Hansen's disease (HD), also known as leprosy, is a bacterial infection caused by Mycobacterium leprae. Although curable, HD has existed for centuries and continues to be a public health problem throughout the world. Globally, the annual number of new cases is approximately 200,000 (World Health Organization, 2021). In the United States the annual number of new cases is roughly 150-200 (Centers for Disease Control and Prevention, 2021). The low prevalence of HD in the United States is one factor why this condition is often initially overlooked and misdiagnosed by many clinicians (Bianchi & Sevigny, 2016).
HD affects mostly marginal populations; patients are often poor and uneducated, thus representing a vulnerable population. Because of fear and misunderstanding of the disease, patients with HD often experience ostracism, stigma, and neglect. Leprosy is considered an orphan disease; quality research is scarce, thereby limiting drug development and treatment options based on rigorous scientific methodology (Lazo-Porras et al., 2020).
Mostly affecting the skin and peripheral nerves, HD is a spectrum of diseases with tuberculoid (TT) and lepromatous (LL) at opposite ends of the spectrum. Patients with TT HD have limited disease with few bacteria in the skin and nerves, whereas LL patients have widespread disease with large numbers of bacteria. In the middle of the spectrum are the intermediate subtypes also known as borderline HD. These subtypes include borderline TT (BT), mid-borderline (BB), and borderline LL (BL). All borderline subtypes are characterized as more extensive than TT with a greater number of lesions and nerve involvement, but not as widespread as LL. This nomenclature, known as the Ridley-Jopling Classification System, reflects the clinical and pathological features along with the immunological responses patients with HD experience.
A major complication in HD is the development of immune reactions. These reactions manifest because of a sudden impairment in the immunological balance between the pathogen and infected individuals. Although reactions can happen before, during, or after treatment, they frequently occur during therapy, which is why they are often incorrectly viewed as complications of treatment. Although antimicrobials effectively kill the bacteria, they do not clear the body of dead bacterial fragments. These lingering fragments increase the risk of patients developing reactions (Sharma et al., 2004). Approximately 30%-50% of patients with HD will develop reactions, classified as Type 1 known as reversal reaction (RR) or Type 2 known as erythema nodosum leprosum (ENL; Costa et al., 2018).
Each reactional type presents differently in the clinical setting. RRs are caused by cell-mediated responses to the bacteria and can present with mild to more severe symptoms (National Hansen's Disease Program, n.d.). Patients diagnosed with TT or BT are more likely to experience RR, which presents in already existing lesions. A mild case manifests as increased edema and erythema with a low-grade fever that may cause body aches and fatigue. More severe symptoms include sensory and motor impairment because of potentially emerging neuritis with or without swelling of the hands and feet (Centers for Disease Control and Prevention, 2018; National Hansen's Disease Program, n.d.).
ENL reactions are caused by antigen-antibody complexes, where an infected individual's circulating antibodies bind to antigens of M. leprae. This type of reaction is seen more commonly in patients with BB, BL, and LL and typically presents as new lesions accompanied by low-grade fever and malaise (Centers for Disease Control and Prevention, 2018). In a more severe and emergent presentation, ENL may be confused with sepsis because of painful ulcerating lesions, elevated white blood count, and sudden loss of sensation (National Hansen's Disease Program, n.d.). Both reaction types are a major cause of disability and deformity in patients with HD that requires immediate treatment with thalidomide or prednisone to lessen the chance of serious consequences.
A third reactional type that is severe and can cause loss of limb or may be life-threatening is Lucio's phenomenon. It is seen in diffuse LL HD and is characterized by deep violet-colored patches and bullous infiltrates. These lesions can ulcerate and become necrotic. However, this phenomenon is rare and occurs in untreated individuals, patients who have not been accurately diagnosed with HD, or those who are being improperly treated. Therefore, for these reasons, patients with Lucio's phenomenon were not included in this study.
Multidrug therapy (MDT) consisting of dapsone, rifampin, and clofazimine is the standard of care and given as first-line therapy (Kar & Gupta, 2015). All the medications included in MDT target different cellular processes of M. leprae and are given in combination to reduce the risk of drug resistance and disease relapse. In addition to its bactericidal effect, clofazimine is added for its anti-inflammatory properties with the intention of minimizing the risk of reactions (Fischer, 2017). However, MDT has limitations such as side effects, drug-drug interactions, potential intolerances, accessibility issues that can lead to treatment delays, questions regarding its ability to prevent reactions, and drug resistance, which has been reported in disease relapse cases (Cambau & Williams, 2019). For these reasons, alternative medications are also used to treat HD.
Minocycline is a well-established second-line drug used as an alternative for any of the MDT medications because of its strong bactericidal effect against M. leprae (Fischer, 2017; Lockwood, 2019). Minocycline may also possess anti-inflammatory and antiapoptotic activity (Garrido-Mesa et al., 2013) and may prevent immune reactions (Narang et al., 2015, 2017, 2019).
Because it is not uncommon for patients to receive second-line medications including minocycline as alternative antileprosy therapy (ALT-M) rather than standard MDT, it is worthwhile to evaluate these different treatment options and assess the occurrence of reactions and time of onset. Reducing the risk of reactions and delivering prompt treatment if reactions occur are critical to providing quality care and improving patient outcomes.
EVIDENCE-BASED LITERATURE REVIEW AND SYNTHESIS
When tetracyclines were reviewed for their nonantibiotic properties, minocycline emerged as the most effective neuroprotective tetracycline in ischemia, traumatic brain injury, neuropathic pain, and several neurodegenerative conditions including Parkinson's, Huntington's, Alzheimer's, multiple sclerosis, and spinal cord injury (Garrido-Mesa et al., 2013). This neuroprotective property prompted researchers to investigate the role of minocycline in patients with HD with a recent onset of nerve function impairment.
Minocycline has neuroprotective properties that may help protect patients from nerve damage, a major cause of disability in patients with HD. In addition, minocycline may play a role in the management and prevention of refractory recurrent and/or chronic ENL, the immune reactions that increase a patient's risk of nerve damage and subsequent disability in HD (Narang et al., 2015, 2017, 2019).
PROJECT RATIONALE
The National Hansen's Disease Program (NHDP) has been using minocycline as an alternative to MDT for several years. The purpose of this project was to evaluate whether there is a difference in the frequency of reactional episodes and time of onset while receiving MDT consisting of dapsone, rifampin, and clofazimine and those treated with ALT-M, which is minocycline in combination with any of the above medications except clofazimine. This is important because the focus of disability prevention in HD is based on treatment of the infection with the premise that earlier diagnosis and treatment will lead to less disability (Scollard, 2019).
Reactions are the leading cause of disability and deformity among patients with HD. Identifying clinical characteristics associated with reactions will allow stakeholders at the NHDP to anticipate more accurately who may be at risk of developing reactions and the likely time of onset. The earlier reactions are identified, the sooner treatment with either thalidomide or prednisone can be initiated, thereby reducing the risk of nerve damage and subsequent physical and permanent disabilities.
Improving quality of care and improving patient outcomes is the cornerstone of this project. The implications for clinical practice are broad. Patients with HD who experience physical and permanent disabilities have a decreased quality of life that affects their psychological, social, and economic outcomes (Queiroz et al., 2015). The information gained from this evaluation may contribute to earlier detection of reactions, which has the potential to minimize the sequelae and disabilities associated with HD.
PROJECT AIMS
The overarching goal of this study is to identify characteristics that allow clinicians to predict, detect, and manage reactions earlier and potentially improve patient outcomes. The specific aims of this project are as follows:
Aim 1: Determine the overall reaction rate and time of onset of reactions experienced among patients with HD. Among those who experience a reaction, identify patient demographic and clinical characteristics that are associated with reaction occurrence and time to onset.
Aim 2: Compare reaction occurrence among patients with HD who are treated with ALT-M relative to those treated with MDT.
Aim 3: Among those who experience a reaction, compare the time to onset of a reaction between those treated with ALT-M relative to those treated with MDT.
Clinicians will use findings from the above aims to develop a patient profile to help guide treatment decisions and tailor patient education based on a better understanding of the expectancy rate of reactions and their time of onset.
METHODS
Duke University Health System Institutional Review Board determined that this project met the definition of exempt research.
Design
This project used an observational design comparing two independent groups. A retrospective chart review from 2020 back to 2011 was conducted to gather patient demographic characteristics and disease status and determine patient group according to treatment received. One group consisted of patients with HD who received ALT-M; another group consisted of those who received MDT. The outcome variable of interest was whether a patient experienced a reaction while receiving ALT-M or MDT treatment. The duration of treatment is either 12 or 24 months based on Ridley-Jopling classification of disease.
Setting
The NHDP is a medical service within the Department of Health and Human Services under Health Resources and Services Administration. The NHDP is in Baton Rouge, Louisiana, and is the only national program devoted exclusively to the management of HD. Individuals living in the United States and its territories are eligible to receive medical care, including medications at no cost, for the diagnosis and treatment of HD and its complications (National Hansen's Disease Program, n.d.).
Sample
G*Power (Faul et al., 2009) was used to determine sample size based on 80% power, alpha set to .05, and an odds ratio of 2. The estimated required sample size was N = 113. The ALT-M group included 56 cases, and the MDT group included 57. The sample was selected based on a quota sampling strategy starting from December 2020 and looking back in time until the appropriate number of patients with HD who met the eligibility criteria were identified. Patients who were diagnosed with reactions before or at the start of treatment with MDT or ALT-M were not enrolled. Inclusion criteria included adult patients aged 18 years and older who were diagnosed with HD using the Ridley-Jopling Classification System and completed a full course of treatment with ALT-M or MDT. Exclusion criteria included patients with pure neural or indeterminate type of HD, patients diagnosed with Lucio's phenomenon, patients who did not complete treatment, patients with allergies and/or contraindications (hepatic and/or renal dysfunction) to ALT-M or MDT drugs, concurrent treatment with disease-modifying antirheumatic drugs such as methotrexate or other biologics, concomitant use of minocycline and clofazimine, and patients who were pregnant.
Measures and Data Collection
Age, gender, race, disease status, treatment regimen (ALT-M or MDT), reaction occurrence, and time of reaction onset measured as months from the start of treatment were collected. Among those with a reaction, the type of reaction, either Type 1 or 2, was recorded. Type 1 reactions were defined as erythematous and edematous changes to existing skin lesions with or without ulceration. The operational definition of Type 2 reactions was new painful erythematous subcutaneous nodules that may ulcerate with or without systemic signs and symptoms such as fever, joint pain, anorexia, and malaise. The occurrence of a reaction and time of onset were measured and reported only while patients were receiving treatment.
Data Analysis
The analysis was conducted using IBM SPSS Statistics (Version 27). Nondirectional statistical tests were performed with the level of significance set at .05 for each two-tailed test. Baseline differences between treatment groups were tested using independent t tests for continuous measures (e.g., age), whereas chi-square/Fisher's exact tests were used to determine differences in categorical measures (e.g., gender, race, and HD type). If a significant treatment difference in a clinical characteristic was found, that variable was included as a covariate in the Aim 2 multivariable analysis to control for confounding factors.
Logistic regression models were used to compare the occurrence of reactions between patients with HD who were treated with ALT-M relative to those treated with MDT.
a. First, a bivariate logistic regression model was conducted to test for treatment differences in the probability of a reaction for individual patient characteristics.
b. Next, a multivariable logistic regression model including significant clinical characteristics as covariates (per above analysis) was used to test for differences in occurrence of reactions between treatment groups. To address effect size, the odds ratio (OR) from the bivariate model and adjusted ORs from the covariate-adjusted model along with their 95% confidence intervals were provided.
A Mann-Whitney U test was performed in the subsample of patients who developed a reaction to test for treatment group difference in time to onset.
Privacy, Data Storage, and Confidentiality
All patients' protected health information was deidentified and therefore remained private. The NHDP electronic health records (EHRs) were queried, and data collected for this project were stored in a password-protected Excel spreadsheet. The encrypted Excel spreadsheet was transmitted as a secure password-protected file via a secure Internet connection and stored on a secure server at Duke University School of Nursing.
RESULTS
The population sample included 114 previously treated patients with HD. One patient, however, did not complete treatment, therefore did not meet inclusion criteria, and was eliminated. As shown in Table 1, 85 (75.2%) of the 113 patients were male and 28 (24.8%) were female. Most patients identified as white (n = 67, 59.3%) followed by Pacific Islander (n = 25, 22.1%). The mean age at the start of treatment was 53.25 years, with a range of 19-90 years and above. The most common HD type was LL (n = 60, 53.1%), followed by BL (n = 26, 23%) and BT (n = 26, 23%). None of the patients had BB-type HD (see Table 1).
Table 1 shows that gender was significantly different between the treatment groups (p = .030). The MDT group had more men (n = 48, 84.2%) than the ALT-M group (n = 37, 66.1%). Whereas more women comprised the ALT-M group (n = 19, 33.93%) than in the MDT group (n = 9, 15.8%). In addition, HD type was significantly different between the two groups (p < .001). More specifically, there were two (3.51%) patients with BT HD in the MDT group and 24 (42.9%) patients with BT in the ALT-M group. Because of these differences between treatment groups, the variables gender and HD type were included as covariates in the multivariable logistic regression model.
Table 1 shows that, overall, 54 (47.8%) patients experienced a reaction during treatment. Fifty-three percent of men had a reaction (45/85), whereas 32% of women experienced a reaction (9/28; see Table 2). The mean age of those who experienced a reaction was 48.15 years, whereas the mean age of those who did not experience a reaction was 57.92 years. Seventy-six percent of Pacific Islanders (19/25) and 37% of whites (25/67) developed a reaction. Fifty-eight percent of patients with LL HD experienced a reaction (35/60). In addition, 46% of the patients with BL type (12/26) and 27% of the patients with BT type (7/26) experienced a reaction. As seen in Table 1, the median time to onset of a reaction was 6 months. The range was from 1 to 24 months, and the middle 50%, that is, the interquartile range, was 10.25 months. Over half (53.7%) of the patients who developed a reaction did so within the first 6 months of treatment, and 13% did so within the first month of treatment. The most frequent reaction type was ENL (n = 46, 40.7%), whereas eight patients (7.1%) developed RR. The patient demographic and clinical characteristics of age, gender, HD type, and race were tested using independent t tests and chi-square/Fisher's exact test for their association with reactions. Age (p = .005), race (p = .008), and HD type (p = .043) were significantly associated with reactions (see Table 2) and therefore added to the final multivariable logistic regression analysis as covariates.
Of the 54 patients who experienced a reaction during treatment, 30 (55.6%) had received MDT and 24 (44.4%) had received ALT-M. This was not a statistically significant difference between the treatment groups (p = .348; see Table 2). Table 1 shows the median time of reaction onset for the MDT group was 6.50 months versus 5.50 months for the ALT-M group. A Mann-Whitney U test did not show a statistically significant difference between treatment groups and reaction time of onset (p = .753).
Bivariate logistic regression was performed to ascertain the individual effects that treatment, age, gender, HD type, or race had on the likelihood of a patient experiencing a reaction. Table 3 shows that age, HD type, and race were significantly associated with the occurrence of a reaction (p = .007, .009, and .002, respectively). To accomplish this, the categorical variables of race and HD type were recoded into dummy variables. Three levels of race that included Pacific Islanders, whites, and Other were created. Next, HD type was recoded as BT or other HD types; there were no patients with BB HD type.
For the final analysis, the multivariable logistic regression model (see Table 4) included reaction as the dependent variable, treatment regimen as the independent variable, and age, HD type, and race as the covariates, as these variables were significant in preliminary analyses, as described above. The comparisons in the model included ALT-M with MDT, nonwhites with whites, Pacific Islanders and whites with all other races, and BT with all other HD types.
The regression model was statistically significant ([chi]2[5] = 16.738, p < .05), explained 18.4% (Nagelkerke R2) of the variance in reactions, and correctly classified 68.1% of cases. As seen in Table 4, results indicated a significant association between BT HD and reactions ([chi]2[1] = 3.898, p = .048). Patients with BT were 68% less likely to develop reactions than any other type of HD while accounting for treatment, age, and race (OR = 0.320, 95% CI [0.103, 0.992]). Although caution must be taken with the interpretation of these data because the variable of interest, that is, reactions, included both Type 1 and Type 2, the results do suggest there is a significant relationship between HD type and reactions.
DISCUSSION
Reactions are a major complication in HD and the most common cause of disability and deformity in leprosy. Early detection and management of reactions are key to minimizing the sequelae and disabilities associated with HD. Identifying clinical characteristics that are associated with reactions may help in the early recognition and treatment of these patients.
The purpose of this project was to evaluate whether differences existed in the frequency of reactional episodes and time of onset in patients with HD receiving ALT-M compared with those receiving MDT. No significant differences were found in reaction rates or onset times between these two treatment groups. A secondary goal of this project was to identify characteristics that allow clinicians to predict, detect, and manage reactions earlier; tailor patient education; and potentially improve patient outcomes. Knowing which characteristics put a patient at risk of developing a reaction allows clinicians to strategize interventions targeted to those individuals at the greatest risk and thereby reduce or prevent reactions from occurring. Because reactions are a major cause of disability and deformity among patients with HD, the earlier reactions are identified, the sooner treatment with either thalidomide or prednisone can be initiated and thus reduce the risk of subsequent disabilities.
Like previous studies, men with HD outnumbered women in this sample by a rate of 3:1 (Mowla et al., 2015). In this analysis, almost half of the patients developed a reaction (47.8%). These findings are consistent with previous literature in which the rate of reactions has been estimated between 30% and 50% (Kamath et al., 2014; Maymone et al., 2020; Scollard et al., 2006). In addition, more reactions occurred in men than in women, confirming other observations (Queiroz et al., 2015; Sharma et al., 2004), yet we found no statistical difference in reaction occurrence by gender. Reactions occurred more frequently in LL-type HD followed by intermediate forms of HD such as BL and BT, as previously reported (Queiroz et al., 2015). Findings from this project identified bivariate associations between reactions and race, HD type, and age, yet only HD type showed statistical significance in multivariable regression analysis. More specifically, patients with BT were 68% less likely to develop reactions than other clinical forms of HD including LL and BL, even while accounting for treatment, age, and race. What is unclear, however, is whether this negative predictive relationship between reactions and HD type holds true for Type 1 reactions, Type 2 reactions, or both. Because of the smaller subset of patients with reactions, this project's design was not adequately powered to test for an association between reaction type and HD classification or whether there was a difference between reaction types between the two treatment groups.
Through a systematic review and meta-analysis, de Paula et al. (2019) set out to identify clinical factors associated with physical disability in patients with HD. The authors showed an association between physical disabilities and gender, multibacillary leprosy, reactions, and LL. Specifically, they found that male patients were almost 2 times more likely than female patients to have physical disabilities. Furthermore, they found that patients with LL had a 5- to 12-fold greater odds of disability. In addition, patients with multibacillary leprosy were found to be more likely to have physical disabilities. Multibacillary leprosy is the World Health Organization classification system and essentially the equivalent of Ridley-Jopling HD classification of BB, BL, and LL. Moreover, de Paula et al. reported that patients with reactions were 2.43 times more likely to have disabilities.
Although this project's goal and objective were not the same as those of de Paula et al. (2019), the rationale for developing a patient profile designed to identify clinical characteristics associated with reactions seems logical given that differences in reactions are not because of differences in treatment. We hypothesize that reactions are more likely related to underlying clinical characteristics like HD type. Reactions are known to increase a patient's risk of developing physical disabilities; if patients at the greatest risk of developing reactions can be identified, clinicians can target strategies and interventions designed to minimize or prevent reactions from occurring.
Findings from this project showed that gender was not significantly associated with reactions. de Paula et al. (2019), however, found that men are almost 2 times more likely to experience physical disabilities than women. Although we did not evaluate disability in this study, documenting disabilities and the corresponding severity grade in the EHR and incorporating it into databases may benefit future studies to elucidate this relationship.
de Paula et al. (2019) also found that patients with BB, BL, and LL HD types were more likely to experience disabilities. Our study supports the finding that patients with LL followed by BL had the highest occurrence of reactions, but again, we did not assess disabilities or severity grade. Our findings indicate that patients with BT HD were less likely to develop reactions than other clinical forms of HD including LL and BL. If patients with BB, BL, and LL HD types are at greater odds of disability, then it may be worth exploring whether reaction rates among these patients are different between the two treatment regimens. This study was not designed to look for such signals; to do so would require a larger stratified sample to ensure proportionately adequate representation of different levels within a category and be matched on clinical characteristics by treatment group. Further exploration of whether patients with TT and BT negatively predict an association with physical disabilities may also be warranted.
Limitations
The major limitation of this analysis is that it was retrospective in nature, using data collected during routine clinical care. The secondary use of healthcare data is limited by the available information within the EHR. The disadvantage of secondary analysis is the lack of control over the original collection of data. Data can be incomplete, missing, or simply inaccurate. For example, documentation on reaction severity and grade was often missing or incomplete and therefore was not included in this analysis. This limitation is outweighed by the availability of the data, the ease of retrieval, and the access it affords to clients with this rare condition. Future prospective studies should be designed with important clinical variables included.
In addition, the sample size was small and nonrandomized. HD is rare in the United States; the power of the data is limited by the lack of numbers. Moreover, the sampling technique used in this project relied on a quota strategy. Patients who met the inclusion criteria were enrolled until the target sample size was reached. The disadvantage of this type of sampling is that there is no way of ensuring every element of the population will be equally represented or included at all. The sampling technique used in this project may explain the lack of equal representation of various clinical characteristics like race and HD type. For instance, there was only one patient with TT-type HD and no patients with BB. In addition, many race categories were underrepresented, like Asian, Black, and Indian. The small sample size coupled with the sampling strategy may not be a true representation of the population. In the future, to address the problem of overrepresentation or underrepresentation, HD clinicians may want to consider using a proportional quota sampling strategy. Stratifying important characteristics can help ensure each segment of the population is proportionately represented in the sample.
Conclusion
To prevent reactions and their associated physical disabilities, the information gained from this project can be used by clinicians to help guide decision making. Patients with HD experienced less reactions with ALT-M (42.9%) than with MDT (52.6%), yet this difference was not statistically significant. Nor was there a significant difference in reaction time of onset (p = .753). Factors other than reactions rates can be considered when deciding treatment regimen, and providers may opt to use either ALT-M or MDT when considering side effects, drug-drug interactions, intolerance, allergies, or simply patient preferences.
Men comprise a greater number of people affected by HD in this sample and experienced a higher incidence of reactions than women, but this difference was not statistically significant. Patients with LL HD have a higher frequency of reactions than other HD clinical forms, and BT is a significant negative predictor of reactions in patients with HD. The rate of reactions was not significantly different between the MDT and ALT-M treatment groups, suggesting that reactions are influenced more by patient characteristics and disease status rather than treatment decisions made by providers. Practitioners should feel comfortable making treatment decisions based on factors other than reactions because these findings indicate there was no significant difference in the reaction rate between the two treatment groups. More studies are needed to help identify clinical characteristics that are associated with reactions and whether a positive or negative predictive relationship exists between these characteristics and reactions. In the future, it is recommended that HD clinicians design prospective studies investigating clinical characteristics that may mitigate or impact treatment outcomes, reaction severity, and disability grade.
In summary, these findings validate and support current practice at the NHDP where MDT has been the standard of care and ALT-M is given when necessary. The goal of this project, to develop a clinical profile of patients with HD, is supported by others (de Paula et al., 2019) where the association between disabilities and clinical characteristics such as reactions was confirmed. However, because of the limitations of this project's data set, no concrete conclusion can be drawn for which patient characteristics should be included in the profile. Reactions are a known risk factor for developing physical disabilities. The rate of reactions in this study was not significantly different between the MDT and ALT-M treatment groups, suggesting that reactions are influenced more by patient characteristics and disease status rather than choice of treatment. These findings allow clinicians to choose between MDT and ALT-M based on client preference and other relevant factors.
REFERENCES