The study protocol was reviewed and approved by the Chulalongkorn University Animal Care and Use Committee (CU-VET IACUC; protocol no. 2031050, 01-09-2021). All animal-related procedures were conducted in accordance with institutional animal welfare requirements and the ARRIVE 2.0 guidelines. Hair and skin lesion samples, including samples from nodules, fistulas, sit-fasts, scabs, and scars, were collected by licensed veterinarians or appropriately trained personnel using procedures intended to minimize unnecessary discomfort to the animals. Before enrollment and sample collection, the study objectives, sampling procedures, and measures for maintaining data confidentiality were explained to the participating cattle farmers. Verbal informed consent was obtained from all participating cattle owners in accordance with the requirements of CU-VET IACUC. Personal protective equipment was used during field sampling to protect personnel and animals, and all data collected during the study were anonymized to protect the identities of participating farmers and animals.

The cross-sectional study was conducted from July to December 2022 on 208 beef cattle farms across 27 provinces representing all regions of Thailand. The study areas comprised Chiang Rai, Phetchabun, Phitsanulok, and Sukhothai in northern Thailand; Ang Thong, Bangkok, Lopburi, Nonthaburi, Phra Nakhon Si Ayutthaya, Saraburi, and Suphanburi in central Thailand; Buri Ram, Chaiyaphum, Khon Kaen, Mukdahan, Roi Et, and Ubon Ratchathani in northeastern Thailand; Sa Kaeo in eastern Thailand; Kanchanaburi, Phetchaburi, Prachuap Khiri Khan, Ratchaburi, and Tak in western Thailand; and Chumphon, Narathiwat, Pattani, and Trang in southern Thailand (Figure 1). The provinces were selected purposively and included areas where LSD outbreaks in beef cattle had previously been reported. The selected provinces were also associated with beef cattle supply chains, including areas adjacent to international cattle movement routes and areas participating in the project to promote beef cattle production throughout the value chain. Participation was coordinated through the DLD network, with the cooperation of provincial livestock personnel [14].

A purposive cross-sectional study design was used to investigate LSDV infection in beef cattle and characterize circulating viral strains. The study integrated epidemiological data collection, conventional polymerase chain reaction (PCR) detection, and Sanger sequencing of the RPO30 and GPCR genes. The minimum sample size was calculated using the single-proportion sample size function for apparent prevalence in EpiTools (Ausvet, Australia). The calculation assumed an expected prevalence of 10%, based on a previous study [15], a precision of 5%, a 99% confidence level, and an estimated beef cattle population of 9,756,642 animals in Thailand. The resulting minimum required sample size was 239 cattle; however, 482 cattle were ultimately enrolled.

Invitations to participate were distributed through the BDCVS to Provincial Livestock Offices nationwide. Provinces that agreed to participate and reported ongoing or recent LSD outbreaks were included. Eligible farms were required to have at least one animal exhibiting clinical signs consistent with LSD, including skin lesions and enlargement of superficial lymph nodes. Sampled cattle were required to have resided on the participating farm for at least 1 month before sample collection. One or more cattle were purposively selected from each farm according to the availability of eligible animals and feasibility of sampling under field conditions. Farms and cattle that did not meet these eligibility criteria were excluded. If the farm did not have animals showing clinical signs, at least one asymptomatic animal was selected.

A total of 523 samples were collected from 482 beef cattle on 208 farms. These comprised 387 hair samples and 136 skin lesion samples, including nodules, fistulas, sit-fasts, scabs, and scars (Table 1). Hair samples were collected by plucking at least 20 hairs with forceps from the skin over suspected lesions. Skin lesion samples of at least 0.5 × 0.5 cm were obtained by biopsy or collected as necropsy tissue. Samples were placed in appropriate collection containers and transported to the laboratory at 4°C within 24 h of collection. Samples were subsequently stored at −20°C until LSDV detection. Animal-level demographic and epidemiological information, including age, sex, clinical signs, and vaccination history, was recorded.

All samples were pretreated using an Omni Bead Ruptor 12 bead mill homogenizer (Omni International, Kennesaw, GA, USA). Approximately 0.5 cm³ of each skin lesion sample or 10–20 hairs were mixed with 300 µL of phosphate-buffered saline (PBS; pH 7.4) and homogenized according to the manufacturer's instructions. Viral DNA was subsequently extracted from 200 µL of each pretreated sample using the DNA/RNA GENTi™ 32 Automated Nucleic Acid Extraction System (GeneAll Biotechnology Co., Ltd., Seoul, Republic of Korea) according to the manufacturer's instructions. Extracted DNA was stored at −20°C until analysis.

LSDV was detected using conventional PCR targeting the RPO30 gene. The forward primer 5′-TCTATGTC TTGATATGTGGTGGTAG-3′ and reverse primer 5′-AGTGATTAGGTGGTGTATTATTTTCC-3′ were used as previously described [16]. The assay was performed according to the validated protocol described by Lamien et al. [16], which had a reported analytical detection limit of 250 copies/mL and high analytical specificity. PCR amplification consisted of initial denaturation at 95°C for 15 min, followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and extension at 72°C for 30 s, followed by a final extension at 72°C for 5 min. PCR products were analyzed by agarose gel electrophoresis, with an expected amplicon size of 172 bp. A DNA sample previously confirmed as LSDV-positive using a validated PCR assay was included as the positive control in each PCR run, and nuclease-free water served as the no-template negative control. DNA extraction and PCR detection were performed at the Center of Excellence for Emerging and Re-emerging Infectious Diseases in Animals (CUEIDAs), Faculty of Veterinary Science, Chulalongkorn University, Bangkok, Thailand.

For molecular characterization, the RPO30 and GPCR genes of LSDV were amplified by PCR using specific primers with minor modifications to a previously described method [5]. Seventeen LSDV-positive samples were selected for sequencing based on DNA quality, geographic location, and collection date. DNA concentration and purity were assessed using a NanoDrop spectrophotometer, with a concentration >20 ng/µL and an A260/A280 ratio >1.80 used as the stated quality criteria.

Each 50-µL PCR mixture contained 5 µL of viral DNA, 0.4 µM each of the forward and reverse primers, 1× HotStarTaq Plus Master Mix (Qiagen, Valencia, CA, USA), and nuclease-free water to a final volume of 50 µL. Amplification consisted of initial denaturation at 95°C for 15 min, followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 45°C for 30 s, and extension at 72°C for 90 s, followed by a final extension at 72°C for 15 min. PCR products were separated on a 1.5% agarose gel prepared in 0.5× Tris-borate-EDTA (TBE) buffer. A 100-bp GeneRuler DNA ladder (Thermo Scientific) was used to verify product sizes. The expected amplicon sizes were 606 bp for RPO30 and 1,134 bp for GPCR.

Following electrophoresis, PCR products were purified using the NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel, Landsmeer, The Netherlands) according to the manufacturer's instructions and subjected to bidirectional Sanger sequencing by U2Bio (Bangkok, Thailand). Forward and reverse chromatograms were manually inspected, edited, validated, and assembled using SeqMan version 5.03 (DNASTAR Inc., Madison, WI, USA) to generate consensus sequences. Ambiguous nucleotide calls were resolved by comparison of the corresponding forward and reverse reads. Consensus sequences were compared with sequences available in the National Center for Biotechnology Information database using the Basic Local Alignment Search Tool (BLAST) to confirm sequence identity and were subsequently exported in FASTA format for further analysis. The nucleotide sequences generated in this study were deposited in GenBank under accession numbers PZ284682–PZ284715.

Phylogenetic analyses were performed by comparing the nucleotide sequences of the RPO30 and GPCR genes generated in this study with reference sequences of members of the genus Capripoxvirus, including goatpox virus (GTPV), sheeppox virus (SPPV), and LSDV strains representing different clades and geographic origins in Africa, Asia, and Europe available in GenBank. Phylogenetic trees based on the RPO30 and GPCR sequences were constructed using MEGA version 11.0 with the neighbor-joining method, Kimura two-parameter model, and 1,000 bootstrap replicates. The Kimura two-parameter model was chosen a priori.

For genetic analysis, the GPCR nucleotide sequences of Thai LSDVs were aligned with reference LSDV sequences representing clades 1.1, 1.2, 2.1–2.4, and 2.5. Multiple sequence alignment and pairwise nucleotide identity analyses were performed using MEGA version 11.0.

Associations between LSDV positivity and animal age, sex, clinical status, and vaccination history were evaluated using individual cattle as the unit of analysis. The statistical model uses individual cattle as the unit of analysis, while farm-level clustering was not considered. Univariable logistic regression and chi-square tests were initially performed. Variables with p < 0.20 in the initial analysis were considered for inclusion in the multivariable logistic regression model. Multivariable logistic regression was subsequently performed to identify factors independently associated with LSDV positivity, and adjusted odds ratio (aOR) with 95% confidence intervals (CIs) were calculated. Statistical significance was established at p < 0.05. Statistical analyses were performed using R software version 4.2 and IBM SPSS Statistics version 29.0.1.0.

Farm characteristics and management practices were assessed using the epidemiological questionnaire, and the findings are presented in Table 2. Among the 208 beef cattle farms surveyed, 129 (62.02%) were classified as small farms (1–10 animals/farm), 54 (25.96%) as medium farms (>10 to <30 animals/farm), and 25 (12.02%) as large farms (≥30–277 animals/farm).

Regarding management actions taken for cattle suspected of LSDV infection (multiple responses were permitted), 200/208 (96.15%) farms reported treating affected cattle with antibiotics. In addition, 13/208 (6.25%) farms reported treatment combined with selling cattle for slaughter, and 8/208 (3.85%) reported treatment combined with selling cattle to another farm. One farm reported all three practices, whereas 8/208 (3.85%) farms reported taking no action.

For insect control, 58/208 (27.88%) farms reported no control measures, whereas 50/208 (24.04%) used nets only, 22/208 (10.58%) used insecticides only, 26/208 (12.50%) used smoke only, and 1/208 (0.48%) used native herbs only. Multiple insect control methods were used concurrently on 50/208 (24.04%) farms. Regarding farm cleaning, 181/208 (87.02%) farms were reported as always being cleaned, whereas 27/208 (12.98%) were reported as never being cleaned. Soil flooring was used on 169/208 (81.25%) farms, concrete flooring on 21/208 (10.10%), and a combination of soil and concrete flooring on 18/208 (8.65%). New animals had been purchased by 24/208 (11.54%) farms, whereas 184/208 (88.46%) reported no recent animal purchases. In addition, 89/208 (42.79%) farms allowed contact between their cattle and cattle from outside the farm (Table 2).

Among the 482 cattle sampled from 208 farms across 27 provinces, 405 (84.02%) exhibited clinical signs at the time of sampling, whereas 77 (15.98%) showed no apparent clinical signs. Clinical manifestations included skin nodules with or without sit-fasts and scabs, fistulas, scars, nasal and/or ocular discharge, edema, and systemic signs (Figure 2). The most frequently recorded clinical manifestations were skin nodules with scabs (80.95%), followed by scars (25.56%), systemic signs (6.27%), fistulas (3.26%), edema (3.26%), and nasal and/or ocular discharge (2.76%). Individual cattle could exhibit more than one clinical manifestation. The reported duration until recovery from clinical signs ranged from 1 to 6 months, with an average of 2 months (Supplementary Table 1).

Overall, 358/482 (74.27%) cattle tested positive for LSDV. Among cattle exhibiting clinical signs, 304/405 (75.06%) were LSDV-positive, compared with 54/77 (70.13%) cattle without apparent clinical signs. According to age group, LSDV positivity was highest among weaners (>6 to <12 months), at 82.35% (70/85), followed by calves (<6 months), at 81.20% (108/133), and adults (≥12 months), at 68.18% (180/264). Positivity was 75.65% (146/193) among males and 73.36% (212/289) among females. According to vaccination status, LSDV positivity was 79.30% (226/285) among unvaccinated cattle, 65.70% (113/172) among vaccinated cattle, and 76.00% (19/25) among cattle with unknown vaccination status.

Multivariable logistic regression identified age and vaccination status as factors significantly associated with LSDV positivity (p < 0.05; Table 3). Compared with adult cattle (≥12 months), weaners (>6 to <12 months) had 1.89-fold higher adjusted odds of LSDV positivity (aOR = 1.89; 95% CI: 1.02–3.69; p = 0.049), whereas calves (≤6 months) had 1.73-fold higher aOR (1.73; 95% CI: 1.01–3.01; p = 0.048). Vaccinated cattle had lower adjusted odds of LSDV positivity than unvaccinated cattle (aOR = 0.62; 95% CI: 0.39–0.98; p = 0.041).

At the sample level, 270/387 (69.77%) hair samples and 125/136 (91.91%) skin lesion samples tested positive for LSDV. The chi-square test showed a significant association between sample type and LSDV detection (χ² = 25.51, p < 0.05). Univariable logistic regression further showed that skin lesion samples had higher odds of testing positive than hair samples (OR = 4.92; 95% CI: 2.56–9.47; p < 0.05) (Table 3).

The RPO30 and GPCR sequences generated from 17 Thai LSDV-positive samples were compared with reference sequences representing members of the genus Capripoxvirus from different geographic regions and collection periods. Pairwise sequence analysis showed 100% nucleotide identity among all 17 Thai LSDVs for both the RPO30 and GPCR genes.

For the RPO30 gene, the Thai LSDVs shared 100% nucleotide identity with representative field strains from China, Indonesia, Hong Kong, Vietnam, and previously reported Thai LSDVs. In comparison, they shared 99.67% nucleotide identity with the classical Neethling and KSGP-O240 vaccine strains. For the GPCR gene, the Thai LSDVs shared 100% nucleotide identity with representative strains from Indonesia, Hong Kong, Thailand, and Vietnam, 99.91% identity with the representative Chinese strain, and 98.83% identity with the classical Neethling and KSGP-O240 vaccine strains.

Phylogenetic analyses based on the RPO30 and GPCR genes separated the analyzed LSDV reference strains into two major clades, clades 1 and 2, with additional subclades within each major group. In the RPO30-based phylogenetic tree, the Thai LSDVs clustered within clade 2.5 together with field strains from China (Xinjiang/19; OP508345), Hong Kong (Hong Kong/20; MW732649), Indonesia (A04224595/22; PQ037207), Thailand (YST/21; OM033705), and Vietnam (Bang-Thanh/20; MZ577076) (Figure 3). Similarly, phylogenetic analysis of the GPCR gene placed the Thai LSDVs within clade 2.5 together with LSDVs reported from China, Hong Kong, Indonesia, Thailand, and Vietnam (Figure 4).

Multiple sequence alignment of the GPCR gene showed that the Thai LSDV sequences were identical and contained a 12-nucleotide insertion (AGCACAATTCTC) at positions 94–105. A corresponding 12-nucleotide insertion was also present in LSDVs belonging to clade 1.1, some clade 1.2 viruses reported from Bangladesh and Kenya, and recombinant LSDV strains belonging to clades 2.1–2.5. In contrast, this insertion was absent from the analyzed wild-type subclade 1.2 LSDVs originating from Africa, the Middle East, and Europe. Among the analyzed subclade 1.2 viruses, the insertion was present in subclade 1.2.2 strains from India but absent from subclade 1.2.1 strains from Africa, the Middle East, Europe, Russia, and Kazakhstan (Table 4 and Figure 5).

This study provides epidemiological and molecular information on LSDV circulation in beef cattle in Thailand during 2022, approximately 1 year after the first reported outbreak in the country. The overall LSDV positivity among sampled cattle, determined by conventional PCR targeting the RPO30 gene, was 74.27%. Skin lesion samples yielded a substantially higher positivity rate (91.91%) than hair samples (69.77%). Comparable positivity rates in skin lesion samples have been reported previously in Thailand and Bangladesh, at 92.79% and 83.89%, respectively [17, 18], whereas another study in Thailand reported an overall LSDV positivity rate of 70.13% [11]. Differences among studies may reflect variation in study populations, outbreak stage, sample type, sampling strategy, and diagnostic approaches; however, these factors cannot be evaluated directly from the present dataset.

Age and vaccination status were significantly associated with LSDV positivity. Weaners (>6 to <12 months) and calves (≤6 months) had higher adjusted odds of LSDV positivity than adult cattle (≥12 months). Similar age-related patterns have been reported previously [19]. Although differences in immune status may contribute to greater susceptibility among younger animals, the present study did not directly assess immune competence, maternally derived antibodies, or individual immunological status. Therefore, the biological mechanisms underlying the observed age association cannot be established from these data.

Vaccinated cattle had lower adjusted odds of LSDV positivity than unvaccinated cattle (aOR = 0.62), supporting an association between vaccination and reduced likelihood of LSDV detection. This finding reinforces the importance of vaccination as a component of LSD prevention and control. Nevertheless, because the study was cross-sectional and information on vaccine type, number of doses, timing of vaccination relative to exposure, and post-vaccination immune response was not incorporated into the reported analysis, vaccine effectiveness should not be inferred directly from the observed odds ratio.

Skin lesion samples demonstrated a higher LSDV positivity rate than hair samples, with 91.91% of skin samples and 69.77% of hair samples testing positive. Univariable analysis further showed that skin lesion samples had higher odds of LSDV detection than hair samples. This finding is consistent with the well-established suitability of skin lesions for molecular detection of LSDV during clinical disease [15]. However, hair samples also yielded a substantial proportion of positive results, indicating their potential utility as an alternative specimen type under field conditions.

Hair samples were collected by plucking hairs to include the roots, where detectable viral material may be present [15]. Compared with skin biopsy, hair collection is less invasive and may be more practical for repeated or large-scale field sampling. Importantly, LSDV was also detected among cattle without apparent clinical signs, suggesting that hair or other minimally invasive samples may have value when lesion material is unavailable. Nevertheless, the present study did not directly evaluate diagnostic sensitivity, specificity, agreement between paired hair and skin samples, or animal-level paired performance. Therefore, the findings support the potential practical utility of hair samples but do not establish equivalence with skin lesion sampling.

Several farm management practices observed during the outbreaks warrant attention. Approximately 28% of farms reported no insect control measures, despite the recognized role of blood-feeding arthropods in LSDV transmission. Strengthening vector control practices may therefore be relevant to integrated outbreak management, particularly during periods of high vector abundance.

Antibiotic administration was reported on approximately 96% of surveyed farms. Because antibiotics do not act against LSDV, their use would generally be relevant only when secondary bacterial infections are suspected or confirmed. The present study did not record antibiotic classes, indications, dosages, treatment duration, withdrawal-period compliance, or veterinary prescription practices. Therefore, inappropriate antimicrobial use or withdrawal-period violations cannot be concluded from these data. Nevertheless, the high frequency of reported antibiotic administration highlights the need for further investigation of antimicrobial use practices during LSD outbreaks and for appropriate veterinary guidance regarding secondary bacterial complications and food-animal withdrawal requirements [4, 19].

LSDV strains have been classified into major phylogenetic clades using whole-genome sequences and commonly analyzed genetic markers, including GPCR and RPO30. Clade 1 includes clades 1.1 and 1.2. Clade 1.1, corresponding to Neethling-type viruses, includes classical vaccine-associated strains reported predominantly from sub-Saharan Africa. Clade 1.2, including Kenya/KSGP-type viruses, comprises field strains reported from Africa, Europe, and Asia and has been further divided into subclades 1.2.1, 1.2.2, and 1.2.3. Subclade 1.2.1 has been associated with outbreaks in the Middle East and Europe, including Balkan and Russian wild-type strains; subclade 1.2.2 includes Kenya-like strains reported in India and Bangladesh; and subclade 1.2.3 includes recently characterized strains reported from West and Central Africa. In contrast, clade 2 comprises recombinant viruses containing genomic regions derived from wild-type and vaccine-like LSDVs. Clades 2.1–2.4 have been reported mainly from Russia and Kazakhstan, whereas clade 2.5 has been reported in East and Southeast Asia, including China, Thailand, and Vietnam [3, 17].

Phylogenetic analyses based on the RPO30 and GPCR genes are commonly used for molecular characterization and evolutionary investigation of LSDV [2, 5, 20]. In the present study, all 17 characterized Thai LSDVs clustered within clade 2.5 based on both gene regions, together with viruses reported from China, Hong Kong, Indonesia, Thailand, and Vietnam [21]. The analyzed Thai sequences also showed complete nucleotide identity with one another in both RPO30 and GPCR, indicating high sequence conservation across these two loci among the sampled viruses. However, because only two gene regions were analyzed, these findings should not be interpreted as evidence of genome-wide genetic stability.

The clustering of Thai LSDVs with clade 2.5 strains reported elsewhere in East and Southeast Asia demonstrates close regional genetic relatedness. This finding is consistent with previous reports describing circulation of recombinant clade 2.5 LSDVs in China, Indonesia, Vietnam, and Thailand [11, 17, 21]. In contrast, LSDVs reported from Bangladesh, India, and Myanmar have been classified within clade 1.2 [4]. These contrasting phylogenetic patterns indicate that genetically distinct LSDV lineages have circulated in different parts of Asia. However, phylogenetic similarity alone cannot determine the direction, timing, or route of virus movement between countries.

Multiple sequence alignment of the GPCR gene demonstrated a 12-nucleotide insertion at positions 94–105 in the Thai LSDVs characterized in this study. The 12-nucleotide insertion was observed in recombinant LSDV lineages and in some non-recombinant strains (Table 4 and Figure 5). Therefore, the insertion represents a useful comparative molecular feature but should not be considered uniquely specific to clade 2.5.

In the present analysis, the insertion was also observed in clade 1.1, some clade 1.2 viruses, and recombinant clades 2.1–2.5, whereas it was absent from several wild-type subclade 1.2 strains from Africa, the Middle East, and Europe as observed in this study (Figure 5). Accordingly, this sequence feature may contribute to genetic characterization and phylogenetic discrimination when interpreted together with additional sequence information. Its biological significance, including any effect on viral phenotype, vaccine performance, or molecular diagnostic assays, remains uncertain. The potential value of this region for the development of lineage-oriented molecular assays would therefore require dedicated analytical and experimental validation.

A major strength of this study was the broad geographic coverage, encompassing 482 cattle from 208 farms across 27 provinces of Thailand, together with the integration of epidemiological assessment, PCR detection, and molecular characterization of representative LSDV-positive samples. The study also generated RPO30 and GPCR sequence data from multiple geographic regions, contributing additional information on the distribution of clade 2.5 LSDVs in Thailand during the 2022 outbreaks. In addition, the identification of age and vaccination status as factors associated with LSDV positivity provides epidemiological information relevant to the prioritization of disease control measures.

Several limitations should be considered when interpreting the findings. First, the cross-sectional design provides a snapshot of LSDV circulation during the study period and does not permit assessment of temporal infection dynamics or causal relationships between explanatory factors and LSDV positivity. Second, provinces and farms were enrolled purposively, with participating farms selected from areas with ongoing or recent suspected LSD outbreaks. Consequently, selection bias is possible, and the observed positivity proportion should not be interpreted as national prevalence. Third, cattle were clustered within farms, and the reported statistical analysis did not indicate whether farm-level clustering was accounted for. Fourth, 523 specimens were obtained from 482 cattle, indicating that some animals contributed more than one sample; therefore, the sample level comparison between hair and skin lesions may include correlated observations. Fifth, molecular characterization was limited to 17 representative LSDV-positive samples and two gene regions, RPO30 and GPCR. Virus isolation and whole-genome sequencing were not performed, limiting inference regarding genome-wide diversity, recombination patterns, and fine-scale transmission relationships.

Future studies should incorporate probability-based or clearly stratified sampling where feasible, longitudinal follow-up of affected farms, and statistical approaches that account for clustering of animals within farms. Direct paired comparisons of hair and skin lesion samples would help determine their relative diagnostic performance. Broader molecular characterization using whole-genome sequencing of viruses from multiple regions and time points would provide greater resolution for evaluating viral diversity, recombinant structure, and regional transmission patterns. Additional investigation of vaccination history, including vaccine type, timing, dose schedule, and post-vaccination status, would also strengthen assessment of the relationship between vaccination and LSDV detection.