title: cd4/cd8 recovery and first-line art: greatest ... · andmortality(serrano-villar, plospathog...

Title:

BACKGROUND• A low CD4/CD8 ratio during antiretroviral therapy (ART) identifies

subjects with heightened immunoactivation andimmunosenescence, and have been shown to predict morbidityand mortality (Serrano-Villar, PLoS Pathog 2014)

• It is unknown whether the different first-line ART regimens havea different impact on the rates of CD4/CD8 ratio normalization.

• We aimed to assess the effects of the INSTI, PI or NNRTI-basedfirst-line ART on long-term CD4/CD8 ratio recovery in a largeprospective cohort.

METHODS• Prospective cohort study in 13,026 HIV-infected individuals

registered in the Spanish HIV Research Network (CoRIS) cohort.

• We included subjects who started triple ART and achieved HIVRNA suppression at 48 weeks.

• We used multilevel mixed models with linear splines to comparelongitudinal changes in the CD4/CD8 ratio and Cox proportional-hazard models to compare the times to CD4/CD8 normalizationby treatment groups (NNRTI, PI, INSTI) at 0.5, 1 and 1.5 cut-offs.

• Analyses were adjusted for sex, country of origin, mode oftransmission, calendar year, educational level, baseline HIV RNA,presence of AIDS, pre-ART nadir CD4, acme CD8 count andbackbone NRTI.

RESULTS• A total of 6,804 individuals contributing to 37,149 persons/years

and 37,680 observations met the inclusion criteria. The medianfollow-up was 49 months (IQR 22-89).

• The study sample was representative of a medium-agedpopulation (median age 36 [IQR 30-44] years) with higherrepresentation of men (85.3%), median nadir CD4+ T-cell count of304 (IQR 178-438) cells/uL, and median baseline CD4/CD8 ratio0.33 (IQR 0.19-0.52).

• The median time from ART initiation to virologic suppression was18 (IQR 10-29) months, and first-line regimens included 2,820subjects starting ART with 2 NRTI+NNRTI (41.5%), 1,574 (23.1%)with 2 NRTI+PI and 2,410 (35.4%) with 2 NRTI+INSTI.

INSTI-based first line ART is associated with a greater CD4/CD8 ratio gain compared to NNRTI and PI-based ART.

Sergio Serrano-Villar1, Javier MarMnez-Sanz1, Raquel Ron1, Alba Talavera2, Borja Fernández1, Francisco Fanjul3, Joaquín Porclla4, Josefa Muñoz5, Concha Amador6, Miguel Alberto de Zárraga7, Maclde Sanchez-Conde1, Sabina Herrera1, Pilar Vizcarra1, María J. Vivancos-Gallego1, Sancago Moreno1

1Hospital Ramón y Cajal, Madrid, Spain; 2Insctuto Ramón y Cajal de Invescgación Sanitaria, Madrid, Spain; 3Hospital Universitario de Son Espases, Palma de Mallorca, Spain;4Hospital General Universitario de Alicante, Alicante, Spain; 5Hospital de Basurto, Basurto, Spain; 6Hospital de la Marina Baixa, Villajoyosa, Spain; 7Hospital San Agusln, Avilés, Spain.

CD4/CD8 Recovery And First-line ART: Greatest Improvement With Integrase Inhibitors 0474

CONCLUSIONS• This study with prospective design, large sample size, and

long follow-up shows that first line ART is associated witha greater CD4/CD8 ratio gain compared to NNRTI and PI-based ART. The INSTI-based ART was associated withhigher rates of CD4/CD8 recovery at the threshold fornormalization.

• Our findings were independent of the immunovirologicalcovariates analyzed (i.e., CD4 nadir, maximum HIV RNA,and development of virological failure during ART).

• This study in real life indicates that ART initiation withINSTI improves immune recovery with respect to otherART classes, which could affect long-term mortality.

FundingThis work was supported by the Instituto de Salud Carlos III (Plan Estatal de I+D+i 2013–2016,projects PI15/00345, PI18/00154, AC17/00019, the Spanish AIDS Research NetworkRD16/0025/0001project) and was co-financed by the European Development Regional Fund ‘‘A wayto achieve Europe’’ (ERDF).

AcknowledgmentsWe acknowledge all study participants who made this research possible.

0.2

5.5

.75

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15Years from ART initiation

395 194 96 37 19 9 5 0 0 0 0 0 0 0 0ART = 2NRTI+1II827 644 486 385 311 239 178 125 93 69 36 14 1 0 0ART = 2NRTI+1IP

1204 916 735 583 477 369 271 177 123 84 39 19 8 4 0ART = 2NRTI+1NNRTINumber at risk

95% CI 95% CI 95% CI

ART = 2NRTI+1NNRTI ART = 2NRTI+1IP ART = 2NRTI+1II

Kaplan-Meier failure estimates for CD4/CD8 ratio normalizationn at cut-off ≥0.4

0.2

5.5

.75

1

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15Years from ART initiation

1234 657 333 138 70 37 24 10 3 0 0 0 0 0 0ART = 2NRTI+1II1477 1277 1115 948 771 629 503 391 296 211 126 71 21 6 0ART = 2NRTI+1IP2656 2248 1899 1548 1236 982 752 567 420 291 173 98 48 9 0ART = 2NRTI+1NNRTI

Number at risk

95% CI 95% CI 95% CI


Kaplan-Meier failure estimates for CD4/CD8 ratio normalization at cut-off ≥1

0.2

5.5

.75

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15Years from ART initiation

1556 926 475 194 106 57 35 14 5 1 0 0 0 0 0ART = 2NRTI+1II1626 1486 1344 1170 980 835 689 552 424 308 193 116 48 13 0ART = 2NRTI+1IP3028 2795 2488 2098 1696 1370 1096 824 618 414 254 147 67 14 0ART = 2NRTI+1NNRTI

Number at risk

95% CI 95% CI 95% CI


Kaplan-Meier failure estimates for CD4/CD8 ratio normalization at cut-off ≥1.5

a

b

c

Figure 2: Subanalysis by INSTI group

Sub-analyses by piecewise indicatedthat these differences were drivenby changes during the first year ofART (Table 1) without significantdifferences in the adjusted CD4/CD8raco trajectories ater the secondyear of ART. Hence, a stronger effectof INSTI during the first year of ARTdetermined greater CD4/CD8 racovalues ater a similar follow-up.

Figure 1: Subjects in the INSTI group experience greater CD4/CD8gain during follow-up than patients in the NNRTI and PI groups.

Figure 3: Kaplan-Meier survival plots of estimated overallCD4/CD8 normalization at cut-offs of 0.4, 1.0, and 1.5.

• No differences between treatment groups for cmes to CD4/CD8 raco normalizacon at 0.4 cut-offwere detected. However, the NNRTI and PI had lower rates of CD4/CD8 raco normalizacon at ≥1 and≥1.5 cut-offs than INSTI (Table 2)

• At year 4, the adjusted mean CD4 count for INSTI, NNRTI and PI was 904, 718 and 696 cells/uL,(p<0.0001), and the adjusted mean CD8 count was 832, 875 and 996 cells/ul, respeccvely (p<0.0001).

Table 3 shows the incidence rates of raconormalizacon by treatment group for each cut-off point.

Subanalyses adjusted for backbone NRTIs did notshow differences in the rates of CD4/CD8 raconormalizacon between raltegravir, dolutegravirand elvitegravir (Table 4).

0.2

5.5

.75

1

1 2 3 4 5 6 7 8Years from ART initiation

114 40 19 2 2 0 0 0Elvitegravir13 6 2 0 0 0 0 0Raltegravir150 55 25 16 8 5 4 0Dolutegravir

Number at risk

95% CI 95% CI 95% CI

Dolutegravir Raltegravir Elvitegravir

Kaplan-Meier failure estimates for CD4/CD8 ratio normalization at cut-off ≥1

a

b

c

0.2

5.5

.75

1



Number at risk

95% CI 95% CI 95% CI


Kaplan-Meier failure estimates for CD4/CD8 ratio normalization at cut-off ≥0.4

0.2

5.5

.75

1



Number at risk

95% CI 95% CI 95% CI


Kaplan-Meier failure estimates for CD4/CD8 normalization at cut-off ≥1.5

Corresponding Author

Javier Martínez Sanz, MD, PhD

Department of Infectious Diseases, Hospital Universitario Ramón y Cajal

M-607 Km. 9.100, 28034 Madrid (Spain)

Email: [email protected]

The observed and adjusted linear trajectories ofCD4/CD8 raco per treatment group significantly differedbetween groups (p<0.001). The differences were drivenby changes in both CD4+ and CD8+ T-cell counts.

Within INSTI, the greatest CD4/CD8 raco gain was observed with EVGfollowed by DTG. This was largely driven by greater declines in CD8+T-cell counts during the first year of therapy (p<0.001).

Table 1

Table 2

Table 4

Table 3

Other sensicvity analyses included the CD4 nadir stratum, thecalendar year period, and the development of virological failureduring follow-up. There were no significant variacons with respect tothe global primary analysis.

title: cd4/cd8 recovery and first-line art: greatest ... · andmortality(serrano-villar, plospathog...

Documents