Introduction
Tuberculosis is a communicable illness and one of the leading causes of death, especially in developing countries like Turkey [1]. One of the problems that must be managed well in the treatment of tuberculosis is drug hypersensitivity [2]. The prevalence of hypersensitivity to tuberculosis drugs was 7.8% in a study conducted with patients hospitalized in a tertiary healthcare institution in Turkey. Among these reactions, delayed drug hypersensitivity was the most common [3]. When we say delayed-type hypersensitivity, we mean maculopapular eruption, fixed drug eruption, acute generalized eczematous pustulosis (AGEP), drug rash with eosinophilia and systemic symptoms (DRESS), erythema multiforme, Stevens-Johnson Syndrome (SJS), toxic epidermal necrolysis (TEN) and lichenoid drug eruption [4]. Type IV or delayed-type hypersensitivity reactions are mediated by T cells and cytokines and can occur at any time 24 h after ingestion [5]. Treatment is interrupted when hypersensitivity develops in anti-tuberculosis therapy. Hypersensitivity is treated, then a new regimen is established after therapeutic tests are started. The first-line agents (isoniazid, rifampicin, ethambutol, pyrazinamide) are very important for the success of treatment. Alternative drugs are more toxic and less successful in treatment [6]. Therefore, it is very important to be able to include first-line drugs in the post-hypersensitivity regimen. At this point, the success of desensitization comes to the fore. Studies on drug desensitization in delayed-type hypersensitivity are less common than in early-type hypersensitivity [4].
There are defined rapid drug desensitization schemes in delayed-type drug hypersensitivity with different antibiotic groups such as β-lactam antibiotics, sulfonamide, co-trimoxazole, chemotherapy and monoclonal antibodies [7–9]. There are fewer studies on rapid drug desensitization in delayed-type drug hypersensitivity to anti-tuberculosis drugs [6]. The mechanism of action of rapid drug desensitization in delayed-type drug hypersensitivity is not clear. However, basophil and mast cell are thought to suppress T cell activity [10].
In this study, the characteristics of the patients with delayed-type drug hypersensitivity developed while being treated for tuberculosis, the desensitization scheme applied and the treatment success were evaluated.
Aim
The primary aim of the study was to determine the prevalence of delayed-type hypersensitivity reactions in drug-sensitive cases; the secondary aim was to determine the appropriate treatment management.
Material and methods
The design of the study was a retrospective study and patients hospitalized in the Tuberculosis department of our hospital between 1.02.2015 and 1.05.2021 were included. Patients aged 18 years and older who developed delayed-type hypersensitivity to antituberculosis drugs and were consulted at the Allergy and Immunology department were examined.
All of the patients were inpatients and the hypersensitivity reactions that developed after drug treatment were confirmed by Allergy and Immunology specialists. Patients who did not receive desensitization in accordance with the scheme and patients with non-tuberculosis mycobacterial infections were not included.
Demographic data of the patients, diagnosis of tuberculosis, clinical features of type 4 immediate hypersensitivity reaction and time of occurrence, drug treatments, and treatment results were evaluated. Age, gender and nationality were noted in demographic data. Tuberculosis diagnoses, organ involvement and treatments were evaluated as determined in the Turkish Ministry of Health Tuberculosis Diagnosis and Treatment Guideline published in 2019. Diagnoses were classified as smear positive, culture positive, molecular test positive, histopathological diagnosis, clinical radiological diagnosis; organs affected by tuberculosis were classified as pulmonary and extrapulmonary; the extrapulmonary group was classified as miliary, lymph node, pleura, kidney, pericardium and larynx [11].
Hypersensitivity reactions were evaluated as defined in the 2019 Approach to Drug Hypersensitivity Reactions: National Guideline Update. Type 4 hypersensitivity reactions, maculopapular eruption, fixed drug eruption, SJS, TEN, AGEP, erythema multiforme, lichenoid drug eruption that developed 6 h or later after taking the drug were evaluated [12]. After the treatment of type 4 hypersensitivity reactions in all patients included in the study, drug desensitization was performed in their stable period.
The development time of the current reaction, the success of the applied desensitization, and the duration of treatment were evaluated. In our desensitization, patients who developed a reaction, isoniazid, rifampicin, ethambutol and finally streptomycin/pyrazinamide were added respectively. If alternative drugs were to be given, isoniazid, rifampicin, ethambutol and streptomycin/pyrazinamide were given first. Other alternative drugs were then given in a non-standard order.
The intradermal test and patch test are the recommended diagnostic approach for those who develop type 4 hypersensitivity reactions [3]. If patients underwent this diagnostic testing process, it was noted.
If an allergic reaction did not develop after each drug was given with desensitization in those who developed a type 4 hypersensitivity reaction, the drug was given with desensitization after the full dose was given for 7 days. Drug doses in desensitization were made as suggested by Buhari et al. [13]. However, unlike what was stated in the protocol, pyrazinamide was given as the last one during the administration of the drugs. When a drug hypersensitivity reaction developed, after the reaction was treated in accordance with international guidelines, the patient was re-evaluated and desensitized. After the regimen was completed, the patient was given the full dose of the regimen in one go in the morning for several days (Table 1).
Table 1
Desensitization scheme for delayed-type hypersensitivity
[i] *Drug doses in desensitization were made as suggested by Buhari et al. [14].
Statistical analysis
In the statistics of the study, all analyses were performed using SPSS 22.0. Differences in the means were evaluated with the Mann-Whitney U test. Relative risk, odd ratios, and 95% confidence intervals were calculated. χ2 and logistic regression analysis were used for categorical parameters.
Results
During the study, 2677 patients were hospitalized in the Tuberculosis inpatient department; delayed-type hypersensitivity reaction was seen in 53 patients. The prevalence of delayed-type hypersensitivity in hospitalized patients was 1.9%. Forty-one patients whose drug desensitization was performed in accordance with the scheme were included. Twenty-six (63.4%) of the cases were male. Mean age (mean ± SD) (min.–max.) was 55.44 ±16.93 (18–87) years; 40 (97.6%) of them were citizens of the Republic of Turkey; 29 (70.7%) of them were diagnosed bacteriologically; 29 (70.7%) of them were diagnosed with pulmonary tuberculosis; 2 (4.8%) of them had previously received antituberculosis treatment. Anti-tuberculosis therapy (isoniazid, rifampicin, ethambutol, pyrazinamide) was initiated in all cases with drug sensitivity (Table 2).
Table 2
Demographic and clinical characteristics of the patients
The most common skin finding was maculopapular drug eruption, which was seen in 26 (63%) cases. Lichenoid drug eruption was observed in 9 (21.6%) cases, erythema multiforme in 2 (4.8%) cases, DRESS in 1 (2.4%) case, and fixed drug eruption in 2 (4.8%) cases. SJS and TEN were not observed (Table 3). The development time (mean ± SD) of the reaction in patients was 34.93 ±39.62 days. The time between index reaction and reintroduction of treatment was 41.3 ±20.4 days.
Table 3
Clinical features of hypersensitivity developed to tuberculosis treatment
| Hypersensitivity reactions | Female N (%) | Male N (%) | Total N (%) |
|---|---|---|---|
| Maculopapular drug eruption (MPE) | 9 (21.6) | 17 (41.4) | 26 (63) |
| Erythema multiforme | 1 (2.4) | 1 (2.4) | 2 (4.8) |
| Lichenoid drug eruption | 3 (7.2) | 6 (14.4) | 9 (21.6) |
| Fixed drug eruption | – | 2 (4.8) | 2 (4.8) |
| Drug rash with eosinophilia and systemic symptoms (DRESS)* | 1 (2.4) | – | 1 (2.4) |
| Exfoliative dermatitis | – | 1 (2.4) | 1 (2.4) |
* (65y, F. In pulmonary tuberculosis, isoniazid, rifampicin, ethambutol, pyrazinamide treatment was started.
She was diagnosed with DRESS on the 60th day of treatment. Her treatment was discontinued. The patient’s treatment compliance was very difficult and she wanted to leave the treatment all the time. The patient was being treated with moxifloxacin, cycloserine, linezolid and protionamide. Each drug was started one by one with desensitization. The treatment was completed in 10 months without any allergic reaction.).
When hypersensitivity developed, it was sufficient to discontinue the current treatment regimen in 11 (26.8%) patients. Oral/parenteral steroids were given to 16 (39%) patients (Table 4).
Table 4
Drugs responsible for hypersensitivity
The responsible agent could be identified in 15 reactions. The most common drug responsible for the reaction was rifampicin in 7 patients. Rifampicin was followed by pyrazinamide and isoniazid. Multiple drug hypersensitivity (moxifloxacin and rifampicin) was detected in only one case. A 25-year-old woman was diagnosed with pulmonary tuberculosis. Initial treatment was HRZE (isoniazid, rifampicin, ethambutol, pyrazinamide). Lichenoid drug eruption developed after treatment. Desensitisation was initially performed to isoniazid, ethambutol, moxifloxacin and rifampicin. Hypersensitivity developed after both moxifloxacin and rifampicin. The patient was considered allergic to these two agents. Treatment was completed with protionamide, cycloserine, para amino salicylic acid and levofloxacin.
In delayed-type hypersensitivity, intradermal test and patch test were recommended to find the responsible drug [3]. In 5 cases, intradermal test early and late readings were made with rifampicin at a concentration of 1/30000 and no positivity was observed. Patch testing was performed in 6 cases. We did not have any patients who underwent in vitro testing before reintroduction. The characteristics of the patients who underwent intradermal test and patch test are as indicated in Table 5. In Table 5, patient number 5 was examined in detail. A 33-year-old female patient was started on isoniazid, rifampicin, ethambutol, and pyrazinamide due to pulmonary tuberculosis. When maculopapular rash developed after 1 week, the treatment was discontinued and she was hospitalized. Prick testing was performed with isoniazid, rifampicin, ethambutol and pyrazinamide and was found to be negative. Intradermal test was done with rifampicin ampoule and it was found negative. Patch tests of isoniazid, rifampicin, ethambutol and pyrazinamide were performed. It was removed from the regimen only when the patch test was positive with rifampicin at the 48th h. Tuberculosis doctors identified the new regimen as isoniazid, ethambutol, pyrazinamide, moxifloxacin, cycloserine, and amikacin. The patient completed the tuberculosis treatment without the development of hypersensitivity.
Table 5
The characteristics of the patients who underwent intradermal test and patch test
| No. | Organ involvement, hypersensitivity reaction | Intradermal testResult | Patch testResult | Initial treatment | Final treatment regimen | Hypersensitivity responsible agent |
|---|---|---|---|---|---|---|
| 1 | Pulmonary, lichenoid drug eruption | Rifampicin Negative | HRZE Negative | HRZE | HRZE | Not determined |
| 2 | Pulmonary, lichenoid drug eruption | Levofloxacin Negative | Amikacin Protionamide Isoniazid PAS Levofloxacin Cycloserine Positive | Amikacin Protionamide Isoniazid PAS Levofloxacin Cycloserine | Cycloserine | |
| 3 | Pleura MPE | Rifampicin Negative | HRZE Negative | HRZE | Not determined | |
| 4* | Lymph node MPE | Rifampicin Negative | HRZE Negative | Isoniazid Ethambutol Pyrazinamide Moxifloxacin Cycloserine | Not determined | |
| 5 | Pulmonary MPE | Rifampicin Negative | HRZE Positive** | Isoniazid Ethambutol Pyrazinamide Moxifloxacin Cycloserine Amikacin | Rifampicin | |
| 6 | Lymph node MPE | Rifampicin Negative | HRZE Negative | Isoniazid Ethambutol Pyrazinamide Moxifloxacin Cycloserine Amikacin | Rifampicin*** |
Although desensitization was initiated with the sensitive regimen in all cases, successful desensitization was achieved in 19 (46.3%) cases. In 3 (7.3%) cases, the regimen was changed due to hepatotoxicity. In 19 (46.3%) cases, it was necessary to change the regimen due to the development of hypersensitivity after the first desensitization, and success was achieved with the changed regimen. In desensitization, drugs were given in accordance with the recommended scheme. If alternative drugs were also planned to be started, first-line drugs were started first, then other drugs were added.
The duration of treatment was 8.97 ±3.44 months (6–18 months). When evaluated in terms of treatment results, cure and treatment completion were accepted as treatment success. In this case, 30 (73.2%) patients successfully completed the treatment: 1 (2.4%) patient was excluded from follow-up, 1 (2.4%) had treatment failure, 1 (2.4%) patient died during the treatment process, 8 (19.5%) patients are still under treatment.
Discussion
In this study, patients who developed delayed-type hypersensitivity to tuberculosis treatment were treated as inpatients. When completing tuberculosis treatment after desensitization was accepted as treatment success; the scheme was successful in 30 (73.2%) patients. Severe reactions such as SJS or TEN were not observed. The most common hypersensitivity was maculopapular eruption and the most common responsible agent was rifampicin.
The prevalence of delayed-type hypersensitivity in hospitalized patients was 1.9% in our study. In a study examining patients who developed DRESS through the hospital registry system, the prevalence was 1.2% [13]. We think that the prevalence was higher, because all patients with delayed-type hypersensitivity were included in our study (not only DRESS) [14].
In this study, the mean age of our patients was 55.44 ±16.93 (18–87) years. In the study of Oh et al., the mean age of the patients was 55 years [15]. The mean age was found to be similar in the study in which delayed-type severe drug reactions were evaluated in Korea [15].
In the study conducted in Australia, 45% [6] of the patients, and in the Korean study, 47% [15] of the patients were male. In our study, 63.4% of the patients were male. Adverse drug reactions are more common in females [16]. In a few publications, it is mentioned that the male gender is at the forefront [17, 18]. These data were found to be inconsistent with the literature. We think that more comprehensive studies to be conducted on the Asian race will clarify the subject.
The frequency of reactions was higher in patients with bacteriological tuberculosis and pulmonary tuberculosis. There are studies supporting that severe delayed-type drug hypersensitivity reactions are more common in patients with pulmonary tuberculosis [6, 14].
The development time (mean ± SD) of the reaction in patients who developed a reaction was 34.93 ±39.62 days. While there are studies indicating that the hypersensitivity reaction occurs earlier [5, 13], there are also studies showing that it occurs later [19]. In general, we can say that such reactions are seen after the 4th week of anti-tuberculosis treatment. It is important to keep this in mind in the treatment follow-up of the patients.
In this study, the most common hypersensitivity reactions were maculopapular and lichenoid drug eruption. Severe-type reactions such as SJS or TEN were not seen. In other studies, the most common type of hypersensitivity reactions was maculopapular eruption [19, 20].
Diagnostic tests recommended in delayed-type hypersensitivity are patch test and delayed control of intradermal test [4]. The situation is the same in tuberculosis [21, 22]. According to Zaiem et al., patch test positivity rates are low in DRESS. However, if the patch test is positive, it provides the advantage of removing the drug from the regimen [23, 24]. Our study is retrospective and patch tests were evaluated according to patient files. Patch testing was performed in 6 patients and was positive in 2 patients. In one of the 2 patients, the drug was removed from the regimen. In one, desensitization was applied.
When hypersensitivity develops during the tuberculosis treatment process, 3 different pathways can be followed. In the first one, especially if the reaction is severe such as SJS or TEN, all drugs can be changed. In the second one, the regimen can be changed by performing diagnostic tests (patch, intradermal test). In the third one, the same drugs can be used again with desensitization [14]. In this study, if the hypersensitivity was maculopapular eruption, fixed drug eruption or lichenoid drug eruption, drugs were given by desensitization. However, if the hypersensitivity was DRESS, SJS, TEN, erythema multiforme, the whole regimen was changed. All changed drugs were started with desensitization.
In our study, each drug was given by rapid desensitization. When each drug was added, patients waited 7 days for new drugs. Ban et al. also used 3-day intervals for each drug addition [6].
In other studies, the target total dose was reached in 3 days [14, 24, 25]. There are also studies suggesting starting each drug in 14 days with slow desensitization [12]. Horne and Grant noticed that isoniazid resistance developed on days 16 and 23 when performing drug desensitization in 2 patients diagnosed with pulmonary tuberculosis in 1963 [26]. Delay in the treatment of tuberculosis both increases the transmission of the disease and causes the development of resistance. These both increase the cost of treatment and impair patient compliance [6, 11]. In our study, all of the patients were hospitalized in the tuberculosis department under the supervision of a doctor, possible allergic complaints could be checked daily. Due to the advantage of not delaying the treatment and catching the reaction that might develop early, the drugs were given at 7-day intervals.
No drug therapy was used for premedication before desensitization. Ban et al. gave antihistamines to patients before desensitization [6]. Moreover, it was observed that the use of antihistamines and steroids before desensitization did not increase the success of desensitization [19]. We think that it would be more accurate to evaluate the success of desensitization without the use of antihistamine treatment.
In a series of patients diagnosed with DRESS, the most common responsible agent was rifampicin [14]. In two other studies including patients who developed drug-induced maculopapular eruption, ethambutol was found to be the most common responsible agent [13, 26]. As in the literature, in our study the most common responsible agent was rifampicin.
When delayed-type hypersensitivity develops in tuberculosis, it may be necessary to interrupt the treatment and give steroids. This situation both makes the disease disseminated and causes drug resistance [27]. Several authors have suggested starting steroid treatment [28]. This issue is controversial and prospective studies with large series are needed.
When completing tuberculosis treatment after desensitization was accepted as treatment success; the scheme was successful in 30 (73.2%) patients. In this study we did not have severe allergic reactions. In another study, the desensitization success rate was 80.7%. It was mentioned that the desensitization success was not related with kinds of anti-tuberculosis medication and clinical manifestation of drug hypersensitivity [6].
This study has some limitations as it is a single-centre and retrospective study. Intradermal test and patch test were not performed in all patients. These limitations should be considered carefully.
Our study is one of the largest series of patients who developed delayed-type hypersensitivity while being treated for tuberculosis and the desensitization scheme is recommended. A practical, easy desensitization scheme had been shared in this paper.
