Abstract: The high attrition rates in glioblastoma (GB) therapeutic development stem largely from preclinical models that fail to adequately recapitulate the dynamic tumor-host ecosystem. Unlike previous reviews that characterize glioma cell lines in isolation, this article integrates tumor biology with the distinct neuro-immune-endocrine landscapes of major laboratory rat strains. We critically evaluate standard rat malignant glioma cell lines (C6, F98, RG2, 9L) alongside transplantable tissue models (GB 101.8, GB 15/47), which offer enhanced translational relevance, demonstrating that the predictive value of any model is contingent upon the specific "glioma model and host strain" pairing and the individual physiological characteristics of the host. We provide evidence that strain-specific hypothalamic-pituitary-adrenal (HPA) axis reactivity (e.g., hyper-reactive Fischer 344 versus normo-reactive Wistar) acts as a decisive, yet often overlooked, modulator of the tumor microenvironment and therapeutic response. The review delineates the utility and limitations of these models, specifically addressing the MHC incompatibilities of the widely used C6 model in immunotherapy research, while contrasting it with the immune-evasive phenotypes of RG2 and the GB 101.8 tissue model. Furthermore, we highlight the superiority of tissue transplants in preserving cellular polyclonality and diffuse infiltration patterns compared to the circumscribed growth often observed in cell line-derived tumors. Consequently, we propose a strategic selection paradigm wherein immunogenic models serve as bioindicators of host immunocompetence, while invasive, non-immunogenic systems (F98, RG2, GB 101.8) are utilized to investigate therapeutic resistance and systemic host-tumor interactions.

Abstract: This study, conducted jointly by diagnostic institutes from Italy and the Republic of Ko-sovo/Albania, aimed to determine the occurrence of relevant infectious and parasitic agents (IPAs) in managed honeybees from these geographical areas to assess the overall health status of local apiaries. Paenibacillus larvae, Melissococcus plutonius, Nosema ceranae, Nosema apis, acute bee paralysis virus (ABPV), black queen cell virus (BQCV), chronic bee paralysis virus (CBPV), deformed wing virus variants DWV-A and DWV-B, and the para-sitoid flies Megaselia scalaris and Senotainia tricuspis were detected by quantitative poly-merase chain reaction (qPCR) and reverse transcriptase qPCR (RT-qPCR) in clinically healthy adult honeybees collected from 131 and 56 apiaries in the Abruzzo and Molise re-gions of Central Italy in 2022 and 2023, respectively, from 140 apiaries in the Republic of Kosovo and 18 in Albania in 2022, and from 66 apiaries in the Republic of Kosovo in 2023. The percentages of positive samples and abundance levels for N. ceranae, P. larvae and DWV-B were significantly higher in the Republic of Kosovo and Albania, while the per-centages of samples positive for M. plutonius, CBPV, DWV-A, and the parasitoid flies were higher in Central Italy. Additionally, P. larvae and some viruses showed significantly dif-ferent occurrence rates between the two years in Italy and the Republic of Kosovo. The co-occurrence of IPAs also differed between the two geographic areas. Their varying dis-tribution could depend on epidemiological dynamics, climatic factors, and management practices specific to each country, whose relative impact should be defined to guide tar-geted interventions to reduce honeybee mortality.

Abstract:

Familial cancers are caused by inherited mutations in specific genes that regulate cell growth, division, and repair. Approximately 5–10% of all cancer cases have a hereditary component, where germline mutations in certain genes increase an individual’s susceptibility to developing cancer. Two major categories of genes are involved in cancer development: tumour suppressor genes and oncogenes. Both play critical roles in regulating normal cell behaviour, and when mutated, they can contribute to uncontrolled cell proliferation and tumour formation. In addition to genetic mutations, epigenetic alterations also play a significant role in familial cancer. Epigenetics refers to changes in gene expression due to DNA methylation, histone modifications, and the dysregulation of non-coding RNAs without alter the underlying DNA sequence. Familial cancer syndromes follow various inheritance patterns, including autosomal dominant, autosomal recessive, X-linked, and mitochondrial inheritance, each with distinct characteristics. Identifying genetic mutations associated with familial cancers is a cornerstone of genetic counselling, which helps individuals and families navigate the complex intersection of genetics, cancer risk, and prevention. Early identification of mutations enables personalized strategies for risk reduction, early detection, and, when applicable, targeted treatment options, ultimately improving patient outcomes.

Abstract:

Background and Objectives: Dichrostachys glomerata and Cissus quadrangularis, two species traditionally used in Cameroon, are recognized for their weight-reducing potential. This study examined the effects of standardized extracts of these botanicals on glucagon-like peptide-1 (GLP-1), dipeptidyl peptidase-4 (DPP-4), and key metabolic outcomes in individuals with excess body weight. Materials and Methods: In this 16-week, randomized, double-blind, placebo-controlled trial, 248 adults (126 women and 122 men; mean age 41.3 ± 0.3 years; BMI 25–34.9 kg/m²) were assigned to receive 400 mg D. glomerata extract (DGE), 300 mg C. quadrangularis extract (CQE), semaglutide (dose-escalated from 3 mg to 14 mg), or placebo, administered once daily. Primary assessments included changes in GLP-1 levels and DPP-4 activity. Secondary evaluations included body composition, caloric intake, satiety response, fasting glucose levels, and lipid profiles. Results: Participants receiving DGE or CQE displayed notable elevations in circulating GLP-1 (+38.6 pg/mL and +42.2 pg/mL, respectively; p < 0.01) and significant reductions in DPP-4 activity (−15.3% and −17.8%; p < 0.01) compared with placebo. Both extracts produced substantial improvements in body weight (−5.2% and −5.8%), body fat (−10.3% and −10.9%), energy intake (−16.2% and −17.5%), and satiety (+25.6% and +27.4%) (p < 0.01). Significant changes in fasting glucose and serum lipid levels were also observed (p < 0.05). These responses are similar to those of semaglutide. Moreover, GLP-1 increments showed strong negative correlations with body fat percentage (r = −0.91 to −0.92; p < 0.001) and DPP-4 activity (r = −0.97 to −0.98; p < 0.001). Conclusion: Supplementation with D. glomerata and C. quadrangularis extracts enhanced GLP-1 secretion and reduced DPP-4 activity, yielding significant benefits for body composition and metabolic parameters. These findings indicate that both botanicals are promising natural agents for managing obesity through incretin-based mechanisms.

Abstract: Background Diabetic foot ulcer (DFU) is a severe complication of diabetes mellitus (DM), contributing significantly to morbidity and healthcare burden. Kerala, a South Indian state, has a high prevalence of diabetes, yet limited data exist on the genetic risk factors for DFU in this population. Vitamin D, beyond its role in calcium homeostasis, is crucial for immune regulation and wound healing. Methods In this study, plasma vitamin D levels were evaluated in individuals with DFU (n=112), DM without ulcers (n=112), and non-diabetic controls (n=112). We also assessed the association of four common vitamin D receptor (VDR) gene polymorphisms—rs7975232 (ApaI), rs731236 (TaqI), rs1544410 (BsmI), and rs2228570 (FokI)—with DFU and DM. Results The arithmetic mean 25(OH)D levels were 17.7 ± 14.2 ng/mL (DFU), 19.5 ± 12.4 ng/mL (DM), and 20.6 ± 13.9 ng/mL (controls). Geometric means were lower across all groups, with 12.7 ng/mL (DFU), 15.4 ng/mL (DM), and 15.7 ng/mL (controls), indicating widespread vitamin D deficiency, especially in DFU cases. Among the SNPs studied, rs731236 and rs1544410 showed statistically significant associations with DFU. The AG genotype of rs731236 was significantly associated with a reduced risk of DFU compared to both DM patients and non-diabetic controls (p < 0.05, OR ≈ 0.5), indicating a strong protective effect. Similarly, the TC genotype of rs1544410 showed a protective association against DFU (p = 0.0219, OR = 0.51). In contrast, the AA genotype of rs731236 and TT genotype of rs1544410 were marginally associated with increased DFU risk. No significant associations were found for rs2228570 and rs7975232. However, the ff genotype of rs2228570, which is associated with a less active VDR protein, was predominantly observed among study participants. VDR polymorphisms, especially rs731236 and rs1544410, may significantly influence DFU susceptibility in the Kerala population and hold promise as genetic risk markers. Routine vitamin D screening could aid early risk assessment in diabetic individuals.

Abstract:

Background: Precise delineation of the rectum is crucial in treatment planning for cancers in the pelvic region, such as prostate and cervical cancers. Manual segmentation is also still time-consuming and suffers from inter-observer variability. Since there are meaningful differences in rectal anatomy between males and females, incorporating sex-specific anatomical patterns can be used to enhance the performance of segmentations. Furthermore, recent deep learning advancements have provided promising solutions for automatically classifying patient sex from CT scans and leveraging this information for enhancing the accuracy of rectal segmentation. However, their clinical utility requires comprehensive validation against real-world standards. Methods: In this study, a two-stage deep learning pipeline was developed using CT scans from 186 patients with either prostate or cervical cancer. First, a CNN model automatically classified the patient’s biological sex from CT images in order to capture anatomical variations dependent on sex. Second, a sex-aware U-Net model performed automated rectal segmentation, allowing the network to adjust its feature representation based on the anatomical differences identified in stage one. The internal validation had an 80/20 train-test split, and 15% of the training portion was held out for validation to ensure balanced distribution regarding sex and diagnosis. Model performance was evaluated using spatial similarity metrics, including the Dice Similarity Coefficient (DSC), Hausdorff Distance, and Average Surface Distance. Additionally, a radiation oncologist conducted a retrospective clinical evaluation using a 3-point Likert scale. Statistical significance was examined using Wilcoxon signed-rank tests, Welch’s t-tests, and Mann-Whitney U test. Results: The sex-classification model attained an accuracy of 94.6% (AUC = 0.98, 95% CI: 0.96–0.99). Incorporation of predicted sex into the segmentation pipeline improved anatomical consistency of U-Net outputs. Mean DSC values were 0.91 (95% CI: 0.89–0.92) for prostate cases and 0.89 (95% CI: 0.87–0.91) for cervical cases, with no significant difference between groups (P=0.12). Surface distance metrics calculated on resampled isotropic voxels showed mean HD values of 3.4±0.8 mm and ASD of 1.2±0.3 mm, consistent with clinically acceptable accuracy. On clinical evaluation, 89.2% of contours were rated as excellent, while 9.1% required only minor adjustments. Automated segmentation reduced the average contouring time from 12.7±2.3 minutes manually to 4.3±0.9 minutes. Conclusion: The proposed sex-aware deep learning framework offers accurate, robust segmentation of the rectum in pelvic CT imaging by explicitly modeling sex-specific differences in anatomical characteristics. This physiologically informed approach enhances segmentation performance and supports reliable integration of AI-based delineation into radiotherapy workflows to improve both contouring efficiency and clinical consistency.

Abstract: Transposable elements (TEs) constitute a significant portion of eukaryotic genomes and play crucial roles in genome evolution, yet their diverse and complex sequences pose challenges for accurate classification. Existing tools often lack reliability in TE classification, limiting genomic analyses. Here we present TEclass2, a software employing a deep learning approach based on a linear Transformer architecture with k-mer tokenization and sequence-specific adaptations to classify TE consensus sequences into sixteen superfamilies. TEclass2 demonstrates improved classification performance and offers flexible model training on custom datasets. Accessible via a web interface with pre-trained models, TEclass2 facilitates rapid and reliable TE classification. These advancements provide a foundation for enhanced genomic annotation and support further bioinformatics research involving transposable elements.

Abstract: Mastitis is one of the major diseases affecting dairy cattle worldwide. Antibiotic therapy remains the most widely used treatment. However, its effectiveness has been com-promised due to the selection of antibiotic-resistant and biofilm-producing pathogenic bacteria. This promotes the search for alternatives that increase the antibacterial and anti-biofilm efficacy of antibiotics such ceftiofur (CFT). Biocompounds that, combined with antibiotics, exhibit these characteristics are gaining attention. Nisin (N) and chi-tosan (CH) may possess these properties. The aim of this study was to evaluate whether N+CFT and CH+CFT enhance the antibacterial activity of the antibiotic on Staphylococcus aureus associated with bovine mastitis, as well as its anti-biofilm effect. Two clinical isolates of S. aureus (AMC-43 and AMC-48) and the reference strain ATCC 27543 re-sistant to CFT were used. The combination of sub-inhibitory concentrations of N (320 µg/mL) and CH (400 µg/mL) with CFT (1, 2, 4, and 8 µg/mL) significantly reduced bac-terial growth (expressed in logarithmic units); however, the CH+CFT mixtures were the most efficient. Anti-biofilm activity was also detected, both in biofilm synthesis and in the reduction of living bacterial cells located inside this preformed structure. These results highlight N and CH as potential agents for the prevention or control of bovine mastitis.

Abstract: This paper details the development and validation of a novel computational framework to address critical brain tissue shifts encountered during image-guided neurosurgery for tumor resection. Utilizing Dynamic Data-Driven Non-Rigid Registration (NRR), this system integrates advanced distributed computing and machine learning paradigms to significantly enhance registration accuracy and speed. We present a robust methodology demonstrating the capability to deliver precise intra-operative image updates within demanding clinical timelines, thereby supporting more complete and safer tumor excisions. Key challenges inherent to real-time integration in the operating room are identified and discussed, providing foundational insights for future advancements in adaptive neurosurgical navigation.

Abstract: Background/Objectives:Calcified thoracic disc herniations are a formidable surgical challenge due to their proximity to the spinal cord and the heightened risk of iatrogenic neurological injury. Traditional anterior and posterolateral approaches, while effective in select scenarios, may not provide adequate exposure for large, centrally located calcified discs.. Methods:We conducted a retrospective case series of seven patients who underwent transpedicular-transdural thoracic discectomy for central or centrolateral calcified disc herniations at our institution over the past years. All patients were followed clinically for a minimum of three months postoperatively. Surgical technique and intraoperative nuances were also documented. Results: The transdural approach enabled direct access to the ventral thoracic spine, allowing for effective decompression of calcified herniations in all cases. No major intraoperative complications were reported, and all patients demonstrated clinical improvement or neurological stability at three-month follow-up. Technical aspects of the microsurgical approach were critical to minimizing risk.Conclusions: The transpedicular-transdural approach is a viable and effective surgical option for select cases of central or centrolateral calcified thoracic disc herniation. When executed with precise microsurgical technique, it offers safe decompression with favorable short-term outcomes.

Abstract: Arteriovenous malformations (AVMs) are fast‑flow vascular malformations formed by direct artery–to-vein shunts without an intervening capillary bed, increasing the risk of hemorrhage and organ‑specific damage. A synthesis of recent advances shows that AVMs arise from interplay between germline susceptibility (ENG, ACVRL1, SMAD4, RASA1, EPHB4), somatic mosaicism (KRAS, MAP2K1, PIK3CA), perturbed signaling (TGF‑β/BMP, Notch, VEGF, PI3K/AKT, RAS/MAPK), hemodynamic stress, and inflammation. Multi-modal imaging—digital subtraction angiography (DSA), MRI/MRA with perfusion and susceptibility sequences, CTA, Doppler ultrasound, and 3D rotational angiography, un-derpins diagnosis and risk stratification, while arterial spin labeling and 4D flow tech-niques refine hemodynamic assessment. Management is individualized and multidisci-plinary, combining endovascular embolization, microsurgical resection, and stereotactic radiosurgery (SRS); a non-surgical approach and monitoring remain reasonable for some asymptomatic AVMs. Device and technique innovations (detachable‑tip microcatheters, pressure‑cooker approaches, newer liquid embolic such as PHIL and Squid) have broad-ened candidacy, and precision‑medicine strategies including pathway‑targeted pharma-cotherapy are emerging for syndromic and somatic‑mutation–driven AVMs. Animal models and computational/radiomics tools increasingly guide hypothesis generation and treatment selection. We outline practical updates and future priorities: integrated ge-nomic–imaging risk scores, genotype‑informed medical therapy, rational hybrid se-quencing, and long‑term outcome standards focused on hemorrhage prevention and quality of life.

Abstract:

Background/Objectives: The equine proximal interphalangeal joint (PIPJ), while often considered a low-motion joint, contributes to flexion/extension of the digit and accommodates multi-planar movements: the adduction/abduction and the internal/external rotation of the middle phalanx in comparison to the proximal phalanx. Its collateral ligaments are primary stabilizers, restricting movements of the equine digit outside the sagittal plane, and their injury often leads to joint instability and degenerative joint disease. End-stage degenerative joint disease in the PIPJ could be managed with surgical arthrodesis of the joint. After surgical arthrodesis of the PIPJ, joint fusion cause loss of PIPJ extension in late stance of the stride during locomotion, which is accommodated by increased extension of the distal interphalangeal joint (DIPJ). To validate joint-preserving repair techniques after a collateral ligament injury in replacement of arthrodesis, a reliable method for assessing PIPJ extrasagittal biomechanics is required. The objective of this study was to develop and validate a three-point bending test setup to quantitatively characterize the mechanical behaviour of the equine PIPJ during abduction/adduction, simulating loading conditions that challenge the collateral ligaments of the joint. Methods: Seven equine PIPJ specimens (five forelimb and two hindlimb) were tested. A customised 3D-printed system was used to stabilise the joints in an MTS testing system. A three-point bending test was performed to evaluate the joint's mechanical response to the mediolateral plane bending. A marker-based video analysis system with a custom MATLAB script was used to calculate the joint abduction/adduction bending angle. Results: The setup provided reproducible load-displacement and time-angle data. The use of a fixed loading rod and custom 3D-printed system minimized slippage and rotation. The forelimb PIPJs exhibited significantly greater maximum medio-lateral bending angles compared to hindlimb joints. No significant differences were observed between fresh and frozen-thawed specimens. Conclusions: The developed three-point bending test setup provides a reliable and reproducible method for the in vitro evaluation of the biomechanical of the equine PIPJ. This study establishes a baseline for healthy joints and is suitable for the future comparative assessment of novel surgical techniques for collateral ligament repair.

Abstract: Objective: Predicting driver fatigue states is crucial for traffic safety. This study develops a deep learning model based on electroencephalogaphy (EEG) signals and multi-time-step historical data to predict the next time-step fatigue indicator percentage of eyelid closure (PERCLOS), while exploring the impact of different EEG features on prediction performance. Approach: A CTL-ResFNet model integrating CNN, Transformer Encoder, LSTM, and residual connections is proposed. Its effectiveness is validated through two experimental paradigms: Leave-One-Out Cross-Validation (LOOCV) and pretraining-finetuning, with comparisons against baseline models. Additionally, the performance of four EEG features differential entropy, α/β band power ratio, wavelet entropy, and Hurst exponent—is evaluated, using RMSE and MAE as metrics. Main Results: The combined input of EEG and PERCLOS significantly outperforms using PERCLOS alone validated by LSTM, CTL-ResFNet surpasses baseline models under both experimental paradigms, In LOOCV experiments, the α/β band power ratio performs best, whereas differential entropy excels in pretraining-finetuning. Significance: This study provides a high-performance hybrid deep learning framework for fatigue state prediction and reveals the applicability differences of EEG features across experimental paradigms, offering guidance for feature selection and model deployment in practical applications.

Abstract: Since the introduction of Advanced Footwear Technology (AFT) in 2017, numerous world records from 5 km to the marathon have been broken. Among these innovations, carbon-plated shoes have received particular attention. Studies report improvements of up to 4% in running economy (RE) and approximately 2% in performance. The rapid progression of performances has generated significant scientific interest; however, a clear understanding of the mechanisms driving the effectiveness of AFT remains limited. Despite widespread adoption and notable outcomes, neither researchers nor manufacturers can fully explain why design features such as increased stack height, embedded carbon fibre plates, and enhanced longitudinal bending stiffness are so effective, making optimisation of their benefits an ongoing challenge. This review summarises current knowledge on AFT and critically evaluates the biomechanical and physiological mechanisms underlying their effects on RE and performance. It also highlights the interaction between shoe design features and individual biomechanics, supporting evidence-based approaches to footwear selection and training strategies tailored to athletes’ needs. A clearer understanding of these mechanisms may provide valuable insights for researchers, coaches, and athletes, and help maximise the potential benefits of AFT.

Abstract: Porcine ear skin is widely used as a surrogate for human skin in in vitro permeation testing (IVPT) due to its structural and physiological similarities. However, commercial sourcing from the food industry often involves scalding, dehairing, and other processing steps that damage key layers of the skin barrier, potentially altering drug diffusion. Despite this, procurement and preparation methods are rarely standardized or reported in detail. In this study, we investigated how pork harvest processing affects the permeation of a topical 6% metformin lotion (ML) formulated with glycerol or glycerol plus propylene glycol. Permeation was evaluated using Franz diffusion cells over 48 hours, and skin structure was examined histologically. Processed porcine skin exhibited dramatically higher metformin flux — up to 20–60× greater within 2 hours — and cumulative diffusion (approximately 80% at 48 hours) compared to intact skin, which retained strong barrier function (< 30% total diffusion). Histology confirmed that pork harvest processing removed the stratum corneum and viable epidermis, facilitating rapid drug penetration. These findings demonstrate that commercially processed porcine skin is not an appropriate surrogate for human skin in permeation studies and highlight the critical need for standardized sourcing, documentation, and barrier integrity assessment in IVPT research.

Abstract: We report a case of hypertrophic pyloric stenosis in a two- months-old infant in a low-resource setting (Vezo Hospital, Madagascar). Since a surgical team was not available, we chose a conservative treatment with atropine and continuous enteral tube feeding, achieving a good clinical response and no need for surgery.

Abstract:

Understanding the drivers of population dynamics in long-lived, K-selected species such as bats is critical for conservation, particularly during vulnerable life-history stages like hibernation. We reviewed winter mortality records from more than 109 monitored hibernacula in Bulgaria. We found that significant die-offs (>30 individuals from at least three species) were recorded at only three sites, suggesting that such events are rare but potentially consequential for local populations. To reveal how habitat use shapes vulnerability during winter, we investigated hibernation preferences of cave-dwelling bats in Bulgaria. We further analysed the age structure of deceased Miniopterus schreibersii (Bonaparte, 1837) from Bulgaria’s largest hibernation colony following mortality events in winter 2022. Carcasses spanned a wide range of age classes, yet younger individuals predominated, consistent with the idea that early-life mortality represents a key demographic filter in bats. These findings emphasise the need for consistent mortality monitoring in bats, using standardised protocols that account for detection biases, scavenger removal, and site-specific variation. Such efforts are essential for clarifying the roles of environmental extremes, disease, and human disturbance in winter mortality.

Abstract: Introduction. Antibiotic resistance is a major global health challenge that disproportionately affects low-resource countries, particularly those in West Africa. E. coli, a major pathogen in childhood diarrhea, is both a prominent infectious agent and a reservoir of resistance genes, including resistance to last-resort antibiotics such as carbapenems. Methodology. The study focused on 98 clinical E. coli isolates collected from stool samples of patients in a hospital setting in Bamako. The analyses included screening for DEC-specific virulence genes, detection of resistance genes across various classes of antibiotics (e.g., beta-lactams, carbapenems, fluoroquinolones), and identification of class I, II, and III integrons. The bla_NDM gene was sequenced to identify mutations associated with carbapenem resistance. Results. Among the isolates, 85.7% carried at least one virulence gene. Of these, half involved co-infections, commonly combining EPEC, EAEC, and ETEC strains. Regarding antibiotic resistance, 94.9% of isolates harbored at least one resistance gene, and 50% were multidrug-resistant. The most frequently detected genes were bla_TEM, qnrS1, and aphA3. Class II integrons were significantly associated with multidrug resistance (p = 0.01). Sequencing of the bla_NDM gene revealed point mutations likely to affect protein function, suggesting an evolution toward increased resistance to carbapenems. Conclusion. The high prevalence of multidrug-resistant diarrheagenic E. coli strains in this study highlights the local antibiotic pressure and the serious health threat it represents. This study shows that the new β-lactamase bla_NDM gene has disseminated in the hospital environment of Bamako. It should be noted that this will become a major challenge for clinicians.

Abstract: Earlier research (2021–2024, Omicron and post Omicron dominant period) established a pronounced seasonal pattern in Italian COVID-19 deaths, with weekly mortality spiking in early winter after increases throughout summer and fall, with a subsequent decline in the late winter and spring periods. Using recent data spanning 53 weeks (late May 2024–late May 2025), a new investigation utilizing piecewise linear regression confirms the continuation of this trend. The data shows the cycle repeating precisely: mortality rose in early summer, peaked at the beginning of wither, and declined steadily through midwinter and spring, tracking the previous year’s trend. This persistence suggests a continued influence of the seasonal trend on COVID-19 mortality, necessitating further exploration into the driving factors.

Abstract: Background/Objectives: Salivary volatile organic compounds (VOCs) are promising noninvasive biomarkers for a wide range of diseases. While glandular saliva, secreted by salivary glands, is a relatively pure biofluid, whole saliva is a complex mixture containing oral microbiota, food debris, and desquamated epithelial cells. Therefore, a comprehensive comparison of the VOC profiles of these two types of saliva is essential to identify biologically relevant compounds. In this study, we aimed to establish a reliable method for VOC profiling from small saliva volumes and identify VOCs that reflect the biological differences between glandular and whole saliva. Methods: We developed a protocol combining MonoTrap extraction with dichloromethane, allowing the analysis of VOCs from just 100 µL of saliva. To address the issue of sampling-derived artifacts, we implemented a two-step blank analysis to systematically exclude compounds originating from the collection device. Results: Our analysis successfully identified 10 VOCs that were significantly higher in whole saliva than in glandular saliva and were confirmed not to be derived from the sampling devices. These compounds included isobutyric acid, isovaleric acid, 4-methylvaleric acid, 3-phenylpropionic acid, indole, skatole, methyl mercaptan, 1-propanol, d-valerolactam, and acetaldehyde. Most of these compounds originate from the metabolic activities of the oral microbiome, suggesting that the distinct VOC profile of whole saliva is predominantly influenced by microbial activity. Conclusions: Our findings demonstrated the effectiveness of this method for identifying biologically relevant VOCs from relatively small sample volumes. The identified VOC profiles highlight the potential of whole saliva as a valuable non-invasive tool for assessing oral health and serve as a solid foundation for future research into clinical applications.