SHORT COMMUNICATION
INT J TUBERC LUNG DIS 18(4):466–469 Q 2014 The Union http://dx.doi.org/10.5588/ijtld.13.0460
Change in serum CXCL10 levels during anti-tuberculosis treatment depends on vitamin D status Z. Hasan,* N. Salahuddin,† N. Rao,‡ M. Aqeel,† F. Mahmood,† F. Ali,† M. Ashraf,* F. Rahman,† S. Mahmood,* M. Islam,* B. Dildar,† T. Anwer,‡ F. Oiighor,† N. Sharif,† A. R. Ullah† *Department of Pathology and Microbiology, and †Department of Medicine, the Aga Khan University, Karachi, ‡ Ojha Institute for Chest Diseases, Dow University of Health Sciences, Karachi, Pakistan SUMMARY
Serum levels of the inflammatory C-X-C motif chemokine 10 (CXCL10) are raised in tuberculosis (TB). CXCL10 gene expression is downregulated in monocytes by metabolically active vitamin D3 (1,25dihydroxy vitamin D). Stratification of patients by serum 25hydroxyvitamin D (25[OH]D) levels at baseline showed that treatment-induced decrease in CXCL10 occurred in those with ‘insufficient’ and ‘deficient’ but not in those
with ‘optimal’ levels. In the deficient group, 25(OH)D showed an inverse correlation with CXCL10 levels. CXCL10 may thus be a useful biomarker for the followup of response to treatment. However, CXCL10 levels should be interpreted taking into account the baseline serum vitamin D levels of the TB patients. K E Y W O R D S : CXCL10; vitamin D; tuberculosis; biomarker
ALTHOUGH the diagnosis and treatment of tuberculosis (TB) is monitored clinically and bacteriologically, it may be difficult to assess, particularly in the case of extra-pulmonary TB. Cytokine and chemokine biomarkers can aid in disease identification and treatment.1 Interferon-gamma (IFN-c) induced protein 10 (IP-10 or C-X-C motif chemokine 10 [CXCL10]), produced by T-cells, is increased in patients with active TB, and decreases with treatment.1 Deficiencies in vitamin D3 (25-hydroxycholecalciferol) levels have been implicated in the development of TB.2 The primary circulating form of vitamin D3, 25-hydroxyvitamin D (25[OH]D), is hydroxylated to physiologically active 1-a,25-dihydroxy vitamin D (1,25[OH]D).2 Supplementation with vitamin D during anti-tuberculosis treatment improves the resolution of inflammation, including CXCL10 levels.3 1,25(OH)D acts through vitamin D receptors, and has a suppressive effect on IFN-c4 and on CXC chemokines such as CXCL105 in macrophages. In the present study, we investigated the relationship between CXCL10 and vitamin D at baseline and during treatment of TB.
ing patients with pulmonary TB,6 approved by the institutional review boards of the Aga Khan University and Dow University of Health Science, Karachi, Pakistan and listed on clinicaltrials.gov (NCT01130311). Consecutive adult patients aged 16 years with smear-positive active pulmonary TB diagnosed and enrolled within 1 week were included. Baseline clinical data, chest radiographs and sputum and blood samples for 25(OH)D and CXCL10 analysis were collected. Sputum microscopy for acid-fast bacilli (AFB) was performed after 0, 4, 8 and 12 weeks of treatment. Drug susceptibility testing of clinical specimens was not performed. All patients received standard DOTS treatment.6 Clinical examination was performed at all visits to calculate a TB score (range 0 – 13), a validated assessment tool to measure changes in the clinical status of TB patients.7 25(OH)D levels were classified as optimal .30 ng/ml, insufficient 20–30 ng/ml and deficient ,20 ng/ml.
STUDY POPULATION AND METHODS We describe results from the placebo arm of a randomised, placebo-controlled clinical trial involv-
RESULTS Among the 127 patients recruited for the study, the default rate was 8.1%, and 119 patients completed the study. Six-month treatment outcome data were available for 115 patients; 110 (95.6%) were successfully treated, while 5 were classified as
Correspondence to: Zahra Hasan, Department of Pathology and Microbiology, the Aga Khan University, Karachi, Pakistan. Tel: (þ92) 21 3486 1927. Fax: (þ92) 21 3493 4294. e-mail: [email protected] Article submitted 26 June 2013. Final version accepted 25 November 2013.
CXCL10 and vitamin D levels in TB
467
treatment failures (4.4%) and recommended for multidrug-resistant TB (MDR-TB) treatment.
between levels at weeks 0 and 12 was not statistically significant.
Association between clinical recovery, CXCL10 and vitamin D levels Microscopy data at 8 weeks were available for 113 patients: 93 (82%) were AFB-negative (converters), while 18 were positive (non-converters). There was no difference in median baseline CXCL10 (converter, 212 ng/ml [IQR 0–561] vs. non-converter, 184 ng/ml [IQR 6–661]) or 25(OH)D levels (converter, 23 ng/ml [IQR 4–28] vs. non-converter, 29 ng/ml [IQR 18–35], P ¼ 0.07) between the two groups. A comprehensive clinical evaluation of response to treatment based on TB scores7 before and after 12 weeks of treatment showed that TB scores decreased, as shown previously.6 Furthermore, individual TB scores were used to stratify patients into ‘responders’ (n ¼ 97) if their scores fell or ‘non-responders’ if their TB scores did not change (n ¼ 10) or if they increased (n ¼ 13) after 12 weeks of treatment (Table). The TB score remained higher in non-responders than in responders at 0 and 12 weeks. Vitamin D 25(OH)D levels were similar among responders and non-responders at 0 and 12 weeks. Following treatment, 25(OH)D levels were significantly decreased in all patients, with a 1.2-fold and 2.5-fold decrease in responders and non-responders, respectively. Serum CXCL10 levels were comparable in responders and non-responders at 0 and 12 weeks; they decreased with treatment in the overall group of patients, primarily in responders, with a 21-fold decrease in magnitude. In non-responders, CXCL10 levels decreased 8-fold; however, the difference
Treatment-induced decrease in CXCL10 levels in patients with sub-optimal vitamin D levels To examine the relationship between CXCL10 and 25(OH)D, we compared changes during treatment in patients classified according to their baseline 25(OH)D levels into ‘optimal’, ‘insufficient’ or ‘deficient’. Treatment induced a decrease in 25(OH)D levels in the optimal (P ¼ 0.001), insufficient (P ¼ 0.003) and deficient (P ¼ 0.019) groups (Figure, A). Post-treatment 25(OH)D levels were marginally different between the insufficient and deficient groups (P ¼ 0.051). Baseline CXCL10 levels were higher in the insufficient (3.8-fold) and deficient (5.5-fold) than in the optimal group (Figure, B). Furthermore, CXCL10 levels in the optimal group were lower than in the deficient group (P ¼ 0.031), with a decreasing trend as compared with the insufficient group (P ¼ 0.067) (Figure, B). After 12 weeks of treatment, a decrease in CXCL10 levels was evident in the insufficient and deficient groups. Posttreatment CXCL10 levels were higher in the deficient group than in the insufficient group (P ¼ 0.011). TB scores were not associated with either CXCL10 or 25(OH)D levels (data not shown). A weak inverse correlation was observed between baseline CXCL10 and 25(OH)D levels in the deficient group (q ¼ 0.324, P ¼ 0.032, Spearman’s rank correlation, Figure, C); however, this association was absent at 12 weeks.
Table Relationship between TBscore, serum CXCL10 and vitamin D levels during antituberculosis treatment Week 0 TBscores* All patients Responders Non-responders Vitamin D level, ng/ml All patients Responders Non-responders CXCL10 levels, pg/ml All patients Responders Non-responders
†
Median [IQR]
Week 12 ‡
P value
Median [IQR]†
P value‡
P value
0.0007
,0.001§ ,0.001§ 0.051
6 [5–8] 7 [5–9] 4 [4–6]
,0.0001
4 [2–6] 3 [1.5-3] 5 [5–6]
23 [14–29] 23 [15–28] 25 [12–32]
0.083
13 [8–22] 20 [14–24] 10 [7–18]
0.248
,0.001§ 0.001§ 0.036§
192 [0.1–621] 255 [0–679] 96 [0–391]
0.17
9 [0.1–151] 12 [0–168] 12 [0–151]
0.96
,0.001§ ,0.001§ 0.363
*TBscore is a validated assessment tool and can range from 0 to 13, from lowest to greatest severity.7 Data for all patients (n ¼ 120) are given, classified into responders (n ¼ 97) and non-responders (n ¼ 23). TBscore was found to decrease in responders after 12 weeks of treatment but did not decrease in non-responders. † Between 25th and 75th percentile for each data set. ‡ Comparison between responders and non-responders at each time interval studied. § Significant difference between week 0 and week 12 values (P , 0.05). Statistical comparisons performed using the Mann-Whitney U non-parametric test. TB ¼ tuberculosis; CXCL10 ¼ C-X-C motif chemokine 10; IQR ¼ interquartile range.
468
The International Journal of Tuberculosis and Lung Disease
Figure Treatment-induced decrease in CXCL10 levels is dependent on serum vitamin D levels. Patients were stratified into those with optimal (.30 ng/ml, n ¼ 22), insufficient (20–30 ng/ml, n ¼ 52) and deficient (,20 ng/ml, n ¼ 44) circulating 25(OH)D levels. 25(OH)D (A) and CXCL10 (B) levels were measured at baseline and after 12 weeks of anti-tuberculosis treatment. The box and whiskers plots depict data at 0 and 12 weeks. The plots define the 25th and 75th quartile range, with median values shown by a horizontal line. Baseline and 12-week data were compared by the Wilcoxon sign rank test, while comparison between groups was performed using MannWhitney U-test analysis. C) Spearman’s rank correlation between CXCL10 and 25(OH)D levels in the deficient group. *P 6 0.05 denotes a significant difference between 0 and 12 week data. †P , 0.05 denotes a significant difference between groups. 25(OH)D ¼ 25hydroxyvitamin D; CXCL10 ¼ C-X-C motif chemokine 10.
(insufficient or deficient) levels of 25(OH)D, but not in those who had optimal levels, likely due to exacerbated CXCL10 levels in the former groups. An inverse association between CXCL10 and 25(OH)D was observed at baseline. Mycobacterium tuberculosis-induced lymphokines activate vitamin D metabolism in host cells, which increase 1,25(OH)D levels.9 Expression of macrophage 25(OH)D hydroxylase (1a-hydroxylase) is upregulated by M. tuberculosis binding to toll-like receptors. 10 Inflammation is resolved during antituberculosis treatment.3 Higher baseline CXCL10 levels in the deficient as compared with the optimal 25(OH)D group would be associated with increased pathology and decreased anti-mycobacterial activity due to lowered 25(OH)D levels. This may explain why the increase in mycobacterial antigen-induced IFN-c responses as a result of vitamin D supplementation during anti-tuberculosis treatment is observed mainly in patients with deficient 25(OH)D levels.6
CONCLUSION Given the increasing interest in immune biomarkers of infection, it is important to consider baseline serum vitamin D levels when using chemokine biomarkers to make predictive or prognostic evaluations in TB. Acknowledgements
DISCUSSION We show that serum CXCL10 levels reflect treatment responses in TB but that these are affected by baseline 25(OH)D levels. In all patients there was a treatment-induced decrease in 25(OH)D levels, possibly as a consequence of rifampicin and isoniazid metabolism in the host.8 A significant decrease in CXCL10 was evident in patients with suboptimal
This work was supported through a University Research Council Grant, the Aga Khan University, Karachi, Pakistan. The authors thank A Raheem for statistical analysis. Conflict of interest: none declared
References 1 Walzl G, Ronacher K, Hanekom W, Scriba T J, Zumla A. Immunological biomarkers of tuberculosis. Nat Rev Immunol 2011; 11: 343–354. 2 Zhang R, Naughton D P. Vitamin D in health and disease: current perspectives. Nutr J 2010; 9: 65. 3 Coussens A K, Wilkinson R J, Hanifa Y, et al. Vitamin D accel-
CXCL10 and vitamin D levels in TB
erates resolution of inflammatory responses during tuberculosis treatment. Proc Natl Acad Sci USA 2012; 109: 15449–15454. 4 Martineau A R, Wilkinson K A, Newton S M, et al. IFNgamma- and TNF-independent vitamin D-inducible human suppression of mycobacteria: the role of cathelicidin LL-37. J Immunol 2007; 178: 7190–7198. 5 Selvaraj P, Harishankar M, Singh B, Banurekha V V, Jawahar M S. Effect of vitamin D3 on chemokine expression in pulmonary tuberculosis. Cytokine 2012; 60: 212–219. 6 Salahuddin N, Ali F, Hasan Z, Rao N, Aqeel M, Mahmood F. Vitamin D accelerates clinical recovery from tuberculosis: results of the SUCCINCT Study [supplementary cholecalciferol in recovery from tuberculosis]: a randomized, placebocontrolled, clinical trial of vitamin D supplementation in
7
8
9
10
469
patients with pulmonary tuberculosis. BMC Infect Dis 2013; 13: 22. Wesje C, Gustafson P, Nielsen J, et al. TBscore: signs and symptoms from tuberculosis patients in a low-resource setting have predictive value and may be used to assess clinical course. Scan J Immunol 2008; 40: 111–120. Brodie M J, Boobis A R, Hillyard C J, et al. Effect of rifampicin and isoniazid on vitamin D metabolism. Clin Pharmacol Ther 1982; 32: 525–530. Rook G A, Steele J, Fraher L, et al. Vitamin D3, gamma interferon, and control of proliferation of Mycobacterium tuberculosis by human monocytes. Immunology 1986; 57: 159–163. Liu P T, Stenger S, Li H, Wenzel L, et al. Toll-like receptor triggering of a vitamin D-mediated human antimicrobial response. Science 2006; 311: 1770–1773.
CXCL10 and vitamin D levels in TB
i
RESUME
Les taux s´eriques de la ch´emokine inflammatoire (la C-XC motif ch´emokine 10) CXCL10 sont augment´es en cas de tuberculose (TB). L’expression du g`ene CXCL10 dans les monocytes est att e´ nu e´ e par la vitamine D 3 m´etaboliquement active (1,25dihydroxy vitamine D). La stratification des patients en fonction de leur taux s´erique de 25hydroxyvitamine D (25[OH]D) ont montr´e que la diminution de la CXCL10 induite par le traitement anti-tuberculeux survenait chez les patients qui avaient
un taux « insuffisant » et « deficient », mais pas chez ceux ayant un taux de d´epart « optimal ». Dans le groupe « deficient », il existait une corr´elation inverse de la 25(OH) D avec la CXCL0. La CXCL10 pourrait donc constituer un biomarqueur utile pour suivre la r´eponse au traitement, mais elle doit eˆ tre interpr´et´ee en tenant compte du taux s´erique initial de vitamine D chez les patients atteints de TB.
RESUMEN
La concentracion ´ s´erica de la quimiocina inflamatoria (la C-X-C motif quimiocina 10) CXCL10 se encuentra aumentada en la tuberculosis (TB). La vitamina D3 metabolicamente ´ activa (1,25 dihidroxi-vitamina D) disminuye la expresi on ´ del gen que codifica la CXCL10 en los monocitos. La estratificacion ´ de los pacientes en funcion ´ de la concentracion ´ s´erica de 25hidroxivitamina D (25[OH]D) al comienzo del estudio revelo´ que la disminucion ´ de la CXCL10 inducida por el
tratamiento ocurre en los pacientes con concentraciones ‘insuficientes’ o ‘deficientes’ de 25(OH)D, pero no en los pacientes con ‘concentraciones optimas’. ´ En el grupo ‘deficiente’, se observo´ una correlacion ´ inversa con la concentracion ´ de CXCL10. Por esta razon ´ la quimiocina CXCL10 podr´ıa ser un marcador util ´ en el seguimiento de la respuesta al tratamiento, pero su interpretacion ´ debe tener en cuenta las concentraciones s´ericas iniciales de vitamina D de los pacientes con TB.
INT J TUBERC LUNG DIS 18(4):466–469 Q 2014 The Union http://dx.doi.org/10.5588/ijtld.13.0460
Change in serum CXCL10 levels during anti-tuberculosis treatment depends on vitamin D status Z. Hasan,* N. Salahuddin,† N. Rao,‡ M. Aqeel,† F. Mahmood,† F. Ali,† M. Ashraf,* F. Rahman,† S. Mahmood,* M. Islam,* B. Dildar,† T. Anwer,‡ F. Oiighor,† N. Sharif,† A. R. Ullah† *Department of Pathology and Microbiology, and †Department of Medicine, the Aga Khan University, Karachi, ‡ Ojha Institute for Chest Diseases, Dow University of Health Sciences, Karachi, Pakistan SUMMARY
Serum levels of the inflammatory C-X-C motif chemokine 10 (CXCL10) are raised in tuberculosis (TB). CXCL10 gene expression is downregulated in monocytes by metabolically active vitamin D3 (1,25dihydroxy vitamin D). Stratification of patients by serum 25hydroxyvitamin D (25[OH]D) levels at baseline showed that treatment-induced decrease in CXCL10 occurred in those with ‘insufficient’ and ‘deficient’ but not in those
with ‘optimal’ levels. In the deficient group, 25(OH)D showed an inverse correlation with CXCL10 levels. CXCL10 may thus be a useful biomarker for the followup of response to treatment. However, CXCL10 levels should be interpreted taking into account the baseline serum vitamin D levels of the TB patients. K E Y W O R D S : CXCL10; vitamin D; tuberculosis; biomarker
ALTHOUGH the diagnosis and treatment of tuberculosis (TB) is monitored clinically and bacteriologically, it may be difficult to assess, particularly in the case of extra-pulmonary TB. Cytokine and chemokine biomarkers can aid in disease identification and treatment.1 Interferon-gamma (IFN-c) induced protein 10 (IP-10 or C-X-C motif chemokine 10 [CXCL10]), produced by T-cells, is increased in patients with active TB, and decreases with treatment.1 Deficiencies in vitamin D3 (25-hydroxycholecalciferol) levels have been implicated in the development of TB.2 The primary circulating form of vitamin D3, 25-hydroxyvitamin D (25[OH]D), is hydroxylated to physiologically active 1-a,25-dihydroxy vitamin D (1,25[OH]D).2 Supplementation with vitamin D during anti-tuberculosis treatment improves the resolution of inflammation, including CXCL10 levels.3 1,25(OH)D acts through vitamin D receptors, and has a suppressive effect on IFN-c4 and on CXC chemokines such as CXCL105 in macrophages. In the present study, we investigated the relationship between CXCL10 and vitamin D at baseline and during treatment of TB.
ing patients with pulmonary TB,6 approved by the institutional review boards of the Aga Khan University and Dow University of Health Science, Karachi, Pakistan and listed on clinicaltrials.gov (NCT01130311). Consecutive adult patients aged 16 years with smear-positive active pulmonary TB diagnosed and enrolled within 1 week were included. Baseline clinical data, chest radiographs and sputum and blood samples for 25(OH)D and CXCL10 analysis were collected. Sputum microscopy for acid-fast bacilli (AFB) was performed after 0, 4, 8 and 12 weeks of treatment. Drug susceptibility testing of clinical specimens was not performed. All patients received standard DOTS treatment.6 Clinical examination was performed at all visits to calculate a TB score (range 0 – 13), a validated assessment tool to measure changes in the clinical status of TB patients.7 25(OH)D levels were classified as optimal .30 ng/ml, insufficient 20–30 ng/ml and deficient ,20 ng/ml.
STUDY POPULATION AND METHODS We describe results from the placebo arm of a randomised, placebo-controlled clinical trial involv-
RESULTS Among the 127 patients recruited for the study, the default rate was 8.1%, and 119 patients completed the study. Six-month treatment outcome data were available for 115 patients; 110 (95.6%) were successfully treated, while 5 were classified as
Correspondence to: Zahra Hasan, Department of Pathology and Microbiology, the Aga Khan University, Karachi, Pakistan. Tel: (þ92) 21 3486 1927. Fax: (þ92) 21 3493 4294. e-mail: [email protected] Article submitted 26 June 2013. Final version accepted 25 November 2013.
CXCL10 and vitamin D levels in TB
467
treatment failures (4.4%) and recommended for multidrug-resistant TB (MDR-TB) treatment.
between levels at weeks 0 and 12 was not statistically significant.
Association between clinical recovery, CXCL10 and vitamin D levels Microscopy data at 8 weeks were available for 113 patients: 93 (82%) were AFB-negative (converters), while 18 were positive (non-converters). There was no difference in median baseline CXCL10 (converter, 212 ng/ml [IQR 0–561] vs. non-converter, 184 ng/ml [IQR 6–661]) or 25(OH)D levels (converter, 23 ng/ml [IQR 4–28] vs. non-converter, 29 ng/ml [IQR 18–35], P ¼ 0.07) between the two groups. A comprehensive clinical evaluation of response to treatment based on TB scores7 before and after 12 weeks of treatment showed that TB scores decreased, as shown previously.6 Furthermore, individual TB scores were used to stratify patients into ‘responders’ (n ¼ 97) if their scores fell or ‘non-responders’ if their TB scores did not change (n ¼ 10) or if they increased (n ¼ 13) after 12 weeks of treatment (Table). The TB score remained higher in non-responders than in responders at 0 and 12 weeks. Vitamin D 25(OH)D levels were similar among responders and non-responders at 0 and 12 weeks. Following treatment, 25(OH)D levels were significantly decreased in all patients, with a 1.2-fold and 2.5-fold decrease in responders and non-responders, respectively. Serum CXCL10 levels were comparable in responders and non-responders at 0 and 12 weeks; they decreased with treatment in the overall group of patients, primarily in responders, with a 21-fold decrease in magnitude. In non-responders, CXCL10 levels decreased 8-fold; however, the difference
Treatment-induced decrease in CXCL10 levels in patients with sub-optimal vitamin D levels To examine the relationship between CXCL10 and 25(OH)D, we compared changes during treatment in patients classified according to their baseline 25(OH)D levels into ‘optimal’, ‘insufficient’ or ‘deficient’. Treatment induced a decrease in 25(OH)D levels in the optimal (P ¼ 0.001), insufficient (P ¼ 0.003) and deficient (P ¼ 0.019) groups (Figure, A). Post-treatment 25(OH)D levels were marginally different between the insufficient and deficient groups (P ¼ 0.051). Baseline CXCL10 levels were higher in the insufficient (3.8-fold) and deficient (5.5-fold) than in the optimal group (Figure, B). Furthermore, CXCL10 levels in the optimal group were lower than in the deficient group (P ¼ 0.031), with a decreasing trend as compared with the insufficient group (P ¼ 0.067) (Figure, B). After 12 weeks of treatment, a decrease in CXCL10 levels was evident in the insufficient and deficient groups. Posttreatment CXCL10 levels were higher in the deficient group than in the insufficient group (P ¼ 0.011). TB scores were not associated with either CXCL10 or 25(OH)D levels (data not shown). A weak inverse correlation was observed between baseline CXCL10 and 25(OH)D levels in the deficient group (q ¼ 0.324, P ¼ 0.032, Spearman’s rank correlation, Figure, C); however, this association was absent at 12 weeks.
Table Relationship between TBscore, serum CXCL10 and vitamin D levels during antituberculosis treatment Week 0 TBscores* All patients Responders Non-responders Vitamin D level, ng/ml All patients Responders Non-responders CXCL10 levels, pg/ml All patients Responders Non-responders
†
Median [IQR]
Week 12 ‡
P value
Median [IQR]†
P value‡
P value
0.0007
,0.001§ ,0.001§ 0.051
6 [5–8] 7 [5–9] 4 [4–6]
,0.0001
4 [2–6] 3 [1.5-3] 5 [5–6]
23 [14–29] 23 [15–28] 25 [12–32]
0.083
13 [8–22] 20 [14–24] 10 [7–18]
0.248
,0.001§ 0.001§ 0.036§
192 [0.1–621] 255 [0–679] 96 [0–391]
0.17
9 [0.1–151] 12 [0–168] 12 [0–151]
0.96
,0.001§ ,0.001§ 0.363
*TBscore is a validated assessment tool and can range from 0 to 13, from lowest to greatest severity.7 Data for all patients (n ¼ 120) are given, classified into responders (n ¼ 97) and non-responders (n ¼ 23). TBscore was found to decrease in responders after 12 weeks of treatment but did not decrease in non-responders. † Between 25th and 75th percentile for each data set. ‡ Comparison between responders and non-responders at each time interval studied. § Significant difference between week 0 and week 12 values (P , 0.05). Statistical comparisons performed using the Mann-Whitney U non-parametric test. TB ¼ tuberculosis; CXCL10 ¼ C-X-C motif chemokine 10; IQR ¼ interquartile range.
468
The International Journal of Tuberculosis and Lung Disease
Figure Treatment-induced decrease in CXCL10 levels is dependent on serum vitamin D levels. Patients were stratified into those with optimal (.30 ng/ml, n ¼ 22), insufficient (20–30 ng/ml, n ¼ 52) and deficient (,20 ng/ml, n ¼ 44) circulating 25(OH)D levels. 25(OH)D (A) and CXCL10 (B) levels were measured at baseline and after 12 weeks of anti-tuberculosis treatment. The box and whiskers plots depict data at 0 and 12 weeks. The plots define the 25th and 75th quartile range, with median values shown by a horizontal line. Baseline and 12-week data were compared by the Wilcoxon sign rank test, while comparison between groups was performed using MannWhitney U-test analysis. C) Spearman’s rank correlation between CXCL10 and 25(OH)D levels in the deficient group. *P 6 0.05 denotes a significant difference between 0 and 12 week data. †P , 0.05 denotes a significant difference between groups. 25(OH)D ¼ 25hydroxyvitamin D; CXCL10 ¼ C-X-C motif chemokine 10.
(insufficient or deficient) levels of 25(OH)D, but not in those who had optimal levels, likely due to exacerbated CXCL10 levels in the former groups. An inverse association between CXCL10 and 25(OH)D was observed at baseline. Mycobacterium tuberculosis-induced lymphokines activate vitamin D metabolism in host cells, which increase 1,25(OH)D levels.9 Expression of macrophage 25(OH)D hydroxylase (1a-hydroxylase) is upregulated by M. tuberculosis binding to toll-like receptors. 10 Inflammation is resolved during antituberculosis treatment.3 Higher baseline CXCL10 levels in the deficient as compared with the optimal 25(OH)D group would be associated with increased pathology and decreased anti-mycobacterial activity due to lowered 25(OH)D levels. This may explain why the increase in mycobacterial antigen-induced IFN-c responses as a result of vitamin D supplementation during anti-tuberculosis treatment is observed mainly in patients with deficient 25(OH)D levels.6
CONCLUSION Given the increasing interest in immune biomarkers of infection, it is important to consider baseline serum vitamin D levels when using chemokine biomarkers to make predictive or prognostic evaluations in TB. Acknowledgements
DISCUSSION We show that serum CXCL10 levels reflect treatment responses in TB but that these are affected by baseline 25(OH)D levels. In all patients there was a treatment-induced decrease in 25(OH)D levels, possibly as a consequence of rifampicin and isoniazid metabolism in the host.8 A significant decrease in CXCL10 was evident in patients with suboptimal
This work was supported through a University Research Council Grant, the Aga Khan University, Karachi, Pakistan. The authors thank A Raheem for statistical analysis. Conflict of interest: none declared
References 1 Walzl G, Ronacher K, Hanekom W, Scriba T J, Zumla A. Immunological biomarkers of tuberculosis. Nat Rev Immunol 2011; 11: 343–354. 2 Zhang R, Naughton D P. Vitamin D in health and disease: current perspectives. Nutr J 2010; 9: 65. 3 Coussens A K, Wilkinson R J, Hanifa Y, et al. Vitamin D accel-
CXCL10 and vitamin D levels in TB
erates resolution of inflammatory responses during tuberculosis treatment. Proc Natl Acad Sci USA 2012; 109: 15449–15454. 4 Martineau A R, Wilkinson K A, Newton S M, et al. IFNgamma- and TNF-independent vitamin D-inducible human suppression of mycobacteria: the role of cathelicidin LL-37. J Immunol 2007; 178: 7190–7198. 5 Selvaraj P, Harishankar M, Singh B, Banurekha V V, Jawahar M S. Effect of vitamin D3 on chemokine expression in pulmonary tuberculosis. Cytokine 2012; 60: 212–219. 6 Salahuddin N, Ali F, Hasan Z, Rao N, Aqeel M, Mahmood F. Vitamin D accelerates clinical recovery from tuberculosis: results of the SUCCINCT Study [supplementary cholecalciferol in recovery from tuberculosis]: a randomized, placebocontrolled, clinical trial of vitamin D supplementation in
7
8
9
10
469
patients with pulmonary tuberculosis. BMC Infect Dis 2013; 13: 22. Wesje C, Gustafson P, Nielsen J, et al. TBscore: signs and symptoms from tuberculosis patients in a low-resource setting have predictive value and may be used to assess clinical course. Scan J Immunol 2008; 40: 111–120. Brodie M J, Boobis A R, Hillyard C J, et al. Effect of rifampicin and isoniazid on vitamin D metabolism. Clin Pharmacol Ther 1982; 32: 525–530. Rook G A, Steele J, Fraher L, et al. Vitamin D3, gamma interferon, and control of proliferation of Mycobacterium tuberculosis by human monocytes. Immunology 1986; 57: 159–163. Liu P T, Stenger S, Li H, Wenzel L, et al. Toll-like receptor triggering of a vitamin D-mediated human antimicrobial response. Science 2006; 311: 1770–1773.
CXCL10 and vitamin D levels in TB
i
RESUME
Les taux s´eriques de la ch´emokine inflammatoire (la C-XC motif ch´emokine 10) CXCL10 sont augment´es en cas de tuberculose (TB). L’expression du g`ene CXCL10 dans les monocytes est att e´ nu e´ e par la vitamine D 3 m´etaboliquement active (1,25dihydroxy vitamine D). La stratification des patients en fonction de leur taux s´erique de 25hydroxyvitamine D (25[OH]D) ont montr´e que la diminution de la CXCL10 induite par le traitement anti-tuberculeux survenait chez les patients qui avaient
un taux « insuffisant » et « deficient », mais pas chez ceux ayant un taux de d´epart « optimal ». Dans le groupe « deficient », il existait une corr´elation inverse de la 25(OH) D avec la CXCL0. La CXCL10 pourrait donc constituer un biomarqueur utile pour suivre la r´eponse au traitement, mais elle doit eˆ tre interpr´et´ee en tenant compte du taux s´erique initial de vitamine D chez les patients atteints de TB.
RESUMEN
La concentracion ´ s´erica de la quimiocina inflamatoria (la C-X-C motif quimiocina 10) CXCL10 se encuentra aumentada en la tuberculosis (TB). La vitamina D3 metabolicamente ´ activa (1,25 dihidroxi-vitamina D) disminuye la expresi on ´ del gen que codifica la CXCL10 en los monocitos. La estratificacion ´ de los pacientes en funcion ´ de la concentracion ´ s´erica de 25hidroxivitamina D (25[OH]D) al comienzo del estudio revelo´ que la disminucion ´ de la CXCL10 inducida por el
tratamiento ocurre en los pacientes con concentraciones ‘insuficientes’ o ‘deficientes’ de 25(OH)D, pero no en los pacientes con ‘concentraciones optimas’. ´ En el grupo ‘deficiente’, se observo´ una correlacion ´ inversa con la concentracion ´ de CXCL10. Por esta razon ´ la quimiocina CXCL10 podr´ıa ser un marcador util ´ en el seguimiento de la respuesta al tratamiento, pero su interpretacion ´ debe tener en cuenta las concentraciones s´ericas iniciales de vitamina D de los pacientes con TB.
