Echinococcosis transmission on the Tibetan Plateau

Abstract

Since the mid-1990s detailed studies and field investigations on the Tibetan Plateau have revealed human echinococcosis to be an under-reported major public health problem, particularly in the dominant pastoral communities in the eastern and central regions. Human prevalence surveys showed that cystic echinococcosis (CE, caused by Echinococcus granulosus) and alveolar echinococcosis (AE, caused by Echinococcus multilocularis) are co-endemic with higher burdens of each disease than other endemic world regions. Epidemiological investigations identified some major risk factors for human CE and AE including dog ownership, husbandry practices and landscape features. Dogs appear to be the major zoonotic reservoir for both E. granulosus and E. multilocularis, but the latter is also transmitted in complex wildlife cycles. Small mammal assemblages especially of vole and pika species thrive on the Plateau and contribute to patterns of E. multilocularis transmission which are influenced by landscape characteristics and anthropogenic factors. Tibetan foxes are a principal definitive host for both E. multilocularis and E. shiquicus. In 2006 a national echinococcosis control programme was initiated in Tibetan communities in northwest Sichuan Province and rolled out to all of western China by 2010, and included improved surveillance (and treatment access) of human disease and regular deworming of dogs with annual copro-testing. Control of echinococcosis in Tibetan pastoral communities poses a difficult challenge for delivery and sustainability.

Introduction

The Tibetan Plateau (also called Qinghai-Tibet plateau, or Qinghai-Xizang plateau) is a vast high altitude ecosystem (average altitude 4000 m, mean average temperature below 0 °C) comprising approximately 2.57 million km² located in southwest China. The Tibetan Plateau is made up of several autonomous regions/prefectures and provincial areas. The Tibet Autonomous Region (TAR, Xizang) is the largest region (1.2 million km²), followed by Qinghai Province (720,000 km²), in addition a substantial area (234,000 km²) of western Sichuan Province (primarily Ganzi Tibetan Autonomous and Aba Tibetan-Qiang Autonomous Prefectures) is located on the Tibetan Plateau as well as three smaller areas (~ 74,230 km² total) including one in southwest Gansu Province (Gannan Tibetan Autonomous Prefecture) and two others in northwest Yunnan Province (Diqing and Niujang Prefectures), and there is also in addition a northern provincial border zone with southern Xinjiang Uygur Autonomous Region (Fig. 1; Supplementary Material). The Tibetan Plateau also has international borders with India, Nepal, Myanmar and Bhutan, and is the source of several great rivers (Yellow, Yangtse, Yarlung Tsangpo, Salween and Mekong). It is framed by mountain ranges to the south (Himalayas), the Kunlun and Arjin ranges to the north, the Qilian mountains to the east and Karakoram/Pamir ranges to the west. The Tibetan Plateau ecosystem is overall a dry and cold tundra-like zone, but distinct topographical regions occur notably associated with dryer (sparsely populated) western regions (< 100 mm rain) and a wetter eastern zone (500–700 mm rainfall) where pastoralism can sustain greater livestock and human population densities. There are several hundred lakes (mostly saline) across the Plateau with Qinghai Lake the largest freshwater lake in China. < 1% of the Tibetan Plateau is cultivated and then only at lower elevations (< 3500 m). Dominant vegetation is made up by grasses and sedges (notably Sipa, Carex, Ceratoides, Festuca and Kobresia spp.) so that rangelands comprise most of this ecosystem and include alpine meadows and high steppe, in total about 8.7 million ha (Shang et al., 2014). There are about 190 species of wild mammal, including ungulates such as Argali, Tibetan gazelle, Tibetan antelope, white-lipped deer, Kiang, wild yak and blue sheep; also carnivores including wolves, Tibetan fox, red fox, brown bear and snow leopard; and a large number of small mammal species including the Tibetan hare, Himalayan marmot, plateau pika (Ochotona curzoniae), and diverse species of microtine and cricetid rodents (Schaller, 1998; Smith and Xie, 2008).

The human population of the Tibetan Plateau is about 6 million of which > 90% are ethnic Tibetan and of those > 70% are semi-nomadic pastoralists with higher population and livestock densities in central and eastern zones. However overall human population density for the Tibetan Plateau is low (< 2 persons/km²). Domestic livestock are principally yak and yak-cattle hybrids (> 13 million head), sheep and goats (> 40 million head) and horses, with free-range pigs and cattle at lower altitudes (< 3500 m) (Shang et al., 2014). Yaks in particular are an important indicator of wealth and status. Domestic dogs are ubiquitous in Tibetan communities across the plateau, while cats are not uncommon in some regions. In Ganzi Tibetan Autonomous Prefecture (GTAP, Tib. Garze), Sichuan Province, the average Tibetan pastoral family contained 5–7 persons, and owned 1–2 dogs, 20–35 yak, 60–80 sheep/goats and 2–3 horses (Hu et al., 2014). In Shiqu (Serxu) County (GTAP) winter pastures are typically located in broad valleys at 4200 m altitude with dispersed community clusters (townships) mainly of mud-brick/sod houses and mud-walled yards. The main fuel for heat and cooking at this altitude is dried yak dung. From May to September families move from winter pastures with livestock and yak felt tents to traditional higher altitude pastures (> 4400 m) usually within a 15–50 km range. Yaks and horses are used for transport. The old, sick and very young persons may be left in winter pastures with old livestock (and a guard dog) during the summer grazing period. Stipa and Festuca grasses provide the higher quality grazing for both wildlife and domestic livestock (Schaller, 1998). Grassland degradation however is now considered a problem which appears to be associated with livestock (trampling and grazing), though very high population densities of plateau pika (O. curzoniae) and smokey vole (Lasiopodomys fuscus) are often blamed for pasture erosion by livestock keepers. As a result indiscriminate poisoning of small mammals occurs in many eastern areas to reduce grazing competition, but also by disease control staff to facilitate plague (Yersinia pestis) control (Wang et al., 2018; Wilson and Smith, 2015). Since the 1990s government initiatives have been implemented to reduce over-grazing including the extensive use of fencing of winter pastures (Qiu, 2016; Wang et al., 2004).

Roasted barley meal or porridge (tsampa) is a staple food for all Tibetans and is occasionally supplemented with meat when available. Livestock may be slaughtered at any time of the year (especially sheep and goats) but higher numbers are killed in early November, particularly yak, in order to store dried meat over the harsh winters. More than 75% of Tibetan livestock keepers practice home-slaughter, though specific slaughter areas may also be used in some townships. There are few abattoirs or controlled slaughter slabs outside large cities. In pastoral townships, owned dogs are used to guard property and protect livestock (mainly from wolf depredation), and may be highly valued and sold or traded at significant cost, including to non-Tibetan buyers from low altitude Chinese cities. In winter settlements dogs are usually tied (leashed) during the day and released at night when they are then free to scavenge, predate on small mammals, mate, roam and form loose packs, usually within a home range of 250 m as observed in Shiqu county (Vaniscotte et al., 2011). Un-owned stray dogs (or ‘community’ dogs) are common, and numbers are usually greater in townships with temples where they may congregate because Buddhist Tibetan monks feed them daily.

In many ways Tibetan regions and autonomous prefectures are still in health transition and slow to adapt to modern biomedical advances (Janes, 1999). Access to healthcare has improved greatly on the Tibetan Plateau since the 1990s, so that most larger townships will have a medical clinic and county towns a basic hospital; however there remains great reliance on traditional cultural-religious health viewpoints, and on the use of traditional Tibetan medicine and practices (Li et al., 2018a). In common with other nomadic pastoral cultures, Tibetan pastoralists, though hospitable, are independent minded and often suspicious of outsiders including healthcare professionals (Zinsstag et al., 2006). Tibetan pastoralists experience challenging health issues in large part because of, or exacerbated by, high altitude effects, a harsh cold climate, poor nutrition, predominantly resource-poor communities, high maternal and infant morbidity and mortality, high illiteracy, and close proximity to domestic and wild animal disease reservoirs.

Common non-infectious chronic conditions of significant incidence in Tibetan communities include: rheumatism, altitude effects (e.g. polycythemia or overproduction of red blood cells), cataracts, vitamin and mineral deficiencies (e.g. endemic Kashin-Beck disease related to iodine/selenium deficiency), gall stones, non-alcoholic fatty liver disease, alcoholic cirrhosis and liver and gastric cancers (Dunzhu et al., 2006; Giordani et al., 2012; Li et al., 2018a; Zhang et al., 2018b). Non-zoonotic infectious diseases of poverty are common including TB, upper respiratory tract infections, microbial and protozoan gastrointestinal infections, ascariasis, scabies, typhus, non-HIV sexually transmitted infections, and hepatitis B and C (Giordani et al., 2012). It is unsurprising that Tibetans, in common with other nomadic pastoral cultures (Macpherson, 1994), are potentially exposed to a number of zoonotic diseases. Endemic bacterial zoonoses include brucellosis, salmonellosis, anthrax and bubonic plague. Viral zoonoses that occur on the Tibetan Plateau include hantavirus infection, influenzas and Japanese encephalitis (Li et al., 2011b; Zhang et al., 2010). Perhaps surprisingly rabies is rare on the Tibetan Plateau region, despite very high frequencies of dog rabies in some southern provinces of China (Wei et al., 2019). Interestingly the rabies strain recently identified in Tibet was of wildlife rather than dog origin (Tao et al., 2015).

In terms of parasitic zoonoses, echinococcosis, the subject of this review, has by far the most public health impact for pastoralists (Budke et al., 2004; Qian et al., 2017; Zheng et al., 2013), and also is one of the most important infectious diseases on the Tibetan Plateau (Barnes et al., 2017; Craig et al., 2008; Li et al., 2010) with a disease burden greater than for most other endemic world regions (Budke et al., 2004, Budke et al., 2006; Torgerson et al., 2010). Of the other parasitic zoonoses endemic in China that have been recorded in Tibetan communities, the helminthic diseases trichinellosis and cysticercosis are probably among the most potentially pathogenic (Li et al., 2006; Liu and Boireau, 2002); however there are a lack of data for these and other parasitic infections such as toxoplasmosis, cryptosporidiosis, amoebiasis, giardiasis and fascioliasis. Domestic pigs are kept by Tibetan households at lower altitudes (< 3500 m) especially in eastern zones, where they are free-roaming and have ready access to garbage and human faeces (Raoul et al., 2013). This results in a high risk for porcine cysticercosis and then via consumption of infested pork risk for human Taenia solium taeniasis, both of which occur at high prevalences in some Tibetan communities which increases the risk for human neurocysticercosis (Li et al., 2006). One community study in Yajiang County (in Ganzi Tibetan Autonomous Prefecture, Sichuan Province) recorded 8.5% occurrence of late-onset epilepsy (a general indicator of neurocysticercosis) and a 4% community sero-prevalence of cysticercosis antibodies against specific glycoprotein antigens; in addition the calculated stool microscopy/coproantigen/copro PCR prevalence of human taeniasis was > 20% (Li et al., 2006). In addition to T. solium, other tapeworm species in Tibetan communities include T. asiatica (also contracted from infected pigs) and the beef tapeworm T. saginata but neither cause human cysticercosis (Craig and Ito, 2007; Li et al., 2013a). Hymenolepis nana is also present but there are no data. Consumption of raw or under-cooked pork is also the main risk factor for human trichinellosis with 85% of cases recorded in southwest China (Yunnan, Guangxi, Tibet Autonomous Region) though consumption of dog and bear meat has been linked to some cases (Wang et al., 2006a).

Human echinococcosis in China and Eurasia presents as two clinical-pathologic forms termed cystic echinococcosis (CE) caused by infection with the larval (metacestode) stage of Echinococcus granulosus (sensu lato), and alveolar echinococcosis (AE) caused by infection with the metacestode of Echinococcus multilocularis (Craig et al., 1996; Deplazes et al., 2017). Both CE and AE are chronic infections that predominantly affect the liver as unilocular or multi-vesiculated cystic lesions respectively; in addition about 20% of CE cyst may also occur in the lungs or other organs or tissues (Kern et al., 2017). Clinical outcomes vary from none or little pathology to life-threatening disease. Hepatic echinococcosis is characterized by long asymptomatic periods of several years, followed by non-specific symptoms usually associated with pressure effects and upper abdominal pain (CE), or jaundice, liver fibrosis and necrosis (AE). Treatment is often complex and expensive requiring surgical intervention (e.g. cyst drainage, cystectomy or liver resection), long-term anthelminthic (albendazole) chemotherapy, or a medico-surgical combination (Junghanss et al., 2008; Kern et al., 2017). Thus the burden of disease in poor endemic or co-endemic communities can be very high in terms of lost disability adjusted life years (Budke et al., 2006). Human treatment has no impact on transmission of echinococcosis because it occurs primarily between dogs and livestock (CE) or between foxes and small mammals (AE), thus predominantly domestic or wildlife cycles respectively. In China however, the role of dogs for zoonotic risk of human AE appears to be significant (Budke et al., 2005a; Craig et al., 1992; Moss et al., 2013; Wang et al., 2016).

In China human CE and AE disease occurs primarily in rural areas of northwest, west and southwest provinces and autonomous regions, though CE in particular has a wider distribution including some central, north and north-eastern provinces (Craig, 2004; Craig et al., 1991; Jiang, 2002; McManus, 2010). In total 20 of China's 33 provinces, autonomous regions and municipalities are endemic for E. granulosus (s.l.); however most CE cases occur in nine provincial regions, i.e., Xinjiang Uygur Autonomous Region (XUAR), Qinghai, Tibet Autonomous Region (TAR), Sichuan, Yunnan, Gansu, Ningxia Hui Autonomous Region (NHAR), Shaanxi and Inner Mongolia (Chai, 1995; Craig, 2004; Deplazes et al., 2017; Wu et al., 2018a). Five of these regions namely, TAR, Qinghai, northwest Sichuan, southwest Gansu and northwest Yunnan are completely or partially on the Tibetan Plateau. Transmission cycles of E. granulosus s.l. in China  utilize domestic dogs (or wild canids) as definitive hosts for the adult tapeworm (located in the gut) and domestic livestock (sheep, goats, cattle, yak, camels, pigs) or wild ungulates, as intermediate hosts for the larval cystic stage (located in organs primarily liver and lungs). Infection of dogs is predominantly associated with home-slaughter of livestock and deliberate feeding of cyst infested offal/viscera to owned dogs, or through dogs scavenging slaughter or butchering garbage sites. Livestock become exposed to infection from dog faeces contaminated pastures, and humans from direct contact with infected dogs or their faecal contaminated environment.

Molecular genotyping of human CE cyst isolates in China has confirmed E. granulosus sensu stricto (s.s.) (or G1 genotype) as the main zoonotic species (Li et al., 2008; Ma et al., 2008), as occurs typically for other CE endemic world regions (see reviews by Alvarez-Rojas et al., 2014; Deplazes et al., 2017). In addition E. canadensis (formerly the G6–G10 genotype complex of E. granulosus s.l.) has been identified from surgical biopsies for four CE cases (G7, G10 genotypes) in northeast China (Heilongjiang Province) and one Tibetan CE case (G6 genotype) in northwest Sichuan (Wang et al., 2015a, Wang et al., 2015b; Zheng et al., 2014). Molecular genotyping of livestock CE isolates indicates E. granulosus s.s. (G1 genotype) infections primarily in sheep but also infection in cattle, yaks and camels (Yang et al., 2005, Yang et al., 2015; Zhang et al., 1999). E. canadensis (G6) has been confirmed from goats, cattle and camels in China (Liu et al., 2013; Zhang et al., 1999). Dog infections with E. granulosus are predominantly caused by E. granulosus s.s. (G1), but adult worms of E. canadensis (G6 genotype) have been confirmed from a dog in Xinjiang (Zhang et al., 2006).

To date CE causing genotypes transmitted on the Tibetan Plateau include E. granulosus s.s. identified (in humans, sheep, yaks, pigs and dogs) in TAR, northwest Sichuan, Qinghai and southwest Gansu, and also E. canadensis (G6 genotype—in goats, yak, dogs) in the last three mentioned Tibetan provincial zones (Hu et al., 2015; Liu et al., 2013; Ma et al., 2008; Wu et al., 2018b; Xiao et al., 2003). There has also been a report of E. canadensis G10 genotype from a single yak in Gannan Tibetan Autonomous Prefecture, Gansu Province (Wu et al., 2018a). DNA typing studies on post-surgical biopsy samples from Tibetan CE patients in Ganzi and Aba Tibetan Autonomous Prefectures, Sichuan Province (n > 100 patients) (Hu et al., 2015; Li et al., 2008; Nakao et al., 2010) and Qinghai Province (n = 23) (Huang et al., 2018; Ma et al., 2012) confirmed all cases to be caused by E. granulosus s.s. (G1). However, one of 108 Tibetan CE cases from northwest Sichuan was confirmed to be caused by E. canadensis G6 genotype (Shang et al., 2018). Haplotype analysis of E. granulosus s.s. (G1) isolates from humans, livestock and dogs on the eastern Tibetan Plateau suggests a relatively conserved genetic profile (96% was of the G01 isolate) and not significantly different from isolates analysed from non-Tibetan endemic provinces (Ma et al., 2008, Ma et al., 2012; Nakao et al., 2010).

Transmission of E. multilocularis and risk of human AE disease on the Tibetan Plateau appears to be more predominant in eastern areas especially in eastern counties of TAR and the Tibetan autonomous counties or prefectures in northwest Sichuan and southeast Qinghai. Transmission involves a range of different wildlife hosts especially small mammals (Table 4 and Section 4.3). The Tibetan fox (Vulpes ferrilata) and the domestic dog are the most important definitive hosts and harbour the largest biomass of adult tapeworms, with the dog posing the greatest zoonotic risk (Budke et al., 2005a, Budke et al., 2005b; Li et al., 2005; Vaniscotte et al., 2011) (Sections 3.2 and 4.1). A large and diverse range of wildlife intermediate hosts include several microtine vole species, cricetid hamsters and lagomorphs (pika, hares) which may be differentially predated on by Tibetan fox, red fox and wolf but also by free-roaming dogs (Jiang et al., 2012; Wang et al., 2018) (Table 4). The distribution of small mammal populations on the eastern Tibetan Plateau is strongly influenced by natural land-cover characteristics and landscape ecology (Giraudoux et al., 2013a, Giraudoux et al., 2013b; Raoul et al., 2006; Wang et al., 2018), but also indirectly by anthropogenic impacts especially grazing practices (Marston et al., 2014; Raoul et al., 2006; Wang et al., 2004, Wang et al., 2010). This in turn may increase or decrease risk of acquiring human AE disease in Tibetan communities (Giraudoux et al., 2013a, Giraudoux et al., 2013b; Marston et al., 2016; Wang et al., 2010) (see Section 5.4).

Globally E. multilocularis appears to exhibit little genetic heterogeneity; for example the genotype is relatively conserved in comparison to E. granulosus s.l. However three forms or lineages of E. multilocularis have been considered based on molecular analysis of human and animal isolates. These have been described as Asian, European and North American forms (Nakao et al., 2006), though this requires further investigation particularly in relation to Asian isolates (Romig et al., 2017; Wu et al., 2017). At the level of parasite populations, haplotypic studies using mitochondrial or nuclear gene sequences have examined local diversity in European (Knapp et al., 2008), North American and Asian E. multilocularis isolates (Ito and Budke, 2015; Nakao et al., 2006). On the Tibetan Plateau at least five haplotypes of E. multilocularis were detected in human and animal isolates with little polymorphism, and Nakao et al. (2010) and Wu et al. (2017) suggested a founder effect. In south Qinghai E. multilocularis isolates from a dog and voles showed identical haplotypes (Ma et al., 2012) and similarly for Microtus spp. and human isolates (Wu et al., 2017) indicative of active transmission. Furthermore, E. multilocularis haplotypes from hosts on the Tibetan Plateau were the same as haplotypes in Xinjiang in northwest China which is far from the Plateau (Wu et al., 2017).

A new Echinococcus species named E. shiquicus was identified in 2005, initially by molecular analysis in Tibetan foxes and plateau pika in Shiqu county (northwest Sichuan) on the eastern Tibetan Plateau (Xiao et al., 2005). To date there is no evidence that it is a zoonosis (Huang et al., 2018; Li et al., 2008). The small adult stage morphologically resembled E. multilocularis, but the cystic larval stage was more like E. granulosus (Xiao et al., 2006a). It is likely therefore that earlier descriptions of plateau pika infected with atypical E. granulosus (Guo et al., 1993) and Tibetan foxes with atypical E. multilocularis (Qiu et al., 1995) were most likely cysts and adults respectively of E. shiquicus (Xiao et al., 2006a). E. shiquicus DNA has now been re-confirmed in V. ferrilata (Jiang et al., 2012) and also detected in dog faeces (Boufana et al., 2013), and further confirmed in O. curzoniae (pika) and the common plateau voles Microtus limnophilus and L. fuscus (Wang et al., 2018) (Table 4) (see Section 4.3). E. shiquicus has most probably evolved within predator-prey cycles of endemic mammal species from the Tibetan Plateau and is thus unlikely to occur in other regions of China or elsewhere (Xiao et al., 2006a).