Global distribution and host range of enzootic bovine leukosis in ruminants: A systematic review and meta-analysis

Highlights

• •The pooled prevalence of Enzootic Bovine Leukosis (EBL) was estimated at 19% (95% CI: 16%–23%) worldwide, with substantial heterogeneity (I² = 99.6%, p < 0.001).
• •The highest EBL prevalence was reported in North America (43%), followed by Asia (17%).
• •Beef cattle had a higher EBL prevalence (26%) compared to dairy cattle (18%).

Abstract

Enzootic bovine leukosis (EBL) causes severe economic losses and is a contagious disease caused by the Bovine Leukemia Virus (BLV), belonging to the delta retrovirus of the Retroviridae family. Thus, this review aimed to estimate the global pooled prevalence of enzootic bovine leukosis and investigate its host range, focusing on cattle and sheep. We used seven databases which include PubMed, Science Direct, HINARI, Scopus, Google Scholar, Web of Science, and AJOL. The included studies (50 articles), conducted between 1992 and 2024, represent diverse geographic regions: Asia, North America, Africa, South America, and Europe. 346,917 animals were involved, of which 99,620 involved positive for BLV. The meta-analysis estimated a pooled BLV prevalence of 19% (95% CI: 16%–23%) using random effect model, though significant heterogeneity existed across studies (I² = 99.6%, p < 0.001). Subgroup analysis revealed that the highest prevalence of EBL was in North America (43%), followed by Asia (17%), Africa (14%), South America (10%), and the lowest in Europe (4%). Analysis by publication year revealed that the prevalence of EBL was 32% after 2016, whereas it was 29% before 2016. Among animal species, beef cattle exhibited the highest prevalence (26%), followed by overall cattle (20%), sheep (19%), and dairy cattle (18%). Geographical and temporal trends revealed varying prevalence rates, with notable high rates in North America and lower rates in regions such as Europe. It is crucial that North America adopt stricter prevention programs, similar to those in Europe, to reduce transmission of EBL and its economic and health impacts on livestock.

Introduction

Enzootic bovine leukosis (EBL) is a chronic infectious disease caused by the bovine leukemia virus (BLV), which primarily affects cattle and other ruminants (Johnson and Kaneene, 1992). Bovine leukemia virus is an oncogenic retrovirus that is a member of the Delta-retrovirus, and the family Retroviridae (Kettmann et al., 1976). It has a significant economic impact due to reduced milk production, early culling, and trade restrictions (Hassan et al., 2020). Naturally, BLV infection has been confirmed in species such as Bos Taurus (domestic cattle), Bos indicus (zebu), Babalus bubalis (water buffalo), and Bos grunniens (domestic yaks). Bovine leukemia virus can be also transmitted experimentally to other species, sheep, goat and water buffalo (Camargos et al., 2014)

BLV was first documented in 1871 in Lithuania, a nation situated on the southeastern shores of the Baltic Sea (Burny et al., 1980). It has been posited that the virus emerged from the coastal areas of the Eastern Baltic nations. Subsequently, it was disseminated to the American continent in the initial decades of the 20^th century via the commerce of live livestock. Thereafter, it made a comeback to Europe. It was initially introduced to additional countries by cattle imported from North America, and it is presently extensively distributed across the globe (Erskine et al., 1996). The European Union recognizes Austria, Czech Republic, Denmark, Estonia, Spain, Finland, Netherlands, Slovakia, and Sweden as EBL-free (Polat et al., 2017). While officially recognized in 2017, Italy continues to manage persistent infection clusters (Righi et al., 2021); however, recent surveillance indicates the country has achieved 100% sensitivity and freedom from EBL, contrasting with Slovenia's 50.5% sensitivity and 81.6% freedom (Fanelli et al., 2024). EBL surveillance programs have also demonstrated freedom from the disease in Norway and Romania (Irimia et al., 2021).

In Argentina, it has been reported that 84% of dairy herds possess these antibodies (Polat et al., 2016), while in Turkey, the prevalence is noted at 2.28% (Şevik et al., 2015), and in Mexico, the range is between 11% and 66% (Heinecke et al., 2017). In Brazil, the seroprevalence of EBL has been documented to fluctuate between 5.64% and 81% in preceding years (Meirelles-Bartoli and Sousa, 2013). In China, a meta-analysis revealed a pooled prevalence of 10% (Ma et al., 2021). Moreover, BLV has also been identified in certain Middle Eastern nations (Rodriguez et al., 2009). In Africa, serological studies have detected BLV in Egyptian dairy cattle with a prevalence of 15.83% (Zaher and Ahmed, 2014). Prevalence rates of 20.8%, 9%, and 0% have been also reported in cattle, buffaloes, and camels, respectively (Selim et al., 2021) . In Tanzania, the disease was first identified in 1997, with a prevalence of 36% among both beef and dairy cattle (Schoepf et al., 1997).

Although traditionally limited to ruminants, recent investigations have raised concerns about the zoonotic potential of BLV, suggesting possible implications for human health (Villalobos-CortÃ, 2017, Buehring et al., 2019). Emerging molecular studies have detected BLV DNA in human tissues, including those from patients with breast cancer (Buehring et al., 2019). This has led to growing interest in the zoonotic implications of EBL, particularly for individuals in close contact with infected animals or those consuming unpasteurized dairy products (Rodriguez et al., 2009). The enzyme-linked immunosorbent assay (ELISA) and the agar gel immunodiffusion test (AGID) is commonly used for serological screening of BLV infection. Due to its high sensitivity and specificity, ELISA is the recommended test for trade purposes (Trono et al., 2001, OIE, 2014).

To eradicate BLV with long incubation periods, all infected animals must be targeted, not just those showing visible symptoms. In dairy herds, rigorous management can reduce within-herd prevalence but cannot fully eradicate this disease. The most effective method for achieving BLV freedom is to eliminate infected animals, primarily through regular screening followed by culling of positive animals (Health and Welfare, 2015). As intra-herd prevalence increases, this approach becomes more challenging. In regions with high BLV prevalence or weaker economies, the high cost of testing and culling makes it unfeasible. Success depends on economic support from regional governments, and countries lacking such support, like the United States, Canada, Argentina, and Japan, have not adopted this strategy. Even in countries that have successfully implemented these strategies and are free of BLV, the virus remains prevalent in many parts of the world. It continues to circulate in various animal production systems, including both beef and dairy cattle. Understanding the prevalence of enzootic bovine leukosis (EBL) is essential for safeguarding cattle health, enhancing productivity, and informing disease control strategies. EBL, caused by the bovine leukemia virus (BLV), poses significant economic challenges, particularly in dairy farming, due to its impact on production performance and trade restrictions.

Additionally, a thorough awareness of the disease's prevalence in cattle and other ruminant animals across the world should be established, as this awareness will help in the development of potential future intervention programs. To identify trends and patterns in EBL prevalence related to different animal production systems (e.g., dairy vs. beef cattle). This systematic review and meta-analysis aimed to estimate the global prevalence of enzootic bovine leukosis in ruminant animals.

Methods

This review was carried out following the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) checklist (Mother (Moher et al., 2010) (S1 Table). The PRISMA checklist was used to ensure all eligible studies were incorporated into the meta-analysis.

Literature search strategy

The literature search was carried out between January 2, 2024, and March 28, 2024. A thorough search technique was developed to explore all relevant studies. The literature search was carried out independently by two researchers (MD and YD). The online databases PubMed, Google, Google Scholar, Scopus, Web of Sciences, and HINARI were explored with English language restrictions. This systematic review used the CoCoPop framework (conditions, context, and population) to search for relevant articles (Munn et al., 2018). The condition was (enzootic bovine leukosis) (Co), the context was globe (Co), and the population was ruminants (Pop). The PubMed search strategy included terms from the Medical Subject Heading (MeSH) and a set of key keywords. The PubMed search engines used were “Prevalence” OR “Epidemiology” OR “Sero-prevalence” OR “Observational studies” AND “EBL OR “Bovine leukemia virus” OR “BLV” AND “Cattle” OR “Bovine” OR “Ruminant animals” AND “Globe”, the detailed search engines and keywords were found supplementary file (S1 file). EndNote was used to remove duplicate articles.

The specific questions of the study were as follows: "What is the pooled global prevalence of enzootic bovine leukosis in domestic ruminant populations? “This question seeks to understand the global prevalence of enzootic bovine leukosis (EBL) in domestic ruminants. Specifically, it aims to measure the prevalence of EBL in the population of domestic ruminants. Since this is a descriptive study, there is no intervention or comparison involved. The focus is purely on assessing the extent of EBL across global domestic ruminant populations.

What is the prevalence of enzootic bovine leukosis in ruminant animals across different continents? This question examines how the prevalence of EBL in cattle varies across different geographic regions, including North America, Europe, Africa, Asia, South America, and Australia. Here, the population is ruminant, and the intervention is the geographic region in which they are located, which serves as an implicit factor. The comparison is not explicitly stated, but the goal is to explore regional variations in EBL prevalence across these different regions.

"Is there a significant difference in the prevalence of enzootic bovine leukosis between dairy and beef cattle populations?" This question also investigates whether there is a significant difference in the prevalence of EBL between dairy and beef cattle populations. The population remains ruminant, but the study looks at two distinct groups: those in dairy farming and those in beef farming. The intervention being considered is dairy farming, with beef farming serving as the comparison group. This question aims to determine if the type of farming influences the prevalence of EBL.

Eligibility criteria

The inclusion was used to confirm the eligibility of the searched paper: Inclusion and exclusion criteria were defined regarding the relevance of the articles to the research questions of interest. The selection of articles was performed by two independent authors (BA &AS).

Inclusion criteria

This systematic review and meta-analysis included only observational studies, specifically cross-sectional, cohort, and case-control studies that reported on the prevalence of EBL in ruminant animals. Studies were considered if they were published after 1990 and provided data on the prevalence or incidence of EBL. Only studies published in English were included, and the animal populations of interest were ruminants, such as cattle, sheep, pig buffalo, and goats, affected by EBL. There were no geographical restrictions, allowing studies from any region to be eligible for inclusion.

Exclusion Criteria

Non-observational studies, including randomized controlled trials, laboratory experiments, reviews, case reports, editorials, commentaries, conference abstracts, and letters to the editor were excluded. Additionally, studies published before 1990 and languages other than English were not considered. Studies focusing on non-ruminant animals or species unrelated to enzootic bovine leukosis are not involved. Furthermore, articles that fail to estimate the prevalence of enzootic bovine leukosis/bovine leukemia virus at the animal level, or where these outcomes are not explicitly defined or measurable were disregarded. Duplicate publications of the same research were also excluded, with only the most comprehensive or recent version being considered.

Data extraction

After the selection of studies that met the previous eligibility criteria, two investigators (MD and GA) independently extracted the relevant data. The extraction process involved both quantitative and qualitative data and was organized into two word tables and an Excel spreadsheet. The final database consisted of 50 articles. The extracted components included information such as the primary author's name, publication year, countries/setting, study animal species (host), type of sample taken, source of animals with their settings, total number of animals examined or tested (sample size), diagnostic methods, ethical considerations and number of positive (test positive). Any disagreements were resolved through discussion and consultation with a third author.

Study quality assessment

In this review, a comprehensive evaluation was conducted to authenticate the methodological excellence of this systematic review. Two authors (BA and YD) independently executed the quality assessment via the tool known as AMSTAR-2 (Shea et al., 2017). The Critical Appraisal Quality Assessment Tool encompasses 16 items (questions) that pertain to randomized or nonrandomized trials of health interventions, or both. Certain quality assessment questions are applicable to both randomized and nonrandomized trials, whereas others are exclusively intended for one or the other. For example, the initial question is solely utilized for randomized trials of health interventions.

Data synthesis and statistical analysis

The random effects model, which employs the restricted maximum likelihood method (REML) to compute within- and between-study variability, was employed to estimate the aggregated prevalence and 95% confidence intervals. REML is often preferred in random-effects meta-analysis of proportions because it provides more accurate, efficient, and unbiased estimates, especially in the presence of substantial heterogeneity and small sample sizes, which are common challenges in such analyses. This model was utilized to conduct a meta-analysis (overall effect size (ES) or often represented as a percentage or proportion of EBL), assess heterogeneity, and determine the weight of each study. Furthermore, graphs and tables were employed to illustrate the prevalence of the EBL on the basis of the geographical distribution, type of animal species, and the study years.

To estimate the pooled prevalence, we extracted data from the number of events and the total number of samples. These data were then subjected to proportional meta-analysis via the "metaprop" function of the 'meta' package version 4.1.3–0 in R statistical software. Pooled proportions of EBL were estimated via logit transformation via a logistic-normal random effects regression model (Viechtbauer, 2010). Subgroup analysis was conducted via the mixed effect logistic regression model. Forest plots were used to visualize the results of the meta-analysis. The plots were generated using the ggplot2 package (Wickham and Wickham, 2016). The plots, however, were not displayed and copied on the manuscript due to the large number of articles.

Investigation of heterogeneity

The Cochran's Q test (indicated by the p value), τ² (representing between-study variance), and the inverse variance index (I²) were employed to assess the sources of heterogeneity. The I² index, as elucidated by Higgins and Thompson, (2002), was calculated to indicate low, moderate, and high levels of heterogeneity, corresponding to I² values of 25%, 50%, and 75%, respectively. Heterogeneity was deemed statistically significant if the I² value exceeded 50%, and the Q test yielded a P value less than 0.10. The extent of study heterogeneity was evaluated through a forest plot diagram, which depicted the weights, effect magnitudes, and 95% confidence intervals for between-study variability.

Publication bias assessment

To assess publication bias, funnel plot diagrams and Egger's regression test were utilized. Egger's regression test is a statistical method used to evaluate the symmetry of a funnel plot (Egger et al., 1998).

Subgroup analysis

Subgroup analysis is a specific type of meta-regression focused on examining the effects of a single categorical subgroup variable across studies (Higgins and Thompson, 2002). In this approach, the entire set of research is divided into two or more subgroups based on the categories of the subgroup variable, often termed the moderator. A moderator represents a particular characteristic of the study that can help explain some of the variation observed among the studies. In this meta-analysis, the moderators include study continents (Asia, Europe, Africa, South, and North America), study year (pre- and post-2015), animal species/ type, ample size category (<200, 200-500 and >500) and methods of diagnosis. To assess variability within and between sources, Cochran's Q statistic and the inverse variance index (I²) were used. A statistically significant Q statistic for a predictor in subgroup analysis indicates that subgroup membership accounts for some or all of the variability observed in the effect sizes.

Meta regression

In meta-regression, evaluating moderators follows a similar approach to regression or multiple regressions used in individual studies (Card, 2019). This method analyses the relationship between covariates (or moderators) and effect sizes across multiple studies, using the studies themselves as the units of analysis (Littell et al., 2008). It accommodates both categorical and continuous moderating variables. We used univariate meta-regression analysis to identify potential sources of variation in pooled prevalence, considering study year, study location, production type and animal species were considered as a moderator. Each variable was analyzed separately in a meta-regression analysis.

Results

Initially, we explored 6, 056 articles by combining keywords in seven databases: PubMed (n = 4,462), ScienceDirect (n = 113), HINARI (n = 14), SCOPUS (n = 126), Google Scholar (n = 755), Web of Sciences (n=124), and AJOL (n = 10) and additional records (n=576). Subsequently, two authors (BAM and GA) independently reviewed the titles, keywords, and abstracts. Articles that focused on general aspects of the virus, were review articles or were written in languages other than English were excluded. Of these, 4,256 were excluded because their titles or abstracts indicated that they were not relevant. The remaining 1,800 studies were further scrutinized, with 556 found to be duplicates. A total of 1, 244 full-text papers were reviewed for eligibility based on pre-established criteria. The remaining 1, 194 articles were excluded because of issues such as the study area and other factors, as detailed in Fig 1. Finally, a total of 50 articles were included in the current review and meta-analysis.

Fig 1.

Flowchart of study selection for EBL in the globe.

Characteristics of the included studies (n=50)

The characteristics of the 50 studies included in this systematic review and meta-analysis (Table 1). All studies were observational and were conducted across various parts of the world, with sample sizes ranging from 44 (Olaya-Galán et al., 2022) in Colombia to 235811 (Sandev et al., 2015) in Canada. The studies were published between 1992 (Cockerell et al., 1992) in the USA and 2024 (Zhao et al., 2024) in China. In general, the table presents a detailed summary of studies examining the prevalence of BLV in livestock across various countries and settings, highlighting important characteristics such as the year of publication, sample collection period, geographic location, sample type, animal source, diagnostic methods, sample size, number of positive cases, calculated prevalence, and whether ethical considerations were reported.

Table 1.

The descriptive of included studies (n=50).

Author &Publication year	Study year	Country/ setting	Sample taken	Source	Diagnostic test	N	Positive	Prevalence	Ethical
(Gjinovci et al., 2020)	2016	UK/ Kosovo/315 Villages	Serum	Dairy cattle	ELISA	5,051	110	3	Yes
(Metwally et al., 2023)	2022	Egypt/Southern, Middle, and Northern	Serum	BD Cattle	ELISA	368	44	12	Yes
(Choi et al., 2002)	2001	USA/California	Semen	Bulls beef	AGID	79	29	36.7	Not stated
(Schoepf et al., 1997)	NA	Tanzania/7 provinces	Serum	BD cattle	ELISA	2849	1041	36	Not stated
(Rukkwamsuk and Rungruang, 2008)	NA	Thailand	Serum	Dairy cattle	ELISA	80	26	32.5	Yes
(Nikbakht et al., 2010)	NA	Iran	Blood sample	Dairy cattle	ELISA	1619	271	16.73	Yes
(Lee et al., 2016)	2013-2014	Thailand	Blood	Dairy cattle	ELISA	744	437	58.7	Yes
(Kathambi et al., 2019)	2016	Kenya	Serum	Pastoral	ELISA	1383	111	8	Yes
(Figueroa et al., 2020)	2015-2016	Colombia	Blood	Dairy Cattle	PCR	289	179	61.9	Yes
Hassan et al., 2020	2019	United Arab Emirates	Serum	Dairy cattle	ELISA	957	246	27.5	Yes
(Asadpour and Jafari, 2012)	2010	Iran/ Tabriz	Semen, blood,	Bull beef/18 herd	PCR	45	6	13.3	Not stated
(Panneum et al., 2009)	2007	Thailand	Serum	Dairy cattle	ELISA	661	250	37.8	Yes
(Morris and Kadish, 1996)	NA	S. Africa/Mpumalanga	Serum	Dairy cattle/10 herd	ELISA	381	14	4	Not stated
(Morovati et al., 2012)	2011	Iran/Isfahan province	Serum	Dairy cattle/21 livestock farm	ELISA	403	303	81.8	Not stated
(Ma et al., 2016)	2013-2014	China	Serum	Yaks	ELISA	1584	334	8.5	Not stated
(LaDronka et al., 2018)	2017	USA/11 states	Milk	Dairy Cattle/103 dairy farms	ELISA	4120	1648	40	Yes
(Batmaz et al., 1999)	-	Turkey	Milk, blood	Elisa	ELISA	717	69	9.6	Yes
(Murakami et al., 2011)	2007	Japan/7 Japanese prefectures	Serum	BBD cattle/209farm	ELISA	5420	1550	28.6	Not stated
(Selim et al., 2021)	2018	Egypt/4 provinces	Serum	Cattle	ELISA	350	73	20.8	Yes
(Selim et al., 2020)	2018	Egypt/4 provinces	Serum	Buffalo	ELISA	100	9	9	Yes
(Selim et al., 2020)	2018	Egypt/4 provinces	Serum	Camels	ELISA	100	0	0	Yes
(Cockerell et al., 1992)	1991	USA/Colorado	Serum	Dairy cattle	ELISA	95	49	52	Not stated
(Selim et al., 2021)	2019	Egypt	Serum	Dairy cattle	ELISA	500	91	18.2	Yes
(Ortega et al., 2020)	2014	Colombia	Serum	Dairy cattle	ELISA	8150	3480	42.7	Yes
(Hsieh et al., 2019)	2016-17	Taiwan/16 cities	Serum	Dairy cattle/110 farms	ELISA	660	540	81.8	Yes
(Khudhair et al., 2016)	2014-2015	Iraq/central/2 governorates	Serum	Dairy cattle	ELISA &PCR	400	28	7	Yes
(Mousavi et al., 2014)	2013	Iran/ 2 provinces	Serum	Dairy herds	ELISA	429	109	25.4	Yes
(Scott et al., 2006)	2002	Canada/Alberta	Serum	Dairy cows	ELISA	2814	756	26.9	Yes
(Olaya-Galán et al., 2022)	2019	Colombia	Blood	Buffalo	PCR	61	12	19.6	Yes
(Olaya-Galán et al., 2022)	2019	Colombia	Blood	Sheep	PCR	44	15	34	Not stated
(Mohammadi et al., 2011)	2010-2011	Iran	Serum	Dairy cattle/2 government and 2 private farm	ELISA	137	41	29.9	Not stated
(Porta et al., 2023)	2019-2020	Argentine	Serum	Beef cattle	ELISA	5827	421	7.23	Yes
(Benitez et al., 2019)	2019	USA	Serum	Beef cattle	ELISA	3146	930	29.5	Yes
(Nekouei et al., 2015)	2012-2014	Iran	Blood	Cattle	PCR	657	144	22.1	Yes
(Nekouei et al., 2015)	2012-2014	Iran	Blood	Sheep	PCR	95	5	5.3	Yes
(Nekouei et al., 2015)	2012-2014	Iran	Blood	Camel	PCR	122	0	0	Yes
(Huser et al., 2023)	2021	USA/kassa state	Blood	Beef cattle	ELISA	2845	1564		Yes
(Sandev et al., 2015)	2012	Canada/Bulgaria	Serum	NA	ELISA	235811	78711	33.38
(Chacón et al., 2023)	2011-2018	Costa Rica	Blood	Beef bull	ELISA	379	165	43.5	Not stated
(Ramalho et al., 2021)	2012-2013	Brazil	Serum	Dairy cattle	ELISA	2067	198	10.8
(Uysal et al., 1998)	1994	Turkey/ Trakya district	Serum	Dairy	ELISA	481	51	11	Not stated
(Chi et al., 2002)	1998	Canada/ Maritimes provinces	Serum	Dairy cattle/90 herd farm	ELISA	2604	700	20.4	Yes
(Walsh et al., 2013)	NA	Canada	Serum, milk	Dairy cattle	ELISA	1229	860	69.9	Yes
(Xu et al., 2016)	NA	China	Blood	Yaks	ELISA	370	22	6	Yes
(Sakhawat et al., 2021)	NA	Pakistan/7 regions	Serum	Water buffalo	ELISA	92	0	0	Yes
(Sakhawat et al., 2021)	NA	Pakistan/7 regions	Serum	Cattle	ELISA	1380	52	3.8	Yes
(Shaukat et al., 2024)	2022	Canada	Milk	Dairy cattle	ELISA	480	417	86.9	Yes
(Yang et al., 2016)	2013-2014	China	Vaginal Swab, milk	Dairy Cattle	ELISA	1963	964	49	Yes
(Yang et al., 2016)	2013-2014	China,21provinces	Blood	Beef cattle	PCR	1390	22	15	Yes
(VanLeeuwen et al., 2006)	NA	Canada	Serum	Dairy Cattle	ELSIA	1530	569	37.2	Not stated
(Fava et al., 2016)	2005-2007	Brazil	Serum	Sheep	AGID	2592	2	0.007	Yes
(Ndou et al., 2011)	2010	S. Africa/ North-west	Serum	Dairy Cattle/9herd	ELISA	340	41	12.6	Not stated
(Kaura and Hübschle, 1994)	NA	Namibia/ keetmanshoop	Serum	Dairy cattle	ELISA	3343	410	12.3	Yes
(Zaghawa et al., 2002)	1996	Upper Egypt,	Blood and serum	Dairy cattle	ELISA	440	230	52.3	Yes
(Şevik et al., 2015)	2013	Turkey/ six provinces	Serum	Dairy cattle, Brown Swiss cows, Holstein	ELISA	28,982	460	15.9	Not stated
(Zhao et al., 2024)	2022-2021	China / seven cities	blood	Dairy cattle/ nine farm	PCR	668	23	3.4	Yes
(Jacobs et al., 1995)	1994	Canada/ Nepean	Serum	Dairy cattle	ELISA	920	236	25.6	Yes
(Puga Torres et al.2012)		Ecuador, 3 provinces	Serum	6-24 months of animals	ELISA	3307	75	2.27	Yes

ELIS= Enzyme-Linked Immunosorbent Assay, PCR= Polymerase Chain Reaction, AGID= Agar Gel Immunodiffusion, BD= Beef and Dairy, N=total number of animals examine

AMSTAR-2 quality assessment results summary

The AMSTAR-2 evaluation revealed variations in the methodological quality of the included systematic reviews. In general, most studies were strong, clearly defining their research questions, and using comprehensive search strategies. Authors like Gjinovci et al. (2016) and Nekouei (2015) even achieved perfect scores across all criteria, showing excellent quality. However, improvements are needed in a few areas. Some studies, including those by Metwally et al.(2023), Choi et al. (2002), and Hassan et al. (2020), didn't disclose their funding sources, which reduces transparency. Additionally, studies like those by Schoepf et al. (1997), Kathambi et al. (2019), and Selim et al. (2021) could have more thoroughly addressed heterogeneity and publication bias, important considerations for statistical rigor. While most reviews did a good job of assessing the risk of bias in the individual studies they used, a few, like Asadpour and Jafari (2012), Panneum et al.(2009), and Hsieh et al. (2019), fell short, which can affect the reliability of their conclusions. The systematic reviews evaluated generally exhibited high methodological quality, adhering to most AMSTAR 2 reporting standards. They effectively addressed key components such as clear research questions, comprehensive literature searches, and appropriate data synthesis methods (S2 Table).

Geographical descriptions of the included studies

The included studies for this review span across Asia, Africa, Europe, South America, and North America. However, the distribution of studies varies significantly: most of the articles were from Asia (n=23), followed by North America (n=13), Africa (n=12), South America (n=8), and Europe (n=4) (Table 1). We noticed that some of the articles were repeated because they included data on different animal species. In the case of Asia, the majority of studies were conducted in China, Japan, Iran, and Thailand. For North America, studies were primarily from states such as Wisconsin, Idaho, New York, Pennsylvania, Texas, Minnesota, Michigan, Ohio, Vermont, Utah, and North Carolina (LaDronka et al., 2018), California, and Colorado. In Africa, the most studies were found in Tanzania, Egypt, Namibia, and South Africa. Similarly, in South America, most of the studies were concentrated in Colombia and Brazil. In European studies, specifically those conducted in the United Kingdom and Turkey's Trakya district.

Description of included animals and diagnostic methods

The review primarily included cattle (both beef and dairy) along with other ruminants such as yaks, buffalo, sheep, and camels. A total of 346,917 animals were included in the analysis, of which 99,620 tested positive for BLV. This data was used to estimate the global prevalence of BLV. In addition to the included studies, one case report from Virginia, USA, highlighted the association between wild animals and BLV, describing a 13-month-old intact male Huacaya alpaca the first reported case of lymphoma in an alpaca that tested positive for BLV using serology and polymerase chain reaction (PCR) (Lee et al., 2016). The diagnostic tools employed across the studies included AGID, ELISA, and PCR, with ELISA being the most commonly used method, primarily utilizing blood serum as the sample. Other samples used were serum, semen, milk, and whole blood. Serological studies based on the detection of antibodies against BLV by ELISA and PCR in animals show conflicting results. Prevalence rates reported in the studies vary considerably, from very low rates such as 0.007% in sheep from Brazil to extremely high rates of up to 86.9% in Canadian dairy cattle, indicating a significant burden of infection in certain regions, particularly among dairy cattle populations in Colombia, Taiwan, and parts of the USA. In contrast, some studies from regions like Namibia and Egypt report relatively lower prevalence rates, including 0% in camels, which may suggest differences in infection control measures or exposure risks. Ethical considerations were reported in many studies, highlighting adherence to research standards, although some studies did not explicitly state whether ethical approval was obtained, an important aspect of research integrity and animal welfare. Overall, the data reveal a complex epidemiological landscape with significant variability in infection rates across different species and geographic areas.

In a study assessing the presence of BLV using both AGID and PCR on the same blood samples, it was found that the AGID test detected a positivity rate of 36.7%, while the PCR test showed 0% positivity and we performed the subgroup analysis. This led the authors to suggest that the presence of antibodies detected by AGID might be due to past exposure to viral antigens rather than indicating an active infection (Choi et al., 2002). Another study in Iran reported that 22.8% of blood samples collected from Tehran province tested positive for BLV using ELISA (40 out of 175 samples), whereas Nested PCR detected the virus in 16.75% of the samples (29 out of 175) (Nikbakht et al., 2010). Similarly, in a specific province in China, PCR showed a positivity rate of 66.7% (188 out of 282 samples), closely matched by an ELISA positivity rate of 67.4% (190 out of 282 samples), demonstrating comparable effectiveness of both methods in detecting BLV in that region (Yang et al., 2016). The study conducted by Asadpour and Jafari, (2012) found a discrepancy between the findings obtained by ELISA and PCR in detecting antibodies against BLV, with ELISA identifying antibodies in 5 out of 45 samples, whereas PCR detected viral antigen in 6 out of 45 samples sourced from blood. Furthermore, the study confirmed that 6 out of 45 bull semen specimens tested positive for BLV through PCR, and this figure of 6/45 (11.1%) was employed for the estimation of the pooled prevalence. The conflict between the outcomes of ELISA and PCR may be ascribed to the disparities in sensitivity and specificity inherent to these two diagnostic methodologies, given that ELISA identifies antibodies while PCR detects viral DNA. Such discrepancies may elucidate the variation in findings, with PCR potentially providing a more precise assessment of infection prevalence. In this scenario, using the PCR result of 6/45 (13.3%) for estimating pooled prevalence appears to be an appropriate approach since PCR directly identifies viral genetic material.

The included studies used different biological samples for diagnostic testing. From these, serum was the most commonly used sample type, appearing in 44 studies, which accounts for approximately 66.7% of the total and blood samples were used in 11 studies (17%). Milk samples were utilized in 4 studies (6.1%), whereas semen was used in 2 studies (3.0%). Additionally, a few studies used a combination of sample types, including blood and serum (1 study, 1.5%), milk and blood (1 study, 1.5%), and milk and vaginal swabs (1 study, 1.5%) (Table 1). The distinction between serum and blood in the dataset is likely due to differences in sample processing and diagnostic application. Serum refers to the liquid portion of blood obtained after coagulation, which is commonly used for antibody detection through methods such as ELISA or AGID. In contrast, whole blood contains both cellular components and plasma, making it suitable for tests of antigen that require intact blood cells or molecular analyses such as PCR. The categorization of these sample types in the dataset reflects the specific methodological approaches used in each study to achieve accurate diagnostic outcomes.

The prevalence of EBL appears to vary annually (Fig 2), showing a cyclical pattern that includes several peaks and troughs. The prevalence indicates a general increasing pattern from 2006 to 2012, reaching a prevalence of approximately 0.8 on average. In 2013, the prevalence fell below 0.2, indicating a significant reduction. Following 2013, there was a slow rise through 2016 and then a big decline in 2017. The prevalence rises again in the next years, peaking about 2019, and then it declines steadily until reaching a low point in 2021. After 2021, the prevalence starts to increase once more, showing a gradual upward trend through 2024. The overall pattern suggests that the prevalence of BLV is highly variable over the years with multiple fluctuations. This variability could be due to a range of factors, including changes in detection practices, management interventions, control measures, or changes in the transmission dynamics of the virus. Notably, periods of high prevalence may indicate lapses in control measures or increased transmission rates, while declines could reflect successful interventions or natural reductions in the virus's spread.

Fig 2.

Yearly Trends in Average Prevalence of Bovine.

Leukemia Virus (BLV) from 2006 to 2024.

Meta-analysis

The current meta-analysis used 50 English-language studies. As a result, the meta-analysis found that the overall random pooled prevalence of BLV across studies was 19% with the confidence interval ranging from 16% to 23%. Statistically significant heterogeneity was found between studies regarding BLV prevalence (I²=99.6%, Q-test = 15687.69, p < 0.001) (Table 2). Despite the high heterogeneity, the studies contributed almost equally to the overall analysis, with individual study weights ranging narrowly from 4.1% to 4.4%. This suggests that each study was considered equally influential in the pooled estimate, with no single study disproportionately affecting the results.

Table 2.

Results of the meta-analysis involving 50 article.

Total examine animals	Test positive	Pooled prevalence	95%-CI	Quantifying heterogeneity		Test of heterogeneity
				I2	tau^2	Q	p-value
346917	99620	0.1948	[16; 23].	99,6	0.68 [0.74;2.63]	15687.69	<0.001

Publication bias assessment

The funnel plot shows some asymmetry (Fig 3a), with a concentration of studies at the top and a lack of studies towards the bottom, particularly on the left side. This suggests that smaller studies or studies with less favorable results may be missing, potentially due to publication bias or selective inclusion of studies with statistically significant results, especially if only the largest and most precise studies were reported. The contour-enhanced plot provides a clear visual indication that publication bias is present, as missing studies would likely fall in the non-significant white region (Fig 3b). The lack of studies in this area further supports the notion of publication bias. The Eggers test: The test for funnel plot asymmetry using a mixed-effects meta-regression model with the predictor as the standard error yielded a z-value of -4.9162 and a p-value of < .0001. A significant p-value (p < .0001) indicates strong evidence of asymmetry in the funnel plot. This statistically significant result suggests that publication bias is likely present. The negative z-value indicates that the bias is skewed towards studies with smaller standard errors, meaning larger studies with more precise estimates are overrepresented, while smaller studies with less precise estimates are underrepresented.

Fig 3.

The normal funnel plot (a) and contour-enhanced plot (b).

Trim and fill

To adjust for publication bias, consider using the trim-and-fill method (Duval and Tweedie, 2000). This was used to estimate the number of missing studies and provide an adjusted pooled estimate that accounts for the potential bias. The solid black circles represent the studies originally included in your meta-analysis. The open circles represent the imputed studies added by the trim-and-fill method to address funnel plot asymmetry (potential publication bias). These imputed studies are placed symmetrically to balance the plot. The imputed studies are added on the left side of the funnel plot, suggesting that smaller might be underreported or missing from the observed data set (Fig 4).

Fig 4.

The Trim and Fill plot.

Subgroup meta-analysis results

The results of this meta-analysis indicate that there are notable variations in effect sizes across different subgroups, moderated by factors such as continent, production type, publication year, species, and sample size category. When examining the continent-wise results, it was observed that studies conducted in Asia reported a prevalence/effect size (ES) of EBL at 0.17 (95% CI: 0.09 to 0.29), while studies from North America showed a higher prevalence of 0.43 (95% CI: 0.33 to 0.55). South America had an effect size of 0.10 (95% CI: 0.02 to 0.37), Africa had 0.14 (95% CI: 0.08 to 0.22), and Europe reported the lowest effect size at 0.04 (95% CI: 0.02 to 0.11). The heterogeneity was extremely high across all continents (I²= 99.0% to 99.8%), indicating substantial variability in the data. The test for subgroup differences was statistically significant (Q = 36.28, p < 0.0001), suggesting that the effect sizes differ significantly between continents, potentially due to varying study conditions, populations, or other region-specific factors.

For production type, the sub-analysis compared dairy and beef cattle. Dairy cattle had a prevalence of 0.18 (95% CI: 0.12 to 0.28), while beef cattle showed a slightly higher prevalence of 0.26 (95% CI: 0.14 to 0.45) EBL in the globe. Both subgroups exhibited high heterogeneity (I² = 99.7% and 99.4%, respectively), yet the differences between the two production types were not statistically significant (Q = 0.96, p = 0.3326). This suggests that the effect of production type on the outcome is minimal, with no substantial variation between dairy and beef cattle in this context.

The publication year-wise analysis, dividing studies conducted before and after 2015, revealed that prior studies had a prevalence of EBL at 0.29 (95% CI: 0.29 to 0.30) with high heterogeneity (I² = 99.0%), whereas studies conducted after 2015 showed a slightly higher prevalence of 0.32 (95% CI: 0.30 to 0.33) with higher source heterogeneity (I² = 99.7%). The subgroup differences were statistically significant (Q = 7.8, p = 0.014), indicating that the time period of the studies significantly influenced the effect sizes.

In terms of species level, the prevalence of EBL was highest in cattle with an effect size of 0.20 (95% CI: 0.13 to 0.27), followed by sheep with an effect size of 0.19 (95% CI: 0.08 to 0.37), buffalo at 0.19 (95% CI: 0.06 to 0.45), camel at 0.07 (95% CI: 0.003 to 0.65), and yak at 0.06 (95% CI: 0.004 to 0.50). Heterogeneity was consistently high across these groups (I² = 97.5% to 99.7%), but the test for subgroup differences was not significant (Q = 1.31, p = 0.8590), indicating no meaningful variation in the occurrence of EBL across different ruminant animals. While the data indicates that cattle have the highest prevalence of EBL with an effect size of 0.20, it is essential to consider the broader implications and potential biases in these findings. The effect sizes for sheep and buffalo are very close at 0.19, suggesting that these species are equally significant in terms of EBL prevalence. Moreover, the relatively low effect sizes for camel and yak, at 0.07 and 0.06 respectively.

Subgroup analysis was conducted to compare the diagnostic performance of ELISA, PCR and AGID. This was achieved by grouping studies based on their diagnostic method, allowing for a direct comparison of prevalence rates and identification of performance differences. ELISA, as the most widely used method, showed the highest pooled prevalence (34.17%), while PCR had a lower prevalence (27.77%) and AGID showed the lowest prevalence (1.16%). Heterogeneity was high across all groups, suggesting variations in study populations, methodologies, and protocols, which necessitates standardization of procedures for more reliable application (S3 Table)

Finally, when considering sample size as a moderator, studies with fewer than 200 participants had an effect size of 0.14 (95% CI: 0.066 to 0.25) with lower heterogeneity (I² = 90.0%, Table 3), studies with 200 to 500 participants reported an effect size of 0.25 (95% CI: 0.13 to 0.44) with high heterogeneity (I² = 99.1%), and studies with more than 500 participants showed an effect size of 0.17 (95% CI: 0.10 to 0.27) with extremely high heterogeneity (I² = 99.8%). The variability across these subgroups was not statistically significant (Q = 1.84, p = 0.397), suggesting that sample size did not significantly influence the prevalence of EBL for the current pooled results. Overall, these findings reveal that geographical region and publication year are significant moderators, while production style, species, and sample size category were found to have less impact. The consistently high heterogeneity across all subgroups points to considerable variability in the occurrence of EBL in the world, which may be due to differences in study design, population characteristics, methods of diagnosis, or other unmeasured factors (confounding ) within the studies included in the included studies for this meta-analysis.

Table 3.

Pooled effect size estimates of BLV, stratified by sub-groups.

Moderators	K	Category	N	Case	ES (95%CI)(RE)	Heterogeneity		Test for subgroup differences (RE)
						I2(%)	τ2	Q	p-value
Continent -wise	23	Asia	19800	5440	17 [0.09; 0.29]	99.1	2.57
	13	N. America	256052	86634	0.43 [0.33; 0.55]	99.3	0.70
	8	S. America	22337	4382	0.10 [0.02; 0.37]	99.8	5.35
	11	Africa	13497	2474	0.14 [0.08; 0.22]	99.0	0.91
	4	Europe	35231	690	0.04 [0.02; 0.11]	99.1	1.06
Production- wise	36	Dairy cattle	77785	10968	0.18 [0.12; 0.28]	99.7	2.66	0.96	0.3326
	10	Beef cattle	244411	82173	0.26[0.14; 0.45]	99.4	1.67
Year- wise	37	Prior 2015	317113	91716	0.29 [0.29; 0.30]	99.0	2.7	7.8	0.014
	23	Post-2015	29804	7904	0.32 [0.30; 0.33]	99.7	2.17
Species- wise	45	Cattle	323487	93253	0.20[0.13; 0.27]	99.7	2.4	1.31	0.8590
	3	Sheep	2768	357	0.19[0.08; 0.37]	97.5	0.58
	5	Buffalo	13018	4988	0.19[0.06; 0.45]	98.0	1.83
	4	Camel	1363	591	0.07[0.003; 0.65]	99.3	10.22
	2	Yak	2974	356	0.06[0.004; 0.50]	99.4	3.9
Sample size	13	<200	13		0.14 [0.066; 0.25]	90.0	2.75	1.84	0.397
	14	200-500	13		0.25[0.13; 0.44]	99.1	2.5
	32	>500	33		0.17[0.10; 0.27]	99.8	1.78

K= Number of included studies, N=Total number of animal population, Case = Test positive, ES= effect

Meta regressions

The results of this Meta-regression provide insights into the effects of various moderators on a given response variable. In publication year-wise, publications before 2015, the estimate is 0.5992, which shows a positive effect, with the p-value=0.0157 indicating this effect is statistically significant in association with the prevalence of EBL in ruminant animals in the world (Table 4). In terms of production type, dairy cattle do not show significant effects on the occurrence (p-value = 0.1824) of EBL compared with beef cattle. However, clinically the prevalence of EBL was higher in beef cattle than in dairy cattle. Regarding the continent, studies from Asia have a positive estimate of 0.3614, the effect was statistically significant (p-value = 0.0487). However, Europe has a non-significant negative effect with an estimate of -1.1810 (p-value = 0.153). On the other hand, North America shows a significant positive effect, with an estimate of 1.6383 and a p-value of 0.0045, indicating that studies from North America are associated with higher values in the prevalence of EBL. South America shows no significant effect with an estimate of -0.1892 (p-value = 0.775).

Table 4.

a clear and concise summary of the Meta-regression results for each model, including the estimates, standard errors, z-values, p-values, and 95% confidence intervals.

Moderators	Component	Estimates	95%CI	St. Error	P-value
Publication Year	Intercept	-1.7494	2.2691 -1.2297	0.2651	<0.0001
	Post-2015	Ref
	Pre-2015	0.5992	-0.2325, 1.4308	0.4243	0.01579
Production Category	Beef cattle	Ref.
	Dairy cattle	-0.4068	-1.5234, 0.7098	0.5697	0.04752
Continent	Africa	Ref.
	Asia	0.3614	-0.6580, 1.3807	0.5201	0.04872
	Europe	-1.1810	-2.8016, 0.4396	0.8269	0.1532
	N. America	1.6383	0.5072, 2.7693	0.5771	0.0045
	S. America	-0.1892	0.5072, 2.7693	0.6628	0.7753
Animal Species	Buffalo	Ref.
	Camel	-0.6475	3.0017, 1.7067	1.2012	0.5898
	Cattle	0.0880	-1.4906, 1.6665	0.8054	0.9130
	Sheep	0.1193	2.2748, 2.5135	1.2215	0.9222
	Yake	-1.1944	-3.9220, 1.5332	1.3917	0.3907

In terms of animal type, none of the other animal types show significant effects. Camel, with an estimate of -0.6475 (p-value = 0.5898), Cattle with 0.0880 (p-value = 0.130), Sheep with 0.1193 (p-value = 0.222), and Yak with -1.1944 (p-value = 0.3907), all show non-significant effects, suggesting no strong evidence that these animal types differ from the buffalo in their impact on the prevalence of EBL. In summary, the model indicates that publication year, production category, continent, and animal type each have baseline effects on the occurrence of EBL, but with notable differences. The only significantly positive effect comes from studies conducted in North America and study years.

Discussion

The current meta-analysis revealed considerable differences in the prevalence of EBL based on geographical, temporal, and species-specific factors, with an overall prevalence estimated at 19%. To the authors' knowledge, this is the first systematic review and meta-analysis to estimate BLV's pooled prevalence in ruminant animals. Compiling findings from multiple studies, which are often fragmented by space and time, is essential for demonstrating the disease’s global burden. This pooled estimate was higher than the 10% reported in China (Ma et al., 2021) and the 15% observed in Africa and Egypt (Zaher et al., 2014). The variation in EBL prevalence may be attributed to differences in management practices, biosecurity measures, and the genetic susceptibility of local cattle populations (Fava et al., 2016). Environmental factors such as climate and herd density, as well as the availability of veterinary services and farmer awareness about EBL prevention, could also play significant roles in influencing the spread of EBL across regions (Kobayashi et al., 2020).

Sub-analysis by publication year revealed that the prevalence of EBL was 32% after 2016, compared to 29% before 2016. The slight increase in the prevalence of enzootic bovine leukosis after 2016 (32%) compared to before 2016 (29%) could be attributed to several factors. These may include improved reporting and surveillance, advancements in diagnostic techniques, changes in farming practices, and more sensitive testing methods that led to better detection of subclinical cases. Additionally, shifts in geographic focus or changes in reporting standards over time might have contributed to the observed increase.

Subgroup analysis revealed that the highest prevalence of EBL was in North America (43%), followed by Asia (17%), Africa (14%), South America (10%), and the lowest in Europe (4%). The higher prevalence in North America may be attributed to the introduction or purchase of unscreened, EBL-infected animals or semen, along with the higher distribution of insect vectors (Ndou et al., 2011) such as stable flies and horn flies, which are known risk factors for BLV transmission. Increased exposure to mechanical transmission vectors could also contribute to the higher prevalence. The movement of cows without knowledge of their infection status into herds is a significant risk factor for the spread of BLV infection (Panneum et al., 2009). In contrast, BLV control programs in regions like Europe, where stricter disease control measures and testing protocols are in place, have likely contributed to the lower prevalence. Studies have reported that 83.9% of U.S. dairy herds were affected by EBL (Zaghawa et al., 2002), while Australia, New Zealand, and most European countries are free of the disease due to rigorous eradication programs (Jacobs et al., 1995). Intensive cattle farming, larger herds, and widespread use of artificial insemination in North America may accelerate the spread of BLV.

Beef cattle had a higher EBL prevalence (26%) compared to dairy cattle (18%). This could be due to less stringent biosecurity measures in beef cattle, with fewer controls on breeding, testing, and culling infected animals (Hassan et al., 2020). Dairy cattle are more frequently screened for BLV due to their close human contact in milking facilities. Additionally, beef cattle are often transported and sold in live markets, increasing the risk of BLV transmission. The higher prevalence of Bovine Leukemia Virus (BLV) in beef cattle compared to dairy cattle is linked to human management practices. This difference is largely driven by the breeding herd, not the fattening animals. Beef breeding cattle have a longer lifespan and less frequent human contact than dairy cows, which makes it harder to detect and isolate infected animals. Unlike the widespread use of artificial insemination in dairy, natural breeding is common in beef herds and can facilitate virus transmission from a bull to multiple cows. In contrast, fattening cattle are younger and have a shorter time on the farm, limiting their potential exposure to the virus. Therefore, differences in management, breeding practices, and the lifespan of the animals are key to understanding why BLV is more prevalent in beef cattle.

This finding contrasts with the results reported by Schoepf et al. (1997), who found higher prevalence rates in dairy herds (41%) and lower prevalence in beef herds (21%). The discrepancy may stem from differences in study methodologies, diagnostic methods, changes in herd management practices, or variations in local disease dynamics. Similarly, the present study found a lower BLV prevalence in beef cattle (11.8%) compared to dairy cows (42.5%) in Taiwan, as reported by Chen et al. (2021).

Among animal species, cattle had the highest EBL prevalence at 20%, consistent with findings from Egypt (20.9%), while yaks had the lowest prevalence at 6%, possibly due to their resistance to BLV infections, as suggested by Nekoei et al. (2015). Sheep had the second-highest prevalence at 19%, while camels showed the second-lowest prevalence at 7%. This variation could be attributed to unbalanced sample sizes across species; for example, the study by Nekoei et al. (2015) included 657 cattle, 92 sheep, and 122 camels. The variation in EBL prevalence across cattle, sheep, buffalo, camel, and yak species can be attributed to several factors, including host susceptibility, viral evolution, species-specific immune responses, management practices, and geographical distribution. Cattle, the primary hosts of the bovine leukemia virus, are particularly vulnerable, which may explain their elevated prevalence. BLV is primarily adapted to cattle, limiting its ability to infect other species like camels and yaks (Juliarena et al., 2016). Furthermore, camels and yaks may have stronger immune systems. Differences in farming systems, such as intensive cattle farming versus less intensive systems for camels and yaks, also affect exposure to the virus. Geographic variations in BLV prevalence are significant, with cattle in regions like North America showing higher infection rates, while camels and yaks are often found in areas where the virus is less common. The evolution of diagnostic tools has significantly influenced reported EBL prevalence over the decades. Early studies commonly used agar gel immunodiffusion (AGID), which has lower sensitivity, potentially underestimating infection rates. The adoption of ELISA allowed for more sensitive and large-scale screening of BLV antibodies, while the emergence of PCR techniques enabled the detection of viral DNA even in the absence of seroconversion. As a result, studies using ELISA and PCR tend to report higher prevalence figures than those relying on older methods. These diagnostic improvements partly explain the observed increase in EBL prevalence in more recent studies and may also contribute to the heterogeneity across studies included in our meta-analysis.

Recent studies published between 2022 and 2025 have significantly advanced our understanding of Bovine Leukemia Virus (BLV) and its role in causing cancer. A survey in Henan province, China, highlighted the diverse genetic makeup of the virus, showing that multiple strains (G1, G4, G6, and G7) are circulating and that infections often cluster within individual farms (Zhao et al., 2024). Other research has focused on the specific mechanisms by which BLV causes disease. Using high-resolution techniques, scientists were able to map where the virus integrates into a host cell's DNA, identifying distinct proviral integration sites one of which was found to be a defective, incomplete version of the virus (Yamanaka., 2022). This work has been further supported by a case study that tracked an infection over time, showing how the virus starts with multiple different clones and eventually progresses to a more dominant, or oligoclonal, state. In one case, a new viral clone was found to have inserted itself into a cancer-related gene called CHEK2, strongly suggesting that this insertional mutagenesis is a key step in the development of the disease (Kobayashi et al., 2022). Similarly, other studies have identified new integration sites in cattle with a chronic form of the infection, with two of these sites located near or within known cancer-driving genes like scfd2 and pgpep1 (Babii et al., 2022). These findings have practical applications, as a method originally developed for a similar human virus has now been adapted for BLV, allowing for the rapid detection of viral clones and the early identification of high-risk cattle. Collectively, these studies have not only provided a more detailed picture of how BLV works but have also paved the way for more effective diagnostic and management strategies.

In general, the overall prevalence of EBL remains considerable, with an increase in recent years. According to the pooled findings, the highest prevalence is seen in North America, among cattle species, and in beef cattle production systems. This emphasizes the importance of epidemiological data in guiding policies to control and prevent the disease. According to the WOAH Terrestrial Animal Health Code, EBL is a notifiable disease, and official control measures must include monitoring, precautions at borders, control of animal movements, and the culling of BLV-infected cattle at government expense. The considerable variation in prevalence underscores the need for targeted interventions tailored to the unique challenges each region faces, as well as increased funding and resources for effective control. Collaborative efforts between veterinarians, farmers, and researchers are crucial to developing targeted interventions and ensuring the adoption of best practices. In conclusion, this study provides critical insights into the global prevalence of EBL, highlighting significant regional disparities and the need for targeted interventions, particularly in non-cattle ruminants.

This systematic review has some limitations. It may not include all relevant studies globally, which could lead to bias, despite our use of publication bias assessment tools. Additionally, the R output was unable to incorporate all articles, which meant that the forest plot could not fully displayed the overall results this is due to high number of articles. Another limitation relates to the risk factors of the disease; this review did not focus on them. Furthermore, the diagnostic tools used across studies varied, with some employing serological techniques like ELISA, while others used PCR. This variability could introduce bias in the pooled results due to the possibility of false positive or false negative outcomes. Lastly, the protocol for this systematic review was not registered with any international registry databases. In general, the pooled prevalence estimate of EBL in this review showed substantial heterogeneity (I² = 99.6%), indicating considerable variability among the included studies. This heterogeneity likely stems from differences in study design, geographical region, time period, diagnostic methods (e.g., ELISA vs. PCR), target animal species, and population characteristics. While subgroup analyses helped to explore some of these sources, the residual heterogeneity suggests that the pooled estimate should be interpreted with caution. Consequently, the generalizability of the findings may be limited, and regional-level estimates may be more appropriate for informing targeted control programs. Future studies using standardized diagnostic criteria and consistent reporting across regions are recommended to reduce variability and improve comparability.

Conclusion

The present pooled prevalence of EBL reveals a significant global prevalence of 19%, highlighting the disease's widespread impact on ruminants. However, there are considerable regional disparities, with North America showing the highest prevalence at 43%, while Europe exhibits the lowest at 4%. Temporal trends indicate that EBL prevalence has increased over time, with rates rising from 29% before 2016 to 32% afterward. Species-specific analysis reveals that cattle are the most affected, with a prevalence of 20%, though sheep and buffalo are also susceptible. Additionally, beef cattle demonstrate a higher prevalence (26%) compared to dairy cattle (18%), potentially due to differences in management practices or environmental exposure. The study also uncovered significant heterogeneity among the included studies, driven by factors such as geographic location, animal species, and production type. This variation emphasizes the need for targeted approaches in combating EBL. Furthermore, publication bias was detected, suggesting that some findings may be underreported or skewed toward significant results. Despite this, the study highlights the growing global burden of EBL and the need for region-specific control measures, particularly in high-prevalence areas like North America. Beef cattle, given their higher infection rates, require more focused interventions. Moreover, the increasing prevalence over time underscores the importance of enhanced surveillance and monitoring. Finally, more research is needed to explore the factors driving heterogeneity and to develop effective control strategies across different regions and production systems.

Consent for publication

All the authors have read and approved the final manuscript.

Availability of data and material

All the data generated or analyzed during this study are available upon the request of the corresponding author.

Funding

Not applicable.

Ethics statement and consent to4 participate

Not applicable.

CRediT authorship contribution statement

Melkie Dagnaw: Writing – review & editing, Writing – original draft, Software, Methodology, Formal analysis, Conceptualization. Getachew Alemu Yilhal: Writing – review & editing, Writing – original draft, Software, Conceptualization. Bemrew Admassu: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis. Yitayew Demessie: Writing – review & editing, Writing – original draft, Validation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.