Volume 24, Number 2—February 2018
Research
Use of Pristinamycin for Macrolide-Resistant Mycoplasma genitalium Infection
, Jørgen S. Jensen, Christopher K. Fairley, Mieken Grant, Jennifer A. Danielewski, Jenny Su, Gerald L. Murray, Eric P.F. Chow, Karen Worthington, Suzanne M. Garland, Sepehr N. Tabrizi, and Catriona S. Bradshaw
Table 3
Characteristics associated with pristinamycin failure in Mycoplasma genitalium infections, Melbourne Sexual Health Centre, Melbourne, Victoria, Australia, 2012–2016*
| Characteristic | Cured, no. (%) | Failure, no. (%) | Unadjusted OR for failure (95% CI) | p value | Adjusted OR for failure (95% CI) | p value |
|---|---|---|---|---|---|---|
| Symptom | ||||||
| Asymptomatic | 15 (94) | 1 (6) | Reference | Reference | ||
| Symptomatic |
70 (71) |
28 (29) |
6.0 (0.8–47.6) |
0.09 |
4.1 (0.5–35.9) |
0.20 |
| Adherence to treatment | ||||||
| Missed no doses | 44 (72) | 17 (28) | Reference | |||
| Missed any doses |
5 (71) |
2 (29) |
1.0 (0.2–5.9) |
0.97 |
||
| No. antimicrobial drugs before pristinamycin | ||||||
| 0–1 | 62 (73) | 23 (27) | Reference | |||
| >2 |
20 (80) |
5 (20) |
0.7 (0.2–2.0) |
0.48 |
||
| Male sexuality | ||||||
| Men who have sex with men | 52 (80) | 13 (20) | Reference | |||
| Heterosexual |
25 (66) |
13 (34) |
2.1 (0.8–5.1) |
0.11 |
||
| Bacterial load, all samples† |
NA |
NA |
1.9 (1.3–2.9) |
<0.01 |
1.9 (1.2–2.9) |
<0.01 |
| 23S known macrolide resistance mutation‡ | ||||||
| Wild-type | 6 (100) | 0 | Reference | |||
| Mutation at 2058 or 2059 |
36 (67) |
18 (33) |
3.9 (0.52–∞) |
0.17§ |
||
| Excluding wild-type cases | ||||||
| Position 2059 | 22 (73) | 8 (27) | Reference | |||
| Position 2058 |
14 (58) |
10 (42) |
2.0 (0.6–6.2) |
0.25¶ |
||
| 23S G2162T‡ | ||||||
| Absent | 21 (66) | 11 (34) | Reference | |||
| Present |
5 (83) |
1 (17) |
0.38 |
0.41 |
||
| 23S T2185G‡ | ||||||
| Absent | 23 (74) | 8 (26) | ||||
| Present |
3 (43) |
4 (57) |
3.8 (0.70–21.0) |
0.12 |
||
| 23S additional G between positions | ||||||
| 2212 and 2213‡ | ||||||
| Absent | 24 (67) | 12 (33) | ||||
| Present | 2 (100) | 0 | NA | 0.32 | ||
| 23S G2362A‡ | ||||||
| Absent | 23 (72) | 9 (28) | Reference | |||
| Present | 3 (50) | 3 (50) | 2.6 (0.40–15.1) | 0.29 | ||
| L4# A515G N172S | ||||||
| Absent | 17 (68) | 8 (32) | Reference | |||
| Present | 4 (50) | 4 (50) | 2.1 (0.4–10.7) | 0.36 | ||
| L22# C430T Q144 stop codon | ||||||
| Absent | 23 (68) | 11 (32) | Reference | |||
| Present | 2 (50) | 2 (50) | 2.1 (0.26–16.9) | 0.49 | ||
*The L4 and L22 mutations result in amino acid changes. Two ribosomal L4 gene mutations were not analyzed because only 1 case was identified for each mutation C94T (P32S) and G166C (E56Q). NA, not applicable; OR, odds ratio.
†OR for each log10 increase in pretreatment bacterial load in 67 urine samples, 26 rectal swab samples, and 3 cervical swab samples.
‡Escherichia coli numbering. 23S mutations at 2058 and 2059 are established causes of macrolide resistance.
§Exact logistic regression.
¶p value for comparison of proportions cured with mutations at 2058 vs 2059, after exclusion of wild-type cases.
#M. genitalium numbering based on reference genome accession number L43967.2.