Immune Discordance on Highly Active Antiretroviral Therapy Can Still be Regarded as a Therapeutic Success Nur F.

1
Immune Discordance on Highly Active
Antiretroviral Therapy Can Still be Regarded as a
Therapeutic Success Nur F. Önen MD, MRCP1 ,
Rachel Presti MD PhD1, E. Turner Overton MD1 ,
Cecilia Blair1 and Kristin Mondy MD1. Washington
University School of Medicine, St. Louis,
Missouri, USA1.
Contact Nur F. Önen. Washington Univ. School
of Med. Campus Box 8051, 660 S. Euclid Ave. St.
Louis, MO 63110 Phone [phone] Fax
314-361-5231 Email nonen_at_im.wustl.edu
Abstract 952

ABSTRACT
RESULTS
RESULTS
Background Risk factors for immune discordance
on highly active antiretroviral therapy (HAART),
and long-term clinical and immunologic outcomes
remain poorly characterized. Methods
Retrospective analysis of 298 HIV patients with
baseline CD4 T cell count lt350 cells/mm3, who
initiated HAART between January 1996 and July
2006 and had viral suppression (HIV RNA lt400
copies/mL pre-1999, lt50 copies/mL thereafter) for
52 weeks. Discordant and concordant immune
responders were defined by CD4 T cell gains of
lt150 or 150 cells/mm3 from HAART initiation
through 52 weeks of viral suppression. Risk
factors for poor CD4 T cell gains were analyzed
by multiple regression and long-term clinical
outcome assessed. Additionally, markers of immune
maturation and activation were prospectively
determined for immune discordant and concordant
responders (total n45) with sustained viral
suppression from the entire cohort. Results 106
(36) patients had immune discordance. In
multivariable analyses, male gender, greater
number of HAART side-effects, use ever of
unboosted indinavir and lower pre-HAART HIV-RNA
viral load were independent predictors of immune
discordance (plt0.05). Outcomes were similar
between groups with regard to mean time on
suppressive HAART (4.5 years), new opportunistic
infections (1) and mortality (5). Higher levels
of memory (p0.04) and activated (p0.08) CD4 T
cells were found among those with immune
discordance, up to a mean of 6.3 years after
viral suppression. Conclusions HIV patients
with long-term immune discordance can still
achieve very low morbidity and mortality,
suggesting immune benefits of HAART may go beyond
gains in CD4 T cells alone. In addition, T-cell
activation may blunt long-term CD4 T cell gains.
Interventions to reduce T cell activation may
facilitate further CD4 T cell recovery in this
patient group.
Table 1. Characteristics of immune discordant vs.
concordant responders.
Figure 1. Absolute CD4 T cell counts after
HAART initiation.
Table 3. Comparison of activation and memory
T-cells between discordant vs. concordant immune
responders.
Characteristics Discordant (n106) Concordant (n192) OR (95 CI) P value
Male gender Age, (years)a Caucasian African American 89 (84) 41.2 0.9 58 (55) 38 (36) 134 (70) 38.2 0.7 75 (39) 108 (56) 2.27(1.24-4.15) - 1.88(1.80-3.05) - 0.01 0.01 0.01 -
Mode of transmission Men who have sex with men Heterosexual 56 (53) 30 (28) 79 (41) 96 (50) 1.78(1.08-2.93) - 0.02 -
Years since HIV diagnosisa Previous OI Prior ARV experience 9.8 0.5 60 (57) 34 (32) 8.3 0.3 104 (54) 62 (32) - 1.11(0.69-1.78) 1.01(0.61-1.68) 0.01 0.72 1.00
CD4 T cell nadir b Highest HIV RNA viral loada 85 (21-180) 4.90 0.06 54(12-189) 5.11 0.07 - - 0.43 0.05
Chronic Hepatitis C Chronic Hepatitis B 17 (11) 9 (6) 16 (8) 21 (11) 1.97(0.96-4.04) 0.72(0.32-1.62) 0.09 0.55
At suppressive HAART initiation CD4 T cell count b HIV RNA level a HAART regimen type NNRTI/ 2 NRTI All PI/ 2 NRTI Unboosted indinavir 126 (37-231) 4.68 0.86 45 (42) 58 (55) 27 (26) 88 (19-222) 4.81 0.06 92 (48) 87 (45) 24 (13) - - 0.86(0.52-1.42) - 0.52(0.24-1.13) 0.15 0.19 0.61 - 0.01
No. () with HAART-related side-effects 36 (34) 35 (18) 1.80(1.09-2.96) lt0.001
Weeks to viral suppression b 12 (6-23) 16 (8-28) - 0.04
One year CD4 T cell gain b 99 (50-133) 269 (211-374) - lt0.001
Data at the time of FACS analysis (mean S.E.M.) Discordant (n20) Concordant (n25) P value
Age Gender Male Female Race Caucasian/Hispanic African American 48.7 2.5 17 (85) 3 (15) 15 (75) 5 (25) 48.7 1.9 19 (76) 6 (24) 10 (40) 15 (60) 0.86 0.71 0.03
CD4 T cell count (cells/mm3) lt200 cells/mm3 201-349 cells/mm3 350 cells/mm3 306 30 20 50 30 683 49 0 0 100 lt0.001
Years of viral suppression 5.6 0.5 7.0 0.4 0.03
CD4 T cellsa CD45RO Naïvememory CD38HLADR 55.0 3.6 1.1 0.3 2.9 0.4 45.1 3.1 1.7 0.3 2.0 0.2 0.04 0.04 0.08
CD8 T cellsa CD45RO Naïvememory CD38HLADR 20.5 2.7 5.2 0.6 1.0 0.2 17.6 1.8 6.4 0.8 1.1 0.1 0.47 0.47 0.38
Table 2. Clinical outcomes during long-term
follow up.
BACKGROUND
aPercentage of total events. CD38HLADR marker
of activation, CD45RO marker of memory cell.
Clinical Outcomes Discordant (106) Concordant (192) OR (95 CI) P value
Years of viral suppression on HAARTa Highest CD4 T cell count attainedb 4.56 0.2 392 (256-566) 4.51 0.2 682 (479-879) - - 0.88 lt0.001
Death Remains in clinical care Lost to follow up 5 (5) 75 (71) 26 (25) 9 (5) 150 (78) 33 (17) 1.00(0.33-3.09) - - 1.00
New opportunistic illness Type of opportunistic illness Years after viral suppression OI prophylaxis at 1 yr viral suppression at highest CD4 T cell count 1 NHL 2 and 5 61 (58) 37 (35) 1 Kaposis sarcoma 8 73 (38) 30 (16) - - - 2.21(1.36-3.58) 2.90(1.66-5.01) - - - lt0.001 lt0.001
Recurrent genital warts Recurrent shingles Recurrent genital herpes 5 (5) 10 (9) 0 1 (1) 7 (4) 6 (3) 9.46(1.09-82.0) 2.75(1.02-7.46) 1.57(1.44-1.71) 0.02 0.06 0.09

• In patients with viral suppression for a mean of
  6.3 years, discordant responders had
  significantly higher levels of memory CD4 T
  cells (CD45RO) with a reduced naïvememory CD4
  ratio (plt0.05 for all).
• Levels of activated CD4 T cells (CD38HLADR)
  were higher among those with immune discordance,
  although this did not reach statistical
  significance.
• Levels of memory and activated CD8 T cells were
  similar between groups.

• Suboptimal immune reconstitution on fully
  suppressive highly active antiretroviral therapy
  (HAART) is frequently observed in clinical trials
  and cohort studies.
• Recent studies suggest that markers of immune
  maturation and activation may predict CD4 T cell
  recovery in patients on suppressive HAART.
• The long-term outcomes of immune discordance on
  HAART in terms of overall morbidity and mortality
  remains poorly studied.

METHODS
CONCLUSION

• Retrospective cohort study of patients initiated
  on HAART between January 1996 and July 2006.
• Inclusion criteria
• HIV-1 infection, age 18 years, baseline CD4 T
  cells lt350 cells/mm3 and viral suppression for
  52 weeks.
• Definitions
• Discordant and concordant responders defined by
  CD4 T cell gains of lt150 or 150 cells/mm3 from
  HAART initiation through 52 weeks of viral
  suppression and followed to death, virologic
  failure (lack of re-suppression of viremia) or
  study termination.
• Data collection
• Socio-demographics, clinical, medication and
  laboratory data.
• Sample collection and analysis
• Blood samples from immune discordant and
  concordant responders (total n45) with sustained
  viral suppression, were obtained to determine the
  percentage of CD4 and CD8 T lymphocytes that
  were memory (CD45RO) or activated
  (CD38,HLADR), using three-color flow-cytometric
  analysis on cryopreserved peripheral mononuclear
  cells.
• Statistics
• Data analyzed using SPSS version 14.0. Continuous
  variables compared using the Students t-test or
  Mann-Whitney U test. Chi square or Fishers
  exact tests used for categorical variables.
• All p values were two-tailed and significant at
  lt0.05 and potential predictive factors for immune
  discordance were evaluated using multivariate
  logistic regression analyses.

• In this cohort with long-term viral suppression,
  immune discordance was high but morbidity and
  mortality remained very low.
• Our findings suggest immune benefits of HAART may
  go beyond CD4 T cell gains.
• T cell-activation may blunt long-term CD4 T cell
  gains and interventions to reduce T-cell
  activation may facilitate further CD4 T cell
  recovery in patients with sub-optimal immune
  reconstitution.
• Low pre-HAART viral load was associated with
  immune discordance and may have implications for
  the use of viral load in future treatment
  guidelines.

aMean standard error of mean, bmedian
(interquartile range), ARV antiretroviral,
NNRTI non-nucleoside reverse transcriptase
inhibitor, NRTI nucleoside reverse
transcriptase inhibitor, PI protease inhibitor.

• Independent factors associated with immune
  discordance on multivariable analyses
• Male gender.
• Lower pre-HAART HIV-RNA viral load.
• Any use of unboosted indinavir.
• Greater number of HAART-related side effects per
  person.

We would like to acknowledge Bristol-Myers Squibb
for the Virology Fellows Research Grant which
enabled this study to be done.
aMean standard error of mean, bMedian and
interquartile range. NHL non-Hodgkins
lymphoma, OI opportunistic infection.