In November 2021, a 10-year-old female neutered German Shepherd dog, regularly vaccinated and with no known previous systemic diseases, was presented to a private veterinary practice in the Province of Bergamo (Lombardy region, Northern Italy) for clinical evaluation in the context of ongoing phenylpropanolamine (Propalin^®) therapy (1 mg/kg) for urinary incontinence. The dog was owned, kept under controlled conditions, and had access to the outdoor environment. Clinical examination revealed a firm, non-painful swelling of the right forelimb in the right front paw. The peripheral lymph nodes were palpably normal. Hematological, serum biochemical, and urinalysis tests as well as a complete X-ray of lateral-lateral thorax and lateral-medial right radio projections, ultrasound examinations of the abdomen, and total body computerized tomography (CT) were performed. Jamshidi needle biopsy samples from the osseous lesions and cystocentesis urine were collected for histological and microbiological investigations. All the collected samples were sent to Veterinary Analysis Laboratory MyLav La Vallonea (Milan, Italy) for microbiological examination. Based on microbiological and histopathological findings, antifungal therapy was initiated with itraconazole (11 mg/kg, orally, once daily). Pantoprazole (1 mg/kg, orally, once daily) was administered as gastrointestinal protection. Given the need for prolonged azole therapy, hepatic support with Glutamax^® and Besame^® was concurrently prescribed as hepatoprotective supplementation, administered according to the manufacturer's instructions. Itraconazole was selected as first-line therapy for suspected systemic fungal infection, in accordance with standard clinical practice, despite the fact that definitive identification of the causative pathogen was not yet available at the time treatment was initiated. Therapeutic follow-up was performed by collecting urine samples every two months. Approximately 190 days after initiation of therapy, the clinical condition worsened, and the dog died. Owner consent for necropsy was not granted. The chronological sequence of clinical procedures, diagnostic investigations, microbiological analyses, and therapeutic interventions performed during the case is summarized in Figure 1.

A comprehensive blood panel was conducted to assess the dog's overall health and to monitor any signs of infection or inflammation. This included a complete blood count (CBC) and the biochemical tests, including the comprehensive metabolic panel (CMP). Urinalysis was conducted to assess kidney function and detect any signs of infection. This included urine chemistry and physical analysis to evaluate color, appearance, specific gravity, pH, and the presence of proteins, glucose, urobilinogen, bilirubin, ketones, erythrocytes, and hemoglobin. Additionally, urine sediment analysis was performed to identify leukocytes, erythrocytes, casts, epithelial cells, crystals, and bacteria.

Preliminary conventional radiography of the right forelimb and thorax was performed using multiple projections (craniocaudal, mediolateral, oblique) to optimally visualize the anatomical structures of interest. The examination allowed for an assessment of the bones, joints, soft tissues, and any morphological or densitometric alterations. A high-frequency linear transducer was employed to perform a comprehensive abdominal ultrasound examination, utilizing a subcostal and suprapubic acoustic window. Multiple imaging planes were acquired during both the filling and emptying phases of the urinary bladder to provide a detailed assessment of the morphology and function of the lower urinary tract organs. Subsequently, to evaluate urinary incontinence and a recently detected lump on the right forelimb, the patient received a thoracic, abdominal, and forelimb computed tomography (CT) scan at the Orobica Veterinary Clinic.

Tissue samples were obtained from the target lesions using a Jamshidi needle biopsy and subjected to histological and microbiological analysis. Urine samples were collected via cystocentesis and analyzed through microbiological culture and cytological evaluation of the urinary sediment. Biopsy tissue and centrifuged urine sediment were air-dried and subjected to Romanowsky staining (May-Grünwald-Giemsa, Merck KGaA, Darmstadt, Germany) for cytological analysis. Biopsy specimens for histopathological analysis were processed utilizing standard histological techniques, which encompassed fixation in 10% buffered formalin, dehydration via a graded alcohol series, clearing with xylene, paraffin embedding, microtomy at 4 μm, and staining with hematoxylin and eosin (H&E), Grocott-Gomori's methenamine silver (GMS), and Periodic acid-Schiff (PAS) stains (12).

Tissue and urine sediment specimens were inoculated onto selective and differential culture media, including Sabouraud Dextrose Agar with Chloramphenicol (SABC), Columbia Sheep Blood Agar, and Brilliance UTI Clarity Agar (Thermo Fisher Scientific-Oxoid, UK). The inoculated plates were incubated under aerobic conditions at 37°C for a period of seven days. Fungal colonies that developed were subjected to a meticulous phenotypic characterization, including macroscopic and microscopic examination, to determine their genus-level identification (13).

Molecular identification of fungal isolates was accomplished through a polyphasic approach involving PCR amplification and Sanger sequencing of the β-tubulin, calmodulin, and nuclear ITS regions. Briefly, DNA extraction from all pure microbial cultures collected and from biopsy samples was carried out by automated DNA extractor (QIAsymphony SP instrument, Qiagen, Milan, Italy), following the manufacturer's instructions. The obtained DNA extracts were processed by PCR to amplify the three different genes, including β-tubulin, calmodulin, and nuclear ITS region (Table 1).

The PCR was done in a final reaction volume of 25 μL containing 12.5 μL of 2x Platinum^TM SuperFi^TM PCR Master Mix (Invitrogen^TM, Milan, Italy), 0.3 μM of each primer, and 2 μL of DNA template; the reaction was brought to the final volume with PCR water. The PCR products were separated by capillary gel electrophoresis using the QIAxcel Advanced (Qiagen, Milan, Italy) and represented as electropherograms by the QIAxcel ScreenGel Software, version 1.5 (Qiagen). Direct sequencing was performed on the PCR products using the BigDye^TM Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems^TM, Milan, Italy) and a SeqStudio^TM 8 Flex Genetic Analyzer (Applied Biosystems^TM, Milan, Italy). The acquired sequences were analyzed using the Sequencing Analysis Software v7.0 (Applied Biosystems^TM, Milan, Italy). The sequences were used for the taxonomic recognition of the fungal isolate by BLASTn search of similar sequences deposited in the GenBank databases. These reference sequences included both clinical and environmental isolates of Penicillium spp., some of which were previously reported in canine and human infections. All accession numbers are provided in Supplementary Figures S9–S11, allowing full traceability of the comparative data. Evolutionary analyses were conducted in MEGA11 (17–19).

Hematological analysis revealed a complete blood count (CBC) within reference intervals, except for a mild leukocytosis characterized by a relative monocytosis. Biochemical profiling demonstrated serum concentrations of electrolytes, renal function markers, and hepatic enzymes within normal reference ranges (20, 21). A slight elevation in total serum protein was noted. Urinalysis indicated a specific gravity below the reference range (1015) but was otherwise unremarkable (Table 2). Microscopic examination of the urine sediment revealed no significant cellular casts or crystals, and the quantitative analysis of formed elements was within normal limits.

Radiographic imaging revealed a constellation of findings initially interpreted as suggestive of a multifocal neoplastic process. A focal osseous lesion measuring approximately 2 cm was identified in the diaphysis of the right radius (Supplementary Figure S1). The lesion showed a proliferative pattern involving both cortical and medullary bone. Thoracic radiographs demonstrated a diffuse reticular pattern consistent with non-specific interstitial lung disease, together with a focal area of increased opacity compatible with a pulmonary nodule or mass (Supplementary Figure S2). At this stage, these findings were considered non-specific, and differentials included neoplastic, inflammatory, or infectious processes. Abdominal ultrasonography did not reveal abnormalities that could explain the reported urinary incontinence. Computed tomography (CT) allowed further characterization of the skeletal and soft tissue lesions. The radial lesion displayed aggressive features, including mixed lytic and sclerotic changes, marked periosteal reaction, and surrounding soft tissue edema (Supplementary Figure S3). Based on imaging alone, primary or secondary bone neoplasia was considered the leading differential diagnosis. Additionally, CT identified multiple lytic lesions affecting the thoracic spine, particularly the T8 and T9 vertebrae (Supplementary Figure S4), associated with a paravertebral soft tissue mass and focal bone destruction. These findings raised concern for metastatic disease or a primary spinal neoplasm, while infectious osteomyelitis was also considered among the differential diagnoses. Increased soft tissue density within the left frontal sinus (Supplementary Figure S5) represented a non-specific finding, with differential diagnoses including sinusitis, inflammatory disease, or fungal infection. Moderate pre-scapular and axillary lymphadenopathy was interpreted as potentially reactive; however, in the context of multifocal skeletal lesions, a neoplastic etiology was also considered. Degenerative changes of the vertebral column, including intervertebral disc space narrowing, sclerosis, and vacuum phenomenon, were consistent with age-related alterations.

Fine-needle aspiration cytology and biopsy specimens were obtained directly from the osseous lesion of the right radius using a Jamshidi needle. No soft tissue involvement was observed during imaging or sampling. Cytological evaluation revealed a mixed inflammatory infiltrate composed predominantly of neutrophils and macrophages, with evidence of fungal hyphae. Microscopically there are septate hyaline hyphae that measure 1.5–3 μm in diameter and from which mono- and bi-verticillate conidiophores arise showing flask-shaped phialides with tapered tip, no metulae and oval/rounded conidia in short chains. Histopathological examination of the bone tissue confirmed chronic fibroplasia and fibrosis associated with mixed inflammation. PAS and GMS staining highlighted fungal hyphae within necrotic areas. Both stains revealed the presence of fungal hyphae, characterized by rounded and elongated structures within the necrotic debris (Supplementary Figure S6).

No bacterial growth was observed on routine culture media. However, after 4–5 days of incubation, a pure fungal culture emerged on SABC medium, while no growth was detected on blood agar. The fungal colony exhibited a characteristic white to grayish appearance with a tendency to develop aerial mycelia (Supplementary Figure S7). Colonies showed a velvety texture and were initially flat and whitish, becoming progressively raised and grayish-brown with age. The reverse surface exhibited pale pigmentation, ranging from whitish to light cinnamon with cream-colored shades. Occasional small hyaline droplets were observed on the colony surface. Microscopic examination revealed the presence of septate hyphae and conidiophores consistent with the genus Penicillium (Supplementary Figures S8A, B). Subsequent urine cultures, collected at two-month intervals during therapy, consistently yielded fungal isolates with identical morphological characteristics to the original Penicillium species, suggesting persistent infection. Molecular identification using PCR amplification and Sanger sequencing of the β-tubulin, calmodulin, and nuclear ITS regions confirmed the persistent presence of the same Penicillium species in these cultures.

The identification of the fungus was confirmed molecularly by PCR amplification and sequencing of β-tubulin (GenBank accession number PV335504), calmodulin (GenBank accession number PV288913), and nuclear ITS region (GenBank accession number PV133825). The same fungus was identified in other biological samples (biopsy and urine samples). Sequencing of β-tubulin, calmodulin, and ITS regions from both urine and bone biopsy samples yielded identical nucleotide sequences across all loci, confirming the persistence of the same P. labradorum strain in different tissues. Briefly, all sequences were queried against the GenBank database. The analysis showed 99.48% nucleotide identity in ITS to Penicillium parvum strain NRRL 2095 (GenBank accession number AF033460.1). The analysis of the calmodulin gene displays an identity of 97% nucleotide identity to P. labradoris strain DI19-20 (GenBank accession number MK887899.1). A nucleotide identity of 100% was detected in the β-tubulin to Penicillium labradorum tt14509 (GenBank accession number ON646037.1). Phylogenetic analyses based on three distinct gene targets demonstrated that the fungal strain isolated from the dog consistently clustered with P. labradorum tt14509 (GenBank accession number ON646037.1), P. labradoris CBS 145775 (GenBank accession number NR173380), and Penicillium striatisporum NRRL 26877 (GenBank accession number NR173380) in the internal transcribed spacer (ITS) region (Supplementary Figure S9). Analysis of the β-tubulin gene revealed that the strain grouped closely with P. labradorum tt14509, while phylogenetic reconstruction based on the calmodulin gene also supported clustering with P. labradorum tt14509 and P. labradoris CBS 145775 (Supplementary Figures S10, S11).