Use of a genotypic assay for the prediction of HIV-1 co-receptor tropism and guiding the use of CCR5 antagonists in clinical practice Malcolm Macartney 1 , Jane Cameron 1 , Angela Strang 1 , Ana Garcia 1,2 , Clare Booth 1 , Neal Marshall 1 , Margaret Johnson 1 and Anna Maria Geretti 1,2 . 1 Royal Free Hospital and 2 Royal Free & University College Medical School, London, UK.
Background • Co-receptor tropism testing is routinely performed prior to the use of Maraviroc (MVC). • Trofile, and later Enhanced Trofile (ESTA), use patient gp120 to determine entry into cells, but: – technically challenging, time consuming; – snapshot of patient at time of sampling, no clinical weighting. • Genotypic tropism testing (GTT),such as g2p, uses the V3 sequence and can also incorporate clinical parameters to determine tropism. • We have previously reported a significant positive correlation between ESTA GTT by g2p. • Recent retrospective analyses of the MOTIVATE and MERIT clinical trials provide clinical support to GTT. Prospective studies of patients treated with MVC based on GTT are limited.
Objective • Determine the co-receptor tropism of patients with either drug-toxicity issues or therapy failure by GTT and prospectively follow patients classed as R5 or Mixed who were started on a MVC-containing regimen to monitor their clinical outcome including: – Viral Load (VL) change, CD4 count change.
Methods • Study population: we introduced GTT as a routine diagnostic service in April 2009. For this analysis we selected patients who started MVC after April 2009 to ensure a consistent testing approach across the cohort. • Specimen selection: – plasma if VL >500 cps/ml; – proviral DNA from PBMCs if VL <500 copies/ml, including <50 cps/ml. • In-house nested PCR and sequencing on 3 replicates/sample. Sequences were processed using the g2p algorithm. • Interpretation: – clonal model at FPR 6%; – Clinical model at FPR 15% was also used if information on the CD4 count, CD4 nadir and current VL was available and reliable. • Results reported as R5, X4 or Mixed. Mixed results included g2p interpretation prediction of R5 by clonal model and X4 by clinical model, and replicate samples or sequences showing mixtures of R5 & X4 by clonal model.
Patient parameters T r e atme nt F ailur e s T oxic ity Numbe r 13 15 Me dian CD4 Nadir (R ange ) 186 (5 – 559) 253 (53 – 617) Me dian CD4 at switc h (R ange ) 477 (44 – 1056) 523 (214 – 1223) Subtype (B/ Non-B*) 8/ 5 7/ 8 Me dian VL at switc h log 10 2.26 (1.69 – 4.72) 1.69 (1.69 – 4.42) (R ange ) Sample type (PBMC/ Plasma) 10/ 3 12/ 3 *Non-B subtypes: A (2), C (4), G (1), CRF_02 (3), cpx (3). Subtypes based on pol sequences where available. For calculations, VL <50cp/ml were given a value of 49 (log 10 1.69). In the Treatment Failures group 2 patients started on MVC showed a Mixed result; 1 patient with a current R5 result had previously shown a Mixed.
Results T r e atme nt F ailur e s T oxic ity Numbe r 13 15 Me dian CD4 at switc h 477 (44 – 1056) 523 (214 – 1223) (R ange ) Me dian CD4 at follow-up 375 (221 – 894) 600 (278 – 1100) (R ange ) L og 10 me dian VL at switc h 2.26 (1.69 – 4.72) 1.69 (1.69 – 4.42) (R ange ) L og 10 me dian VL at follow-up 1.69 (1.69 – 3.43) 1.69 (1.69 – 2.44) (R ange ) Me dian MVC follow-up 19 we e ks (4 – 35) 25 we e ks (4 – 42) (R ange ) Follow-up period is defined as the period between the switch to a regimen containing MVC and the date of the last VL test while still on MVC.
Toxicity switches (1) • Overall, 15 patients replaced one or more ARVs with MVC due to toxicity, 12 based upon a proviral DNA GTT. • 11 of the 15 patients switched therapy with a VL <50cps/ml, and the remaining 4 patients changed HAART before achieving VL suppression. • Median number of active drugs, including MVC in all patients, was 3 (range 2 – 5); 11 of 16 patients (73%) received MVC without a PI/r. • At follow-up, 14 of 15 (93%) patients either maintained or achieved a VL <50 cps/ml. – no patient who switched with a VL <50 cps/ml rebounded to >50 cps/ml. – 1 patient has not yet suppressed but has achieved a 2.5 log 10 VL drop 4 weeks after switching Truvada/EFV for Truvada/MVC.
Toxicity Switches (2) In the 15 patients switched due to toxicity, MVC replaced the following drugs; AT V/ Ralte gravir, 1 T ruva da , 2 F P V/ r, 1 DRV/ r, 1 E F V, 2 SQV/ r, 2 T DF , 2 L P V/ r, 2 T 20, 1 AT V/ r, 1
Treatment failures (1) Of the 13 patients with treatment failure who received a new regimen containing MVC , 7 (54%) showed a VL <50 cps/ml at follow-up. The patients who achieved <50 cps/ml had a similar median number of active drugs in their regimen compared to patients who failed to achieve <50cps/ml. F ailure gr oup <50c p/ ml F ailure gr oup >50c p/ ml Me dian numbe r of ac tive dr ugs (R ange ) 2.75 (2 – 5) 2.50 (0.5 – 4) Pa tie nt L og 10 VL at L og 10 VL at follow- MVC tr e atme nt Ac tive dr ugs g2p r e sult switc h up time (We e ks) in r e gime n 2 3.24 2.18 35 3 CCR 5 5 4.64 3.14 18 1.5 Mixe d 23 2.47 2.58 27 4 CCR 5 24 4.72 3.5 14 0.5 Mixe d 25 2.04 2.21 8 4 CCR 5* 26 2.03 2.07 9 2.5 CCR 5 *Patient 25 had a previous tropism sample interpreted as Mixed. Active drug scores include MVC
Treatment failures (2) • Of the 6 patients who had not achieved a VL of <50cps/ml at the end of follow-up: – 2 patients had a Mixed tropism result at testing and received either 1 (pt 5) or 0 (pt 24) other active drugs in their background regimen; – Of the 4 remaining patients: • 3 had dual-class resistance (pt 2 NRTI & NNRTI; pt 25 NRTI & NNRTI) or triple-class (pt26) resistance; • the 4 th patient has had a low-level VL, between 50 and 300 cps/ml in the absence of detectable drug resistance, for over 2 years.
Sequence data V3 Sequencing data for the 2 Mixed patients, both subtype B. Ref. Sequence C T R P N N N T R K S I R I G P G Q - - A F Y A T G D I I G D I R Q A H C C T R P N N N T R K S I X1 I G P G R/S - - A F Y A T G D I I G D I R Q/E A X2 C Prediction Clonal FPR 56%, Clinical FPR 91% X1 is a potential mixture of H, Y, P & S (codon is ymt). X2 is a potential mixture of H, P, Y & S (codon is ymt). C T R P N N H T R K R I H I G T G R Y S K V Y A T E D I I G D V K Q A H C Prediction Clonal FPR 0.2%, Clinical FPR 0.1% Reference Sequence C T R P N N N T R K S I R I G P G Q A F Y A T G D I I G D I R Q A H C CCR5 C T R P N N N T R K G I N M G P G R A F Y A T T D I V G D I R Q A H – Prediction Clonal FPR 32%, Clinical FPR 20% CXCR4 C T R P N N N T R K G I T M G P G R A F Y A T T D I V G D I R Q A H – prediction Clonal FPR 13%, Clinical FPR 14%
Conclusions • From this limited patient data set it can be seen that a g2p R5 interpretation is highly predictive of successful use of MVC in treating HIV-1, with 21/26 patients (81%) maintaining or achieving and maintaining <50 cps/ml during the follow-up period. • The use of proviral DNA for GTT appears to be reliable in guiding treatment switches in suppressed patients with toxicity. • Both patients with a Mixed result achieved a reduction in VL (1.5 and 1.22 log 10 ) while on MVC with 1 or 0 other active drugs in their background regimen. This may suggest that MVC may achieve a degree of virological suppression in some patients classified as Mixed. • While a longer follow-up period and larger numbers are desirable, these findings provide encouraging evidence of GTT for guiding the use of CCR5 antagonists in clinical practice.
Our decision criteria Sequence into geno2pheno Minimum clinical data set No / insufficient / unreliable clinical details (VL, nadir CD4 count and %) If clonal result at If clonal result at VL>1000 cp/ml: VL<1000 cp/ml, unknown FPR 6% or FPR 6% or lower is Use clonal interpretation or <50 cp/ml: higher is R5…… X4: result is X4 only at FPR 6% Use clonal interpretation only at FPR 15% ..and clinical at FRP ..and clinical at FRP 15% or lower is X4 15% or higher is R5: result is R5 but with a clinical comment : result is R5 Please note that the clinical characteristics of this patient suggest that the presence of low frequency or archived X4 virus cannot be excluded. A CCR5 antagonist should be used in the context of a regimen with good supportive activity.
Recommend
More recommend
