Background
Tuberculosis (TB) remains a leading cause of morbidity and mortality, and a major public health problem worldwide [
1]. World Health Organization (WHO) has estimated that 10.4 million new cases and 1.8 million deaths from TB occur every year [
2,
3]. Approximately, one-third of the world’s population is infected with
Mycobacterium tuberculosis (
M. tuberculosis), and about 1.7 billion people have latent tuberculosis infection (LTBI) [
4,
5], especially in developing countries [
1]. China is one of the countries with a high burden of TB and LTBI. Although the prevalence of TB is gradually decreasing [
6], it still remains an important threat due to its high infectiousness. The Chinese government is making strenuous efforts to prevent further spread of infection and provide rapid detection and treatment facilities through national support and project grants [
7].
Currently, the most common method used to detect LTBI implies testing the hosts’ immune response to
M. tuberculosis derived antigens [
8,
9]. Tuberculin pure protein derivative (PPD) skin test and Interferon-γ (IFN-γ) release assays (IGRAs) are commonly used for detecting
M. tuberculosis infection [
10‐
12], but the tuberculin skin test (TST) has a high false positive rate [
13,
14]. IGRAs is an immunodiagnostic test that measures effector T cell-mediated IFN-γ response to
M. tuberculosis-specific antigens. Both 6 kDa early secretory antigenic target (ESAT6) and 10 kDa culture filtrate protein (CFP10) are co-transcribed and specifically secreted by
M. tuberculosis [
15,
16], which is absent in most non-tuberculous mycobacteria (NTM) and Bacillus Calmette-Guérin (BCG) [
17‐
20]. IGRAs can limit the false positive reaction [
21,
22] and overcome the operational drawbacks of TST [
23]. However, the detection reagents used for IGRAs are expensive, and the operation process is difficult to be automated [
24]; thus, IGRAs are not suitable for use in remote areas and wide-range screening.
According to the characteristics of M. tuberculosis-specific antigens, China has recently developed a recombinant fusion protein ESAT6-CFP10 (EC) as a new skin test reagent, which was expected to have high sensitivity and specificity. In this study, we conducted large sample research to clarify whether induration and erythema diameter could be used as diagnostic indicators for EC skin test to detect M. tuberculosis infection.
Methods
Clinical trial units
The study on TB patients was completed by Shanghai Public Health Clinical Center, Tianjin Haihe Hospital, Wuhan Tuberculosis Prevention and Treatment Center, Beijing Chest Hospital, First Affiliated Hospital of Chongqing Medical University, Wuxi Infectious Disease Hospital, Fuzhou Pulmonary Hospital, Zhenjiang Third People’s Hospital, Anhui Provincial hospitals and Shenzhen Third People’s Hospital; Shanghai Public Health Clinical Center was the main responsible unit. The research was performed from December 17, 2015 to March 02, 2018. Jiangsu Provincial Center for Disease Control and Prevention completed the research on healthy volunteers from February 27, 2016 to June 18, 2016.
The Human Research Ethics Committee granted the research protocol. All participants read and signed the informed consent.
Inclusion and exclusion criteria
Healthy volunteers aged 18 to 65 were recruited and included in this study if they met the following criteria: (a) no history or family history of TB; (b) no X-ray chest abnormality; (c) normal physical condition and temperature. Healthy volunteers were excluded if they (a) had serious diseases such as advanced cancer, acute exacerbation of the chronic obstructive pulmonary disease, acute or progressive liver disease, acute or progressive kidney disease or congestive heart failure; (b) had a mental illness; (c) had known or suspicious immune function impairment or abnormality; (d) had an acute febrile illness and infectious disease; (e) were currently participating in other drug clinical trials or have participated in any other new drug clinical trials within 3 months before the testing; (f) had a history of drug allergy; (g) were pregnant or lactating; (h) had a clear history of hypertension; (i) had other reasons that investigators believed might affect the evaluation of the test.
TB patients’ inclusion criteria were: patients 18 to 65 years old who were definitively diagnosed according to the criteria for pulmonary TB of the Ministry of Health of China (WS 288–2008,
supplementary). TB patients were excluded if they met the (a) (b) (e) (f) (g) (i) six items in the health volunteers’ exclusion criteria.
Finally, 743 TB patients and 1514 healthy volunteers were enrolled in this study.
Skin test reagent
ESAT6-CFP10 included the following: a recombinant fusion protein of ESAT6 and CFP10, which was manufactured by Anhui Longcom Biologic Pharmacy Co. Ltd. This reagent was supplied as a liquid formulation (0.3 ml/bottle), consisting of ESAT6-CFP10 antigen (50 U/10μg/ml), phosphate-buffered saline (PBS, 1.0 mmol/liter), phenol (3%) and Tween 80 (0.0005%).
Skin test operation
Before starting the skin test, skin test reagent, rescue medicine and rescue equipment were prepared. Each numbered skin test reagent was only be used by one subject and could not be mixed with others. The authorized medical staff checked the drug number/subject number of each subject, and then checked the expiration date, bottle wall, and drug contents.
The skin was sterilized with ethanol in the first 1/3 of the subjects’ forearm volar side without scars or lesions. Then, 0.1 ml skin test reagent was inhaled into a 1 ml disposable syringe, the scale and the tip hole slope were consistently upward. The medical staff’s left thumb tightened the skin of the subject where the injection was required, and the right hand hold the syringe, which penetrated into the skin at 5° ~ 10°from the skin. The reagent was injected slowly; the white rounded ridge was visible, the pores were revealed, and the boundary was clear after the injection. After the injection was completed, the needle was pulled out while rotating 90 degrees. Subjects were instructed not to get wet or rub the injection site for 72 h. Medical staff was required to change gloves for each patient to avoid contamination.
Outcome and measurement
Previous literature reports and small sample clinical trial suggested that erythema can be used as a specific response to the skin test. In this study, outcome measures were induration and erythema diameters in every subject. All patients were observed at 24 h, 48 h, and 72 h after a skin test, and healthy volunteers were observed at 24 h and 48 h after skin test. Firstly, the medical staff analyzed the induration (touching the skin with fingers) and observed the erythema outer ring to determine edges of induration and erythema. The upper and lower edges (longitudinal diameter), left and right edges (transverse diameter) of the induration were marked by a short line with a black pen. The boundary of the erythema was marked by a short line with a red pen. The longitudinal and transverse diameters of induration and erythema were measured with Vernier caliper. Two medical staff measured and read the size of the skin test response to ensure the reading error between the two people controlled within ±0.2 mm. The averages of the longitudinal and transverse diameters was the sizes of induration and erythema diameter. A previous clinical trial showed that the area under the ROC curve was the largest when the cutoff value of the diameter of induration and erythema was 5 mm. Therefore, in this study, the diameter of induration and erythema of 5 mm was the positive criterion for the EC skin test response. This clinical trial expanded the sample size for further verification. So, the results of the EC skin test were considered positive when the diameter of induration or erythema was ≥5 mm, and strongly positive if a local blister, necrosis, or lymphangitis were observed.
Statistical analysis
Statistical analysis was conducted using SAS 9.3 software. The diameters of induration and erythema after EC skin test in all subjects were described with mean ± SD, number, median, minimum and maximum. Sensitivity, Specificity, Positive predict value (PV+), Negative predict value (PV−), Youden index (YI), consistency, and Kappa value (95% confidence interval) were used to compare the consistency of induration / erythema and clinical diagnostic results. Positive coincidence rate (CR+), Negative coincidence rate (CR−), consistency, and Kappa value (95% CI) were carried out to compare the consistency of induration and erythema results. A Chi-square test was applied to compare the rates in different groups. P < 0.05 was considered to be statistically significant.
Discussion
Currently, there is no golden standard for LTBI diagnosis. Therefore, in order to detect
M. tuberculosis infection, TB patients were used instead of LTBI [
28‐
30]. All enrolled patients underwent a bacteriological test of
M. tuberculosis. Based on the bacteriological results, imaging results, and clinical symptoms, patients were divided into bacteriologically positive and negative patients, and extrapulmonary tuberculosis patients. In previous clinical trials, we validate the consistency and safety of the recombinant fusion protein EC for detecting
M. tuberculosis. In this phase III clinical trial, we expanded the number of subjects and found that induration or erythema diameter not less than 5 mm could be used as a diagnostic indicator for detecting
M. tuberculosis infection.
Combined with the results of previous clinical trials [
31,
32], these data suggested that the best sensitivity, specificity, and other diagnostic results are obtained 48 h after testing. Furthermore, we found that diagnostic results of induration and erythema had higher sensitivity, specificity, consistency, and Kappa values compared with clinical diagnosis. Therefore, we believed that induration and erythema could be used as diagnostic indicators for EC skin test to detect
M. tuberculosis infection. On the one hand, the recombinant fusion protein is more stable than recombinant mixed protein [
33] and could produce a stronger cellular immune response. Thus, the erythema response could not be ignored. On the other hand, EC skin test reagent contains only
M. tuberculosis-specific antigens proteins; skin responses should be specific. Therefore, induration and erythema as skin response should be measured.
Previous small sample clinical trial suggested that erythema was a specific response to the EC skin test. After the test, the sensitivity of the erythema was higher then 80%, the specificity was higher then 90%, and the test results were consistent with clinical diagnosis results (Kappa value ≥0.78). In the BCG vaccination population, more than 90% of the subjects had a negative erythema response. A large prospective study of PPD skin test conducted in Japan confirmed that the specificity of erythema measurement was closely related to induration, and erythema was more likely to produce a specific response without being confused by induration [
34]. Moreover, another study showed that erythema and induration had the same response trend. In this study, subjects were divided into
M. tuberculosis infection negative and positive groups according to the erythema response [
35]. We believed that erythema should be a specific response to the EC skin test according to the above results. Therefore, we included erythema as a judgement indicator and further verified its detection effect by expanding the sample size. In TB patients, the sensitivity of erythema measurement was higher than that of induration. If only the induration is used, many patients may be misdiagnosed, which could reduce the sensitivity of the test. In this study, erythema could improve the sensitivity of the test in TB patients. Although the erythema response had a certain false-positive rate in healthy volunteers, the false positive rate was low, and its specificity was still high. Especially in BCG vaccination volunteers, whether induration or erythema, its specificity was higher than 90%, which explained that EC skin test response would not be affected by BCG. In addition, the analysis found that the consistency of erythema and clinical diagnosis was high. Overall, the erythema response in different types of TB patients kept the sensitivity of about 90% after 48 h; its sensitivity was higher than that of induration. In healthy volunteers, the erythema had a specificity of about 90%. Above all, the sensitivity and specificity of the erythema response were both high. Erythema used as the judgment indicator can improve the positive detection rate in TB patients without affecting the specificity of healthy volunteers. Therefore, erythema was a specific response to the EC skin test and could be used for detecting
M. tuberculosis infection. We believed that erythema could be used as a diagnostic indicator of the EC skin test.
The skin test response produced in subjects showed that the diameter of erythema was larger, smaller, or equal to the diameter of induration. Therefore, it was one-sided and unreasonable if uniformly stipulated that it used induration alone as the diagnostic indicator for the EC skin test. The CR+ and the consistency of induration and erythema were higher than 90%, and the Kappa value was higher than 0.66 in different types of TB patients at 48 h after testing, which indicated the consistency of induration and erythema results was better in patients. The CR− of induration and erythema was more than 99% in healthy volunteers. Notably, the consistency was 98.28%, and Kappa value was 0.90 at 48 h after testing, indicating that the consistency of induration and erythema results was excellent in healthy volunteers. Since induration and erythema had a good consistency, both of them could be used as diagnostic indicators for EC skin test.
In terms of sensitivity, the use of induration and erythema in parallel was higher than that when the indicators used alone or in tandemly; In terms of specificity, although the two indicators used in parallel were lower than the two indicators used alone or in tandemly, there was no statistical difference. In terms of consistency with clinical diagnosis, the consistency of the four diagnostic indicators was about 90%. In summary, we believed that the parallel combination of induration and erythema should be used to evaluate the responses of the EC skin test.
In this study, only yellow-skinned people were included. People with darker skin will have a specific effect on the interpretation of erythema results. Therefore, in the darker-skinned population, we mainly take the diagnosis of induration, supplemented by the diagnosis of erythema.
The above content was only part of the phase III clinical trial for the EC skin test. In phase III clinical trials, we performed a PPD skin test, EC skin test, and T-SPOT test in TB patients and healthy volunteers, respectively. Our study indicated that an EC skin test can be used for screening M. tuberculosis infection and was an effective skin test reagent for clinical auxiliary diagnosis of tuberculosis. No adverse events and serious adverse events leading to withdrawal occurred in a clinical trial. In terms of sensitivity, the EC skin test was equivalent to the PPD skin test and T-SPOT test. In terms of specificity, the EC skin test was superior to the PPD skin test and non-inferior to the T-SPOT test. Compared the consistency of the three tests and clinical diagnosis, it was concluded that the consistency of PPD skin test and clinical diagnosis was lower than that of EC skin test or T-SPOT and clinical diagnosis.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.