Abstract
Histoplasmosis is an infectious disease caused by the dimorphic fungus Histoplasma capsulatum, which in chronic conditions, is generally difficult to distinguish from pulmonary tuberculosis (TB) based on its clinical appearance; therefore, diagnostic errors could occur. Meanwhile, the prevalence of multidrug-resistant pulmonary TB (MDR-TB) in Indonesia remains high. Study determining the incidence of histoplasmosis in MDR-TB is unavailable worldwide. The aim of this study was to determine the risk factors of histoplasmosis incidence in MDR-TB patients in Indonesia. A cross-sectional was conducted at H. Adam Malik General Hospital, Medan, Indonesia and the ELISA platform (semi-quantitative) was used to detect histoplasma antibodies. Factors associated with histoplasmosis incidence among MDR-TB were determined using a Chi-squared test. A total of 50 MDR-TB patients were included this study of which 14 of them (28%) had histoplasmosis. The majority of histoplasmosis occurred in males, in MDR-TB patients with a history of TB treatment and among who had chest x-rays with far-advanced lesions. However, statistical analyses indicated none of those factors (sex, TB treatment history, status of the lung) as well as age group, acid-fast bacillus result, Mycobacterium tuberculosis culture result, having pet, living in damp house, working in the field or plantation, having HIV infection and smoking status were associated with histoplasmosis incidence. This study highlights that the incidence of histoplasmosis is relatively high and therefore further studies are important to be conducted in Indonesia that has a high MDR-TB cases.
Keywords: Tuberculosis, resistant TB, MDR-TB, histoplasmosis, histoplasma
Introduction
Histoplasmosis is a recognized disease caused by Histoplasma capsulatum, with soil serving as an environmental reservoir [1,2]. Mapping of histoplasmosis in Southeast Asia reported a total of 407 cases which is from Thailand (233 cases), Malaysia (75 cases), Indonesia (48 cases), and Singapore (21 cases) [2]. A study reported 13.61% cases of H. capsulatum in Medan, North Sumatra, Indonesia based on histoplasmin skin test [3]. The most common precipitating factors of positive histoplasmin test results are associated with pets, caves and forests [4]. Studies using the intradermal reaction test (IDR) with histoplasmin demonstrated a high prevalence of Histoplasma exposure in Americas general population and parts of Asia and Africa [5-8]. Autopsy studies in Brazil and Peru identified 7–28% of deaths related to tuberculosis (TB) and 12%–21% of deaths related to histoplasmosis [9,10].
According to The European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium (EORTC/MSGERC), the gold standard criterion for histoplasmosis diagnosis is positive culture or histopathology or direct histoplasmic microscopy of clinical specimens (bone marrow, blood, biopsied tissue, sterile fluid, or respiratory specimens) [11]. However, EORTC/MSGERC also states that a diagnosis can be made with evidence of environmental exposure, compatible clinical disease and the presence of histoplasmic antigens in body fluids [12]. Culture has low sensitivity in acute and subacute cases while PCR showed a sensitivity of 98%, a specificity of 99%, and a positive and negative predictive value of 82% and 99%, respectively [13].
Several environmental conditions have been documented to support the presence of H. capsulatum in an area including high temperature, high humidity and soil pH above 10 and below 5 [14]. The areas with a high number of reported cases of histoplasmosis and significant skin reactivity to histoplasmin had soil types (pH ≥7) that overlapped [14]. The cities of Benin (15%) and Calabar (6%) which reported the highest rates were both in humid rainforest zone [14]. Individuals with a home or workplace adjacent to the garden had higher levels of skin reactivity than those without fruit trees [14]. The skin test survey showed a significantly higher positivity rate among people working/living around the forest, most of whom were from rural areas, a finding consistent with a previous report reporting 8.8% histoplasmin reactivity in patients visiting rural clinics [14].
Histoplasmosis has been considered a co-infective disease with acquired immunodeficiency syndrome (AIDS). A previous study found that the majority of histoplasmosis patients were men with a mean age of 35 years old and immunodeficiency status was found in the majority of patients [15]. Another study found that among histoplasmosis patients, 58% had human immunodeficiency virus (HIV) infection, 28% of patients had other immune deficiencies, and only 14% had no immune deficiencies [16]. Other than HIV infection, histoplasmosis is associated with environmental and toxin exposures, autoimmune diseases, malignancies and their treatment, chronic lung disease, immunosuppressive therapy, pancytopenia, T-cell deficiency, and malnutrition [17]. A study demonstrated that variables associated with decreased incidence of histoplasmosis were antiretroviral treatment for more than 6 months, fluconazole treatment, and pneumocystosis among HIV-positive patients [18]. A study found that the mortality was 13.5% at 1 month, 17.5% at 3 months, and 22.5% at 6 months after the diagnosis of histoplasmosis [18].
The clinical presentation of acute pulmonary histoplasmosis is milder and often asymptomatic in immunocompetent patients. After an acute pulmonary histoplasmosis infection, chronic pulmonary histoplasmosis (CPH) may develop. CPH is generally difficult to distinguish from pulmonary TB based on its clinical appearance; therefore, diagnostic errors often occur [19]. Considering that Indonesia is a Second World country with the largest number of TB cases and the success rate of therapy is below the target, the possibility of multidrug-resistant pulmonary tuberculosis (MDR-TB) with histoplasmosis can occur. Pulmonary TB, especially MDR-TB, can coexist with CPH. A study at the Persahabatan Hospital showed fungal colonization in patients with former MDR-TB was 68.9% (42 out of 61 patients) [20].
The similarity of factors influencing the incidence between histoplasmosis and TB makes it highly likely that histoplasmosis may also be associated with MDR-TB. However, study assessing the association between MDR-TB and histoplasmosis is limited. Therefore, the aim of this study was to determine the risk factors of histoplasmosis incidence in MDR-TB patients in Indonesia.
Methods
Study setting and patients
A cross-sectional study was conducted at MDR-TB Polyclinic of H. Adam Malik General Hospital, Medan, Indonesia from December 2022 to April 2023, to determine factors associated with histoplasmosis incidence in MDR-TB patients. The sampling method employed was total sampling.
All MDR-TB patients who: (a) aged >18 years; (b) having confirmed diagnosis of MDR-TB based on molecular rapid tests; (c) willing to participate in the study by signing an informed consent form were considered eligible. All patients with MDR-TB with extensive resistance, having fungal infection, pregnancy, or having severe psychosis were excluded.
Data collection and study variables
Some potential associated factors were collected during the study through interviews, medical records and direct assessment. The interview asked the patients related to comorbid diseases, history of previous TB treatment, history of having pets, condition of the house, history of working in the field or plantations, and the patient’s smoking history. Data regarding anti-TB therapy history, Mycobacterium Growth Indicator Tube (MGIT) culture results, and the extent of the lesion on the chest x-ray were also collected.
The dependent variable of the study was the incidence of histoplasmosis. Briefly, approximately 5 mL of venous blood was collected. Histoplasma IgG antibody analysis was then performed using an MVista Histoplasma Quantitative Enzyme Immunoassay (EIA) kit (MiraVista Diagnostics, Indianapolis, US). This semi-quantitative platform was able to detect the presence of IgG Histoplasma antibody in MDR-TB patients.
Statistical analysis
The data obtained was processed using the SPSS computer program version 26.0 (SPSS Inc., Chicago, USA). Chi-squared test or Fisher exact test was used to assess risk factors associated with histoplasmosis incidence in MDR-TB patients as appropriate. A p-value less than 0.05 was considered statistically significant.
Results
Patients’ characteristics
A total of 50 MDR-TB patients were included in this study and the characteristic data of the patients are presented in Table 1. The majority of the patients were male and aged between 46– 60 years old age. Among the patients, most were non-smokers, 22% had a history of working in the field or plantations and 48% lived in damp houses. The most common comorbidity recorded was diabetes mellitus (DM) and 50% of them had no comorbid. There were 38% of patients that had far-advanced extensive lesions based on their chest x-rays.
Table 1. Demographic and clinical characteristics of MDR-TB patients included in the study (n=50).
| Subject characteristics | Frequency | Percentage |
|---|---|---|
| Gender | ||
| Male | 36 | 72 |
| Female | 14 | 28 |
| Age (years old) | ||
| 16−30 | 7 | 14 |
| 3−45 | 16 | 32 |
| 46−60 | 21 | 42 |
| >60 | 6 | 12 |
| Comorbidity | ||
| Diabetes mellitus | 22 | 44 |
| HIV | 2 | 4 |
| Hypertension | 3 | 6 |
| Asthma | 1 | 2 |
| Cancer | 1 | 2 |
| Subject characteristics | Frequency | Percentage |
| No comorbidities | 25 | 50 |
| Tuberculosis treatment history | ||
| No | 18 | 36 |
| Yes | 32 | 64 |
| Acid-fast bacillus result | ||
| Positive | 11 | 22 |
| Negative | 39 | 78 |
| Mycobacterium tuberculosis culture | ||
| Positive | 34 | 68 |
| Negative | 16 | 32 |
| Pets | ||
| Yes | 15 | 30 |
| No | 35 | 70 |
| Damp house | ||
| Yes | 24 | 48 |
| No | 26 | 52 |
| History of working in the field or plantation | ||
| Yes | 11 | 22 |
| No | 39 | 78 |
| History of smoking | ||
| Smoker | 0 | 0 |
| Ex-smoker | 15 | 30 |
| Non smoker | 35 | 70 |
| Extent of the lesion on chest X-ray | ||
| Minimal advanced | 15 | 30 |
| Moderately advanced | 16 | 32 |
| Far-advanced | 19 | 38 |
| Histoplasmosis | ||
| Negative | 36 | 72 |
| Positive | 14 | 28 |
Factors associated with histoplasmosis incidence
Out of the total MDR-TB patients, 14 (28%) patients tested positive for histoplasmosis. Our univariate analysis indicated that sex (p=1.000), age group (p=0.116), TB treatment history (p=0.744), acid-fast bacillus result (p=0.148), Mycobacterium tuberculosis culture result (p=0.330), having pet (p=1.000), living in damp house (p=0.352), working in the field or plantation (p=1.000), having HIV infection (p=1.000), smoking status (p=1.000), and the status of the lung lesion (p=0.069) were not associated significantly with the incidence of histoplasmosis (Table 2).
Table 2. Association between factors and histoplasmosis incidence in MDR-TB patients.
| Risk factor | Histoplasmosis | p-value | |||
|---|---|---|---|---|---|
| Negative (n=36) | Positive (n=14) | ||||
| n | % | n | % | ||
| Gender | 1.000 | ||||
| Male | 26 | 72.2 | 10 | 71.4 | |
| Female | 10 | 27.8 | 4 | 28.6 | |
| Age (years old) | 0.116 | ||||
| 16−30 | 7 | 19.4 | 0 | 0 | |
| 31−45 | 13 | 36.1 | 3 | 21.4 | |
| 46−60 | 13 | 36.1 | 8 | 57.1 | |
| >60 | 3 | 8.4 | 3 | 21.4 | |
| TB treatment history | 0.744 | ||||
| No | 14 | 38.9 | 4 | 28.6 | |
| Yes | 22 | 61.1 | 10 | 71.4 | |
| Acid-fast bacillus result | 0.148 | ||||
| Positive | 10 | 27.8 | 1 | 7.1 | |
| Negative | 26 | 72.2 | 13 | 92.9 | |
| Mycobacterium tuberculosis culture | 0.330 | ||||
| Positive | 10 | 27.8 | 6 | 42.9 | |
| Negative | 26 | 72.2 | 8 | 57.1 | |
| Pet | 1.000 | ||||
| Yes | 11 | 30.6 | 4 | 28.6 | |
| No | 25 | 69.4 | 10 | 71.4 | |
| n | % | n | % | ||
| Damp house | 0.352 | ||||
| Yes | 17 | 47.2 | 9 | 64.3 | |
| No | 19 | 52.8 | 5 | 35.7 | |
| History of working in the field or plantation | 1.000 | ||||
| Yes | 17 | 47.2 | 6 | 42.9 | |
| No | 19 | 52.8 | 8 | 57.1 | |
| HIV infection | 1.000 | ||||
| Yes | 2 | 5.6 | 0 | 0 | |
| No | 34 | 94.4 | 14 | 100 | |
| History of smoking | 1.000 | ||||
| Yes | 26 | 72.2 | 9 | 64.3 | |
| No | 10 | 27.8 | 5 | 35.7 | |
| Extent of the lesion on chest X-ray | 0.069 | ||||
| Far advanced | 11 | 30.6 | 8 | 57.1 | |
| Moderate advanced | 11 | 30.6 | 5 | 35.7 | |
| Minimal advanced | 14 | 38.8 | 1 | 7.1 | |
Discussion
MDR-TB, defined as resistance to isoniazid and rifampicin, is a growing threat and a significant burden on healthcare systems worldwide. A meta-analysis study demonstrated that a history of previous TB disease, incomplete TB treatment, and failure of TB treatment were strongly associated with MDR-TB [24]. Our study found the majority of MDR-TB patients who tested positive for histoplasmosis were male and aged 46−60 years old. A study reported a male-female ratio of 1.2:1 among suspected pulmonary TB with histoplasmosis and the mean age at presentation was 39 ± 14 years. The study also reported that 35% of the patients were people living with HIV, compared to 44.1% who had been diagnosed with TB [17]. Our study only found two HIV-positive individuals among the MDR-TB patients. This implies that there is a lower prevalence of MDR-TB among HIV patients in comparison to those with pulmonary tuberculosis. Unfortunately, this study could not find any associations of sex, age group, TB treatment history, having HIV infection among MDR-TB patients with the incidence of histoplasmosis.
Indonesia’s tropical climate, with its high humidity and warm temperatures, provides an ideal environment for the growth of H. capsulatum, where the primary reservoir is soil [4,21]. Our study found that 28% of the MDR-TB patients tested positive for histoplasmosis and 42.9% of them had a history of working in the field or plantation, indicating that they were from agricultural areas. However, a study in Jakarta, a city in Indonesia, reported histoplasmosis was present in 11% of adults and 2.7% of children [22]. These similarities suggest that the occurrence of histoplasmosis in Indonesia remains relatively high. Other fungal pulmonary infections, such as coccidioidomycosis and paracoccidioidomycosis, are not endemic to Indonesia. Consequently, the occurrence of these infections is possible only through international travel [23]. Moreover, in Latin America, the estimated histoplasmosis incidence is equivalent to that of TB in people living with HIV, and the highest overall prevalence of prior histoplasmosis exposure was observed in Guatemala, Mexico [25].
The true burden of HIV-associated histoplasmosis is not fully known or overlooked, mainly due to its non-reportable disease. Histoplasmosis is still widely mistaken for MDR-TB, leading to numerous preventable deaths [28]. The clinical presentation of histoplasmosis is also very similar to that of TB, including primary pulmonary involvement that is self-limited and a latent phase that can cause coin lesions on chest radiographs; chronic secondary progressive lung disease localized at the apex of the lung and causing fever, coughing, and night sweats; and then progressing with widespread involvement [27]. A study reported that 44 of 45 patients had cavitation in the upper lobe, more often occurring in the right apex (84%). The cavity often had thick walls with some thickening of the apical pleura. Although calcified pulmonary granulomas are common (77.8%), hilar lymphadenopathy is rare (2.2%) compared with acute lung disease [19,29]. A cavity on a chest x-ray with a lesion diameter of >4 cm is called a far-advanced lesion. The majority of patients having this far-advanced lesion suffer from histoplasmosis.
Conclusion
This study found that 28% of MDR-TB patients tested positive for histoplasmosis. Interestingly, all factors, including sex, age group, TB treatment history, acid-fast bacillus result, Mycobacterium tuberculosis culture result, having pet, living in damp house, working in the field or plantation, having HIV infection, smoking status, and the status of the lung lesion had no significant association with the incidence of histoplasmosis. Although multiple risk factors have been included, this study suggests that there could be other factors associated with the incidence that were missed from our study and warrant further study.
Acknowledgments
The authors would like to thank Haji Adam Malik General Hospital as our attending hospital and Integrated Lab Universitas Sumatera Utara. The authors would also like to thank MiraVista Diagnostics in Indianapolis, USA, for delivering the EIA kit. The authors thank the supervisors, Robiatul Adawiyah and Anna Rozaliyani, who have assisted in the preparation of this study. The author also expresses gratitude to all parties who have helped in the realization of this research.
Ethics approval
The study was approved by H. Adam Malik General Hospital, Medan, Indonesia (LB.02.02/XV.III.2.2.2/802/2023).
Competing interests
The authors declare that there is no conflict of interest.
Funding
This study received a research grant from Indonesian Science Fund (Dana Ilmu Pengetahuan Indonesia)/Lembaga Pengelolaan Dana Pendidikan (DIPI/LPDP) and DWD from the UK Medical Research Council with contract number MR/P017622/1.
Underlying data
Derived data supporting the findings of this research are available from the corresponding author on request.
How to cite
Khairunnisa S, Soeroso NN, Abdullah M, et al. Factors influencing histoplasmosis incidence in multidrug-resistant pulmonary tuberculosis patients: A cross-sectional study in Indonesia. Narra J 2023; 3 (3): e403 - http://doi.org/10.52225/narra.v3i3.403.
References
- 1.Cano MV, Hylorajjeh RA. The epidemiology of histoplasmosis: A review. Semin Respir Infect 2001;16(2):109–118. [DOI] [PubMed] [Google Scholar]
- 2.Norkaew T, Ohno H, Sriburee P, et al. Detection of environmental sources of Histoplasma capsulatum in Chiang Mai, Thailand, by Nested PCR. Mycopathologia 2013;176(5–6):395–402. [DOI] [PubMed] [Google Scholar]
- 3.Baker J, Setianingrum F, Wahyuningsih R, Denning DW. Mapping histoplasmosis in South East Asia-implications for diagnosis in AIDS. Emerg Microbes Infect 2019;8(1):1139–1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tanjung A, Mardianto. Prevalensi histoplasmosis pada mahasiswa kedokteran Universitas Islam Sumatera Utara dan hubungan hewan peliharaan dengan tes histoplasmin. Berkala Ilmu Kedokteran. 1997;3:139–144. [Google Scholar]
- 5.Caceres DH, Valdes A.. Histoplasmosis and tuberculosis co-occurrence in people with advanced HIV. J Fungi 2019;5(3):73:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Edwards PQ, Klaer JH, Principe A, et al. Worldwide geographic distribution of histoplasmosis and histoplasmin sensitivity. Am J Trop Med Hyg 1952;5:235–257. [PubMed] [Google Scholar]
- 7.Oladele RO, Ayanlowo OO, Richardson MD, Denning DW. Histoplasmosis in Africa: An emerging or a neglected disease? PLoS Negl Trop Dis 2018;12(1): e0006046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bahr NC, Sarosi GA, Meya DB, et al. Seroprevalence of histoplasmosis in Kampala, Uganda. Med Mycol 2016;54(3):295–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Souza SL, Feitoza PV, Araujo JR, et al. Causes of death among patients with acquired immunodeficiency syndromeautopsiedat the Tropical Medicine Foundation of Amazonas. Revista da Sociedade Brasileira de Medicina Tropical 41(3):247–251. [DOI] [PubMed] [Google Scholar]
- 10.Eza D, Cerrillo G, Castro C, et al. Resultados post mórtem e infecciones oportunistas en pacientes VIH-positivos de un hospital público del Perú. Rev Peru Med Exp Salud Publica 2006;23(4):270–274. [Google Scholar]
- 11.Peter Donnelly J, Chen SC, Kauffman CA, et al. Revision and update of the consensus definitions of invasive fungal disease from the european organization for research and treatment of cancer and the mycoses study group education and research consortium. Clin Infect Dis 2020;71(6):1367–1376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wijaya M, Adawiyah R, Wahyuningsih R.. Histoplasmosis: Diagnostic and therapeutic aspect. Indonesian J Trop Infect Dis 2021;9(2):66–77. [Google Scholar]
- 13.Agossou M, Turmel JM, Aline-Fardin A, et al. Acute pulmonary histoplasmosis of immunocompetent subjects from Martinique, Guadeloupe and French Guiana: a case series. BMC Pulm Med 2023;23(1):95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Oladele RO, Toriello C, Ogunsola FT, et al. Prior subclinical histoplasmosis revealed in nigeria using histoplasmin skin testing. PLoS One 2018;13(5):20196224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Damasceno LS, Teixeira M de M, Barker BM, et al. Novel clinical and dual infection by Histoplasma capsulatum genotypes in HIV patients from Northeastern, Brazil. Sci Rep 2019;9(1):11789. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Camous L, Surel A, Kallel H, et al. Factors related to mortality in critically ill histoplasmosis: A multicenter retrospective study in Guadeloupe and French Guyana. Ann Intensive Care 2023;13(1):30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ekeng BE, Oladele RO, Emanghe UE, et al. Prevalence of Histoplasmosis and molecular characterization of Histoplasma species in patients with presumptive pulmonary tuberculosis in Calabar, Nigeria. Open Forum Infect Dis 2022;9(8):1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nacher M, Adenis A, Blanchet D, et al. Risk factors for disseminated histoplasmosis in a cohort of HIV-infected patients in French Guiana. PLoS Negl Trop Dis 2014;8(1):7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wheat LJ, Wass J, Norton J, et al. Cavitary Histoplasmosis occurring during two large urban outbreaks: Analysis of clinical, epidemiologic, roentgenographic, and laboratory features. Medicine 1984;63(4):201–209. [DOI] [PubMed] [Google Scholar]
- 20.Rozaliyani A, Jusuf A, Priyanti ZS, et al. Pulmonary mycoses in Indonesia: Current situations and future challenges. J Resp Indonesia 2019;39(3):210–214. [Google Scholar]
- 21.Delima SM. Berbagai kasus histoplasmosis di Indonesia 1932-1988. Medika 1999;4(16):312–318. [Google Scholar]
- 22.Lie KL, Tjokronegoro S, Njo-Injo TE, Lim GF. Histoplasmosis, mainan tentang sebuah peristiwa. MKI 1956;6(2):45–52. [Google Scholar]
- 23.Setianingrum F, Rozaliyani A, Adawiyah R, et al. A prospective longitudinal study of chronic pulmonary aspergillosis in pulmonary tuberculosis in Indonesia (APICAL). Thorax 2022;77(8):821–828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pradipta IS, Forsman LD, Bruchfeld J, et al. Risk factors of multidrug-resistant tuberculosis: A global systematic review and meta-analysis. J Infect. W.B. Saunders Ltd; 2018:469–478. [DOI] [PubMed] [Google Scholar]
- 25.Reinhardt SW, Spec A, Meléndez J, et al. AIDS-defining illnesses at initial diagnosis of HIV in a large Guatemalan cohort. Open Forum Infect Dis 2016;3:1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Dewi IMW, Fauziah N, Ekawardhani S, et al. Antibodies against Histoplasma capsulatum and Aspergillus fumigatus among chronic TB patients in Indonesia: A cross-sectional study. Med Mycol 2023;61(5):1–15. [DOI] [PubMed] [Google Scholar]
- 27.Qureshi A. A case of Histoplasmosis mimicking tuberculosis. JPMA 2008; 58(8):457–458. [PubMed] [Google Scholar]
- 28.Adenis AA, Aznar C, Couppie P.. Histoplasmosis in HIV-infected patients: A review of new developments and remaining gaps. Curr Trop Med Rep 2014;1:119–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Baker J, Kosmidis C, Rozaliyani A, et al. Chronic pulmonary histoplasmosis-A scoping literature review. Open Forum Infect Dis 2020;7(5):ofaa119. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Derived data supporting the findings of this research are available from the corresponding author on request.