Short Communication: Plasma Levels of Vitamin D in HIV Patients Initiating Antiretroviral Therapy Do Not Predict Immune Restoration Disease Associated with Mycobacterium tuberculosis

Abstract

Immune restoration disease associated with Mycobacterium tuberculosis (TB IRD) is clinically important among HIV patients commencing antiretroviral therapy in countries where tuberculosis is endemic. Vitamin D affects dendritic cell and T cell function and the antimicrobial activity of monocytes. Plasma levels of vitamin D and polymorphisms in the vitamin D receptor may affect tuberculosis, and HIV infection associates with vitamin D deficiency. Here we assess whether plasma vitamin D levels may predict TB IRD. Samples were available from prospective studies of TB IRD in Cambodia (26 cases), India (19 cases), and South Africa (29 cases). IRD cases and controls from each site were similar in age and baseline CD4⁺ T cell count. Plasma samples were assessed using 25(OH) vitamin D immunoassay plates. DNA samples were available from a subset of patients and were genotyped for the VDR FokI (F/f) [C/T, rs10735810] SNP. When data from each cohort were pooled to assess ethnic/geographic differences, 25(OH)D levels were higher in Cambodian than Indian or South African patients (p<0.0001) and higher in South African than Indian patients (p<0.0001). TB IRD was not associated with differences in levels of 25(OH)D in any cohort (p=0.36–0.82), irrespective of the patients' prior TB diagnoses/treatment. Carriage of the minor allele of VDR FokI (F/f) was marginally associated with TB IRD in Indian patients (p=0.06) with no association in Cambodians. Neither plasma levels of vitamin D nor the vitamin D allele will usefully predict TB IRD in diverse populations from TB endemic regions.

I mmune restoration disease associated with coinfection with Mycobacterium tuberculosis (TB IRD) causes significant morbidity and mortality among HIV patients who commence antiretroviral therapy (ART) in resource-poor countries where tuberculosis is endemic. TB IRD may present as a “paradoxical” worsening of TB following initial improvement on TB therapy, after initiation of ART (denoted TB-IRIS) or as “unmasking” of TB not clinically apparent pre-ART (denoted ART-TB). Risk factors include a higher mycobacterial burden, shorter duration of TB therapy prior to ART initiation, advanced immunodeficiency, and effective immune reconstitution.¹ IRD probably represents an excessive immune response triggered by efficient immune reconstitution in the setting of persistent microbial antigen. Prediction of IRD at the initiation of ART would open options for better clinical management.

The active form of vitamin D, 1α,25-dihydroxyvitamin D₃ [1,25(OH)₂D], inhibits dendritic cell maturation and migration² and influences cell cycle progression and expression of FoxP3 by regulatory T cells.³ Vitamin D also mediates antimycobacterial activity of monocytes via cathelicidin.⁴ A role for vitamin D in the control of mycobacteria in vivo is postulated as levels of the more stable form [25(OH)D] in plasma are low in TB patients⁵ and rise during therapy,⁶ though intervention studies have obtained variable results.⁷ Associations between tuberculosis and polymorphisms in the vitamin D receptor (VDR) have been demonstrated in populations of different ethnicity and geographic location.^8,9

HIV infection is associated with deficiency of vitamin D cautiously associated with dietary deficiency, malabsorption, and/or reduced exposure to sunlight.^10,11 Considering its immunological and antimicrobial properties, Conesa–Botella et al. ¹² proposed that low levels of vitamin D may predispose HIV patients to TB IRD in the context of immune reconstitution on ART. This would open up a number of simple therapeutic options.

To establish whether low plasma vitamin D levels at ART initiation are a risk factor for development of IRD, we compared plasma levels of 25(OH)D in HIV patients beginning ART in Cambodia, India, and South Africa. These populations differ in their diet, skin color, and the climate prevailing in their region, but all exhibit high rates of tuberculosis. 25(OH)D was chosen in place of the active form 1,25(OH)₂D as it is more stable and readily assessed with commercial reagents. This is important in the selection of a predictive test for resource-constrained settings. 25(OH)D is accepted as a reliable marker of vitamin D status.¹¹

Plasma samples collected at or before the initiation of ART were available from three prospective cohort studies of TB IRD. CD4⁺ T cell counts were performed locally in routine clinical laboratories. IRD cases and controls from each site were similar in age and baseline CD4⁺ T cell count (see Table 1). All studies were approved by the relevant Institutional Ethics Committees and patients provided written informed consent.

Table 1.

Immune Restoration Disease Patients and Controls Were Matched at Baseline

Study group	n	Age (years)	CD4 ⁺ T cells/μl	Time between ART and IRD (days)
Durban, South Africa
Paradoxical TB-IRIS	11	33 (24–50)	51 (7–143)	10 (5–90)
HIV⁺TB⁺ controls	11	32 (25–49)	53 (3–254)	N.A.
ART-associated TB	18	34 (25–44)	102 (4–261)	10 (1–104)
HIV⁺TB⁻ controls	38	34 (23–57)	81 (2–348)	N.A.
New Delhi, India
Paradoxical TB-IRIS	19	36 (18–51)	78 (14–189)	17 (5–109)
HIV⁺TB⁺ controls	43	35 (18–65)	100 (2–198)	N.A.
Phnom Penh, Cambodia
Paradoxical TB-IRIS	15	35 (26–54)	45 (1–133)	10 (7–89)
HIV⁺TB⁺ controls	30	35 (25–63)	44 (2–160)	N.A.
ART-associated TB	11	35 (28–45)	37 (1–227)	10 (1–28)
HIV⁺TB⁻ controls	22	38 (24–56)	38 (1–231)	N.A.

Data are presented as median (range).

ART, antiretroviral therapy; IRD, immune restoration disease; TB, tuberculosis; N.A., not applicable.

1. Patients commencing ART at the King Edward VIII and RK Khan Hospitals in Durban, South Africa. From 102 HIV⁺TB⁺ patients beginning ART after antitubercular therapy, 14 (14%) developed paradoxical TB-IRIS including 11 with pre-ART plasma samples available for analysis. A total of 23 of 396 patients (5.8%) with no prior history of TB developed ART-TB, including 18 with pre-ART plasma samples. Controls were matched by age group, gender, pre-ART CD4⁺ T cell count (in bands of 50 cells/μl), and TB history in a 1:1 ratio with paradoxical TB-IRIS cases and approximately 2:1 with ART-TB cases.^13,14 Post hoc analyses confirmed that IRD patients and controls had similar body mass indices, plasma albumin levels, prevalence of renal disease and diabetes, and use of multivitamin supplements. The occurrence of IRD did not segregate with gender or the ART regimen (data not shown).

2. Patients (n=62) commencing ART while receiving antitubercular therapy at the L.R.S. Institute of Tuberculosis and Respiratory Disease in New Delhi, India. Nineteen patients (31%) experienced paradoxical TB IRIS. This outcome did not segregate with receipt of vitamin supplements (B₆ only), gender, or the ART regimen (data not shown). L.R.S. is a tertiary referral center for tuberculosis taking patients with severe disease. Six patients died in the first year of therapy (three cases and three controls). This may contribute to the high incidence of TB IRIS.

3. Patients commencing ART at the National Centre for HIV/AIDS, Dermatology and Sexually Transmitted Diseases (NCHADS) Social Health Clinic in Phnom Penh, Cambodia.¹³ Of the 75 HIV⁺TB⁺ patients beginning ART after antitubercular therapy, 15 (20%) developed paradoxical TB-IRIS. Eleven of 231 patients (4.8%) with no prior history of TB developed ART-TB. Each IRD case was matched with two controls from the cohort, by age, gender, pre-ART CD4⁺ T cell count, and TB history.

Plasma samples were assessed using 25(OH)D immunoassay plates (Immunodiagnostic Systems, United Kingdom) which detect 25(OH) vitamin D₂ and D₃.¹¹ Each cohort was run on a separate plate, but six samples from Australian controls were run on all three plates to confirm standardization. Blood samples from Cambodia had been incubated overnight without stimulation under a Quantiferon protocol. A trial of two Australian donors established that overnight incubation did not affect levels of 25(OH)D (data not shown). Levels of 25(OH)D were compared using the Mann–Whitney U test and presented as median (range).

DNA samples available from a subset of the Cambodian patients and all Indian patients were genotyped for the single nucleotide polymorphism (SNP) known as VDR FokI (F/f) (C/T, rs10735810, C_12060045_20) as this polymorphism associates with tuberculosis in several populations.^8,9 We used TaqMan FAM and VIC-labeled probes and Universal PCR Master Mix (Applied Biosystems, Foster City, CA). Thermocycling involved 10 min at 95°C, followed by 40 cycles of 92°C for 15 s with annealing and extension at 60°C for 1 min. Fluorescence was read on an ABI 7900 Sequence Detector (Applied Biosystems). The genotyping success rate was ∼100%.

IRD cases and controls from each site were similar in age and baseline CD4⁺ T cell count (see Table 1). When data from each cohort were pooled to assess ethnic/geographic differences, median (range) levels of 25(OH)D were 80 (24–200) nmol/liter, 25 (6–65) nmol/liter, and 48 (18–121) nmol/liter in Cambodian, Indian, and South African patients (Fig. 1A). Levels in Cambodian patients were higher than in Indian or South African patients (p<0.0001) and levels were higher in South African than Indian patients (p<0.0001). The higher levels in Cambodians are in accordance with the literature and may be attributed to light skin color, high sun exposure and/or a diet rich in fish.¹⁵ The low levels in Indians (many of whom are vegetarian) fit with evidence that associations between low plasma levels of 25(OH)D and tuberculosis (without HIV) are clearer in Indian than Vietnamese donors.^5,16 Among South African patients, all TB-IRIS cases and HIV⁺TB⁻ controls, 16/18 ART-TB cases and 10/11 HIV⁺TB⁺ controls were of black African (Zulu) ethnicity. The remaining three individuals were ethnically Indian or of mixed race. Samples were drawn from all cohorts throughout the year, with no seasonal differences noted between groups. Seasonal variation was checked formally in the Indian cohort, as Delhi has relatively cold winters. No variation was detected [median (range): 25 (8–61) nmol/liter in winter vs. 24 (13–65) nmol/liter in summer, p=0.79].

FIG. 1.

Plasma levels of 25(OH) vitamin D in HIV patients commencing antiretroviral therapy (ART) in South Africa, India, and Cambodia (A). Levels in immune restoration disease (IRD) cases and controls within the South African (B), Indian (C), and Cambodian (D) cohorts are denoted as follows: TB-IRIS cases (•), TB-IRIS controls (○), ART-TB cases (▴), and ART-TB controls (▵). Median values and significant p-values (Mann–Whitney U test) are shown.

In general, neither TB-IRIS nor ART-TB was associated with differences in levels of vitamin D in any cohort (p=0.36–0.82). The one possible exception was paradoxical TB IRIS, which approached significance (p=0.067) in Cambodian patients (Fig. 1B–D). However, this trend was reversed in South Africans and not evident in Indians, so it is not considered meaningful.

Carriage of the minor allele of VDR FokI (F/f) was similar in combined TB IRD patients and their controls from the Cambodian cohort [13/17 (76%) vs. 36/54 (67%); Fisher's exact test, p=0.55; OR (95% CI)=1.6 (0.5–5.7)], but marginally associated with paradoxical TB IRIS in Indian patients [17/43 (40%) vs. 13/19 (68%); p=0.06; OR (95% CI)=3.3 (1.1–10)]. Further studies should determine whether genetic associations exist in other populations with low levels of plasma 25(OH)D. Other polymorphisms in the vitamin D receptor (VDR) may be more informative, as VDR haplotypes affect both HIV disease and tuberculosis in Indian donors.¹⁷

While the small size of the cohorts precludes the identification of small differences, the results suggest that neither plasma levels of 25(OH)D nor VDR FokI (F/f) will prove clinically useful in the prediction of paradoxical TB IRIS or ART TB. The data do not support the development of an intervention study.

Footnotes

Acknowledgments

The authors thank Dr. Silvia Lee for assistance with the vitamin D assays, and all patients and controls who donated blood for this study.

Author Disclosure Statement

No competing financial interests exist.

References

Price

, Agarwal

, Murdoch

, Elliot

, Lewin

, French

. Immune restoration diseases reflect diverse immunopathological mechanisms. Clin Microbiol Rev, 2009; 22:651–663.

Gauzzi

, Purificato

, Donato

, Jin

, Wang

, Daniel

et al. Suppressive effect of 1 α,25-dihydroxyvitamin D3 on type I IFN-mediated monocyte differentiation into dendritic cells: Impairment of functional activities and chemotaxis. J Immunol, 2005; 174:270–276.

Morales-Tirado

, Wichlan

, Leimig

, Street

, Kasow

, Riberdy

. 1α,25-Dihydroxyvitamin D3 (vitamin D3) catalyzes suppressive activity on human natural regulatory T cells, uniquely modulates cell cycle progression, and augments FOXP3. Clin Immunol, 2011; 138:212–221.

Liu

, Stenger

, Tang

, Modlin

. Cutting edge: Vitamin D-mediated human antimicrobial activity against Mycobacterium tuberculosis is dependent on the induction of cathelicidin. J Immunol, 2007; 179:2060–2063.

Sasidharan

, Rajeev

, Vijayakumari

. Tuberculosis and vitamin D deficiency. J Assoc Physicians India, 2002; 50:554–558.

Tostmann

, Wielders

, Kibiki

, Verhoef

, Boeree

, van der Ven

. Serum 25-hydroxy-vitamin D3 concentrations increase during tuberculosis treatment in Tanzania. Int J Tuberc Lung Dis, 2010; 14:1147–1152.

Yamshchikov

, Desai

, Blumberg

, Ziegler

, Tangpricha

. Vitamin D for treatment and prevention of infectious diseases: A systematic review of randomized controlled trials. Endocr Pract, 2009; 15:438–449.

Motsinger-Reif

, Antas

, Oki

, Levy

, Holland

, Sterling

. Polymorphisms in IL-1beta, vitamin D receptor Fok1, and Toll-like receptor 2 are associated with extrapulmonary tuberculosis. BMC Med Genet, 2010; 11:37.

Gao

, Tao

, Zhang

, Jin

. Vitamin D receptor genetic polymorphisms and tuberculosis: Updated systematic review and meta-analysis. Int J Tuberc Lung Dis, 2010; 14:15–23.

10.

Haug

, Aukrust

, Haug

, Mørkrid

, Müller

, Frøland

. Severe deficiency of 1,25-dihydroxyvitamin D3 in human immunodeficiency virus infection: Association with immunological hyperactivity and only minor changes in calcium homeostasis. J Clin Endocrinol Metab, 1998; 83:3832–3838.

11.

Welz

, Childs

, Ibrahim

, Poulton

, Taylor

, Moniz

, Post

. Efavirenz is associated with severe vitamin D deficiency and increased alkaline phosphatase. AIDS, 2010; 24:1923–1928.

12.

Conesa-Botella

, Mathieu

, Colebunders

, Moreno-Reyes

, van Etten

, Lynen

, Kestens

. Is vitamin D deficiency involved in the immune reconstitution inflammatory syndrome? AIDS Res Ther, 2009; 6:4.

13.

Elliott

, Vohith

, Saramony

, Savuth

, Dara

, Sarim

et al. Immunopathogenesis and diagnosis of tuberculosis and tuberculosis-associated immune reconstitution inflammatory syndrome during early antiretroviral therapy. J Infect Dis, 2009; 200:1736–1745.

14.

Haddow

, Moosa

MYS

, Easterbrook

. Validation of a published case definition for tuberculosis-associated immune reconstitution inflammatory syndrome. AIDS, 2010; 24:103–108.

15.

Mithal

, Wahl

, Bonjour

, Burckhardt

, Dawson-Hughes

, Eisman

et al. Global vitamin D status and determinants of hypovitaminosis D. Osteoporos Int, 2009; 20:1807–1814.

16.

Ho-Pham

, Nguyen

, Bui

, Nguyen

. Association between vitamin D insufficiency and tuberculosis in a Vietnamese population. BMC Infect Dis, 2010; 10:306.

17.

Alagarasu

, Selvaraj

, Swaminathan

, Narendran

, Narayanan

et al. 5' regulatory and 3' untranslated region polymorphisms of vitamin D receptor gene in south Indian HIV and HIV-TB patients. J Clin Immunol, 2009; 29:196–204.