between standard and reduced inoculums for MGIT960 PZA susceptibility test (p >0.05). The failure rate for standard and reduced inoculums was 1.8% and 6.4% respectively. The 1 false susceptible and 1 false resistant isolates reported by TB reference Laboratory (PHLSB) was probably accounted by the drug susceptibility test other than MGIT960. Since 2014, PHLSB has adopted the MGIT960 system with standard inoculums size for routine service to all Hospital Authority Hospitals in Hong Kong. For routine PZA MGIT 960 susceptibility testing using standard inoculums, clinical isolates reported as resistant to PZA should be confirmed by PZase activity and pncA gene mutation analysis using PCR-sequencing.