Prediction of Liver Disease,AIDS,and Mortality Based on Discordant Absolute and Relative Peripheral CD4 T Lymphocytes in HIV/Hepatitis C Virus-Coinfected Individuals

Abstract

Hepatitis C virus (HCV)-induced liver fibrosis and splenomegaly may lead to discordance between absolute numbers and percentages of lymphocytes and subpopulations because of sequestration. We investigated lymphocyte discordance in HIV/HCV-coinfected individuals and its relationship to progression to liver disease, AIDS, and all-cause mortality. This is an observational retrospective cohort study. Adjusted hazard ratios (aHRs) with 95% confidence intervals (95% CIs) associated with liver disease, AIDS, or mortality were computed by time-updated Cox proportional hazards regression. Of 380 HIV/HCV-coinfected adult individuals followed for a median of 8.2 years, 360 individuals had a median of 11 discordant measurements corresponding to 5,080 of 9,091 paired samples (56%). Discordance alone was not associated with any of the outcomes. By multivariable analysis, a doubling of absolute or percentage CD4 cells was associated with comparable lower risks of mortality (aHR: 0.60, 95% CI: 0.53–0.67, p < .0001 and aHR: 0.67, 95% CI: 0.56–0.79, p < .0001, respectively). Higher CD4/CD8 ratio was associated with a lower mortality risk (aHR: 0.39, 95% CI: 0.22–0.71 per doubling, p = .002). Only absolute CD4 cell measurements predicted AIDS. Development of liver disease was not predicted by total lymphocyte count or subpopulations. Despite a high prevalence of lymphocyte-subpopulation discordance with HIV/HCV coinfection, absolute CD4 cell count predicted mortality and AIDS, whereas CD4 percentage only predicted mortality. Neither CD4 T lymphocyte count nor CD4 percentage was associated with liver disease in this cohort. These findings may be necessary and useful in countries where antiretroviral treatment is not initiated for all HIV-infected individuals.

Introduction

An accurate evaluation of immune status in HIV-infected individuals is critical to assess HIV disease progression and mortality.^1,2 Owing to shared transmission routes, hepatitis C virus (HCV) and HIV coinfection is frequent.³ Liver fibrosis and cirrhosis, through possible splenic sequestration of lymphocytes, was suggested to cause discordance between the absolute CD4 T lymphocyte count and CD4 T lymphocyte percentage.^4,5 A previous study has shown that cirrhotic individuals without HIV had a low absolute CD4 T lymphocyte count, but the majority of these had a normal CD4 percentage.⁴ This discordance between absolute CD4 T lymphocyte count and CD4 lymphocyte percentage was also shown to be true for HIV-infected individuals with cirrhosis.^6
–8 The absolute CD4 T lymphocyte count was identified as an important prognostic marker for AIDS and mortality,^1,2,9 but it was suggested that the CD4 T lymphocyte percentage might be a more appropriate HIV disease predictor in individuals with liver cirrhosis.^10,11 Debate still persists as to whether absolute CD4 T lymphocyte count or CD4 T lymphocyte percentage is most accurate in predicting HIV diseases and mortality.^10

–13

Low CD4/CD8 ratio was shown to be associated with increased morbidity and mortality in both HIV/HCV-coinfected and monoinfected individuals and therefore was suggested to be useful in monitoring disease progression.^14
–16

Total lymphocyte count was found to be a cost-effective surrogate marker for CD4 T lymphocytes and therefore of great importance and benefit in resource-limited settings.^17
–19

Use of CD4 T lymphocyte measurements in clinical decision making is of limited value in high-income countries where it is recommended to initiate antiretroviral treatment (ART) for all HIV-infected individuals.²⁰ However, in low-income countries CD4 T lymphocyte indices, total lymphocyte count, and CD4/CD8 ratio unfortunately may be necessary and useful for prioritized care.

Furthermore, high discordance has been associated with end-stage liver disease and liver fibrosis^7,8; thus, discordance might be considered as an indication to search for prevalent hepatic disease in patients who have not already been diagnosed with hepatic fibrosis.

The aim of this study was to investigate CD4 T lymphocyte discordance in HIV/HCV-coinfected adult individuals and its relationship to progression to AIDS, liver disease, and all-cause mortality. Secondary aims were to study the relationship between changes in CD4/CD8 ratio and total lymphocyte count and progression to AIDS, liver disease, and death.

Materials and Methods

Hvidovre Hospital clinic cohort

Patients eligible for this study were those older than 18 years of age in care from January 1994 to May 2015 at the outpatient clinic of the Department of Infectious Diseases and who had a positive HCV antibody test.

All patients had a clinical follow-up every 3–6 months. AIDS events during follow-up were captured through the prospective, ongoing nationwide Danish HIV Cohort Study.²¹

Information regarding dates of first positive HIV test, start of ART, HIV exposure group, and diagnosis of AIDS and liver disease were retrieved from the Danish HIV Cohort Study and patient files. Information regarding total lymphocyte count, CD4 T lymphocyte count and percentage, platelet count, plasma HIV-RNA, HCV antibody, plasma HCV-RNA, and hepatitis B surface antigen (HBsAg) were retrieved manually from patient files or extracted electronically from laboratory data files.

The study was approved by the Danish Data Protection Agency (record no. 2014-41-3492). Informed consent is not required by Danish legislation for register-based studies.

Databases

The Civil Registration System is a national registry that stores information on vital status, residency, and migration. Every Danish resident is given a personal identification number at birth or immigration that enables tracking and eliminates multiple registrations.²²

The Danish National Patient Registry (NPR) contains information about admissions to Danish hospitals and discharge diagnosis codes according to the International Classification of Diseases (ICD) and is updated monthly.²³

Outcomes

The outcomes of interest were liver disease, progression to AIDS, or death.

Liver disease was defined as any diagnosis with the ICD-10 codes K70–K76, I85, and C22. Liver biopsy and fibroscan were not routinely performed and therefore alanine transaminase (ALT) and platelet count were included in the analysis as proxy for severity of liver diagnosis.

AIDS was defined by the presence of an opportunistic infection or AIDS-defining malignancy in an HIV-positive individual and therefore not based on CD4 T lymphocyte count.

Death was defined as all-cause mortality.

Independent variables

Independent variables of interest were discordance, gender, age, race, HIV exposure group, ART, HIV-RNA, platelet count, ALT, HBsAg, and HCV coinfection.

The definition of concordance and discordance between CD4 T lymphocyte count and CD4 T lymphocyte percentage was extrapolated from previous studies evaluating the relationship between absolute CD4 T lymphocyte count and CD4 T lymphocyte percentage.^6,7,15 Concordant pairs were defined as follows: CD4 T lymphocyte cell count <100 cells/μL and CD 4 T lymphocyte percentage <7%, CD4 T lymphocyte between 100 and 200 cells/μL and CD4 T lymphocyte percentage between 7% and 14%, CD4 T lymphocyte between 200 and 300 cells/μL and CD4 T lymphocyte percentage between 14% and 121%, CD4 T lymphocyte between 300 and 500 cells/μL and CD4 T lymphocyte percentage between 21% and 35%, CD4 T lymphocyte >500 cells/μL and CD4 T lymphocyte percentage >35%. Discordance was defined as either high (CD4 T lymphocyte percentage >CD4 T lymphocyte count) or low (CD4 T lymphocyte count > CD4 T lymphocyte percentage).

HCV coinfection was determined by a positive HCV antibody test and patients were further categorized into three groups: chronic HCV, spontaneously resolved HCV, and HCV resolved after treatment. Chronic HCV was defined by positive HCV-RNA in two consecutive tests taken at least 6 months apart. Resolved HCV was determined by nondetectable HCV-RNA in two tests taken at least 6 months apart. Resolved HCV as a result of HCV treatment was determined by nondetectable HCV-RNA found in two tests taken at least 6 months apart after receiving HCV treatment. Antibody to HCV was determined by third-generation enzyme-linked immunosorbent assay after 1997 and plasma HCV-RNA was determined by reverse transcription polymerase chain reaction after 2005, as previously described.²⁴ Individuals with resolved HCV infection assayed before 1997 were retested for HCV antibody in the early 2000s and for HCV-RNA in the late 2000s. The retesting was performed on stored plasma samples retrieved from our plasma repository.

ART is provided free of charge and ART approved by the European Medicines Agency was available throughout the study period. The national guidelines for recommended first-line regimens of ARTs have been published by the Danish Infectious Diseases Society since 1996. Adherence to these have been evaluated.²⁵ In the study period the national criteria for ART initiation were presence of one of the following: acute HIV infection, an AIDS defining disease or HIV-related disease, pregnancy, CD4 T lymphocyte count <300 cells/μL (until 2009) or <350 cells/μL (from 2009 and until 2015), and, until January 1, 2001, a plasma HIV-RNA load >100,000 copies/mL.

Thrombocytopenia was defined as a platelet count <150 × 10⁹/L in two consecutive tests taken at least 6 months apart.

Statistics

Follow-up was calculated from the date of first positive HCV antibody to the date of death, emigration, or May 22, 2015, whichever came first. Patients were censored for development of outcome and loss to follow-up. Loss to follow-up was defined as 365 days without contact to the HIV care system or migration.

Baseline characteristics of patients stratified by HCV status were calculated for the overall population and for those who died during follow-up. Categorical variables were given as number and percentage and continuous variables as median and interquartile range (IQR). The number and proportion of discordant pairs of absolute CD4 and CD4 percentage was determined, as was the median and IQR for number of pairs per individual. Time-updated adjusted Cox proportional hazards models producing hazard ratios (HRs) with 95% confidence intervals (95% CIs) were used to determine the association between the independent variables of absolute CD4 T lymphocyte count, CD4 T lymphocyte percentage, and CD4:CD8 ratio, and the outcomes of progression to AIDS, liver disease, and all-cause mortality. These models were adjusted for gender, age, race, HIV exposure group, time-updated ART, time-updated HIV-RNA, time-updated platelet count, time-updated ALT, time-updated HBsAg, and time-updated HCV status.

Additional Cox proportional models were produced examining the association between absolute lymphocyte count, ALT, and platelet count and the outcomes of progression to AIDS, liver disease, or death. These models were adjusted for gender, age, race, HIV exposure group, time-updated ART, time-updated HIV-RNA, time-updated HBsAg, and time-updated HCV status.

We included predefined variables previously associated with outcome and novel variables that were statistically significantly associated with outcome by univariate analysis. Statistical significance was defined as p < .05.

Individual follow-up time was split into smaller time periods defined by dates of new blood samples enabling inclusion of time-updated values using the Stratify macro for SAS (SAS Institute, Inc. Cary, NC).²⁶ All analyses were performed by SAS statistical software 9.4.

Results

Study participants

Baseline characteristics of the study cohort at entry are given in Table 1. We included 380 individuals in this study. We found that 231 had persistent HCV-RNA viremia, 119 had resolved spontaneously by study entry, 18 resolved during follow-up, and 12 had resolved after HCV treatment. Of the HIV/HCV-coinfected individuals 65% were men, 94% were whites, and 74% were infected by injecting drug use. Median follow-up time was 8.2 years (3.5–13.2) for all participants, 8.7 years (4.0–13.6) for individuals with chronic HCV infection, and 7.6 years (2.3–12.7) for individuals with resolved HCV infection.

Table 1.

Baseline Characteristics of Study Subjects According to Hepatitis C Virus Status

	Number of individuals at baseline (% of total), N = 380
Characteristics	Chronic HCV, n = 231	Resolved HCV, spontaneous, n = 137	Resolved HCV, treated, n = 12
Sex
Male	154 (66.7)	85 (62.0)	9 (75)
Female	77 (33.3)	52 (38.0)	3 (25)
Age at inclusion, years
<35	79 (34.2)	52 (38.0)	4 (33.3)
35–55	146 (63.2)	80 (58.4)	7 (58.3)
>55	6 (2.6)	5 (3.6)	1 (8.3)
Median (IQR)	38 (33–44)	37.18 (32.9–43.1)	40.11 (34.1–45.9)
Race
Caucasian	217 (93.9)	129 (94.2)	11(91.6)
Other	14 (6.1)	8 (5.8)	1 (8.3)
HIV exposure
IDU	180 (77.9)	102 (74.5)	4 (33.3)
MSM	13 (5.6)	16 (11.7)	3 (25)
HSX	29 (12.6)	16 (11.7)	3 (25)
Other	9 (3.9)	3 (2.2)	2 (16.7)
AIDS at baseline
Yes	20 (8.7)	16 (11.7)	1 (8.3)
No	211 (91.3)	121 (88.3)	11 (91.6)
AIDS during follow-up	n = 211	n = 121	n = 11
Yes	48 (22.7)	41 (33.9)	2 (18.2)
No	163 (77.2)	80 (66.1)	9 (81.8)
ART at entry
Yes	54 (23.4)	31 (22.6)	4 (33.3)
No	177 (76.6)	106 (77.4)	8 (66.7)
ART during follow-up	n = 177	n = 106	n = 8
Yes	131 (74)	68 (54)	8 (100)
No	46 (26)	38 (36)	0 (0)
Nadir CD4
Median (IQR)	145 (63–234)	117 (45.5–244)	227 (126.5–281)
Baseline CD4 cell count	n = 141	n = 70	n = 10
≥500	39 (27.7)	23 (25.8)	3 (30)
200–500	66 (46.8)	33 (37.1)	6 (60)
<200	36 (25.5)	33 (37.1)	1 (10)
Median (IQR)	334 (197–508)	282 (131–509)	366 (252–638)
Latest CD4 cell count
≥500	78 (33.8)	41 (29.9)	7 (58.3)
200–500	85 (36.8)	54 (39.4)	4 (33.3)
<200	68 (29.4)	42 (30.7)	1 (8.3)
Median (IQR)	365 (167–592)	343 (159.5–536)	590 (354.5–918)
Latest CD4:CD8 ratio	n = 228	n = 132	n = 12
<0.4	1 (0.4)	0 (0)	0
0.4–1.0	27 (11.8)	17 (12.9)	3 (25)
>1.0	200 (87.7)	115 (87.1)	9 (75)
Median (IQR)	2.3 (1.3–3.9)	2.3 (1.4–4.3)	1.63 (1.1–1.9)
Discordant pairs
Median (IQR)	16 (7–31)	11 (5–15)	6 (3.5–13)
Latest HIV-RNA	n = 116	n = 65	n = 10
≥400	74 (63.8)	57 (75)	6 (60)
<400	42 (36.2)	19 (25)	4 (40)
Median (IQR)	5,165 (39–37,500)	15,150 (399.5–188.300)	9,400 (19–99,300)
Latest HCV-RNA
Median (IQR)	16,750,000 (265,306–5,740,000)	0	1286.5 (95–30612.5)
HBsAg	n = 129	n = 52	n = 9
Positive	6 (4.6)	13 (13.3)	0
Negative	123 (95.4)	85 (86.7)	9 (100)

IDU, injection drug use; MSM, men who have sex with men; HSX, heterosexual; HCV, hepatitis C virus; HBsAg, hepatitis B virus surface antigen; IQR, interquartile range; ART, antiretroviral treatment.

CD4 T lymphocyte count and percentage discordance

A total of 9,091 paired samples were collected. Discordance was found in 5,080 pairs (56%) from 360 individuals. The median number of discordant pairs per individual was 11 (4.5–20.0). The distribution of discordant pairs is given in Table 2.

Table 2.

Relationship Between CD4 T Lymphocyte Count and Percentage Discordance in 380 Individuals Positive for Hepatitis C Virus

	<100 CD4 cells/μL, n (%)	100–200 CD4 cells/μL, n (%)	200–300 CD4 cells/μL, n (%)	300–500 CD4 cells/μL, n (%)	>500 CD4 cells/μL, n (%)
<7% CD4	414 (4.55)	111 (1.22)	59 (0.65)	85 (0.93)	64 (0.70)
7%–14% CD4	245 (2.69)	572 (6.29)	344 (3.78)	177 (1.95)	24 (0.26)
14%–21% CD4	88 (0.97)	369 (4.06)	614 (6.75)	789 (8.68)	269 (2.96)
21%–35% CD4	17 (0.19)	180 (1.98)	436 (4.80)	1,407 (15. 48)	1,460 (16.06)
>35% CD4	3 (0.03)	19 (0.21)	75 (0.82)	266 (2.93)	1,004 (11.04)
Number of discordant pairs					5,080 (55.55)
Median number of pairs per individual (IQR)					11 (4.5–20)

Bold values represents concordant values.

Analyses of high discordance, low discordance, and the different outcomes were made (Table 3). We found no statistically significant association between high or low discordance and the development of AIDS, liver disease, and death. However, there was an association between platelet count <150 × 10⁹/L and discordance; low discordance (CD4 N > CD%) was associated with 44% lower risk of thrombocytopenia [adjusted hazard ratio (aHR): 0.56, 95% CI: 0.35–0.92, p = .02] and high discordance (CD4% > CDN) was associated with 162% increased risk of thrombocytopenia (aHR: 2.62, 95% CI: 1.70–4.04, p < .0001) compared with concordant individuals.

Table 3.

Time-Updated Analysis of Progression to Liver Disease, AIDS, and Death According High and Low Discordance

	Liver disease		AIDS		Death
	HR (95% CI)	p Value	HR (95% CI)	p Value	HR (95% CI)	p Value
Concordant values	1		1		1
Low discordance	1.01 (0.47–2.20)	.97	0.53 (0.25–1.12)	.09	0.77 (0.55–1.10)	.1563
High discordance	0.49 (0.66–2.06)	.49	1.15 (0.58–2.26)	.69	1.29 (0.91–1.83)	.1498

Adjusted for age, gender, race, HIV exposure group, antiretroviral treatment, resolved HCV, hepatitis B surface antigen status, and HIV viral load.

HR, hazard ratio; CI, confidence interval.

Progression to AIDS

During 2,982 person-years of follow-up, 91 individuals developed AIDS. The median time to AIDS was 1.99 (0.68–6.09) years and the median time from AIDS to death was 2.5 (0.6, 5.4) years. Table 4 gives the adjusted analysis on the effect of CD4 cells on the risk of AIDS. Neither nadir CD4 nor baseline CD4 T lymphocyte count was associated with risk of AIDS during follow-up. However, a doubling of latest CD4 T lymphocyte count was associated with 25% lower risk of AIDS (aHR: 0.75, 95% CI: 0.58–0.98, p = .03). The latest absolute CD4 T lymphocyte count <200 cells/μL was associated with a 134% higher risk of AIDS than a latest CD4 T lymphocyte count between 200 and 500 cells/μL (aHR: 2.34, 95% CI: 1.08–5.07, p = .03). A CD4 cell count >500 cells/μL was not associated with a lower risk of AIDS.

Table 4.

Time-Updated Analysis of Progression to AIDS According to CD4 T Lymphocyte Count, Percentage, and Ratio

	Absolute CD4		CD4%		CD4:CD8 ratio
	Adjusted HR (95% CI)	p Value	Adjusted HR (95% CI)	p Value	Adjusted HR (95% CI)	p Value
Baseline value	1.00 (0.99–1.00)	.60	0.96 (0.92–1.01)	.10	0.19 (0.04–1.05)	.06
Time-updated value per doubling	0.75 (0.58–0.99)	.03	0.79 (0.54–1.15)	.22	0.63 (0.17–2.34)	.49
Time-updated value^a
Low	2.35 (1.09–5.07)	.03	2.20 (0.98–4.93)	.06	0.54 (0.19–1.56)	.26
Intermediate	1		1		0.31 (0.10–0.98)	.047
High	0.64 (0.24–1.71)	.38	1.08 (0.43–2.72)	.87	1

Adjusted for age, gender, race, HIV exposure group, antiretroviral treatment, resolved HCV, hepatitis B surface antigen status, and HIV viral load.

Latest absolute CD4 divided in <200, 200–500, >500. Latest CD4% divided in <14, 14–28, >28. Latest ratio divided in <0.4, 0.4–1.0, >1.0.

A doubling in latest CD4 T lymphocyte percentage was not statistically significantly associated with risk of AIDS and there was no association with the risk of AIDS whether the latest CD4 T lymphocyte percentage was low (<14), intermediate (14–28), or high (>28) (Table 4).

A CD4:CD8 ratio of 0.4–1.0 compared with a CD4:CD8 ratio of >1.0 reduced the risk of AIDS by 69% (aHR: 0.31, 95% CI: 0.10–0.98, p = .047). None of the other CD4:CD8 ratio variables were associated with the risk of AIDS (Table 4).

An increase in latest absolute lymphocyte count was associated with 44% lower risk of AIDS (aHR: 0.56, 95% CI: 0.34–0.91, p = .021) (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/aid).

Neither baseline platelet count, baseline ALT values, time-updated platelet count, or time-updated ALT values were associated with the risk of AIDS (Supplementary Table S2).

Liver disease

Of the 380 participants included in this study, 11 were diagnosed with liver disease at entry and 38 people developed liver disease during follow-up. The univariate analysis showed no association between change in CD4 T lymphocyte count, CD4 percentage, CD4:CD8 ratio, total lymphocyte count, baseline or time-updated platelet count, baseline and time-updated ALT values, and the development of liver disease (Table 5).

Table 5.

Univariate Analysis of Progression to Liver Disease According CD4 T Lymphocyte Count, Percentage, or Ratio

	Absolute CD4 T lymphocyte count		CD4 T lymphocyte percentage		CD4:CD8 ratio
	HR (95% CI)	p Value	HR (95% CI)	p Value	HR (95% CI)	p Value
By 1 U increase	1.00 (0.99–1.00)	.68	0.98 (0.95–1.01)	.27	1.01 (0.99–1.03)	.39

All-cause mortality

During 3,340 person-years of follow-up, 226 individuals died (Supplementary Table S2). The median time to death was 4.9 (2.1–9.2) years.

Table 6 shows the time-updated analysis on the effect of CD4 cells on the risk of death. A single CD4 cell/μL increase was associated with 0.2% lower risk of death (aHR: 0.997, 95% CI: 0.997–0.998, p < .0001). A doubling of latest CD4 T lymphocyte count was associated with a 40% lower risk of death (aHR: 0.60, 95% CI: 0.53–0.67, p < .0001). The risk of death given by latest absolute CD4 category was 162% higher for a CD4 cell count <200 cells/μL than a cell count of 200–500 cells/μL (aHR: 2.62, 95% CI: 1.81–3.78, p < .0001). A high CD4 cell count (>500 cells/μL) was associated with 39% lower risk of death than a cell count of 200–500 cells/μL (aHR: 0.61, 95% CI: 0.40–0.94, p = .02).

Table 6.

Time-Updated Analysis of Progression to Death According CD4 T Lymphocyte Count, Percentage, or Ratio

	Absolute CD4		CD4%		CD4:CD8 ratio
	Adjusted HR (95% CI)	p Value	Adjusted HR (95% CI)	p Value	Adjusted HR (95% CI)	p Value
Baseline value	0.99 (0.99–1.00)	.03	0.98 (0.96–0.99)	.02	1.02 (0.77–1.37)	.87
Time-updated value per 1 U increase	0.99 (0.99–0.99)	<.0001	0.97 (0.96–0.99)	.0006	0.48 (0.29–0.81)	.006
Time-updated value per doubling	0.56 (0.53–0.67)	<.0001	0.67 (0.56–0.79)	<.0001	0.39 (0.22–0.71)	.002
Time-updated value by groups^a
Low	2.62 (1.81–3.78)	<.0001	1.66 (1.12–2.46)	.01	1.98 (1.04–3.78)	.04
Intermediate	1		1		1.33 (0.69–2.55)	.40
High	0.61 (0.40–0.96)	.02	0.82 (0.56–1.22)	.33	1

Adjusted for age, gender, race, HIV exposure group, antiretroviral treatment, resolved HCV, hepatitis B surface antigen status, and HIV viral load.

Latest absolute CD4 divided in <200, 200–500, >500. Latest CD4% divided in <14, 14–28, >28. Latest ratio divided in <0.4, 0.4–1.0, >1.0.

A single percentage-point increase in CD4 T lymphocyte percentage was associated with 3% lower risk of death (aHR: 0.97, 95% CI: 0.96–0.99, p = .0006). A doubling of latest CD4 T lymphocyte percentage was associated with a 32% lower risk of death (aHR: 0.67, 95% CI: 0.56–0.79, p < .0001). A latest CD4 percentage of <14 was associated with 66% higher risk of death than a latest CD4 percentage of 14–28 (aHR: 1.66, 95% CI: 1.12–2.46, p = .01). A CD4 T lymphocyte percentage of >28 was not statistically significantly associated with the risk of death compared with a CD4 T lymphocyte percentage of 14–28 (Table 6).

An increase in latest CD4:CD8 ratio by one was associated with 52% lower risk of death (aHR: 0.48, 95% CI: 0.29–0.81, p = .006). A doubling in latest CD4:CD8 ratio was found to be associated with 61% lower risk of death (aHR: 0.39, 95% CI: 0.22–0.71, p < .002). Low (<0.4) CD4:CD8 ratio was associated with a 4% higher risk of death than a high (>1.0) CD4:CD8 ratio, but intermediate (0.4–1.0) CD4:CD8 ratio was not statistically significantly associated with death compared with high CD4:CD8 ratio (Table 6).

Baseline values of absolute CD4 T lymphocyte count, CD4 T lymphocyte percentage, and CD4:CD8 ratio did not predict death (Table 6).

We found that a 10⁹ cell/L increase in latest absolute lymphocyte count was associated with a 44% lower risk of death (aHR: 0.56, 95% CI: 0.34–0.91, p = .02) (Supplementary Table S3).

Neither baseline platelet count, baseline ALT values, or time-updated platelet count values were associated with the risk of all-cause mortality; however, an increase by 10 × 10⁹/L in time-updated platelet count was associated with 5% lower risk of death (aHR: 0.95, 95% CI: 0.93–0.97, p < .001) (Supplementary Table S3).

Stratified analyses of individuals with and without persistent HCV-RNA viremia were performed for all groups of lymphocyte and lymphocyte subsets without significant changes in associations to AIDS and mortality (data not given).

Furthermore, an analysis of the interaction between CD4 T lymphocyte count and CD4 percentage was performed, but no interaction effect was found (data not given).

Discussion

Our study showed that despite frequent discordance, the absolute CD4 T lymphocyte count and CD4 percentage independently predicted all-cause mortality in HIV/HCV-coinfected individuals. The associations persisted in adjusted analysis including other important predictors of survival such as gender, age, HIV exposure group, platelet count, HBsAg, and HCV-RNA status.

Results on the use of absolute CD4 T lymphocyte count and CD4 T lymphocyte percentage as prognostic indicators for disease progression of HIV in coinfected individuals are conflicting. Bongiovanni et al. ¹² and Gebo et al. ¹³ found a greater prognostic value of absolute CD4 T lymphocyte count compared with CD4 T lymphocyte percentage. Hulgan et al. showed that CD4 T lymphocyte percentage was an important indicator of when to initiate treatment in patients with an absolute lymphocyte count >350 lymphocytes/mL.¹⁰ Hulgan et al. also showed that CD4 T lymphocyte percentage was associated with either disease progression or death in patients initiating ART.¹¹ These studies investigated the association of lymphocyte and subpopulation changes with disease progression, but none of the studies investigated a possible association with mortality. Our study showed that absolute CD4 T lymphocyte count and CD4 T lymphocyte percentage were equally reliable markers to predict mortality. We found that absolute CD4 T lymphocyte count predicted the development of AIDS, but we were unable to confirm an association between progression to AIDS and CD4 T lymphocyte percentage except for the baseline value. This is consistent with the findings of Bongiovanni et al. ¹² and Gebo et al. ¹³ We found no association between baseline or nadir CD4 T lymphocyte count and the development of AIDS. This can be explained by most of the patients receiving ART during follow-up. Studies have shown that the latest CD4 T lymphocyte count is a much more reliable predictor of disease outcome than baseline and nadir CD4 T lymphocyte count in patients receiving ART.^27
–29

Hull et al. found an association between high discordance and end-stage liver disease.⁷ We were unable to confirm an association between high discordance and either of the outcomes. A possible reason for this could be that the proportion of subjects with portal hypertension is actually low and therefore the analysis cannot provide helpful guidance for interpretation of CD4 T lymphocyte or percentage in this population of interest. However, we did perform an analysis on the association between discordance and thrombocytopenia, which showed that high discordance was associated with 162% increased risk of developing thrombocytopenia, suggesting that high discordance identifies individuals with cirrhosis and splenic sequestration resulting in thrombocytopenia.^30
–32

A study showed that the risk of liver-related death increased ninefold for HCV-RNA-positive individuals compared with HCV-RNA-negative individuals.³³ Another study performed on the same cohort showed that HCV viral load increases over time in coinfected individuals not receiving ART treatment.³⁴ A recent study by our group that included the majority of patients in this study showed that there was an association between HCV viral load and all-cause mortality.³⁵ Therefore we did stratified analysis of HCV antibody-positive individuals with and without HCV viremia, but this did not significantly alter the results (data not given). Thus, absolute CD4 T lymphocyte count or CD4 percentage may be a useful predictor regardless of HCV viremia. However, our ability to detect a difference may be limited by sample size.

Kuniholm et al. ¹⁴ found that coinfected patients who had spontaneously resolved HCV had a significantly lower CD4:CD8 ratio than patients with chronic HCV and HCV-negative patients. In our study, we found no difference in the latest median CD4:CD8 ratio between individuals with chronic HCV and individuals who had spontaneously resolved HCV. However, we found that individuals who resolved their HCV infection after antiviral treatment had a lower median CD4:CD8 ratio, but the CD4:CD8 ratio was not associated with liver disease, AIDS, or death, when adjusted for HCV status.

Liver disease is among one of the most common causes of death among HIV/HCV-coinfected individuals and therefore a predictor of liver disease is of great importance. Previous studies have found an association between discordance and fibrosis and end-stage liver disease, suggesting discordance could potentially serve as a predictor for progression to liver disease. However, we found no association between discordance and liver disease. We also found no association between lymphocyte or subpopulation cell count or percentage and the development of liver disease. This is somewhat conflicting with the findings of Grint et al.,³⁶ who found that a low baseline CD4 cell count was associated with liver-related death among individuals with HIV and chronic HCV, and Anderson et al. ³⁷ who found that a low CD4 cell count was a strong indicator of clinical liver disease progression among HIV/HCV-coinfected drug abusers. The low number of individuals developing liver disease in this cohort could explain the reason for these findings. In our study, only 11% of the individuals in the cohort had liver disease diagnosed during follow-up, thus making the statistical power to study such an association low. Furthermore, our data on liver disease were retrieved from NPR and to our knowledge no studies have investigated the validation of liver diagnosis. Therefore it is possible that liver disease was prevalent, but undiagnosed, thus underestimating the number of study subject with liver disease. This could also have affected our interpretation of the time course of CD4 T lymphocyte evolution.

Transient elastography has been shown to be the most accurate noninvasive test of liver fibrosis in coinfected individuals.³⁸ Pérez-Latorre et al. ³⁹ showed that transient elastography was a valid predictor of liver-related events among HIV/HCV-coinfected individuals. Further studies looking into transient elastography as a predictor for liver disease among coinfected individuals are needed.

Our study is strengthened by the consecutive inclusion of study subjects, complete follow-up, a large number of paired blood samples, and the time-updated modeling allowing assessment of more recent values as predictors of outcome. Our study has some limitations. Our study did not adjust for fibrosis or cirrhosis. Liver biopsy was only performed when indicated and transient elastography has only been performed routinely in recent years. Other noninvasive surrogate tests of fibrosis and cirrhosis such as aspartate aminotransferase-to-platelet ratio index, fibrosis-4 index, and Fibrotest were unfortunately not available because we rely exclusively on ALT. Therefore we were unable to determine whether every participant actually had liver fibrosis or cirrhosis.

In conclusion, CD4 T lymphocyte count and CD4 T lymphocyte percentage predicted death, but only absolute CD4 T lymphocyte indices were associated with the risk of AIDS. Therefore these finding suggest that absolute CD4 T lymphocyte count may be used as an indicator of disease progression and death in HIV/HCV-infected individuals.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Phillips

, Lee

, Elford

, Janossy

, Kernoff

: The cumulative risk of AIDS as the CD4 lymphocyte count declines. J Acquir Immune Defic Syndr, 1992; 5:148–152.

Mellors

, Muñoz

, Giorgi

, et al.: Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med, 1997; 126:946–954.

Operskalski

, Kovacs

: HIV/HCV co-infection: Pathogenesis, clinical complications, treatment, and new therapeutic technologies. Curr HIV AIDS Rep, 2011; 8:12–22.

McGovern

, Golan

, Lopez

, et al.: The impact of cirrhosis on CD4+ T cell counts in HIV-seronegative patients. Clin Infect Dis, 2007; 44:431–437.

Mandorfer

, Reiberger

, Payer

, Peck-Radosavljevic

, Vienna HIV & Liver Study Group: The influence of portal pressure on the discordance between absolute CD4+ cell count and CD4+ cell percentage in HIV/hepatitis C virus-coinfected patients. Clin Infect Dis, 2013; 56:904–905.

Gandhi

: Cirrhosis is associated with low CD4+ T cell counts: Implications for HIV-infected patients with liver disease. Clin Infect Dis, 2007; 44:438–440.

Hull

, Rollet

, Odueyungbo

, et al.: Factors associated with discordance between absolute CD4 cell count and CD4 cell percentage in patients coinfected with HIV and hepatitis C virus. Clin Infect Dis, 2012; 54:1798–1805.

Claassen

, Diener-West

, Mehta

, Thomas

, Kirk

: Discordance between CD4+ T-lymphocyte counts and percentages in HIV-infected persons with liver fibrosis. Clin Infect Dis, 2012; 54:1806–1813.

Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services. 2015. Available at http://aidsinfo.nih.gov/contentfiles/lvguidelines/AdultandAdolescentGL.pdf, accessed August 6, 2015.

10.

Hulgan

, Raffanti

, Kheshti

, et al.: CD4 lymphocyte percentage predicts disease progression in HIV-infected patients initiating highly active antiretroviral therapy with CD4 lymphocyte counts >350 lymphocytes/mm3. J Infect Dis, 2005; 192:950–957.

11.

Hulgan

, Shepherd

, Raffanti

, et al.: Absolute count and percentage of CD4+ lymphocytes are independent predictors of disease progression in HIV-infected persons initiating highly active antiretroviral therapy. J Infect Dis, 2007; 195:425–431.

12.

Bongiovanni

, Gori

, Lepri

, et al.: Is the CD4 cell percentage a better marker of immunosuppression than the absolute CD4 cell count in HIV-infected patients with cirrhosis?. Clin Infect Dis, 2007; 45:650–653.

13.

Gebo

, Gallant

, Keruly

, Moore

: Absolute CD4 vs. CD4 percentage for predicting the risk of opportunistic illness in HIV infection. J Acquir Immune Defic Syndr, 2004; 36:1028–1033.

14.

Kuniholm

, O'Brien

, Prokunina-Olsson

, et al.: Association of hepatitis C virus infection with CD4/CD8 ratio in HIV-positive women. J Acquir Immune Defic Syndr, 2016; 72:162–170.

15.

Taylor

, Fahey

, Detels

, Giorgi

: CD4 percentage, CD4 number, and CD4:CD8 ratio in HIV infection: Which to choose and how to use. J Acquir Immune Defic Syndr, 1989:114–124.

16.

, Mehraj

, Vyboh

, Cao

, Li

, Routy

: CD4:CD8 ratio as a frontier marker for clinical outcome, immune dysfunction and viral reservoir size in virologically suppressed HIV-positive patients. J Int AIDS Soc, 2015; 18:20052.

17.

Abdollahi

, Saffar

, Shoar

, Jafari

: Is total lymphocyte count a predictor for CD4 cell count in initiation antiretroviral therapy in HIV-infected patients?. Niger Med J, 2014; 55:289–293.

18.

Kakar

, Beri

, Gogia

, et al.: Absolute lymphocyte count: A cost-effective method of monitoring HIV-infected individuals. Indian J Pathol Microbiol, 2011; 54:107–111.

19.

Agrawal

, Rane

, Jadhav

: Absolute lymphocyte count as a surrogate Marker of CD4 count in monitoring HIV infected individuals: A prospective study. J Clin Diagn Res, 2016; 10:EC17–EC19.

20.

Lundgren

, Babiker

, Gordin

, et al.: Initiation of antiretroviral therapy in early asymptomatic HIV infection. N Engl J Med, 2015; 373:795–807.

21.

Obel

, Engsig

, Rasmussen

, Larsen

, Omland

, Sørensen

: Cohort profile: The Danish HIV cohort study. Int J Epidemiol, 2009; 38:1202–1206.

22.

Pedersen

, Gøtzsche

, Møller

, Mortensen

: The Danish Civil Registration System. A cohort of eight million persons. Dan Med Bull, 2006; 53:441–449.

23.

Lynge

, Sandegaard

, Rebolj

: The Danish national patient register. Scand J Public Health, 2011; 39(7 Suppl):30–33.

24.

Clausen

, Weis

, Astvad

, et al.: Interleukin-28B polymorphisms are associated with hepatitis C virus clearance and viral load in a HIV-1-infected cohort. J Viral Hepat, 2011; 18:e66–e74.

25.

Petersen

, Andersen

, Gerstoft

, et al.: Adherence to national guidelines for initiation of antiretroviral regimens in HIV patients: a Danish nationwide study. Br J Clin Pharmacol, 2011; 72:116–124.

26.

Rostgaard

: Methods for stratification of person-time and events—a prerequisite for Poisson regression and SIR estimation. Epidemiol Perspect Innov, 2008; 5:7.

27.

Achhra

, Amin

, Law

, et al.: Immunodeficiency and the risk of serious clinical endpoints in a well-studied cohort of treated HIV-infected patients. AIDS, 2010; 24:1877–1886.

28.

Achhra

, Petoumenos

, Law

: Relationship between CD4 cell count and serious long-term complications among HIV-positive individuals. Curr Opin HIV AIDS, 2014; 9:63–71.

29.

Tarwater

, Margolick

, Jin

, et al.: Increase and plateau of CD4 T-cell counts in the 3(1/2) years after initiation of potent antiretroviral therapy. J Acquir Immune Defic Syndr, 2001; 27:168–175.

30.

Aster

: Pooling of platelets in the spleen: Role in the pathogenesis of “hypersplenic” thrombocytopenia. J Clin Invest, 1966; 45:645–657.

31.

Mitchell

, Feldman

, Diakow

, Sigal

: The pathophysiology of thrombocytopenia in chronic liver disease. Hepatic Med Evid Res, 2016; 8:39–50.

32.

Qamar

, Grace

, Groszmann

, et al.: Incidence, prevalence, and clinical significance of abnormal hematologic indices in compensated cirrhosis. Clin Gastroenterol Hepatol, 2009; 7:689–695.

33.

Rockstroh

, Peters

, Grint

, et al.: Does hepatitis C viremia or genotype predict the risk of mortality in individuals co-infected with HIV?. J hepatol, 2013; 59:213–220.

34.

Grint

, Peters

, Reekie

, et al.: Stability of hepatitis C virus (HCV) RNA levels among interferon-naive HIV/HCV-coinfected individuals treated with combination antiretroviral therapy. HIV Med, 2013; 14:370–378.

35.

Clausen

, Astvad

, Ladelund

, Larsen

, Schønning

, Benfield

: Hepatitis C viral load, genotype 3 and interleukin-28B CC genotype predict mortality in HIV and hepatitis C-coinfected individuals. AIDS, 2012; 26:1509–1516.

36.

Grint

, Peters

, Rockstroh

, et al.: Liver-related death among HIV/hepatitis C virus-co-infected individuals: Implications for the era of directly acting antivirals. AIDS, 2015; 29:1205–1215.

37.

Anderson

, Horsburgh

Jr. , Williams

, et al.: CD4 recovery on antiretroviral therapy is associated with decreased progression to liver disease among hepatitis C virus-infected injecting drug users. Open Forum Infect Dis, 2015; 2:ofv019.

38.

Schmid

, Bregenzer

, Huber

, et al.: Progression of liver fibrosis in HIV/HCV Co-infection: A comparison between non-invasive assessment methods and liver biopsy. PLoS One, 2015; 10:e0138838.

39.

Pérez-Latorre

, Sánchez-Conde

, Rincón

, et al.: Prediction of liver complications in patients with hepatitis C virus-related cirrhosis with and without HIV coinfection: Comparison of hepatic venous pressure gradient and transient elastography. Clin Infect Dis, 2014; 58:713–718.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.03 MB

0.02 MB