The state of Karnataka comprises 31 districts, which according to the Karnataka State Remote Sensing Applications Centre are classified into ten distinct agro-climatic zones namely, North-Eastern Transition Zone, North-Eastern Dry Zone, Northern Dry Zone, Central Dry Zone, South-Eastern Dry Zone, Southern Dry Zone, Southern Transition Zone, Northern Transition Zone, Hilly Zone, and Coastal Zone. The sampling strategy was designed to ensure broad representation across these agro-climatic zones, facilitating a comprehensive epidemiological analysis of the Lumpy Skin Disease Virus (LSDV) throughout the state ( Supplementary Table 1 ).

Geographical data were used to investigate the epidemiology of Lumpy Skin Disease (LSD), assess the impact and severity of outbreaks, and evaluate the effectiveness of vaccination programs in controlling the disease in Karnataka from 2021 to 2024. Retrospective epidemiological data, including district-wise LSD cases, outbreaks, deaths, and vaccination details, were obtained from the Department of Animal Husbandry and Veterinary Services in Karnataka. Spatial analysis was conducted using Quantum Geographic Information System (QGIS) software (version 3.32.2) to create district-level maps illustrating the cattle population, mortality rate, and vaccination coverage. The mortality rate was measured to assess the impact of Lumpy Skin Disease on the cattle population.

The mortality rate and vaccination coverage were calculated using the following formulas:

The Case Fatality Rate (CFR) was analyzed to assess the severity of Lumpy Skin Disease (LSD) across different agro-climatic zones for the periods 2021–2022, 2022–2023, and 2023–2024. CFR was determined by calculating the proportion of confirmed cases that resulted in mortality, providing a quantitative measure of disease severity. The CFR was computed using the following formula:

Exploratory Spatial Data Analysis (ESDA) was employed to assess the spatiotemporal distribution and patterns of LSD cases in Karnataka from 2021 to 2024. The temporal progression of the disease was analyzed through spatial cluster analysis using GeoDa software version 1.22. Thematic maps were generated to illustrate the district-wise distribution of LSD cases over three time periods: 2021-2022, 2022-2023, and 2023-2024, utilizing data from the Department of Animal Husbandry and Veterinary Sciences, Karnataka.

Temporal analysis was performed by monitoring spatial clustering trends over successive years (2021–2022, 2022–2023, and 2023–2024), enabling an understanding of disease progression across districts. A spatial weight matrix was used to assess the relationships between the 31 districts, employing a queen contiguity matrix to define neighborhood interactions.

Moran's I statistic was used to assess the strength of spatial clustering and disease heterogeneity in the study area. The Moran's I value ranges from -1 to +1, where values near zero indicate random distribution, +1 signifies strong positive clustering, and -1 indicates strong negative clustering. Local Indicators of Spatial Association (LISA) clustering was performed to further identify spatial clusters and outliers. LISA categorizes districts into high-high, low-low, low-high, and high-low clusters, indicating varying patterns of spatial association. For further spatial evaluation, the Getis-Ord-Gi* statistic was used to measure local spatial autocorrelation and identify hot and cold spots of LSD cases.

During the LSD outbreak, field investigations were carried out with approval from the Department of Animal Husbandry and Veterinary Sciences, Karnataka. Clinical samples were collected from affected cattle in accordance with the guidelines set by the World Organisation for Animal Health (WOAH) from 21 districts across eight agro-climatic regions of Karnataka. Nasal swabs were collected by carefully inserting a sterile, flexible, fine-shafted swab into the nasopharyngeal cavity and gently rotating it to absorb secretions. The swabs were subsequently placed into sterile collection tubes containing 1 mL of viral transport medium (VTM). Additionally, 2–3 mL of blood was drawn via jugular vein puncture and collected in ethylene diamine tetra acetic acid (EDTA) vacutainers. Skin nodules or scabs from cattle exhibiting clinical signs of the disease were aseptically collected by first washing and cleaning the affected area with sterile saline and removing the hair using sterile scissors. The site was then punctured, and a portion of the nodule or scab was excised after disinfection with 70 percent isopropyl alcohol.

Additionally, a structured questionnaire was employed to gather essential information regarding the sampled animals, including age, sex, breed, and demographic details such as district and region. A total of 1,353 samples were collected from 21 districts across eight agro-climatic regions of Karnataka. To ensure sample integrity, all specimens were transported in insulated shipping containers with dry-ice packs, maintaining a strict cold chain. Samples were analyzed within 4–8 hours of receipt at the BSL-2+ virology laboratory, Indian Council of Agricultural Research-National Institute of Veterinary Epidemiology and Disease Informatics (ICAR-NIVEDI), under the approval of the Institute Animal Ethics Committee (No. NIVEDI/IAEC/2022/06), Bengaluru. All samples were subsequently stored at -20°C except for blood samples, which were preserved at 4°C.

The collected samples were thawed to room temperature before processing. Nasal swabs were vortexed thoroughly and aliquoted for extraction, while 150 μL of blood was suspended in 50 μL of 1X phosphate-buffered saline (PBS, pH 7.2). Approximately 30 mg of skin nodules or scab samples were finely chopped using a sterile Bard Parker blade, and a tissue homogenate was prepared by triturating the sample with 1 mL of chilled 1X PBS using a mortar and pestle. The total genomic DNA was extracted from the processed clinical samples using the DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany), following the manufacturer's protocol. DNA concentration and purity were assessed using a Nanodrop spectrophotometer (NABI, Microdigital Co., Ltd., South Korea). The extracted DNA was stored at -20°C until further use. All biological materials were disposed of in accordance with the standard operating procedures (SOPs) established by the Institute Biosafety Committee (IBSC).

Diagnostic screening for the presence of lumpy skin disease virus (LSDV) was conducted on 1,353 clinical samples, including nasal swabs(n=451), whole blood (n=451), and skin nodules or scabs (n=451) from the same cattle. DNA was extracted and initially amplified for the Capripoxvirus-specific partial P32 gene, followed by amplification of the LSDV-specific extracellular enveloped virion (EEV) glycoprotein gene (LSDV126) using respective forward and reverse primers, as detailed in Table 1 .

Conventional polymerase chain reaction (PCR) assay was performed using a Bio-Rad thermal cycler (Hercules, CA, USA) with a total reaction volume of 25 μL. The reaction mixture comprised 12.5 μL of PCR master mix (DreamTaq Green PCR Master Mix 2X, Thermo Scientific, Waltham, MA, USA), 1 μL each of forward and reverse primers (10 pmol/μL), 8.5 μL of nuclease-free water, and 50 ng of template DNA. Following amplification, PCR products were analyzed via agarose gel electrophoresis on a 1.5% agarose gel prepared in 1X Tris-acetate-EDTA (TAE) buffer, stained with ethidium bromide, and visualized under a UV transilluminator using a Gel Documentation System (Vilber Bio-Print, Vilber Lourmat, France).

The Madin-Darby Bovine Kidney (MDBK) cell line was used for virus adaptation and isolation. Nasal swab samples, skin nodules, and scab tissue suspensions that tested positive by PCR were syringe-filtered using a 0.45 μm filter and subsequently used to infect a confluent monolayer of MDBK cells. Virus-infected MDBK cell lines were maintained in Minimum Essential Medium (MEM; HiMedia, Mumbai, India) supplemented with an antibiotic-antimycotic solution (1X; HiMedia, Mumbai, India) and 2% fetal bovine serum (FBS; MP Biomedicals, Santa Ana, CA, USA). The infected cells were incubated at 37°C in a 5% CO₂ incubator (Thermo Scientific, Waltham, MA, USA) and monitored daily for five days to observe cytopathic effects (CPE). Following infection, the virus was harvested by subjecting it to three freeze-thaw cycles and stored at -80°C. The viral suspension was collected and passaged. Upon the appearance of CPE, genomic DNA was extracted from the harvested virus and confirmed by PCR.

PCR-positive amplicons of full-length genes (P32, RPO30, and GPCR), resolved by agarose gel electrophoresis, were excised and purified using the NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel, Düren, Germany), following the manufacturer's instructions. The purified DNA was stored at -20°C and subsequently subjected to Sanger sequencing at Eurofins Genomics India Private Limited, Bengaluru, India. Sequencing data were analyzed and edited using GeneTool software (Informer Technologies, Inc.), and the sequences were submitted to GenBank database and are available under the accession numbers as follows: (P32 gene: OR472969-OR472971, OR472973, OR472975, OR863389, OR602866, OR472977, OR699297, OR699298; RPO30 gene: OP903442, OP903444-OP903458, OM362830, OM362845, OR863389, OR602866; GPCR gene: PP530467, PP530477, PP530466, PP530475, OR602866, OR863389, OR699291).

Comparative sequence analysis was performed using Capripoxvirus-associated nucleotide sequences of P32, GPCR (G-protein coupled receptor), and RPO30 genes retrieved from GenBank, incorporating in-house Karnataka isolates, sequences from neighboring states, and other regions of India, as well as global LSDV strains, vaccine strains, and representative Sheep pox and Goat pox virus sequences. Multiple sequence alignment was performed using the MUSCLE algorithm in Molecular Evolutionary Genetics Analysis (MEGA) software, version 11. Phylogenetic relationships were inferred using the Maximum Likelihood method, employing the Tamura 3-parameter model for P32 and GPCR genes and the Hasegawa-Kishino-Yano model for RPO30, with bootstrap analysis of 1000 replicates to assess branch reliability. The tree with the highest log-likelihood value was selected, and the percentage of replicate trees supporting each cluster was displayed alongside the branches. Initial heuristic search trees were automatically generated using the Neighbor-Joining (NJ) and BioNJ algorithms, based on pairwise distance matrices computed from the selected model, and the topology with the highest log-likelihood was retained. The constructed phylogenetic tree was visualized using Tree Explorer and further annotated and refined using Interactive Tree of Life (iTOL) v7.

The collected samples were categorized based on the information collected during the outbreak on various epidemiological criteria. For the statistical analysis, age was classified into three groups: less than 1 year (n = 336), 1 year - 5 years (n = 498), and ≥ 5 years (n = 519). Sex-based stratification divided the samples into male (n = 528) and female (n = 825). Breed classification included local breeds (n = 576), exotic breeds (n = 570), and non-descript breeds (n = 207). Spatial classification based on agro-climatic zones included: North-Eastern Dry Zone (n = 147), Northern Dry Zone (n = 24), Central Dry Zone (n = 237), South-Eastern Dry Zone (n = 426), Southern Dry Zone (n = 114), Southern Transition Zone (n = 198), Northern Transition Zone (n = 159), and Hilly Zone (n = 48). Additionally, samples were grouped based on the year of collection: 2021 (n = 513), 2022 (n = 462), 2023 (n = 276), and 2024 (n = 102).

A chi-square analysis was conducted using GraphPad Prism 8.0.1 to assess the association between categorical variables and the proportion of positive cases for 451 samples each of blood, nasal swabs, and skin tissue or scab. The chi-square test for proportions was applied to evaluate statistical significance and determine whether a significant association exists between the analyzed variables and the percentage of positive cases.

Statistical analyses were conducted using the Statistical Package for the Social Sciences (IBM^® SPSS^® version 27). In this study, clinical samples of lumpy skin disease (LSD), including nasal swabs, whole blood, and skin nodules or scabs, acted as dependent variables, representing the diagnostic outcomes. Independent variables (predictors) included various risk factors such as agro-climatic region, age, sex, breed, and year of sample collection, as these factors could influence disease occurrence or be associated with diagnostic results.

A univariate analysis was initially performed to evaluate each risk factor individually. This was followed by a multivariate logistic regression analysis to assess the combined effects of multiple risk factors on disease outcomes. In the logistic regression model, the odds ratio, also known as Exp(B), was calculated to quantify the strength and direction of associations between predictor variables and disease outcomes. A 95% confidence interval (CI) was estimated to determine the likely range of the true population parameter, helping assess uncertainty and precision. Statistical significance was determined using a p-value threshold of < 0.05, indicating a significant association, while a p-value ≥ 0.05 suggested no statistically significant relationship.

According to the 20th Livestock Census (2017), Karnataka's cattle and buffalo population is approximately 12 million, representing 4.3% of the national livestock population. The total cattle population in the state is 8,469,004, with the highest population observed in districts such as Belagavi (549,540), Hassan (548,185), Shivamogga (518,653), and Mysore (492,598). Conversely, districts with the lowest cattle populations include Kodagu (71,684), Gadag (136,311), Dharwad (172,219), and Bidar (173,634) ( Figure 1A ).

Vaccination coverage data from 2021 to 2024 indicated that several districts, including Bengaluru Urban, Chitradurga, Davangere, Kolar, Ramanagara, Shivamogga, Tumakuru, Chikkamagaluru, Mandya, and others, achieved 100% vaccination coverage in 2022. However, vaccination coverage varied across the years, with an initial rate of 0.99% in 2021, peaking at 100% in 2022 and 2023, followed by a slight decline to 94.8% in 2024 ( Figure 2 ). Additionally, districts such as Bengaluru Rural (93.89%), Chikkaballapur (89.87%), Chamarajanagar (96.42%), Dakshina Kannada (97.95%), Mysuru (96.92%), Udupi (95.71%), Uttara Kannada (97.51%), and Kodagu (91.24%) exhibited slightly lower vaccination coverage ( Figure 1C ).

The disease spread across 4,380 villages in 160 taluks, affecting over two million cattle, spanning 30 districts of Karnataka. Mortality rates from 2021 to 2024 indicated that districts such as Belagavi (1.45%), Gadag (1.07%), and Haveri (1.15%) exhibited higher mortality rates. In contrast, Kodagu (0.03%), Udupi (0.04%), Mysuru (0.07%), and Bidar (0.07%) reported the lowest mortality rates ( Figure 1B ). A temporal analysis of mortality rates revealed no reported fatalities in 2021, followed by an increase to 0.29% in 2022, a subsequent decline to 0.08% in 2023, and no reported mortality rate in 2024 ( Figure 2 ).

The case fatality rate (CFR) across Karnataka's agro-climatic zones from 2021 to 2024 exhibited distinct temporal variations ( Figure 3 ). During 2021–2022, the overall CFR was 0.15%, with most zones reporting no fatalities, except for the Central Dry Zone (1%) and the Northern Transition Zone (0.5%). By 2022-2023, the CFR increased substantially to an average of 7.6%, with the highest rates recorded in the North Eastern Transition Zone (12.37%), Northern Transition Zone (12.33%), Northern Dry Zone (10.13%), and Central Dry Zone (9.71%). The Southern Dry Zone (4.83%) and Coastal Zone (1.36%) exhibited the lowest CFR during this period ( Figure 3 ).

In 2023–2024, the CFR further increased to an average of 13.8%, with certain regions experiencing pronounced spikes. The Northern Transition Zone, including Belagavi and Haveri districts, recorded the highest CFR at 27.20%, aligning with their elevated mortality rates in 2022. Similarly, the Northern Dry Zone (21.6%) and Central Dry Zone (18.29%) showed substantial increases. The North Eastern Transition Zone, which had the highest CFR (12.37%) in 2022-2023, exhibited a 4.07% decline in 2023-2024. Meanwhile, the North Eastern Dry Zone maintained a relatively lower CFR of 6.47%, whereas the Coastal Zone (7.50%), Hilly Zone (12.02%), and Southern Transition Zone (13.03%) continued to show an upward trend ( Figure 3 ).

At the peak of the LSD outbreaks, an extensive epidemiological investigation was conducted in the affected areas of Karnataka. Infected cattle showed typical clinical signs of LSD, such as an elevated body temperature, reduced milk production, anorexia, lymphadenopathy, lacrimation, hypersalivation, and characteristic nodular skin lesions about 50 mm in diameter ( Figure 7 ). Diagnostic evaluation using PCR demonstrated variations in positivity across different sample types. Statistical analysis using the Chi-square test demonstrated a significant association between sample type and PCR outcome with a Pearson chi-square value of χ² = 6.438 (df = 2, p = 0.040) and a likelihood ratio test result of χ² = 6.409 (df = 2, p = 0.041). Among the tested samples, skin nodules or scabs showed the highest positivity rate at 83.6%, followed by nasal swabs (80.9%) and blood samples (76.9%) within the category.

Representative samples, including scabs and nasal swabs collected during LSD outbreaks, were used for virus isolation. The isolates were subjected to two to three blind passages to induce characteristic cytopathic effects (CPE) in cell cultures. The observed CPE, indicative of LSDV infection, included the formation of multinucleated syncytia, cellular rounding, vacuolation, and ballooning, culminating in cytolysis and cell detachment. The presence of foci, characterized by clusters of infected cells, indicated robust viral replication. Consistently, the LSDV field isolates were successfully adapted to the MDBK cell line, with CPE becoming apparent between the sixth and tenth passages. This progressive adaptation and foci formation, as illustrated in Figure 8 underscores the pronounced cytopathic nature of LSDV, reflected in distinct morphological alterations in infected cells ( Figure 8 ).

For the molecular epidemiology of the disease, cell culture isolated LSDV from affected cattle, and along with PCR-positive diagnostic samples, full-length sequencing and phylogenetic analysis were employed on key genes, including P32 (envelope protein), RPO30 (RNA polymerase gene), and GPCR (integral membrane protein). The sequences were edited and submitted to GenBank and received accession numbers for P32, GPCR, and RPO30. Maximum likelihood phylogenetic trees were constructed, yielding log likelihood values of -1489.21 for P32, -942.71 for RPO30, and -2096.13 for GPCR, as illustrated in Figures 9A–C respectively. The analysis revealed distinct clustering patterns, depicting the genetic relationships and evolutionary trajectory of LSDV over time. Karnataka isolates are distributed across multiple clades, indicating regional dissemination and genetic variability. These isolates exhibited close phylogenetic clustering with LSDV field strains circulating in India, Kenya, Ismailia, Serbia, Russia, Israel, Turkey, Kazakhstan, Bangladesh, Pakistan, Nepal, Myanmar, and neighboring regions of Karnataka, suggesting extensive geographical transmission networks. Notably, Karnataka isolates demonstrated phylogenetic proximity to Middle Eastern and African strains, particularly from Kenya and Egypt, suggesting historical transmission links. Meanwhile, recent isolates exhibited slight genetic divergence, indicative of local adaptation and ongoing viral evolution. The phylogenetic trees also clearly differentiated field isolates from vaccine strains, with vaccine-derived LSDV forming distinct clades in each tree, emphasizing the genetic divergence between wild-type and vaccine strains. Furthermore, Karnataka LSDV isolates were distinctly separated from other Capripoxviruses, such as goat pox and sheep pox viruses, reaffirming LSDV's genetic specificity within the Capripoxvirus genus and its unique evolutionary trajectory as a distinct lineage ( Figure 9 ).

The univariate and multivariate regression analyses revealed several demographic risk factors associated with the prevalence of LSD in Karnataka, with both sets of analyses revealing key findings summarized in Tables 2 and 3 . Age emerged as a significant determinant in both analyses. Cattle aged 1 to 5 years and those less than one year showed a higher risk of LSD infection compared to cattle aged over 5 years. Specifically, cattle in the age group of 1 to 5 years exhibited 2.5–3.5 times greater odds of infection, a trend that persisted in multivariate models. Further, younger cattle (<1 year) demonstrated a significantly moderate risk of infection (OR: 0.951-2.017) when compared to cattle aged over 5 years.

Breed analysis indicated that exotic cattle were more susceptible to LSDV infection compared to indigenous cattle. In univariate analysis, exotic breeds had 2.1–2.6 times higher odds of infection compared to indigenous breeds, a finding corroborated by multivariate regression. Gender differences were evident, with females showing a higher likelihood of infection than males. This trend was particularly notable in all the different types of samples, where females showed 7% to 76% more chance of LSD incidence than males. Geographical location significantly influenced LSD prevalence. The Southern Transition Zone covering Chikkamagaluru, Hassan, and Shivamogga districts exhibited the highest infection risk across all sample types. Blood samples from this zone had 12.5 times (OR: 12.5; 95% CI: 1.559 - 100.198) greater odds of testing positive compared to the Hilly Zone, with nasal swabs and skin samples also showing elevated risks. Conversely, the Northern Dry Zone was identified as a low-risk area, with significantly reduced odds in both univariate and multivariate analyses. Temporal analysis revealed a marked reduction in LSD risk over time, with 2021 representing the peak outbreak year. Skin scab samples collected in 2021 showed 21.3 times greater odds of infection (OR: 21.338; 95% CI: 5.559–81.903) compared to 2024. Subsequent years, including 2022 and 2023, showed a progressive decline in infection risk, reflected by steadily decreasing odds.