| Background Nipah virus(NiV)is an enveloped,single-stranded,negative-sense RNA virus.It was first discovered in Malaysia in 1998-1999 and has since caused outbreaks in several countries in South Asia and South-East Asia.Bats serve as the natural hosts of NiV,and humans can be infected through various transmission routes,including direct contact with infected animals or their secretions,consumption of food or drinks contaminated by infected animals,or contact with infected individuals or their bodily fluids.The case fatality rates of NiV infection were estimated ranges from 40% to 75%by WHO,and its main clinical symptoms manifest as severe respiratory and central nervous system symptoms.NiV infections pose a significant threat to public health in South Asia and South-East Asia,attracting global attention.Currently,there are no vaccines or specific antiviral drugs available for the prevention and treatment of NiV infections.The World Health Organization(WHO)has identified NiV as one of the priority infectious disease pathogens requiring research and intervention in the Research and Development Blueprint.Since its discovery,extensive studies have been conducted on the epidemiology,pathogenesis,clinical symptoms,diagnosis,and treatment of NiV.However,most of these studies have focused on individual countries and limited time periods,lacking comprehensive research on the transmission and infection risks of NiV from multiple perspectives.This study aims to comprehensively and longitudinally investigate NiV based on heterogeneous data from multiple sources,considering the influence of factors such as the environment,animal,and human activity.By utilizing multidisciplinary techniques such as epidemiology,geographic information science,phylogenetics,evidence-based medicine,ecology,and machine learning techniques,this research aims to gain a comprehensive understanding of the spatial distribution,phylogenetic evolution,clinical outcomes,and transmission risks of NiV at individual,population,molecular,and ecological levels.The findings will provide scientific evidence for the control and prevention of NiV transmission.Objective(1)To describe the epidemiological characteristics of NiV infections,create spatial distribution maps of human NiV cases and animal NiV detections at regional and national scales.(2)To explore the phylogenetic evolution and phylogeographic migration routes of NiV,trace the spread process of NiV at the molecular level.(3)To conduct subgroup analysis of the clinical features and outcomes of NiV infection,based on evidence-based medicine combined with the results of phylogenetic evolution,revealing factors associated with increased risk of NiV infection.(4)To map the ecological niche distribution of potential reservoir hosts for NiV and study the relationship between ecological niche distribution probability and environmental,animal,and human-activity variables.(5)To quantify the transmission risk of NiV and create potential ecological niche maps for zoonotic transmission of NiV,identifying high-risk areas for pathogen spillover and cross-species transmission.Methods(1)We integrated heterogeneous data sources to establish a foundational NiV database.All available data on NiV infections in humans and animals were collected through literature searching and gray literature,extracting geographic information,clinical information,and epidemiological information.NiV sequence data was collected from Gen Bank,extracting background information such as source country,host species,collection time,and specimen type.Distribution data of NiV reservoir hosts was collected from species distribution databases.(2)We conducted the epidemiological analysis by describing the epidemiology of NiV through three-dimension distribution(the distribution by people,time,and place)and identifying the demographic characteristics of NiV infection cases in different countries.Geographic epidemiology focuses on describing the spatiotemporal changes in human NiV infection cases at regional and national scales.(3)The phylogeny of NiV was analyzed by using maximum likelihood methods to analyze the evolutionary relationships of NiV nucleotide sequences.Phylogeographic analysis was conducted using the Nextstrain framework,and fixed-effects likelihood models,single likelihood ancestor counting models,fast unconstrained Bayesian approximation models,and mixed-effects evolutionary models were used to analyze positive selection pressures on NiV genes,describing the molecular evolution of NiV,the migration routes of sequences,and adaptive changes in NiV gene evolution.(4)Single-arm proportion meta-analysis was applied to study the case fatality rate(CFR)and clinical symptoms of NiV infections based on evidence-based medicine concept.Subgroup analysis and meta-regression were used to explore the sources of heterogeneity in NiV infection outcomes.Binary meta-analysis using the MantelHaenszel method was performed to analyze the risk factors for NiV infections.Through the meta-analysis,clinical characteristics,clinical outcomes,and risk factors of NiV infections were analyzed at the population level.(5)We built the ecological niche models of NiV reservoir hosts by using the Boosted Regression Tree(BRT)model.We employed a targeted background sampling method to extract false-negative points from the Chiroptera distribution datasets as controls and NiV animal host species distributions as positive points.The covariates included environmental,animal,and human activity variables in a 10 km × 10 km spatial grid.The model was fitted using cross-validation and an optimal number of trees was selected.Spatial predictions were made for the suitable distribution of NiV reservoir hosts’ ecological niches.(6)We used the BRT model to analyze the risk prediction of NiV transmission.All geographical information related to the occurrence of NiV infections in humans and animals were extracted.Incorporating covariate data of animals,environment,and human activity,we developed three separate models: the human model based on the locations of human NiV infection cases,the reservoir host model based on the locations where NiV was detected in host animals,and the zoonotic model based on these two types of location information.Using the cutoff value,we calculated the area of high-risk regions and the number of affected populations predicted by the zoonotic model.We comprehensively evaluated the impact range of NiV transmission from both spatial and demographic perspectives.Results(1)The NiV basic database was based on 98 seaching literature and 22 gray literature,spanning from 1999 to 2021.It included a total of 749 NiV human cases and 658 NiV positive-detection animals.NiV has been primarily detected in nine countries,namely two countries in South Asia(India and Bangladesh)and seven countries in South-East Asia(Malaysia,Singapore,Philippines,Thailand,Cambodia,Indonesia,and East Timor).Human cases of NiV infections have been mainly reported in Malaysia,Singapore,Bangladesh,India,and Philippines.Among these,Bangladesh has reported a total of 322 cases,distributed across 7 divisions and 33 districts,primarily located in the northwest and central regions of the country.Faridpur district in Dhaka division had the highest number of reported cases among all districts(71 cases),followed by Rajbari district in Dhaka division(30 cases),and Naogaon in Rajshahi division(25 cases).Records of NiV positive-detection animals were mainly distributed among 7 countries,including Bangladesh,India,Malaysia,Thailand,Cambodia,Indonesia,and East Timor.In terms of the epidemiological characteristics of NiV infection,the highest proportion of cases occurred in adults aged 15-59(89.05%),and the number of male cases(73.77%)exceeded the number of female cases.In terms of temporal distribution,the outbreak in Malaysia and Singapore in 1998-1999 accounted for the highest proportion of cases among all years(39.25%).The peak months varied by country,with Malaysia experiencing a peak in March,Bangladesh in January,and India in February.(2)Phylogenetic analysis revealed that NiV could be divided into two clades: the NiV Malaysia clade(NiV-MY)and the NiV Bangladesh clade(NiV-BD).The NiV-MY clade mainly included sequences from Malaysia,Thailand,and Cambodia,while the NiVBD clade mainly included sequences from Bangladesh,India,and Thailand.The phylogeographic inference estimated the time of the most recent common ancestor(MRCA)of NiV to be 1916(95% CI: 1852-1931).The emergence of the NiV-MY clade was estimated to have occurred in 1970(95% CI: 1951-1979),followed by introductions into Cambodia and Thailand around 2000 and 2005,respectively.The NiV-BD clade emerged later than the NiV-MY clade,with an estimated time of emergence in 1995(95%CI: 1990-2001),followed by introductions into Thailand and India around 2005 and 2012,respectively.Site analysis suggested that 7 amino acid sites in the NiV genome may be associated with positive selection pressure.(3)The overall CFR of NiV infections,after adjusted by meta-analysis,was 70.45%.By subgroup meta-analysis,the national CFR of NiV infections,ranked from high to low,were as follows: 77.75% in India(95% CI: 52.73%-95.82%),76.72% in Bangladesh(95%CI: 68.41%-84.22%),52.94% in Philippines(95% CI: 28.80%-76.44%),39.48% in Malaysia(95% CI: 24.92%-55.01%),and 9.09% in Singapore(95% CI: 0.00%-35.01%).The meta-regression analysis showed that the phylogenetic clade,source country,proportion of male cases,and travel time to healthcare facilities had an impact on the heterogeneity of the fatality rate of NiV infections.The main clinical symptoms of NiV infections,in descending order,were fever(100.00%,95% CI: 99.81%-100.00%),altered consciousness(77.05%,95% CI: 57.71%-92.30%),and headache(67.63%,95% CI:59.97%-74.91%).Subgroup analysis revealed differences in four clinical symptoms between the two phylogenetic clades,namely fever,chills,cough,and dyspnea.The binary meta-analysis of NiV showed that male(OR=1.5470,95% CI: 1.0888-2.1980),contacting with pigs(OR=22.0996,95% CI: 1.7827-273.9594),contacting with or sighting of bats(OR=1.6172,95% CI: 1.1249-2.3250),consumption of raw date palm sap(OR=3.8141,95% CI: 1.7813-8.1667),harvesting of raw date palm sap(OR=5.3820,95%CI: 1.6743-17.3009),and climbing trees(OR=1.5009,95% CI:1.0972-2.0530).(4)The seven potential reservoir host niche models based on Boosted Regression Trees(BRT)showed good predictive performances,with AUC ranging from 0.98 to 1.00.The predicted results indicated that Pteropus lylei were mainly distributed in central Thailand,southern Cambodia,and southern Vietnam,exhibiting an elongated distribution pattern.Pteropus giganteus were primarily distributed in the Indian subcontinent,including India,Bangladesh,and Sri Lanka.According to the ecological niche model,Pteropus vampyrus were mainly distributed in the Malay Archipelago,including southern Thailand,coastal areas of Cambodia and southern Vietnam,the Philippines,Malaysia,Indonesia,and Papua New Guinea;Pteropus hypomelanus were mainly distributed on islands in the South-East Asian tropical region;Rousettus leschenaultia were primarily distributed in coastal areas of southern India,the Indochinese Peninsula,the Malay Archipelago,and certain regions in southern China;Rousettus amplexicaudatus were mainly distributed in the Malay Archipelago,the Indochinese Peninsula,and Papua New Guinea,while Hipposideros larvatus were mainly distributed in the Indochinese Peninsula,the Malay Archipelago,and southern regions of China.Relative contribution analysis indicated that Chiroptera richness had a significant impact on all the reservoir host niche,ranging from 9.09% to 46.34%.Sensitivity analysis showed that the species distribution models based on Maxent were highly consistent with the ecological niche models established by BRT in predicting the distribution range and probability of risk,confirming the stability and reliability of the predictive models.(5)The three NiV transmission risk prediction models based on Boosted Regression Trees(BRT)showed good performances,with AUC ranging from 0.87 to 0.99.The human model revealed that the risk regions of NiV spillover were primarily distributed in Bangladesh(almost nationwide),the southern Kerala in India,and the bordering areas with Bangladesh,as well as the southern region of the Malay Peninsula(referred to as West Malaysia).The reservoir host model showed that the regions favorable for carrying NiV by reservoir hosts were in the bordering areas between southern Thailand and Cambodia and the coastal areas of southern India.Additionally,the entire Indian subcontinent had a moderate probability of risk distribution.The zoonotic model indicated that the regions favorable for NiV zoonotic transmission were found in Bangladesh,the coastal areas of southern India,the bordering areas with Bangladesh,the bordering areas between southern Thailand and Cambodia,and the southern region of the Malay Peninsula.The predicted area of zoonotic transmission risk was 219,326.96 km2,with the countries in descending order of risk size being Bangladesh(121271.66 km2),India(48558.85 km2),Thailand(24848.56 km2),Malaysia(14120.09 km2),and Cambodia(3368.56 km2).There was a total population of 204 million living in these high-risk areas,with the countries in descending order of population being Bangladesh(124.68 million),India(41.25 million),Thailand(19.58 million),Malaysia(11.47 million),and Indonesia(3.19 million).Bangladesh was the country most severely threatened by NiV transmission risk in terms of both area and population.Relative contribution analysis indicated that reservoir host factors and temperature-related bioclimatic variables played important roles in the NiV transmission risk prediction models.Sensitivity analysis based on Maxent species distribution models showed similar risk range and probability predictions to the transmission risk predictions based on BRT,indicating robust and reliable results.Conclusion(1)Based on the foundational NiV database,the spatial distribution,and epidemiological characteristics of NiV in human and animals revealed NiV has a wide range of hosts,including bats,pigs,and dogs,in addition to humans.Human cases of NiV infections in Bangladesh were mainly concentrated in central Dhaka division and northwestern Rangpur and Rajshahi divisions,indicating a potentially higher risk of NiV exposure in these areas.Outbreaks of human NiV infections in South-East Asia and South Asia exhibited strong seasonal patterns,possibly associated with bat breeding and the harvesting of date palm sap.Attention should be focused on hotspot areas affected by NiV and surveillance of NiV host animals and human infection cases should be strengthened during peak bat activity seasons.(2)Sequence analysis found that after a century of evolution,NiV gradually diverged into two clades,and NiV showed local migration characteristics at the molecular level in South Asia and South-East Asia.The evolution type of NiV sequences was found to be associated with their national attributes.This indicates regional evolution within NiV sequences and suggests that NiV outbreaks in different regions may be indicative of geographical restrictions on the virus.Seven sites validated by two or more models suggest the presence of positive selection pressure,providing reference clues for understanding viral adaptive changes,predicting viral gene function,and future development of anti-NiV drugs and vaccines.(3)There were significant differences in clinical outcomes caused by NiV infection at the phylogenetic clade level.Additionally,multivariable meta-regression revealed that the heterogeneity of NiV fatality rate was also associated with source country and travel time taken to health facilities.Lower economic levels and fragile healthcare facilities are clearly unfavorable for the outcome of this highly infectious disease and may lead to more severe consequences.(4)Several potential NiV reservoir hosts,including Pteropus lylei,Pteropus giganteus,Pteropus vampyrus,Pteropus hypomelanus,Rousettus leschenaultia,Rousettus amplexicaudatus and Hipposideros larvatus,were widely distributed in SouthEast Asia,South Asia,and China,indicating closer contact and interaction opportunities between them.This natural advantage in viral connectivity or horizontal transmission within the same genus promotes viral sharing and spread.Virologists,epidemiologists,ecologists,and wildlife biologists should collaborate to conduct animal host monitoring.(5)Machine learning models suggested that regions such as India,Bangladesh,Thailand,Cambodia,and Malaysia had NiV transmission risks that extend far beyond the currently reported geographical range,with the risk areas bordering China.Meanwhile,China also harbored NiV host species,posing pressure on cross-border infectious disease prevention and control.Global attention should be paid to NiV outbreaks,and through close cooperation,a global human health community should be collectively established. |
PDF Full Text Request