Circulating levels of insulin-like growth factor-I (IGF-I) correlate with disease status in leprosy

The study population consisted of 21 healthy controls (HC) from endemic laboratory staff volunteers (3 males, 18 females, ranging in age from 19 to 62) and 93 patients (50 males, 43 females, in a 18-65 age range) referred to the Souza Araújo Ambulatory (Reference Center for Leprosy Diagnosis and Treatment, Oswaldo Cruz Foundation, Rio de Janeiro, RJ, Brazil) for the diagnosis and treatment of leprosy. Healthy controls and patients were carefully screened by anthropometric measurements (body mass index, BMI) and biochemical analysis; and those with diabetes, obesity or cholesterol disorders were excluded from the study. Each patient was assessed clinically throughout treatment and detailed analyses of medical and dermatological examinations were routinely carried out. Bacteriological examinations of slit-skin smears were performed to determine BI. The patients were categorized according to the Ridley and Jopling's classification scale [3] into borderline tuberculoid (BT), borderline lepromatous (BL), or lepromatous (LL). The diagnosis of leprosy Type 1 reaction (RR) was determined by the appearance of erythema and edema in either existing or new skin lesions. The diagnosis of leprosy Type 2 reaction (ENL) was primarily based on the acute occurrence of nodular skin lesions, accompanied by fever with or without peripheral nerve pain and/or nerve dysfunction. Nonreactional (NR) groups were selected among leprosy patients who were rendered the same histopathological classification as reactional (R) patients but who did not experience reaction at initial diagnosis or during follow-up. These NR patients were distributed into the following groups: 13 nonreactional BT (NR BT); 19 nonreactional BL (NR BL); 18 nonreactional LL (NR LL) patients, 25 BL who had RR during treatment (R BL); and 18 LL who developed ENL during treatment (R LL). Blood was collected at three different time points: i) at the beginning of leprosy treatment, ii) at diagnosis of a reactional episode if one occurred; and iii) 3-5 years post-treatment. Blood samples were taken from 1995 to 2005; and patient sera were extracted and stored at -20°C until retrieval for analysis. The baseline characteristics of each group of individuals included in the study are shown in Table 1.

Patients received World Health Organization-recommended leprosy multidrug therapy (MDT). Reactions were treated with prednisolone, starting at 40 mg/day, and gradually tapering off over a 12-week period.

This study was approved by the Ethics Committee of the Oswaldo Cruz Foundation. Informed written consent was obtained from all patients or their guardians and the healthy controls prior to specimen collection.

The serum levels of IGF-I, IGFBP-3, and TNF-α were measured by the IMMULITE 2000 Analyzer (EuroDPC Med Limited, Llanberis, UK). Assays were carried out using the solid phase, enzyme-labeled, chemiluminescent-immunometric method in accordance with the manufacturer's instructions. IGF-I and IGFBP-3 levels were considered normal according to the reference range by age provided by the kit manufacturer [14].

Data are expressed as median, mean ± S.D. values. Group comparisons were evaluated by the analysis of a variance (ANOVA) test, using age as a covariate, followed by Bonferroni's test for multiple comparisons. Paired t test was employed to compare IGF-I, IGFBP3, and TNF-α levels before and during reaction. Pearson partial correlation coefficient (controlled for age) was calculated between BI and IGF-I and IGFBP-3 levels. All statistical calculations were done via the SPSS software program; and p values lower than 0.05 were considered statistically significant.

Evidence in the literature has shown that the somatotropic axis is responsible for modulating the immune system, directly influencing both the humoral and cellular immune (CMI) responses [15, 16]. Since the spectral clinical forms of leprosy occur as a result of the capacity of the host to mount anywhere from low- (LL)-to-high (TT) CMI responses against ML, the serum levels of IGF-I and IGFBP-3 were compared among patients having the different clinical forms at the pre-MDT stage. Only those who did not undergo reactional episodes during the course of the disease were included in this analysis, and the levels found in leprosy patients were compared with those of the HCs. The IGF-I levels were significantly lower in NR BL and NR LL patients when compared to the NR BT (P = 0.001) and HC (P < 0.001) groups (Figure 1a). A decrease in IGFBP-3 levels was also observed in NR LL patients when compared to the NR BT (P < 0.001) and HC (P = 0.001) individuals (Figure 1b). According to the reference range by age, 83,3% (10/12) and 75% (12/16) of the BL and LL patients, respectively, showed below-normal levels of serum IGF-I. Moreover, via Pearson's partial correlation analysis, a weak inverse correlation was found between elevated BI and reduced IGF-I (r = -0.56, P < 0.001) and IGFBP-3 (r = -0.47, P < 0.001) levels in these patients (data not shown).

Next, to reinforce the notion that the low IGF-I levels observed in the LL patients resulted from active modulation by ML infection, their hormone levels were measured at the conclusion of MDT. Interestingly, when serum IGF-I was evaluated in a group of nonreactional BL/LL patients 3-5 years post-MDT treatment (n = 9), significantly higher levels (P < 0.001) were observed in comparison to those found in the group evaluated at diagnosis (n = 28) (Figure 1c), with 78% of this post-MDT nonreactional BL/LL fitting into the normal range. Increment of serum IGF-I levels after treatment can be clearly seen in paired pre- and post-MDT serum samples taken from the same patient (Figure 1d).

Since the somatrotropic axis plays an important regulatory role during host immune-inflammatory responses to infection, we then investigated circulating IGF-I and IGFBP-3 levels in the context of ENL episodes. This type of inflammatory episode occurs principally in LL patients [6, 17], the main concern being the management of leprosy care to prevent disabilities [18]. For this purpose, a group of LL patients who suffered ENL episodes during MDT (referred to as R LL; n = 18) were enrolled in the study. Serum samples of the R LL patients were obtained at two different time points: i) at the onset of MDT, at which time no signs of reaction were detected (referred to as R LL_{t = 0}); and ii) during reactions (referred to as R LL_ENL) taking place between 3 weeks and 2 years after the beginning of treatment (mean average of 11.5 ± 9.3 months).

It was first investigated whether reactional LL patients (clinically unstable patients) already display distinct levels of IGF-I and IGFBP-3 at diagnosis in comparison to nonreactional LL patients (a clinically stable group). Notably, the R LL_{t = 0} samples presented significantly higher levels of IGF-I and IGFBP-3 than the NR LL group (P < 0.01), as can be seen in Figure 2a, b, respectively. The behavior of these proteins was subsequently monitored in the R LL group during ENL. Interestingly, these levels decreased dramatically (R LL_ENL; P < 0.05) during reaction, reaching values similar to those detected in the NR LL group, as observed in Figure 2c, d.

TNF-α was also quantified in the same serum samples due to the well-established role of this pro-inflammatory cytokine in ENL [19, 20] along with its antagonistic activity in IGF-I [21]. Similar range levels of TNF-α were observed in NR LL and R LL_{t = 0} serum samples (data not shown). However, consistent with previous published data [19, 20], TNF-α levels were higher during ENL (P < 0.05) in comparison to those measured prior to reaction (Figure 2e). Taken together, these results demonstrate that the IGF-I and IGFBP-3 circulating levels at disease diagnosis were significantly higher in the LL patients who developed ENL than among the LL patients who did not experience reaction during MDT. In the course of ENL episodes, however, these levels tended to decrease, attaining levels similar to those found among the NR LL patients.

Next, the IGF-I and IGFBP-3 levels in the context of RR episodes, another frequent type of acute inflammatory episode in leprosy, were investigated. Type 1 reaction, or RR, more common among BL patients, affects approximately one-third of these patients, especially in the first year of MDT [22]. For this reason, a group of BL who underwent RR episodes during MDT (referred to as R BL; n = 25) was included in this analysis. Serum samples were obtained at 2 different time points in R BL patients: i) at the onset of MDT, when no signs of reaction were detected (referred to as R BL_{t = 0}); and ii) during reaction (referred to as R BL_RR) occurring between 2 weeks and 18 months after the beginning of treatment (mean average of 6.6 ± 4.8 months).

As a first step, circulating levels of IGF-I and IGFBP-3 in R BL vs. NR BL patients at disease diagnosis were compared. In contrast to the differences observed for LL patients in the context of ENL, similar levels of IGF-I and IGFBP-3 were detected in NR BL and R BL patients (Figure 3a; data not shown). The next step involved measuring the levels of IGF-I and IGFBP-3 before and during RR episodes in R BL patients (n = 15). Curiously, in contrast to the decreasing levels observed during ENL, IGF-I levels significantly increased during RR (P < 0.05), as shown by the individual behavior of each patient displayed in Figure 3b. However, no significant alterations were observed in IGFBP-3 and TNF-α levels during RR (data not shown).