Table 4. Univariate and multivariable logistic regression analysis of factors associated with treatment failure over 48 weeks.
| Univariate analysis | Multivariablee analysis | ||||||
|---|---|---|---|---|---|---|---|
| N | Failure (%) | p-value | AOR | 95% CI | p-value | ||
| Age (years) | <35 | 151 | 50.3 | 0.04 | 1.00 | 0.05 | |
| ≥35 | 1212 | 40.1 | 0.68a | 0.46–0.99 | |||
| 35–44 | 458 | 38.0 | |||||
| 45–54 | 532 | 42.7 | |||||
| ≥55 | 222 | 38.3 | |||||
| Ethnicity | White Caucasian | 718 | 40.8 | 0.17 | NI | ||
| Black African | 231 | 42.9 | |||||
| Asian | 18 | 33.3 | |||||
| Other | 29 | 51.7 | |||||
| Nadir CD4 count (cells/mm3) | <50 | 285 | 51.9 | <0.01 | 1.00 | ||
| 50–199 | 469 | 41.4 | 0.70 | 0.50–0.97 | 0.03 | ||
| 200–349 | 290 | 39.7 | 0.73 | 0.51–1.05 | 0.09 | ||
| ≥350 | 148 | 35.1 | 0.48 | 0.30–0.77 | <0.01 | ||
| Baseline CD4 count (cells/mm3) | <50 | 51 | 58.8 | <0.01 | NI | ||
| 50–199 | 236 | 47.9 | |||||
| 200–349 | 257 | 39.3 | |||||
| ≥350 | 782 | 37.7 | |||||
| Baseline HIV-1 RNA (cps/mL) | <50 | 630 | 32.5 | <0.01 | 1.00 | ||
| 50–199 | 147 | 48.3 | 1.82 | 1.01–3.27 | 0.04 | ||
| 200–4,999 | 253 | 51.4 | 2.45 | 1.46–4.11 | <0.01 | ||
| 5.000–99,999 | 189 | 39.7 | 1.80 | 1.02–3.33 | 0.04 | ||
| ≥100,000 | 120 | 54.2 | 2.98 | 1.67–6.03 | <0.01 | ||
| Drugs experienced (n) | <3 | 329 | 36.5 | 0.11 | NI | ||
| 3–6 | 345 | 43.8 | |||||
| 7–10 | 320 | 47.8 | |||||
| ≥10 | 333 | 39.9 | |||||
| ART classes experienced, n | 1 | 455 | 36.9 | 0.01 | NI | ||
| 2–3 | 926 | 43.6 | |||||
| Maraviroc daily dose | ≥300 mg | 167 | 50.9 | <0.01 | 0.72 | 0.49–1.06 | 0.09 |
| <300 mg | 992 | 38.8 | 1.00 | ||||
| Maraviroc schedule | Once daily | 276 | 46.0 | <0.01 | NI | ||
| Twice daily | 661 | 38.6 | |||||
| Context of maraviroc use | Naive | 46 | 41.3 | <0.01 | 0.48 | 0.19–1.20 | 0.12 |
| Switch, suppressed VL | 409 | 32.3 | 1.00 | ||||
| Switch, detectable VL | 441 | 49.0 | 1.21 | 0.68–2.16 | 0.51 | ||
| Intensify, suppressed VL | 50 | 42.0 | 1.52 | 0.79–2.94 | 0.21 | ||
| Intensify, detectable VL | 50 | 56.0 | 1.42 | 0.63–3.23 | 0.40 | ||
| Previous PI experience | Yes | 1218 | 42.9 | 0.04 | 1.62 | 1.05–2.48 | 0.03 |
| No | 163 | 30,7 | 1.00 | ||||
| Viral tropism | R5 | 1061 | 39.1 | <0.01 | 1.00 | ||
| Non-R5 | 165 | 50.3 | 1.65 | 1.20–2.27 | <0.01 | ||
| FPR | ≤10 | 51 | 51.0 | <0.01 | NI | ||
| (10–20] | 68 | 48.9 | |||||
| (20–40] | 156 | 37.3 | |||||
| ≥40 | 330 | 40.0 | |||||
| GSS of accompanying regimen | <1 | 83 | 27.7 | <0.01 | 1.00 | ||
| 1–2 | 254 | 37.8 | 1.33 | 0.64–2.76 | 0.34 | ||
| 2–3 | 228 | 32.9 | 1.08 | 0.52–2.26 | 0.69 | ||
| ≥3 | 96 | 55.2 | 2.47 | 1.07–5.66 | 0.01 | ||
| Year of maraviroc start | ≥2012 | 681 | 60.9 | <0.01 | 3.20 | 2.39–4.30 | <0.01 |
| <2012 | 127 | 37.3 | 1.00 | ||||
Note: Table refers to 1,381 subjects; cases with missing values for each variable are not reported in the table; in 264 cases viral tropism was collected but no FPR was available at the data analysis
NI: Not included in the final model
In the univariate analysis only variables with p-value <0.20 are shown. Additional variables tested but not found to be associated (p ≥0.20) in univariate analysis comprised: gender, previous experience with NRTI, NNRTI, integrase inhibitors or enfuvirtide, and viral subtype.
a(≥ 35 vs <35 years)
All variables shown in the univariate analysis were included in the multivariable model. Variables retained in the final step of the backward elimination procedure (see methods) are shown, all the other were excluded, except the viral tropism result than was manually included. All AOR shown are simultaneously adjusted. The FPR of the geno2pheno[coreceptor] genotypic tropism interpretation was not included in the multivariable model due to co-linearity with tropism result.