Manifestation, diagnosis, and management of foodborne trematodiasisBMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e4093 (Published 26 June 2012) Cite this as: BMJ 2012;344:e4093
- Thomas Fürst, research fellow12,
- Somphou Sayasone, research fellow3,
- Peter Odermatt, associate professor and research group leader12,
- Jennifer Keiser, professor and head of the helminth drug development unit24,
- Jürg Utzinger, professor and head of the ecosystem health sciences unit12
- 1Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, PO Box, CH-4002 Basel, Switzerland
- 2University of Basel, Basel, Switzerland
- 3National Institute of Public Health, Ministry of Health, Vientiane, Lao People’s Democratic Republic
- 4Department of Medical Parasitology and Infection Biology, Swiss Tropical and Public Health Institute, Basel, Switzerland
- Correspondence to: J Utzinger
- Accepted 19 April 2012
Foodborne trematodiasis is a cluster of zoonotic infections caused by parasitic trematodes ingested in undercooked, mainly aquatic, food
Prevalence is increasing because of the growth of inland fish production; most cases are in Asia and Latin America, but infections in migrants and returning travellers are reported elsewhere, including Europe and North America
Foodborne trematodes are grouped as liver, lung, and intestinal flukes and—depending on the species, the duration and intensity of infection, and host susceptibility—inflammatory lesions and damage to tissues and organs occur with various clinical manifestations
The most serious clinical consequences are cancer of the bile duct (in clonorchiasis and opisthorchiasis) and ectopic infections (mainly in paragonimiasis, but also in fascioliasis and intestinal fluke infections)
Direct parasitological diagnosis via the detection of eggs in faeces (all flukes) and sputum (lung flukes only) is the most common approach
Praziquantel and triclabendazole are safe and efficacious treatments but other drugs are being investigated
Foodborne trematodiasis is a cluster of zoonotic infections caused by parasitic worms (class: trematoda; also known as flukes), which are transmitted via the ingestion of contaminated, mainly aquatic, food. More than one billion people are at risk of infection according to a systematic review from 2005.1 Another systematic review and meta-analysis suggests that 56 million people were infected in 2005, mainly in Asia and Latin America, with a global burden of 665 000 disability adjusted life years.2
Depending on the fluke species, foodborne trematodiasis is associated with a variety of signs, symptoms, and pathological consequences. The non-specificity of the clinical manifestations, the wide range of fluke species, and shortcomings in current diagnostic techniques are some of the reasons why foodborne trematodiasis is underestimated.3 4
This review introduces the most important foodborne trematode species and describes their geographical distribution. It also discusses pathological consequences, clinical manifestations, diagnosis, treatment, and control of foodborne trematodiasis. Our review is based on the limited evidence obtained from the peer reviewed literature, textbooks, reports, and international guidelines.
Sources and selection criteria
Information for this clinical review was obtained from a database that we established for a project to estimate the global burden of foodborne trematodiasis.2 The database originated from a systematic review of 11 electronic datasources: PubMed, WHOLIS, FAOBIB, Embase, CAB Abstracts, LILACS, ISI Web of Science, BIOSIS Preview, Science Direct, African Journals Online, and SIGLE. It included all available literature from 1 January 1980 to 31 December 2008 and had no language restrictions. Details on the initial search strategy and database have been presented elsewhere.2 For this review, the database was updated to include all available information until 30 September 2011. The data were complemented by personal reference archives and the authors’ experience.
What causes foodborne trematodiasis?
A recent systematic review listed more than 80 different trematode species that have been identified from human infections.2 According to the target organ in the definitive host, they are grouped as liver, lung, or intestinal flukes. However, on the basis of recent biomedical reviews and still incomplete national prevalence data, only a dozen species are of public health importance (box 1).2 5 6 7 8
Box 1 Foodborne trematode species of public health importance
The life cycles of the foodborne trematodes are species specific, with distinct snail species being first intermediate hosts, and fish, mollusc, crustacean, amphibian, and insect species being second intermediate hosts (fig 1⇓). Notable exceptions are Fasciola spp and F buski, which do not need a second intermediate host, as their infectious stages (so called metacercariae) adhere directly to aquatic plants.3 A detailed list of known first and second intermediate hosts can be found in the annex of a comprehensive technical report published by the World Health Organization.9
Humans usually acquire an infection through the ingestion of second intermediate hosts or, in the case of Fasciola spp and F buski, aquatic plants that contain viable metacercariae. With the exception of Fasciola spp and F buski, metacercariae are not released from their intermediate hosts into water, so the risk of infection from drinking untreated water is small.9 10 It is unclear which aspects of food processing (such as heating, freezing, smoking, acidification, salting, drying, washing, disinfection, irradiation, and pressure treatment) inhibit the infectivity of metacercariae, but properly cooked or deeply frozen food is considered safe.9 11 In the duodenum of an infected human, hermaphroditic juvenile flukes develop from the metacercariae, migrate to their target organ, mature, mate, and start producing eggs. Parasite eggs are then released via the human host’s faeces (all foodborne trematodes, including coughed up and swallowed eggs of Paragonimus spp) or sputum (only coughed up eggs of Paragonimus spp) and have to reach appropriate water bodies with suitable intermediate hosts to complete their life cycles.3 12
Where does foodborne trematodiasis occur?
Foodborne trematodiasis is commonly classified as a tropical disease, even though the endemic area is not limited to the tropics.9 11 C sinensis is endemic in East and South East Asia; O viverrini in South East Asia; and O felineus in central, northern, and western Eurasia. Fasciola spp exist worldwide, but most endemic areas are in the Andean region, North Africa, and the Caspian Sea region. Paragonimus spp occur in parts of the Andean region, West and Central Africa, East and South East Asia, and North America. Intestinal flukes are found worldwide, with most endemic areas in East and South East Asia. A comprehensive literature review identified allochthonous cases (diagnosed in countries where disease transmission does not naturally occur) all over the world, probably as a result of increasing international travel, human migration, and the food trade.13 A series of recent reviews suggests that human foodborne trematodiasis is increasing, mainly because of the exponential growth of inland fish production (aquaculture).1 2 9 11 w1 w2
At a regional level, foodborne trematodiasis usually shows a focal distribution, which is governed by social-ecological contexts (such as specific eating habits and environmental conditions that favour maintaining the parasites’ life cycles).3 11 At the individual level, a meta-analysis of the proportion of infected humans shedding high numbers of eggs indicates that distribution is highly aggregated: a few people harbour most of the parasites.2 Hospital based and community based cross sectional surveys in endemic areas show that infection with multiple species of foodborne trematodes is common because they are all acquired through consumption of raw food.14 w3 w4
What are the pathological consequences and clinical manifestations?
Morbidity depends on the species involved and also the host’s susceptibility, duration of infection, and the number of worms harboured (infection intensity). These parameters govern the occurrence and severity of inflammatory lesions and damage to tissues and target organs. The severity of infection is usually determined by the number of parasite eggs per gram of faeces or per 5 mL of sputum in paragonimiasis.3 4 5 6 7 8 9 11 15 16 17 w5-w19 Because the egg laying capacity of foodborne trematodes varies greatly—from fewer than 100 eggs per day per worm (Haplorchis taichui)18 to 13 000-26 000 (F buski)19—species specific thresholds are used to differentiate between light, moderate, and heavy infections. However, these thresholds have never been standardised. Furthermore, some studies challenge a direct association between egg counts and worm burden because crowding effects may lower egg production, obstructions in the hosts’ organs may affect the excretion of eggs, and the distribution of eggs in faeces can be uneven.15 16 20 w20-w24
A patient’s perspective: opisthorchiasis in Lao People’s Democratic Republic
I am a 46 year old teacher and often go fishing in the Mekong River with friends. We eat the caught fish uncooked.
For more than a year I felt unwell—tired and without energy. I had many gastritis-like symptoms, such as abdominal pain, bowel rumbling, nausea, and bloating. I also had itchy rashes on my arms, stomach, and legs. I lost about 5 kg in weight but always felt hungry. After meals, particularly dinner, I often had stomach cramps and sometimes vomiting.
After repeated visits to different health services, I went to the hospital and had an abdominal ultrasound scan, in addition to blood and stool tests. The ultrasound results were normal and my blood was negative for hepatitis. However, liver fluke eggs were detected in my stool. I learnt that this parasite is acquired by consumption of raw fish and that after treatment I could be reinfected if I continue to eat undercooked fish. I received medicine called praziquantel. The rash on my arms, legs, and abdomen disappeared two weeks after treatment and I generally felt better. One month later my stool was free of parasite eggs. However, the abdominal discomfort disappeared only slowly. Three months after treatment, I occasionally have bowel rumbling, nausea, and abdominal bloating. My doctor assures me that the cure takes time and that it is most important not to eat raw fish.
Clonorchiasis and opisthorchiasis
The pathological consequences and clinical manifestations of infection with C sinensis and Opisthorchis spp are similar. After ingestion by the host, the metacercariae excyst in gastric juice and migrate via the duodenum, the ampulla of Vater, and the extrahepatic biliary system to the intrahepatic bile ducts.11 Pathological changes are mainly confined to the bile duct, liver, and gallbladder.7 Tissue damage is caused directly by the parasite via mechanical and chemical irritation and indirectly via the immune response, and it can lead to cancer of the bile duct (cholangiocarcinoma; box 2).11 21 Hepatomegaly, gallbladder enlargement, gallstones, sludge, and periductal fibrosis along the intrahepatic biliary tree and periportal vein are often seen with ultrasonography (figs 2⇓ and 3⇓).4
Box 2 Liver fluke induced cholangiocarcinoma
Pathological consequences and clinical manifestations
Cholangiocarcinoma is a malignant tumour of the bile duct epithelium. Although the exact mechanism of liver fluke induced carcinogenesis is unclear, chronic biliary infection increases the susceptibility of the bile ducts to the action of carcinogens. Patients usually present with non-specific symptoms, which are similar to those of liver fluke infections. Liver fluke induced cholangiocarcinoma cannot be differentiated from other forms of cholangiocarcinoma. Timely diagnosis is rare and prognosis is poor, even when patients receive appropriate treatment.
The diagnosis, localisation, and staging of cholangiocarcinoma are challenging and require a combination of imaging, biochemical investigations, and cytological techniques (ultrasonography, cholangiography, choledochoscopy, computed tomography, positron emission tomography, magnetic resonance imaging, biopsy and cytological analysis, and laboratory analysis of serum tumour markers).
Consultation with an infectious disease specialist, hepatologist, and oncologist is recommended. Response to chemotherapy is poor. Complete resection with negative histological margins or transplantation are the only curative interventions. Currently, no effective adjuvant treatment exists, but radiotherapy may be indicated postoperatively. Often, only palliative treatment remains.
Acute and light infections are mostly asymptomatic,3 9 11 21 but an acute onset with hepatitis-like symptoms, including high fever and chills, has been reported, particularly for infection with O felineus.3 7 11 w25 Some moderately infected people may have mild symptoms,3 21 w8 but even chronic infections may remain asymptomatic, and often only heavily infected people have symptoms, signs, and complications (table 1⇓).3 7 9 11 21
Unlike other liver flukes, after ingestion and excystment in the duodenum Fasciola spp migrate through the intestinal wall into the body cavity and then through the liver into the bile ducts.11 15 Major pathological changes are associated with migration and the related destruction, focal bleeding, and inflammation in the host’s body (acute phase). Some migrating flukes may die on their way, leaving cavities filled with necrotic debris, or deviate from their usual route to cause ectopic infections (box 3). In the bile ducts (latent phase), the parasites may cause inflammation, resulting in thickening and expansion of the ducts and fibrosis. Imaging may depict these lesions as “tunnels and caves” in peripheral parts of the liver.4
Box 3 Ectopic foodborne trematode infections
Pathological consequences and clinical manifestations
Paragonimus spp, and less often Fasciola spp and intestinal flukes of the families heterophyidae and microphallidae, can cause ectopic infections. The reasons for incomplete migration are unclear. Migratory tracks and cavities filled with trapped dead parasites or eggs often cause tissue damage with inflammatory reactions and fibrosis. Clinical manifestations depend on the exact location, and such infections can even cause death. Paragonimus spp have been reported in the central nervous system, eyes, skin, heart, abdominal organs, and reproductive organs; Fasciola spp in the central nervous system, orbit, subcutaneous tissue, abdominal wall, heart, genitals, spleen, muscles, gastrointestinal tract, blood vessels, lungs, and pleural cavity; and heterophyidae and microphallidae in the central nervous system and heart.
Suspect ectopic foodborne trematodiasis in a patient with a diagnosis of foodborne trematode infection and unexplained, often severe, manifestations (such as neurological disorders, abscesses) that could result from damage in ectopic locations mentioned above. To confirm the diagnosis, a combination of ultrasonography, radiography, endoscopy, computed tomography, positron emission tomography, magnetic resonance imaging, and biopsy is needed.
Consultation with infectious disease specialists and other relevant specialists is recommended. Treatment with praziquantel or triclabendazole may help in the early phases of infection but should be used with care. Parasite death may lead to antigen release and increased risk of inflammation. Anti-inflammatory drugs (such as corticosteroids) may help to reduce oedema and inflammation during treatment. Parasites, eggs, and lesions may need to be surgically removed. Adjuvant therapy should be used as appropriate.
Early acute manifestations are typical (table 1).8 11 15 Although some infected people are asymptomatic in the latent phase, others may experience repeated relapses of the acute manifestations.8 11 15 When the irritation in the biliary system is severe enough, a permanent obstructive phase, which has additional symptoms, signs, and complications (table 1), may develop.3 4 8 9 11 15 Rarely, fascioliasis can be fatal.15 w26
In the classic natural course of pleuropulmonary paragonimiasis, swallowed metacercariae of Paragonimus spp excyst in the duodenum, penetrate the intestinal wall, and migrate over several days to the pleural cavity and into the lungs where they mature. After several weeks, adults become encapsulated in fibrotic tissue, where they mate or reproduce parthenogenetically. Eggs pass via the bronchioles into the sputum or, if swallowed, the faeces, and then into the environment. Some of the eggs may be trapped in tissue and, together with aberrantly migrating flukes, provoke further irritation and ectopic infections (box 3).6 9 11 22 A recent cross sectional study from India found cavity and cyst formations, nodular lesions, fibrotic infiltrates, calcifications, bronchiectasis, pulmonary consolidations, pleural thickening and effusion, mediastinal lymphadenopathy, and patchy ground glass opacity as pleuropulmonary radiological features in paragonimiasis (fig 4⇓).23
Unless the patient is heavily infected, early stages of pleuropulmonary infection tend to be asymptomatic. Heavy pleuropulmonary infections may present with many different bronchitis-like, asthma-like, and tuberculosis-like symptoms and signs (table 1). The similarity of the clinical manifestations often result in pleuropulmonary paragonimiasis being misdiagnosed as bronchitis, asthma, or (non-responsive) tuberculosis,3 6 9 11 22 even though patients with paragonimiasis usually present in comparatively better general health.
Intestinal fluke infection
More than 70 species of intestinal flukes are implicated in human infection.2 5 The morphology of these flukes is diverse, and their life cycles and geographical distributions not well studied. After ingestion of viable metacercariae, flukes excyst and adhere to the intestinal wall, where mechanical irritation and inflammation may lead to the manifestations described in table 1.3 5 9 11 Similar to other foodborne trematode infections, mild intestinal fluke infections are mainly asymptomatic, but heavy infections can be severe.3 5 9 11 w27
What are the other complications of foodborne trematodiasis?
Recent reviews have summarised the association between infection with C sinensis and O viverrini and cholangiocarcinoma, as well as the mechanisms of carcinogenesis, diagnosis, and management (box 2).21 24 25 w8 w27-w33 On the basis of the epidemiological, experimental, and pathological evidence, these two trematodes have been classified as definite carcinogens (group 1) by the International Agency for Research on Cancer (IARC).26 27 There is still insufficient evidence of the oncogenic potential of O felineus and Fasciola spp for them to be classified.26 27 w28 w34
Ectopic infections are severe and potentially fatal complications of foodborne trematodiasis (box 3). As highlighted in other reviews, Paragonimus spp,6 9 11 22 w5 w11 w35-w39 and more rarely Fasciola spp and intestinal flukes of the heterophyidae and microphallidae families,9 11 15 w26 w27 do not always enter their usual target organs in the host’s body, but continue to migrate to ectopic locations.
How can foodborne trematodiasis be diagnosed?
Accurate diagnosis of foodborne trematodiasis remains a challenge. The three main diagnostic approaches are direct parasitological diagnosis, immunodiagnosis, and molecular diagnosis. Direct parasitological diagnosis is facilitated by the detection of eggs in faeces (all flukes), sputum (only lung flukes), and, more rarely, other biofluids such as bile or duodenal content. Egg detection in faecal samples is the most common approach, and methods include Kato-Katz thick smear, formalin-ethyl-acetate technique, Stoll’s dilution egg count method, and sedimentation techniques. However, differential diagnosis is difficult because parasite eggs from different species resemble each other.3 4 5 6 7 8 11 Furthermore, the small number of eggs discharged by some fluke species, crowding effects, obstructions in the patients’ organs, heterogeneous distribution of eggs in the patients’ faeces (or sputum in Paragonimus spp), and light infections are additional challenges for an accurate diagnosis and require multiple sampling and testing. Occasionally, it is possible to demonstrate adult parasites—for example, when the flukes are excreted in the faeces, coughed up (lung flukes), or removed during surgery.5 7 8 9 11 The website of the US Centers for Disease Control and Prevention features photographs of parasite eggs, which might help in the diagnosis of foodborne trematode infections.
Immunodiagnostic methods such as intradermal tests, indirect haemagglutination assays, indirect fluorescent antibody tests, and indirect enzyme linked immunosorbent assays (ELISAs) aim to detect specific antibodies.3 6 7 8 11 Immunodiagnosis is especially useful during the prepatent phase and ectopic infections, or if direct parasitological diagnosis is ambiguous. However, false positive results after the infection has resolved, cross reactivity, high costs, and lack of availability at the point of care in remote rural areas are important problems that need to be tackled.3 6 7
Molecular diagnosis—the detection of trematode DNA in samples using the polymerase chain reaction—has a high sensitivity and specificity.3 6 7 8 11 However, for the foreseeable future, molecular diagnosis is unlikely to be used at the point of care in endemic areas.3 6
Radiological methods such as ultrasound, computer tomography, and magnetic resonance imaging are complementary tools for an accurate diagnosis. Although these techniques might be available in well equipped laboratories in the developed world and undoubtedly help characterise pathological changes caused by foodborne trematodiasis, they are currently out of reach in resource constrained areas.3 6 7 8 9 11
How can foodborne trematodiasis be treated?
Praziquantel and triclabendazole are the two drugs of choice; however, triclabendazole is currently registered for human use in only four countries. Various treatment regimens are suggested for the treatment of foodborne trematodiasis.3 5 6 7 8 11 28 w40 For clonorchiasis and opisthorchiasis, the recommended treatment schedule is praziquantel 25 mg/kg three times a day for two consecutive days.3 7 11 28 The same treatment regimen has been successfully used in paragonimiasis.3 6 11 28 The first choice treatment against intestinal fluke infections is also praziquantel, and a single dose of 10-20 mg/kg is efficacious against all intestinal fluke species,29 but higher dosages of three times 25 mg/kg in one day have also been proposed.3 5 8 11 28 The drug of choice against fascioliasis is a single (10 mg/kg) or double dose (two times 10 mg/kg) of triclabendazole.3 8 11 28 Triclabendazole (single dose of 10 mg/kg, or two doses of 10 mg/kg each within 12-24 hours) is also efficacious against paragonimiasis.3 In general, all treatments are safe with no serious adverse events. Table 2⇓ summarises the dosages of praziquantel, triclabendazole, and alternative drugs, including reported adverse events and contraindications. Figure 5⇓ shows a diagnostic and treatment algorithm.
Towards control and elimination: challenges and opportunities
New diagnostic techniques that have a high sensitivity and specificity and are simple and inexpensive are key to understanding the extent of the problem.2 Improved point of care diagnostics could conceivably improve the management of cases and avoid severe sequelae.
The development of new drugs is a low priority for drug companies.28 The treatment of foodborne trematodiasis currently relies on two drugs, and two small drug intervention studies reported unexpectedly low cure rates against clonorchiasis and fascioliasis.30 31 A recent review recommended pursuing promising drug candidates (such as tribendimidine, the artemisinins, and synthetic trioxolanes) and combination treatments.32 The development of vaccines for animals against infections with Fasciola spp is at an advanced stage. Vaccination of animals and humans may become an important means of interrupting transmission.33 w43
Drugs are currently the main method of controlling the morbidity associated with foodborne trematodiasis, but integrated programmes are vital for sustainable disease control and eventual elimination.11 15 19 w22 w27 w44 w45 Several follow-up studies highlighted the complexity of the epidemiological settings and showed high reinfection rates after drug based interventions.w21 w46-w51 Hence, integrated control strategies should also include improved sanitation, food inspections, information, education, and communication campaigns (also for travellers) and, as far as feasible, control of intermediate, reservoir, and non-human definitive hosts. These additional interventions aim to change human behaviour and interrupt disease transmission.3 6 7 8 9 11 21 w2 w30 However, considering deeply rooted eating habits in humans and the myriad non-human hosts these aims pose formidable challenges. Only a few endemic countries have embarked on national control programmes against foodborne trematodiasis. Integration with control programmes targeting other infectious diseases as well as collaborations beyond the health sector (such as with the agricultural, environmental, and educational sectors) may offer largely untapped opportunities for prevention and control.9 34 w2
Additional educational resources for patients and healthcare professionals
US Centers for Disease Control and Prevention (www.cdc.gov/parasites/clonorchis; www.dpd.cdc.gov/dpdx/html/clonorchiasis.htm; www.cdc.gov/parasites/opisthorchis; www.dpd.cdc.gov/dpdx/html/opisthorchiasis.htm; www.cdc.gov/parasites/fasciola; www.dpd.cdc.gov/dpdx/html/fascioliasis.htm; www.cdc.gov/parasites/paragonimus; www.dpd.cdc.gov/dpdx/html/paragonimiasis.htm; www.cdc.gov/parasites/fasciolopsis; www.dpd.cdc.gov/dpdx/html/fasciolopsiasis.htm; www.dpd.cdc.gov/dpdx/html/echinostomiasis.htm; www.dpd.cdc.gov/dpdx/html/heterophyiasis.htm; www.dpd.cdc.gov/dpdx/html/metagonimiasis.htm)—One of the few fully functional, up to date, open access sources of information. It contains a wealth of data on the most important foodborne trematode infections and is suitable for healthcare professionals and patients
Areas for future research
More precise data on the extent of the human and veterinary disease burden of foodborne trematode infections (geographical distribution, prevalence, infection intensity, coinfection, subtle morbidity, and societal impact) are needed to increase awareness (in patients and clinicians) and raise political commitment to foster control and elimination efforts
Additional studies should verify the taxonomy of foodborne trematodes and associated intermediate, reservoir, and definitive hosts
New diagnostic methods with a high sensitivity and specificity that are inexpensive and can be used at point of contact are needed
New trematocidal drugs and alternative treatment regimens (for example, multiple dosing, combination treatment) that are safe and efficacious need to be developed
Molecular research on the antigenic structure and immunology of foodborne trematodes and research into the immune mechanisms of infected humans and animals may be useful for vaccine development
A better knowledge of liver fluke induced carcinogenesis may improve the diagnosis, treatment, and prevention of cholangiocarcinoma and provide fundamental insights into carcinogenesis in general
Mechanisms of ectopic foodborne trematode infections should be further explored to improve diagnosis, treatment, and prevention of severe complications
Improved knowledge on physical and chemical parameters to inhibit infectivity of metacercariae could help to advance safe food processing methods and guidelines
The cost effectiveness of integrated control programmes for foodborne trematodes—which include different stakeholders, a variety of interventions (such as chemotherapy, improved sanitation, food inspections, information, education and communication campaigns, and control of non-human parasite life cycles), and that take advantage of synergy between different sectors (public health, livestock production, food industry, water and sanitation, socioeconomic development, education)—should be determined
Cite this as: BMJ 2012;344:e4093
We thank P Steinmann, H Zhou, M Tsai, P Ayé Soukhathammavong, K Phongluxa, G Casagrande, H Immler, R Hirsbrunner, M Hellstern, R Gutknecht, and H Walter (all Swiss Tropical and Public Health Institute, Basel, Switzerland) for their invaluable help in obtaining and translating foreign language literature. Sincere thanks to P Ayé Soukhathammavong for providing figs 2 and 3 and M Strobel for providing fig 4. We are also grateful to S Becker and S Odermatt-Biays for carefully checking the signs, symptoms, and pathological implications of foodborne trematodiasis.
Contributors: TF, JK, and JU conceived and wrote the first draft. SS and PO provided the patient perspective, organised figs 2-4, and revised the manuscript. All authors had access to all data. TF and JU are guarantors.
Funding: TF is associated to the National Centre of Competence in Research (NCCR) North-South and received financial support through a Pro*Doc Research Module from the Swiss National Science Foundation (SNSF; project number PDFMP3-123185). JK acknowledges a personal career development grant from SNSF (projects no PPOOA3–114941 and PPOOP3_135170). The donors had no influence on study design, data collection, analysis, interpretation, writing, or submission.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: all authors completely disclosed their funding for the submitted work in the funding statement above; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Not commissioned, externally peer reviewed.
Patient consent obtained.