Click to edit Master text styles
MVC has a novel target
it targets a host receptor, CCR5 coreceptor, rather than
a viral target.
As such it has unique resistance issues. Resistance to maraviroc can occur in the classical
sense: where the phenotype
changes by virus mutation resulting in viral entry and replication in the presence of
HOWEVER since this drug targets the
CCR5 receptor and will
mechanistically work against only CCR5-tropic virus, the virus can bypass MVC by a tropism
switch of the virus to use the X4 coreceptor or outgrowth of an already existing X4-tropic virus
And There are concerns that X4-tropic virus is more pathogenic
Our analysis examined both changes in tropism and the geno/pheno changes to MVC as
mechanisms of viral resistance
GSS and PSS scores were used to determine susceptibility.
A score of 1 was given for each susceptible drug in OBT. Therefore, the higher the GSS and
PSS score the more
susceptible drugs in OBT
and PSS scores were balanced across the three treatment groups in both studies with Median GSS=1 and PSS=2
of subjects had on overall susceptibility scores <2 meaning they had 2 or less active drugs their in
had only one potentially active drug in OBT and
of subjects had no potentially active drug in their OBT.
with a heavily treatment-experienced population.
2560 subjects were screened; 56% had
About 10% of viruses changed from
R5-tropic to Dual-mixed or
were non-phenotypable in the time period from screening until baseline.
So 90% of the subjects enrolled had
CCR5-tropic virus at Baseline
Throughout these trials, it appears
that the percentage of isolates
which are non-phenotypable using the Monogram tropism assay ranged from 5 to 15%.
Why Did Subjects Fail MVC Treatment in Studies 1027 and 1028?
Reasons for Failure on
MVC could include:
Co-receptor switch from CCR5-tropic to CXCR4-tropic
virus by mutation in the
Outgrowth of CCR5-tropic viruses that are resistant to
Outgrowth of CXCR4-tropic viruses present at baseline
but not detected with
the standard tropism assay (because one of the limitations of the assay is that it is not able to
detect with 100%
sensitivity X4 when present below 10% levels in a viral mixture)
Resistance to OBT
Host CCR5 genotype although not known at this point,
it might be possible
that MVC may not bind efficiently to some host CCR5 receptor genotypes
For resistance analyses, we use as treated analyses. Therefore, subjects are censored from the analysis if they
discontinued with ≤400 copies/mL or
if they discontinued with >400 copies/mL between
Baseline and Week 4 or
if they discontinued between Baseline
and Week 8 with at least
0.5 log decrease and no rebound (previous ≥2 log decrease with 1 log increase).
Forty-nine and thirty-nine subjects were censored from the analysis of studies 1027 and
1028, respectively = total of 88 censored
We first determined the percentage of virologic failures that had CCR5-tropic and CXCR4-tropic
virus at time of failure.
The analysis was done using two definitions of treatment failure
1) the protocol-defined
treatment failure (PDTF) definition and
2) subjects with PDTF plus
subjects with >400 copies/mL at Week 24
Regardless of the definition of treatment failure, more subjects (~50-60%) failed with
CXCR4- or dual/mixed-tropic
virus in the MVC arms, whereas >80% of the subjects in the placebo arm failed
with CCR5-tropic virus.
A high percentage of treatment failure on MVC appears
to be driven by
tropism change from CCR5-tropic to CXCR4- or dual/mixed-tropic virus.
This supports the mechanism of action of MVC and suggests emergence of CXCR4-tropic
virus is a prominent
reason for failure on MVC.
Another reason for treatment failure could be resistance to the other drugs in the OBT.
Most subjects typically had low phenotypic and genotypic susceptibility scores at screening,
indicating reduced susceptibility
to their OBT.
Overall in studies 1027 and 1028, As the number of susceptible drugs in the OBT
increased, reflected in an increased overall susceptibility score, the percent of subjects that achieved <400 copies/mL
increased in the MVC arms (orange and blue) to 70% if 3 or more susceptible drugs in OBT.
In the placebo arm (red), response
rates were <20% with less than 2 active drugs, but increased to 61% when subjects had 3 or more susceptible drugs in OBT.
When examining the OBT in the Subjects who were Treatment Failures:
28% of treatment failure subjects had no susceptible drugs at baseline. There was no difference between the MVC and placebo treatment arms.
When looking at changes in susceptibility to OBT on therapy, 43% of treatment failure subjects
lost susceptibility to drugs in their OBT on treatment and again there was no difference between treatment arms.
Looking specifically at ENF use, ENF use in the failures was comparable between arms at 45%.
About 70% of subjects developed ENF
resistance mutations on
treatment in the MVC QD and placebo arms.
And There were significantly fewer ENF resistance
mutations that developed
in the MVC BID arm than in the placebo arm (p=0.01) or the MVC QD arm (p =0.02).
An examination of overall susceptibility scores of the treatment failures by tropism shows
80% of subjects who failed with
X4-tropic virus had susceptibility
scores of 0-1 compared to only 3% that had 3 or more active drugs in OBT.
50-60% of subjects who failed with R5 or dual/mixed tropic virus had susceptibility scores of
0-1 compared to less than 20% with susceptibility scores of 3 or more.
comprehensive analysis was requested for subjects who had experienced treatment failure
and/or had a changes in their
Given the complexity and exploratory
nature of the tropism and
resistance analyses, substudies from Clinical Studies A4001027 and A4001028 were
A subset of subjects failing with CCR5-tropic virus were analyzed to identify possible
phenotypic and genotypic markers
associated with MVC resistance in vivo.
Including determining MVC susceptibility in cell
Nucleotide sequence of gp120 to identify aa
substitutions that might
contribute to MVC resistance
sequence of protease and RT to identify resistance to drugs in OBT
A subset of subjects failing with X4-tropic virus were
analyzed to ascertain
whether the CXCR4-tropic virus emerged from undetected X4-tropic virus at screening or as a result
of mutations in a CCR5
tropic virus cauing a tropism switch while on MVC: The evaluation included
Clonal evaluation of virus at baseline and on treatment samples to determine the relative
number of CXCR4-tropic and
CCR5-tropic viral isolates
sequence analysis of the gp120 region to identify amino acid changes that may
contribute to a coreceptor switch from R5 to X4
analysis to determine the relationship of emerging CXCR4-tropic virus to the CCR5-tropic virus at baseline (to determine if X4 virus
was distinct from R5 virus at baseline)
In addition, PR and RT were sequenced
to determine resistance
In the virology subgroup analysis the sponsor selected 267 subjects on blinded therapy from
studies 1027 and 1028.
From the pool of 267 subject, there were 38 who failed with CCR5-tropic virus
13 subjects were randomized to
receive MVC (6 BID: 7 QD) and 25 subjects were on placebo.
The genotype and phenotypic susceptibility to MVC of Env
pseudoviruses was analyzed from these 38 subjects
Virus from 2 failure subjects
had 3-fold shifts in MCV susceptibility
at the time of failure.
All other subjects on MVC had
FC <2-fold within the normal range of the Monogram assay.
Viruses from 5 subjects failing MVC treatment with CCR5 tropic virus showed evidence of a
lower plateau in maximum percentage
inhibition rather than fold changes in EC50 values
The results support previous findings
from selection of MVC-resistant
virus in cell culture and isolates from Phase 2 MVC studies that lower plateaus in the
maximum percentage inhibition
(MPI) were associated with subjects failing a MVC-containing regimen rather than
changes in EC50 values.
All 5 subjects had amino acid changes in the V3 loop of gp120 which were present at failure
timepoints but were not present
These 5 subjects also had lower Cmin values <75 ng/mL
The V3 loop sequences reflected the heterogeneity associated with the V3 region of
failure clones had multiple different V3 amino acid changes
However, Changes at either amino
position 13 or 26 were seen
in the V3 loop of gp120 in all five of the subjects with MVC-associated lower plateaus in
role of the V3 Loop Amino Acid Substitutions in MVC Resistance was confirmed by
amino acids in baseline clones to those seen in MVC resistant clones resulted in the MVC resistance phenotype of <95% MPI
Showing that these mutations contributed to MVC
Back mutation of aa changes of failure clones resulted
in MVC sensitive
Not all subjects failing MVC treatment with CCR5-tropic virus had phenotypic markers of
of the subjects receiving MVC did not show phenotypic markers of MVC resistance in the Monogram assay
The majority of these subjects (5/7) had evidence of reduced susceptibility to one or more drugs
within their OBT at screening
and/or at failure.
CCR5 receptor genotype should be
examined from these subjects
to see if MVC can bind to these receptors
A subset of 20 subjects failing with CXCR4-tropic virus were analyzed:
16 were in a MVC treatment group and
4 in the placebo group.
pre-treatment and 48 on-treatment clones from each of the 20 subjects with
X4-Tropic Virus were analyzed for
here is the clonal evaluation of one subject
square is a clone, pink=R5, blue=dual/mixed, green=X4, blank were non-functional clones
proportion of R5 and X4 clones at pretreatment and on treatment
Most of the pretreatment clones were R5 consistent
with this subject being classified as
having R5-tropic virus
Then more of the on-treatment clones are X4 or
dual/mixed consistent with this subject
failing with X4-tropic virus
Each of these clones was sequenced region
sequenced was 290 bases encompassing
the V3 loop of the env gene
Phylogenetic trees were generated using these
sequences in order to investigate
possible ancestry of the CXCR4-using clones - To determine if green X4 clones were related to any of
these pink pretreatment R5 clones
(which would suggest tropism switch) or if they were related to green X4 pretreatment clones (which would suggest outgrowth of this X4-tropic virus)
of clones was confirmed in the validated format of the Trofile assay
14 subjects, CXCR4-tropic clones in the on-treatment samples shared a common ancestor with
a pre-treatment X4-tropic
virus indicating outgrowth of X4-tropic virus
the remaining 6 subjects, CXCR4-using env clones identified
in the on-treatment samples did not have a pretreatment X4 ancestor but they were also genetically distinct from both the
pre-treatment and on-treatment R5 population.
V3 loop sequences of the these on-treatment X4-tropic clones differed by 7-17 amino acid residues from
the V3 loop of the
nearest R5 sequence on the phylogenetic tree.
tropism assay has a sensitivity of 100% for detecting CXCR4-tropic or dual/mixed HIV if present at 10%
or more. If present at <10%, the sensitivity of the assay
goes down (83% at a 5%
Although a tropism switch whereby CXCR4-using virus emerged on MVC resulting from these
7-17 amino acid substitutions
in the CCR5 precursor cannot be ruled out from the review of the data submitted,
From the amino acid sequence
differences and phylogenetic
tree data, the most likely explanation is the CXCR4-using clones in these 6
subjects emerged from a pre-existing CXCR4-using virus not detected by the assay at BL.
So why did
subjects fail on MVC?
Most (~50-60%) subjects failed
with CXCR4- or dual/mixed-tropic
virus in the MVC arms.
The data from the virology
substudies suggest that the most prominent reason for failure in these studies was outgrowth of
CXCR4-using viruses not detected at screening.
Treatment failure on MVC with
CCR5-tropic virus also occurred and resulted from phenotypic and genotypic resistance to MVC and resistance
It remains to be determined if
host CCR5 genotype also plays a role in MVC failure
The evolution to a CXCR4-utilizing HIV has been proposed to result in a more virulent virus.
Therefore, there is
a concern that using MVC will cause outgrowth of CXCR4-tropic virus and result in worse outcomes for patients.
An examination of CD4+ cell counts
by Tropism at Failure
showed that the mean and median change in CD4 cell counts from baseline using
last observation carried forward was lower in subjects with X4 and dual/mixed tropic virus than those with
So we asked for follow-up on the subjects who failed with X4-tropic virus in studies 1027 and
Long term follow-up on the
subjects viral loads, CD4+ cell counts, HIV coreceptor usage and AIDS defining events
subjects failed with CXCR4 virus and were followed by the sponsor, 20 had at least one
2/3 had changed tropism back to CCR5 or dual-mixed and 35% still had X4-tropic virus
For the subjects with R5- or dual/mixed-tropic virus at end of
follow-up, the median
time to last follow-up was approximately 5 months (range 18 days to 8 months).
In contrast, the follow-up time for the subjects who remained X4-tropic at the last follow-up visit
was one month or less (median
time was approximately 11 days).
that over a longer period of follow-up, CCR5 viruses will outgrow X4 viruses in these subjects
In half of the subjects, CD4 cell counts also declined (mean change -21, median change -3)
consistent with the ongoing viremia.
Viral loads remained similar to the value at treatment
No new category C AIDS-defining events were reported for
any of the 20
If subjects went on a new ARV
treatment, viral loads decreased.
CD4 increases were seen concomitant with the reduction in viral load.
Most (50-60%) subjects failed with CXCR4- or dual/mixed-tropic virus in the MVC arms
Most prominent reason for failure in these studies was outgrowth of a minor CXCR4-tropic
virus population not detected
Treatment failure on MVC with CCR5-tropic virus also occurred and resulted from
phenotypic and genotypic resistance to MVC and resistance to OBT
The role of the CCR5 host receptor
genotype is yet to be determined
Lower plateaus in maximum plateau inhibition were detected in viruses from 5 subjects failing
maraviroc and fold changes in EC50 values to maraviroc
were only seen in two subjects
Changes in the V3 sequence of gp160 correlated with the presence of lower plateaus and
There is heterogeneity of the envelope protein and likely multiple pathways to MVC resistance.