MONTENEGRO SKIN TEST AND AGE OF SKIN LESION AS PREDICTORS OF TREATMENT FAILURE IN CUTANEOUS LEISHMANIASIS

Abstract

A case-control study was conducted to examine the association among the Montenegro skin test (MST), age of skin lesion and therapeutic response in patients with cutaneous leishmaniasis (CL) treated at Evandro Chagas National Institute of Infectious Diseases (INI), Oswaldo Cruz Foundation (FIOCRUZ), Rio de Janeiro, Brazil. For each treatment failure (case), two controls showing skin lesion healing following treatment, paired by sex and age, were randomly selected. All patients were treated with 5 mg Sb⁵⁺/kg/day of intramuscular meglumine antimoniate (Sb⁵⁺) for 30 successive days. Patients with CL were approximately five times more likely to fail when lesions were less than two months old at the first appointment. Patients with treatment failure showed less intense MST reactions than patients progressing to clinical cure. For each 10 mm of increase in MST response, there was a 26% reduction in the chance of treatment failure. An early treatment - defined as a treatment applied for skin lesions, which starts when they are less than two months old at the first appointment -, as well as a poor cellular immune response, reflected by lower reactivity in MST, were associated with treatment failure in cutaneous leishmaniasis.

INTRODUCTION

American tegumentary leishmaniasis (ATL) is an anthropozoonosis widely spread throughout Brazil. There were 21,866 new cases in 2010^5,8, with two main forms: cutaneous leishmaniasis (CL) and mucosal leishmaniasis. The localized form of CL, presenting as single or multiple ulcers, is the most typical one⁵.

Culture of biopsy specimens of suspected lesions and the direct smear of material obtained from the lesion are the main confirmatory tests. Diagnosis may also be established by detection of parasites in histological specimens and by the polymerase chain reaction¹³. Due to high cost and technical complexity of tests, culture parasite and molecular detection of Leishmania in Brazil are restricted to the reference centers for this disease^5,13. The Montenegro skin test (MST), with a sensitivity rate of 86.4% up to 100%^6,15,23,29, is the main diagnostic test in primary care. In the presence of a suspicious cutaneous lesion, MST supports the diagnosis of Leishmania infection. Skin reactions to MST ≥ 5 mm are considered positive and < 5 mm are considered negative⁵. Patients with negative MST and diagnostic confirmation by other tests are more prone to relapse²². MST is a marker of cellular immune response¹⁰, evaluating type IV hypersensitivity response mediated by T cells²⁸. However, it may be influenced by clinical presentation¹⁸ and by disease duration^7,15.

The first choice for CL drug treatment in Brazil is meglumine antimoniate (Sb⁵⁺)⁵. The Brazilian Ministry of Health (2010) recommends 10 - 20 mg Sb⁵⁺/kg/day for 20 days. However, clinical studies in Rio de Janeiro suggest that 5 mg Sb⁵⁺/kg/day for 30 days may be easier to deliver, allowing effective treatment of the disease, with lower toxicity and cost^19–21.

Therapeutic failure is defined by DEPS et al. (2000) as lack of clinical response during and after the treatment, with stabilization or worsening of the ulcerated lesion¹². In individuals treated with meglumine antimoniate with 10-20 mg Sb⁵⁺/kg/day, therapeutic failure varies from 46% to 75%^12,25,31,32. Factors related to therapeutic failure include: three or more cutaneous lesions, prior treatment to leishmaniasis, weight exceeding 68 kg and irregular treatment²⁷.

In this study, the association among MST, age of cutaneous lesion and therapeutic response in patients with CL was evaluated.

MATERIALS AND METHODS

1. Study design: This was an observational retrospective case-control study, with localized CL patients who attended the Leishmaniasis Surveillance Laboratory (Lab Vigileish) at Evandro Chagas National Institute of Infectious Diseases (INI), Oswaldo Cruz Foundation (FIOCRUZ), Rio de Janeiro, Brazil, between January 1989 and December 2009.

2. Inclusion criteria: Patients with CL diagnosis made by parasitological criteria (culture of Leishmania in appropriate media, parasites seen at histological exam or direct smear and/or imprint of a biopsy specimen; or Leishmania DNA detection by polymerase chain reaction) and/or epidemiological (patients living or travelling to endemic areas) and/or clinical (suggestive cutaneous lesions) and/or immunological criteria (Leishmania serology – indirect immunofluorescence reaction and/or immunoenzymatic assay) were included. They were treated with intramuscular meglumine antimoniate 5 mg Sb⁵⁺/kg/day for 30 consecutive days²⁹.

3. Exclusion criteria: Patients with other clinical forms of CL, without register of MST in millimeters, lack of post-treatment follow-up for at least two years, and/or those presenting relapse after an initial response (temporary healing) were excluded.

4. Selection of cases and controls: Therapeutic failures, defined as the non-occurrence of progression to definitive scar formation after the initial treatment, were considered as cases.

Patients with permanent clinical cure, defined as the epithelization of the ulcerated lesions with progressive disappearance of crusts, scaling, infiltration and erythema, with a final establishment of an atrophic, hypertrophic or imperceptible scar, as well as the lack of mucosal lesion after the initial treatment, were considered as controls. Controls were randomly selected among the patients with clinical cure and were matched by sex and age range to the cases in the proportion of 2:1. Matching was performed according to the following age ranges: 5-9; 10-14; 15-19; 20-24; 25-29; 30-34; 35-39; 40-45; 46-50; 51-55 years.

5. Data analyses: Description of the data was made using simple frequencies of categorical variables (most probable area of infection; circumstances of exposure; results of MST, parasitological, serological and histopathological exams; co-morbidities; number of lesions; site of lesions; and age of lesions), and with average and minimum/maximum of continuous variables (number of lesions; MST in millimeters; duration of disease before treatment; and time until the occurrence of therapeutic failure).

The McNemar test was used for comparison of categorical variables between cases and controls, and the Wilcoxon test was used for continuous variables, when normality hypothesis was rejected by the Kolmogorov-Smirnov test.

Simple models of binary conditional logistic regression were used to evaluate the effect of each variable in the occurrence of therapeutic failure (age of cutaneous lesion, number of lesions, site of lesions, co-morbidities, MST in millimeters, and positivity of serology, histopathology and culture). After that, a multiple model was performed to therapeutic failure, adjusted by the same characteristics. The likelihood ratio test was used to indicate the best adjusted model. Odds ratio (OR) and its respective confidence interval (CI) was used as a measure of the association among the variables in the regression models.

MST in millimeters was used and, due to the lack of normal distribution, it was performed using its logarithm. Consequently, the application of anti-logarithm was used to the analysis of the results.

p-values < 0.05 indicated significant association in the performed tests.

Data were analyzed with the Statistical Package for the Social Sciences (SPSS-WIN) version 16.0 and STATA 10.0 softwares.

RESULTS

From 800 ATL patients followed-up in the studied period, 634 (79.25%) presented CL. Four hundred and seven individuals were excluded from the study (Fig. 1).

Fig. 1 -. Selection of cases and controls after analysis of the Data Bank of Lab Vigileish, INI/FIOCRUZ, 1989 to 2009, Rio de Janeiro, Brazil.

From 227 eligible patients, 32 (14.1%) evidenced therapeutic failure, and 64 matched controls were selected, with 96 patients being included in the study. Studied population included 59.4% men, with age varying between five and 55 years. For 89.6%, the probable place of infection was Rio de Janeiro State and 64.6% were from rural areas. Ninety seven per cent of the studied patients lived and/or had previously been in endemic areas for the disease. All patients with negative MST resided in endemic areas. Three of the subjects had no data concerning the place of residence or travels to endemic areas, and all of them showed strong MST reaction.

Main exposure motive in 93.8% of the patients was domicile. Co-morbidities were observed in 34.4% of the patients, the most frequent being anemia, arterial hypertension, thyroid disease, atopic disease, smoking and tuberculosis. Data regarding HIV-serology, use of drugs or other diseases that could compromise immunity were not found in the studied records.

Culture, serology and histopathology for leishmaniasis were positive in 78.1%, 62.5% and 63.5%, respectively. Among the case group, all patients had parasitological confirmed diagnosis. In the control group, 60 patients had parasitological confirmation, two had epidemiological, clinical and serological diagnosis, and two had epidemiological and clinical diagnosis. As a prerequisite for inclusion in the control group, all patients had had a good response to antimonial treatment. All of the four control subjects who had no parasitological confirmed diagnosis underwent mycological examination with negative results; all of them had a positive MST; and two had chronic granulomatous infiltrate in histopathology of the cutaneous lesion, suggestive of leishmaniasis. No significant difference among these variables and the occurrence of therapeutic failure was found.

Average time to detection of therapeutic failure after the end of the treatment was 93 days, varying from 18 to 316 days. In the control group, the average in days until definitive scar formation was 194 days and varied from 51 to 418 days.

There was a significant difference in cases regarding the age of the skin lesion prior to the diagnosis. Cases with lesions that were less than two months old at the first appointment were 71.9%, compared to 35.9% for controls (p < 0.001).

MST reactivity in millimeters also exhibited significant difference between cases and controls (p = 0.027). In cases, MST varied from 0 to 60 mm, with median of 13.5 mm; and in controls, it varied from 0 to 50 mm, with median of 17.5 mm. However, significant difference between groups was not observed when comparing MST using 5 mm as a cut-off (positive or negative) (p = 0.289).

Age of skin lesion and the LogMST were significant regarding the occurrence of therapeutic failure in the simple model of conditional logistic regression, p < 0.001 and p = 0.026, respectively. Other variables (number of lesions, site of the lesion, co-morbidities, serology, histopathology and culture) were not statistically significant in the simple regression model (Table 1).

Table 1. Odds ratio (non-adjusted) of the conditional simple logistic regression model.

Variables	ORa	95% CI	p-value
Evolution time before treatment	5.64	[2.35 - 13.53]	< 0.001
LogMST b	0.32	[0.12 - 0.87]	0.026
Number of lesions	1.19	[0.53 - 2.68]	0.679
Site of the lesion	1.15	[0.56 - 2.37]	0.712
Co-morbidities	0.60	[0.25 - 1.47]	0.268
Serology positivity	1.76	[0.55 - 5.64]	0.345
Histopathology positivity	1.75	[0.70 - 4.34]	0.229
Culture positivity	1.23	[0.26 - 5.69]	0.793

^aodds ratio;

^bLog skin test Montenegro results in millimiters.

The likelihood ratio test indicated a final multiple logistic model constituted by the LogMST in millimeters (adjusted OR: 0.26 / CI: [0.09 - 0.77] / p = 0.015) and by the age of the skin lesion (adjusted OR: 6.33 / CI: [2.52 - 15.90] / p < 0.001), and there was no significant interaction between these two variables. Patients with skin lesions less than two months old had 5.33 more chances of therapeutic failure than those with skin lesions more than two months old. For each 10 millimeters of increase in MST, there was a decrease of 26% in the chances of occurrence of failure.

DISCUSSION

This study evaluated the association between therapeutic response and clinical and laboratory parameters, defined as age of skin lesion and MST.

An inverse association between the intensity of MST reaction and the occurrence of therapeutic failure was observed, with strong reactors to MST being less susceptible to the occurrence of treatment failure.

There was a significantly higher chance of development of therapeutic failure in patients with lesions less than two months old at the first appointment.

The percentage of therapeutic failure (14%) was lower than reported in previous studies with meglumine antimoniate treatment (46% to 75%)^12,27,29,33. Low failure rates found in this study support the effectiveness of low-dose treatment (5 mg Sb⁵⁺/kg/day) previously reported in studies at INI/FIOCRUZ^19,29.

The fact that the majority of the studied population acquired the infection at their domicile is consistent with the peri-urban epidemiological profile of the disease in Rio de Janeiro⁶. Most of the subjects were from the state of Rio de Janeiro, where Leishmania (Viannia) braziliensis is responsible for almost all cases¹¹.

MST has been used since the 1920s, when it was developed by MONTENEGRO¹⁷. The importance of the standardization of the technique (application and reading) of MST and of the used antigen for its performance has been highlighted by several authors. The test must be performed by a trained technician, in order to assure comparable results and to avoid common causes of variation in the execution and interpretation of cutaneous tests, as follows: the amount of injected antigen; the site and deepness of the injection; the physiological status of the patient; the used vehicle; and the observer performing the measure^3,16,35. Patients seen at Lab Vigileish, who performed MST in the institution, are always submitted to a standard protocol, ensuring the quality of the assistance in the test.

The safety and quality of the used antigen was ensured with the use of the antigen delivered by the Brazilian Ministry of Health, produced on a large scale within the standards of good practices¹. Although the used strain is Leishmania (Leishmania) amazonensis, it is produced to be used throughout the country, regardless of the species of Leishmania causing disease.

The main difficulty found in the study was the incomplete patients' records. The use of electronic records, however, turned easier the process of data collection.

Study population of referral centers have characteristics which may differ from the general population, and this is a possible selection bias in this study. An assessment of all patients with poor response to treatment followed-up during the study, in addition to the matching between groups, was the strategy chosen to decrease biases. However, internal validity of the study was guaranteed by homogeneity in the distribution of the groups.

Previous studies demonstrated higher rates of therapeutic failure in patients treated before the development of the cutaneous lesion; in those ones, treatment did not prevent the occurrence of the lesion^14,34.

Reactivity to MST reflects the development of specific Th1 cellular immune response to the Leishmania parasite²⁸. The development of adaptive immune response against leishmaniasis is the basis for the resistance to infection⁴. Usually, MST is used as an indicator of protection in vaccine evaluation studies against Leishmania ^9,10.

MST indicates the establishment of a cellular immune response against Leishmania antigens^2,4,23. Low percentage of negative response to MST in this study confirms the high sensitivity of the test^7,17,25,30. The proportion of patients with negative MST in the case group was twice that of the control group. The lack of statistical significance in MST analysis when classified in positive or negative may be related to the low proportion of negative tests²⁷. PASSOS et al. suggested that a negative response to MST may function as a predictor of post-treatment relapse²².

A defect of the host cellular immune response associated with the persistence of the parasite is a hypothesis to the occurrence of therapeutic failure or relapse^5,26,32. A negative response to MST after two months of disease evolution may be related to a defect of the host immune response. However, patients presenting definitive cure with negative MST indicate that other factors may also be related to the occurrence of cure or therapeutic failure.

Additional studies with the standard dose of meglumine antimoniate for CL (10 - 20 mg Sb⁵⁺/kg/day during 20 continuous days) and with different populations in other locations should be performed to confirm these results.

CONCLUSION

Early treatment, reflected by an evolution time of the lesion of less than two months at the first appointment, and a poor cellular immune response, reflected by a less intense MST, both demonstrated to contribute to the occurrence of treatment failure in cutaneous leishmaniasis in CL patients with intramuscular meglumine antimoniate in the dose of 5 mg Sb⁵⁺/kg/day for 30 consecutive days. Clinicians should be aware of the possibility of therapeutic failure in the follow-up of negative MST patients and those who undergo treatment before the skin lesion is two months old.

FUNDING

This work received support from INI/FIOCRUZ and the National Council of Scientific and Technicological Development - Brazil (CNPq), which did not participate in the design, execution, or report.

ETHICS CONSIDERATIONS

This study follows National Health Council guidelines and was approved by the Ethics Committee on Human Research (CEP) of INI/FIOCRUZ, number CAAE 0016.0.009.000-02.