OETDK.

Introduction: Adrenal scintigraphy using [131I]6ß-iodomethyl-19-norcholesterol has beenavailable since 1975, primarily for the diagnosis of hyperaldosteronism helping to distinguishbetween unilateral adenoma and bilateral hyperplasia. First line treatment of ACTH-independent form of Cushing syndrome is surgery after localization of the cortisol-producingadrenal tumor. Case report: A 31-year-old woman has been referred to the endocrine clinic with suspectedCushing syndrome in 2020. Tonsillectomy, lactose intolerance, allergic rhinitis and epilepsywere present in her medical history. She complained of weight gain (30 kg during one year),secondary amenorrhea, and elevated blood pressure. Physical examination revealed roundface, central obesity, oedema on the legs and striae on the skin of the abdomen. Laboratoryexaminations proved hypercortisolism with suppressed ACTH level. We diagnosed her withhypertension, osteoporosis and impaired glucose tolerance as complications of Cushingsyndrome. Adrenal CT showed bilateral adrenal adenomas, with 30 mm diameter on the rightside and 17 mm on the left side. Iodo-methyl-norcholesterol SPECT/CT proved cortisol-secreting adenoma in the right adrenal gland. Coronavirus pandemic-driven restrictionsprevented immediate surgery; this was postponed to an uncertain date in the future. Wedecided to start metyrapone treatment until surgery, which partially controlled herhypercortisolism. Laparoscopic adrenalectomy on the right side was performed one year later.Histology confirmed cortisol secreting adenoma. After operation, the patient had transientadrenal insufficiency for a few weeks, then normal cortisol levels resumed spontaneously asexpected. Clinical features of Cushing syndrome gradually disappeared; her body weightdecreased from 126 kg to 78 kg.Conclusion: In Cushing syndrome with bilateral adrenal lesions, it may be challenging tolateralize the hormone source. Iodo-methyl norcholesterol SPECT/CT is a feasible non-invasive method to localize the cortisol-secreting adrenal adenoma preoperatively.

Témavezető: Erdei Annamaria

IntroductionThe developing brain exhibits high plasticity and ongoing myelination, which may increase its susceptibility to environmental influences. Although diagnostic ultrasound (US) is widely used and generally considered safe, earlier findings suggest that it may influence the dendritic arborization and spine morphology of CA1 neurons. Ultrasound exposure has also been shown to exert a preventive effect on age-related hippocampal atrophy; particularly influencing the development of white matter tracts.ObjectiveThis study focuses on whether repeated diagnostic-range ultrasound alters the morphology of selected key white matter tracts of the central nervous system connectome, specifically the fimbria (fimbria-fornix pathway) and the corpus callosum, including its genu.MethodsTime-mated pregnant CD1 mice, were exposed to 10-min sessions of diagnostic-type ultrasound, at embryonic day 18.5 and then once weekly during the first four postnatal weeks (five sessions total). Surviving offspring were followed until 1 year of age, when brains were collected and imaged using high-resolution SkyScan 1272 micro-computed tomography (micro-CT) system. The longest diameter of fimbria and the thickness of the corpus callosum and its genu were measured and compared with age-matched sham-handled control.ResultsUltrasound-treated mice showed a significantly greater fimbria diameter (p = 0.0025, control 566.22 ± 53.09, US treated 725.81 ± 75.72), whereas the thickness of the corpus callosum and its genu did not differ significantly from controls (p = 0.18, control 279.30 ± 40.20, US treated 328.29 ± 57.75 and p = 0.52, control 384.89 ± 50.07, US treated 424.33 ± 57.12, respectively).ConclusionsRepeated diagnostic-range ultrasound did not induce generalized white-matter enlargement but produced a specific increase in fimbrial size. Because fimbria carries hippocampal outputs, this finding suggests that some hippocampal efferent pathways may be more sensitive to ultrasound during development than interhemispheric callosal fibers. Together with our earlier observation on magnetic resonance imaging (MRI) in 1-month-old mice of ultrasound-related hippocampal changes, these data demonstrate that the structural effect may be long-lasting. The enlargement in the fimbria size could potentially enhance neuronal activity and hence improve memory and learning.

Témavezető: Dr. Papp Tamás

Introduction:Insulin and IGF-1 are important for metabolic regulation and homeostasis.Anotherimportant regulator of metabolism is pituitary adenylate cyclase-activating polypeptide(PACAP-38).It plays a multifunctional role in various systems in the body, including the pancreas,where it serves as a pleiotropic endocrine regulator of insulin release.While the roles of these hormones in the blood and tissues are well researched,their functions in saliva have yet to be investigated.Aim of the study:By measuring salivary PACAP-38 levels in children and adolescents,we can utilize these findings to investigate the role of PACAP-38 in insulin and IGF-1 homeostasis,as well as its relationship with BMI percentile markers, with the aim of gaining new insights into their potential roles in non-invasive metabolic monitoring.Material and Methods:We analysed saliva samples from 48 young patients following orthodontic anamnesis and measured insulin,PACAP-38,and IGF-1 levels via radioimmunoassay(RIA).Height(m)and mass(kg)were measured to calculate BMI (weight divided by height squared, kg/m²).BMI percentiles were then determined for individuals aged 2–19 years using CDC growth charts.Results:Pateints had the following mean values:BMI percentile of 41.61±30.94%,insulin level of 6.990±5.175µIU/ml,PACAP-38 level of 37.74±29.17fmol/ml and IGF-1 level of 0.9231±0.2877 ng/ml.Pearson correlation analysis showed positive correlation between BMI percentile and both insulin(r=0.3212;p=0.0260)and IGF-1 levels(r=0.3357;p=0.0197).No correlation was found between BMI percentile and PACAP-38(r=–0.01729;p=0.9071).In contrast,PACAP-38 showed significant negative correlation with insulin(r=–0.4508;p=0.0013)and IGF-1(r=–0.4398;p=0.0018) levels in saliva samples.Positive correlation was seen between insulin and IGF-1(r=0.3844;p=0.0070).Discussion:The findings suggest that high insulin and IGF-1 levels are observed in patients with higher BMI Percentile.Negative correlation was observed between PACAP-38 levels and insulin,IGF-1,highlighting its potential role in metabolic control.These findings suggest saliva-based assessment could serve as a non-invasive tool for monitoring metabolic status in pediatricpopulations.Funding:The work was supported by GINOP-2.3.3-15-2016-00021,GINOP2.3.4- 15-2016- 00002,GINOP-2.3.1-20-2020-00004 and TKP2021-EGA-18 projects.

Témavezető: Dr. Bombicz Mariann és Görögh Panna

IntroductionPositron emission tomography–computed tomography (PET-CT), a nuclear medicine imaging technique is preferably used for lung tumor detection. Nevertheless, manual image analysis may be highly time-consuming. Deep learning-based segmentation algorithms offer automated alternatives; however, few studies have compared their performance using radiomic parameters derived from PET-CT images.ObjectiveThis research addresses this gap by comparing the performance of two deep learning-based segmentation algorithms No-new-net (nnU-Net) and Research Consortium for Medical Image Analysis Lung Cancer FDG-PET/CT v1.5 (RECOMIA) for solitary lung tumor segmentation, using eight groups of radiomic parameters calculated by PyRadiomics software.MethodsPET-CT images of 221 patients at the University of Debrecen’s Hospital were collected and anonymized. Predicted tumor masks were compared with the manual segmentations using the Wilcoxon signed-rank test. Combined p-values were calculated with Brown’s method, and absolute values of feature correlations were summarized to assess agreement between manual and automatic segmentations.ResultsFor nnU-Net and RECOMIA, the raw combined p-values were 0.481 and 4.708 x 10-24. After false discovery rate (FDR) adjustment, the combined p-values were 0.91 and 4.708 x 10-24 and after Bonferroni correction, they were 1 and 4.708 x 10-24. The summed absolute correlation values were 37.259 for nnU-Net and 16.358 for RECOMIA.ConclusionsnnU-Net demonstrated superior segmentation accuracy compared to RECOMIA. However, its greater computational and training demands may limit its clinical adaptation. Despite lower accuracy, RECOMIA remains advantageous for rapid, efficient and user-friendly implementation. However, many recent lightweight pretrained nnU-Net models such as TotalSegmentator and MOOSE, as well as streamlined 3D Slicer module integration with simplified deployment may narrow the accessibility gap between RECOMIA and locally deployed nnU-Net segmentation algorithms, making the nnU-Net increasingly accessible in both research and clinical settings. We suggest ongoing research applying this method to other algorithms to assist in the selection of specific clinical and research needs.

Témavezető: Dr. Opposits Gábor

Cardiovascular diseases (CVDs) remain the leading cause of death worldwide. Early and accurate diagnosis is essential for improving patient outcomes and lowering mortality rates. Recent literature highlights mitochondrial dysfunction as a promising target for developing PET radiotracers. Generally, these tracers rely on cyclotron-produced radioisotopes. To address this limitation, this study aimed to synthesise an alternative agent, by conjugating a mitochondria-targeting peptide with a generator-produced radionuclide to evaluate its in vivo efficacy. SS-31 peptide was conjugated with NODAGA chelator and labelled with Gallium-68 to produce [68Ga]Ga-NODAGA-SS-31. The conjugation process was optimised to deliver a high radiochemical purity through the adjustment of temperature, concentration of the precursor and reaction time. Thin layer chromatography (TLC) was used for quality control, and serum stability was tested by incubation in rat serum for 120min at 37 °C. After synthesis, the radiotracer was intravenously injected into healthy rats (n = 9, injected dose = 30 ± 4 MBq/kg of body weight). The evaluation of the 120-min dynamic dataset enabled the determination of optimal post-injection time window to be used for static imaging. Ex vivo biodistribution studies were performed and comparison with the most common tracer FDG was conducted to evaluate uptake intensity.Radiochemical purity exceeded 99% and >95% of the molecule remained intact for a two-hour period, showcasing high stability. PET imaging showed accumulation of the radiotracer in the mitochondria-rich heart. Rapid renal clearance was observed, confirming the hydrophilic nature of the molecule. Comparative studies revealed that the signal intensity of [68Ga]Ga-NODAGA-SS-31 was lower than that of FDG. While it may not match FDG in sensitivity, the novel radiotracer offers advantages in not depending on an on-site cyclotron and for its simplified synthesis.The newly synthesised [68Ga]Ga-NODAGA-SS-31 radiotracer exhibited excellent radiochemical purity and demonstrated efficient tissue uptake; however, its structure may still require further optimisation to enhance imaging performance. Despite this, the approach facilitates mitochondrial imaging using generator-produced isotopes, providing a viable alternative to cyclotron-dependent tracers. Therefore, this tracer serves as a promising lead compound for non-invasive assessment of mitochondrial function.

Témavezető: Dr. Arató Viktória Zsófia

[177Lu]Lu-DOTATATE peptide receptor radionuclide therapy is an established treatment for advanced somatostatin-receptor-positive neuroendocrine tumours. The red bone marrow and kidneys are the principal dose-limiting organs. This retrospective study evaluated cycle-specific (subacute) and cumulative hematologic toxicity in 120 patients treated with [177Lu]Lu-DOTATATE between May 2022 and October 2025 and examined correlations between absorbed doses and peripheral blood cell counts.Complete blood counts (RBC [T/L], WBC [G/L], platelets [G/L]) were obtained ≤2 weeks before and ≤13 days after each cycle. Bone marrow and whole-body effective doses were calculated using quantitative post-therapy SPECT/CT and standard kinetic models. Statistical analysis was performed with IBM SPSS v.31.Cycle-specific analysis showed a small but significant mean RBC decrease (one-sample t-test, p = 0.039); WBC and platelets exhibited no significant subacute change (both p > 0.1). After Hochberg-corrected Spearman correlations, no dose metric significantly correlated with cycle-specific blood count changes.Cumulative analysis from pre-first-cycle baseline revealed significant time-dependent decreases in all three cell lines (p < 0.001), with platelets showing the strongest correlation with elapsed time. Cumulative whole-body effective dose correlated significantly with RBC and platelet decreases (both p < 0.001), but not with WBC (p > 0.1). Correlations with cumulative bone marrow absorbed dose were consistently weaker.[177Lu]Lu-DOTATATE therapy causes minimal subacute hematologic effects and no detectable short-term dose–response relationship. Over multiple cycles, significant cumulative toxicity occurs, predominantly affecting platelets and red blood cells. Cumulative whole-body effective dose is a stronger predictor of long-term hematologic toxicity than bone marrow absorbed dose alone, highlighting the need for cumulative dosimetry and prolonged blood count monitoring.

Témavezető: Dr. Barna Sándor Kristóf

Introduction: Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive treatment for severe aortic stenosis. Self-expanding valves may shift between the pre-release and post-release phases, making accurate positioning essential to prevent complications such as paravalvular leakage or coronary obstruction.Aim: To evaluate the deployment motion of two self-expanding valves Evolut (EV) and Navitor (NAV), and determine whether valve type or patient anatomy affects final implantation depth.Methods: A retrospective analysis was conducted on 90 TAVI procedures, including 45 EV and 45 NAV cases performed at the University of Debrecen between March 2021 and May 2025. Data were obtained from angiographic and computed tomography (CT) records. Using the Cardiac X-ray Application, the distance between the inflow plane of the valve and both the non-coronary cusp (NCC) and left-coronary cusp (LCC) was measured before and after valve deployment to assess implantation depth and displacement. Morphometric parameters of the left ventricular outflow tract (LVOT), annulus, sinotubular junction (STJ), ascending aorta (AAo), angle of the annulus, and calcification scores of the NCC, LCC, and right-coronary cusp (RCC) were collected from the CT reports. Statistical analyses including t-test, Mann–Whitney U-test, Absolute Sum (AS), and Annulus–LVOT (An-LVOT) were calculated.Results: Both valves migrated toward the LVOT during release. EV moved significantly toward the LVOT on the NCC side (mean 1.48 ± 3.45 mm; p = 0.006), but not on the LCC side (mean 0.80 ± 3.33 mm; p = 0.112). NAV moved significantly toward the LVOT on both the NCC (mean 1.21 ± 3.78 mm; p = 0.037) and LCC (mean 1.38 ± 3.18 mm; p = 0.006). Predictors of valve displacement >5.5 mm in the NAV group included a more negative An-LVOT value indicating a relatively larger LVOT compared to the annulus, and lower annular eccentricity. No significant predictors were observed for EV.Conclusion: Valve migration during TAVI is a measurable and device-dependent phenomenon. NAV demonstrated significant bilateral displacement toward the LVOT, while EV showed motion on the NCC side, with the LCC acting as a hinge point. Despite these differences, both valves achieved similar final implantation depths. Anatomical factors, particularly An-LVOT and annular eccentricity, appear to influence displacement and should be considered during pre-procedural planning to ensure accurate valve positioning.

Témavezető: Dr. Kertész Attila