Fulminant Aeromonas caviae bacteremia with migratory myalgia: a case report of a CTX-M-3-producing multidrug-resistant strain Authors: Lijing Guo, Ziyu Zhao, Kuo Li, Weichen Huang, Ningni Ma, Xinyi Jin, Baihui Gui, Qianhui Li, Chang Li, Chaogui Tang, Hongmei Zhao, Jingjing Huang, Sumin Gao, Ning Lin Published: 2026 ABSTRACT AbstractBackgroundAeromonas caviae is an emerging waterborne pathogen of increasing clinical concern, particularly regarding antimicrobial resistance. While extended-spectrum β-lactamase (ESBL) genes such as blaCTX−M are well-documented in Enterobacterales, they remain uncommon in A. caviae. We report a rare case of fulminant A. caviae bacteremia with atypical clinical manifestations and a complex genomic resistome.Case presentationA 71-year-old man without formal immunodeficiency but with predisposing vulnerabilities (advanced age, prior subtotal gastrectomy) presented with fulminant bacteremia progressing to septic shock and unusual migratory myalgia following a minor paddy field injury. Initial empirical ertapenem therapy at a referring hospital was clinically ineffective. Blood cultures isolated A. caviae with multidrug resistance. Whole-genome sequencing (WGS)–based phylogenomic analysis confirmed species identity (clustering with A. caviae type strain NCTC 12244; bootstrap 100%; average nucleotide identity 98.09%), distinguishing it from closely related species (A. dhakensis, A. hydrophila, A. veronii). WGS identified ESBL genes blaCTX−M−3, blaTEM−1B, and blaOXA−10, alongside aac(6’)-Ib-cr and tet(A). The isolate remained susceptible to amikacin and representative carbapenems (imipenem, meropenem). Therapy was successfully switched to biapenem, achieving rapid clinical recovery and microbiological clearance.ConclusionsA. caviae, an environmental organism, can harbor clinically significant ESBL genes and cause life-threatening infections even in individuals without formal immunodeficiency. Ertapenem failure despite carbapenem class susceptibility suggests that pharmacokinetic/pharmacodynamic factors may contribute to treatment failure in critically ill patients with hypoalbuminemia and acute kidney injury, although direct evidence is lacking. These findings highlight the need for integrated One Health surveillance to monitor antimicrobial resistance gene flow between environmental and human populations, though confirmation through prospective multicentre studies is warranted.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-026-13808-3.Keywords: Aeromonas caviae, Bacteremia, Multidrug resistance, blaCTX−M−3, Whole-genome sequencing, Case report Introduction Aeromonas is a genetically diverse genus of Gram-negative bacilli ubiquitous in aquatic environments [1, 2], with 42 species described [3]. However, human clinical infections are predominantly caused by four major species: Aeromonas caviae, A. dhakensis, A. veronii, and A. hydrophila, which together account for over 95% of isolates [4]. A. caviae is notably prevalent among these, accounting for 37.26% of clinical isolates [5]. Concurrently, antimicrobial resistance in Aeromonas poses a serious therapeutic challenge. Intrinsic resistance to ampicillin is common, and the emergence of acquired resistance to broad-spectrum cephalosporins and fluoroquinolones is rising [6]. A key mechanism is the acquisition of extended-spectrum β-lactamase (ESBL) genes, particularly those of the CTX-M family [7]. CTX-M-type ESBLs, especially those of the CTX-M-1 group such as CTX-M-3 and CTX-M-15, are frequently plasmid-encoded and have disseminated globally among Enterobacterales [8]. However, their occurrence in Aeromonas, and specifically in A. caviae, remains relatively uncommon. To date, only a few reports have documented CTX-M-3-producing Aeromonas isolates, including an A. caviae strain from a pneumonia patient in China that co-harbored MOX-4 and TEM-1 [9], and an A. salmonicida strain selected during medicinal leech therapy in Switzerland [10]. Herein, we report a rare case of fulminant bacteremia caused by a multidrug-resistant A. caviae strain in an elderly male without formal immunodeficiency following minor trauma in a paddy field. The clinical course was notable for rapid progression to septic shock accompanied by severe migratory myalgia—an unusual manifestation of Aeromonas infection. Whole-genome sequencing further characterized the isolate’s resistome, revealing the presence of the ESBL gene blaCTX−M−3, which remains uncommon in this species. Case presentation Patient information and clinical history A 71-year-old male patient was admitted to our hospital on Hospital Day 1 (in late summer), presenting with a 5-day history of persistent fever. The clinical course began five days earlier (Hospital Day 1), when the patient fell in a muddy paddy field during agricultural work, sustaining a traumatic laceration (approximately 1.5 cm in length) on the anterior aspect of his left knee, with active bleeding (Supplementary Fig. 2). He subsequently sought care at a local external hospital for dizziness, during which he developed chills and a high-grade fever peaking at 39.5 ℃, accompanied by generalized fatigue. He denied any respiratory symptoms (cough, sputum production) or gastrointestinal disturbances (diarrhea, nausea). His past medical history was notable for atrial fibrillation and hyperuricemia spanning over two years, both managed with irregular medication compliance. His surgical history included a subtotal gastrectomy and appendectomy performed in October 2023 secondary to gastric perforation and appendicitis. He explicitly denied any history of chronic systemic diseases, including hypertension, diabetes mellitus, or coronary artery disease, and was not on any antihypertensive therapy. His social history included alcohol consumption and a smoking history of approximately 20 cigarettes per day for 40 years. His family history was unremarkable. Diagnostic assessment The patient’s initial diagnostic workup commenced at the external hospital, where preliminary blood cultures flagged positive for Gram-negative bacilli. Initial laboratory results revealed markedly elevated inflammatory markers (Table 1), and a chest computed tomography (CT) scan demonstrated bilateral pulmonary infiltrates with concomitant pleural effusion. Despite initiation of ertapenem, the patient exhibited a suboptimal clinical response, with recurrent high-grade fevers and progression to acute respiratory failure by Hospital Day 2. He was briefly managed in the intensive care unit at the referring hospital, receiving non-invasive ventilation, central venous catheterization, and supportive care. During this period, he developed severe paroxysmal, migratory pain affecting the lumbar, sacrococcygeal, right gluteal, and right lower limb regions, notably sparing the left knee injury site. Suspecting localized necrotizing fasciitis or severe deep tissue infection refractory to standard analgesia, the patient was transferred to our institution’s emergency department and admitted with a primary diagnosis of sepsis. Table 1Serial laboratory data of the patient with A. caviae bacteremiaTime point & ParameterHospital Day 1Hospital Day 4Hospital Day 5Hospital Day 9Hospital Day 13Reference RangeInflammatory Markers WBC (×10⁹/L)2.52 ↓10.74 ↑12.81 ↑11.62 ↑8.373.5–9.5 Neutrophils (%)89.3 ↑81.7 ↑92.8 ↑81.6 ↑80.1 ↑40–75 PCT (ng/mL)9.13 ↑3.77 ↑3.509 ↑0.369 ↓0.145 ↓< 0.5 IL-6 (ng/mL)0.497 ↑0.155 ↑0.070 ↑NA0.143 ↑< 0.007 hs-CRP (mg/L)NA> 5 ↑> 5 ↑63.34 ↑77.63 ↑< 3Liver Function TP (g/L)NANANA51.359.165.0–85.0 ALB (g/L)NANANA29.135.140.0–55.0 TBIL (µmol/L)51.2 ↑NA80.4 ↑38.1 ↑40.6 ↑4–20 ALT (U/L)174.4 ↑NA439.0 ↑122.7 ↑53.4 ↑9–50 AST (U/L)NANA532.0 ↑71.5 ↑21.1 ↓15–40Kidney Function CREA (µmol/L)146 ↑NA141.0 ↑94.4 ↓79.8 ↓57–111 UA (µmol/L)644.8 ↑NA681.9 ↑548.0 ↑269.0 ↓202.3–416.5 eGFR (mL/min/1.73 m²)43 ↓NA45 ↓69 ↓82 ↓≥ 90Note: Data are presented as patient’s laboratory values. Abbreviations: TP, total protein; ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CREA, creatinine; hs-CRP, high-sensitivity C-reactive protein; IL-6, interleukin-6, eGFR, estimated glomerular filtration rate; NA, not available, NA values for specific inflammatory markers were due to tests not being routinely performed at the external hospital or at certain time points during the recovery phase when clinical stabilization was observed; PCT, procalcitonin; TBIL, total bilirubin; UA, uric acid; WBC, white blood cell count. Symbols: ↑, above the reference range; ↓, below the reference range. The hs-CRP values on Hospital Day 4–5 were reported as > 5 mg/L because they exceeded the upper detection limit of the high-sensitivity assay employed at that time. Subsequent measurements (Hospital Day 9–13) used a standard CRP assay to provide quantitative results Serial laboratory data of the patient with A. caviae bacteremia Note: Data are presented as patient’s laboratory values. Abbreviations: TP, total protein; ALB, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CREA, creatinine; hs-CRP, high-sensitivity C-reactive protein; IL-6, interleukin-6, eGFR, estimated glomerular filtration rate; NA, not available, NA values for specific inflammatory markers were due to tests not being routinely performed at the external hospital or at certain time points during the recovery phase when clinical stabilization was observed; PCT, procalcitonin; TBIL, total bilirubin; UA, uric acid; WBC, white blood cell count. Symbols: ↑, above the reference range; ↓, below the reference range. The hs-CRP values on Hospital Day 4–5 were reported as > 5 mg/L because they exceeded the upper detection limit of the high-sensitivity assay employed at that time. Subsequent measurements (Hospital Day 9–13) used a standard CRP assay to provide quantitative results On Hospital Day 5, the patient was lethargic and confused, with altered mental status and providing irrelevant responses. Vital signs were: temperature 36.5 °C, heart rate 115 bpm (regular), respiratory rate 15 breaths/min, blood pressure 156/110 mmHg, and oxygen saturation 98% on supplemental oxygen. There was no scleral icterus or mucocutaneous jaundice. Auscultation of the lungs revealed bilateral coarse breath sounds, without rales or wheezes. Abdominal examination showed mild distension without palpable tenderness. Inspection of the primary inoculation site on the left knee demonstrated a completely healed scar with no evidence of active local infection. However, significant tenderness to palpation was present over the right gluteal region and right lower extremity. Intensive laboratory investigations at our hospital confirmed severe systemic infection and systemic inflammatory response syndrome (SIRS), complicated by multiple organ dysfunction syndrome (MODS). Significant abnormalities included profound leukocytosis (12.81 × 10⁹/L) with severe neutrophilia (92.8%), and soaring acute-phase reactants, including procalcitonin (PCT, 3.509 ng/mL) and interleukin-6 (IL-6, 0.070 ng/mL). It is noteworthy that a dissociation was observed between the biomarkers: PCT and IL-6 had declined since their peaks on Hospital Day 1 (PCT 9.13 ng/mL, IL-6 0.497 ng/mL), likely due to empiric antibiotic therapy. However, the failure of CRP to decline (rising to 77.63 mg/L) confirmed a state of sustained systemic inflammation. Comprehensive metabolic panels indicated acute hepatic injury (total bilirubin 80.4 µmol/L, alanine aminotransferase 439.0 U/L) and acute kidney injury (creatinine 141.0 µmol/L). Additionally, the patient exhibited mild hypoalbuminemia (serum albumin [ALB] 35.3 g/L; reference 40.0–55.0 g/L) and decreased total protein (TP 59.1 g/L; reference 65.0–85.0 g/L), reflecting systemic inflammatory catabolism superimposed on pre-existing nutritional compromise related to his prior subtotal gastrectomy. Coagulation profiles demonstrated a prolonged prothrombin time (16.7 s) and elevated D-dimer (2.76 µg/mL). Arterial blood gas (ABG) analysis revealed metabolic acidosis with a compensatory respiratory alkalosis (Table 1). Radiological assessment included a non-contrast head CT which was unremarkable. A repeat chest CT confirmed bilateral pulmonary infiltrates, pleural effusions, and cardiomegaly (Fig. 1a, b). Crucially, a pelvic magnetic resonance imaging (MRI) scan elucidated minimal bilateral hip joint effusion alongside extensive edema permeating the gluteal muscles, proximal thigh musculature, and surrounding pelvic soft tissues (Fig. 1c, d). Based on these collective findings, the principal admission diagnoses were established as: sepsis, bacteremia, extensive soft tissue infection, MODS (involving hepatic, renal, and cardiac systems), atrial fibrillation, anemia, and thrombocytopenia. Radiological findings of the chest and pelvis. (a) Axial chest CT scan (lung window) demonstrating bilateral pulmonary infiltrates and ground-glass opacities (blue arrows) with concomitant pleural effusion (yellow arrows). (b) Axial chest CT scan (mediastinal window) highlighting bilateral pleural effusion (yellow arrows) and cardiomegaly. (c, d) Axial pelvic magnetic resonance imaging (MRI) scans revealing extensive edema and high-signal inflammatory changes permeating the gluteal muscles, pelvic musculature, and surrounding soft tissues (red arrows). These findings are consistent with infectious myositis or fasciitis. Minimal bilateral hip joint effusion is also visible Microbiological identification and genomic analysis Prior to transfer (Hospital Day 4), the external hospital’s automated blood culture system provisionally identified the isolate as the A. hydrophila/A. caviae complex. Upon admission to our facility, both repeat blood cultures and subsequent sputum cultures definitively yielded pure growth of A. caviae. For precise taxonomic placement, the isolate was subjected to whole-genome sequencing (WGS)–based phylogenomic analysis, which provides the resolution required for definitive species-level identification of closely related Aeromonas species. A maximum-likelihood phylogenomic tree was reconstructed using core-genome single-copy orthologous genes extracted from the WGS data of our clinical isolate (BioProject PRJNA1391908, GenBank assembly accession: JBYDFT000000000) together with the publicly available genomes of 39 representative Aeromonas type strains and the outgroup Tolumonas auensis DSM 9187, retrieved from NCBI RefSeq. In the resulting phylogenomic tree, our clinical isolate clustered tightly with the A. caviae type strain NCTC 12,244 with maximum bootstrap support (100%), while clearly diverging from the closely related and clinically important species A. dhakensis CIP 107,500, A. hydrophila ATCC 7966, and A. veronii ATCC 35,624, which segregated into distinct, well-supported clades (Fig. 2). To further substantiate species-level identification, average nucleotide identity (ANI) analysis was performed using FastANI between our isolate and the A. caviae type strain NCTC 12,244, yielding an ANI value of 98.09%, well above the universally accepted ≥ 95% threshold for prokaryotic species delineation, thereby unambiguously confirming the isolate’s identity as A. caviae and excluding misidentification as a closely related Aeromonas species. Whole-genome–based maximum-likelihood phylogenomic tree of the clinical Aeromonas caviae isolate and 39 reference Aeromonas type strains. The tree was constructed from concatenated core-genome single-copy orthologous genes using the maximum-likelihood method, with Tolumonas auensis DSM 9187 included as the outgroup. The clinical A. caviae isolate from this study (BioProject PRJNA1391908, GenBank assembly accession JBYDFT000000000, marked with “#”) clustered with the A. caviae type strain NCTC 12,244 with 100% bootstrap support, and was clearly distinguished from the closely related species A. dhakensis CIP 107,500, A. hydrophila ATCC 7966, and A. veronii ATCC 35,624. Numbers at the internal nodes represent bootstrap support values (1,000 replicates). The scale bar represents 0.01 nucleotide substitutions per site. Strain names are color-coded according to their source of isolation (Aquatic Environments, Animal Specimens, Human, Shellfish, Fish, Unknown), as indicated in the legend. NCBI RefSeq assembly accession numbers are provided after each strain name To guide targeted therapeutic interventions, comprehensive in vitro antimicrobial susceptibility testing (AST) was performed using the broth microdilution method with the Vitek 2 Compact automated system (bioMérieux, Marcy-l’Étoile, France) in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines [11]. MIC results reported as “≥” or “≤” reflect values outside the system’s validated detection range. The organism exhibited a multidrug-resistant (MDR) phenotype, displaying resistance to ampicillin/sulbactam, piperacillin, cefepime, aztreonam, gentamicin, tobramycin, ciprofloxacin, levofloxacin, chloramphenicol, and trimethoprim-sulfamethoxazole. Additionally, the isolate showed intermediate susceptibility to ceftazidime and piperacillin/tazobactam. Conversely, susceptibility was retained against imipenem, meropenem, amikacin, polymyxin B, and minocycline (Table 2). Table 2Antimicrobial susceptibility profile and key resistance determinants of the A. caviae isolateAntimicrobial AgentMIC (µg/mL)InterpretationKey Resistance Gene(s) DetectedGenotype-Phenotype Consistencyβ-Lactams Ampicillin/sulbactam≥ 32/16RblaCTX−M−3, blaTEM−1B, blaOXA−10Concordant Piperacillin≥ 128RblaCTX−M−3, blaTEM−1B, blaOXA−10Concordant Piperacillin/tazobactam64/4IblaTEM−1B, blaOXA−10Concordant (Partial) Cefepime≥ 32R bla CTX−M−3 Concordant Ceftazidime8I bla CTX−M−3 Concordant (Partial) Imipenem≤ 1SNot detectedConcordant Meropenem≤ 1SNot detectedConcordant Aztreonam32RblaCTX−M−3, blaOXA−10ConcordantAminoglycosides Amikacin16S aac(6’)-Ib-cr** Discordant* Gentamicin≥ 16RNot detectedDiscordant* Tobramycin≥ 16R aac(6’)-Ib-cr** ConcordantFluoroquinolones Ciprofloxacin≥ 4R aac(6’)-Ib-cr** Concordant Levofloxacin≥ 8R aac(6’)-Ib-cr** ConcordantOthers Polymyxin B≤ 2SNot detectedConcordant Trimethoprim/sulfamethoxazole≥ 8/152RdfrA12/14, sul1Concordant Chloramphenicol≥ 32RfloR (Match = 1)†Concordant Minocycline≤ 4S tet(A) Discordant*Note: MIC, minimum inhibitory concentration; S, susceptible; I, intermediate; R, resistant. Interpretive categories were determined according to the CLSI 2016 guidelines [11]. Discordant*: Indicates genotype-phenotype inconsistencies. For gentamicin, phenotypic resistance in the absence of known resistance genes suggests potential unidentified aminoglycoside-modifying enzymes (AMEs) or overexpression of efflux systems. Susceptibility to minocycline is retained because Tet(A) primarily confers resistance to first-generation tetracyclines (e.g., tetracycline and doxycycline) but exhibits reduced activity against it. For amikacin, aac(6’)-Ib-cr can elevate MICs, though the value remained below the clinical breakpoint. †, The floR gene was detected with a Match = 1 status (partial sequence identity or length < 100%), indicating a possible variant. **aac(6’)-Ib-cr is a bifunctional enzyme conferring co-resistance to aminoglycosides and fluoroquinolones. Ertapenem and biapenem were not tested; carbapenem class susceptibility was inferred from imipenem and meropenem results (both MIC ≤ 1 µg/mL), consistent with standard laboratory practice Antimicrobial susceptibility profile and key resistance determinants of the A. caviae isolate Note: MIC, minimum inhibitory concentration; S, susceptible; I, intermediate; R, resistant. Interpretive categories were determined according to the CLSI 2016 guidelines [11]. Discordant*: Indicates genotype-phenotype inconsistencies. For gentamicin, phenotypic resistance in the absence of known resistance genes suggests potential unidentified aminoglycoside-modifying enzymes (AMEs) or overexpression of efflux systems. Susceptibility to minocycline is retained because Tet(A) primarily confers resistance to first-generation tetracyclines (e.g., tetracycline and doxycycline) but exhibits reduced activity against it. For amikacin, aac(6’)-Ib-cr can elevate MICs, though the value remained below the clinical breakpoint. †, The floR gene was detected with a Match = 1 status (partial sequence identity or length < 100%), indicating a possible variant. **aac(6’)-Ib-cr is a bifunctional enzyme conferring co-resistance to aminoglycosides and fluoroquinolones. Ertapenem and biapenem were not tested; carbapenem class susceptibility was inferred from imipenem and meropenem results (both MIC ≤ 1 µg/mL), consistent with standard laboratory practice To characterize the molecular basis of the observed multidrug resistance, we performed whole-genome sequencing (WGS) on the isolate. Using the genomic data, we then conducted resistome analysis to screen for resistance-associated genes against the ResFinder (v4.1), CARD (v3.2.8), and NCBI AMRFinderPlus databases. The genomic assay revealed a highly diverse repository of acquired antimicrobial resistance (AMR) genes. The β-lactamase content included the ESBL gene blaCTX−M−3, the penicillinase gene blaTEM−1B, the Class D β-lactamase gene blaOXA−10, and the AmpC-type gene blaMOX−6. Aminoglycoside resistance was mediated by a diverse set of modifying enzymes, including aac(6’)-Ib-cr, aph(3’)-Ia, aph(6)-Id, aph(3’’)-Ib, and aadA1/2. Notably, the bifunctional gene aac(6’)-Ib-cr also conferred plasmid-mediated quinolone resistance, complementing the phenotypic resistance to fluoroquinolones. Other significant AMR determinants identified included sul1 and dfrA12/14 (sulfonamides/trimethoprim), floR (phenicols), mph(A) (macrolides), and tet(A) (tetracyclines). This complex resistome precisely correlated with the observed MDR phenotype, highlighting the pathogen’s high clinical significance (Table 2). Therapeutic intervention and outcome Guided by the comprehensive AST and genomic profiling results, the initial empirical ertapenem therapy was discontinued, and targeted pathogen-directed therapy was initiated with intravenous biapenem (300 mg q12h, adjusted for renal function). This antibiotic strategy was augmented with aggressive, multidisciplinary supportive care. Clinical management encompassed hepatoprotective therapy, prophylactic gastrointestinal mucosal protection, intravenous human albumin supplementation, strict hemodynamic and blood pressure control, and targeted urate-lowering therapy. Particular clinical attention was devoted to the severe, migratory musculoskeletal pain affecting the patient’s right extremities. This was aggressively managed via a multimodal analgesic approach, including intramuscular diclofenac/lidocaine and the application of fentanyl transdermal patches. Although the patient experienced transient recurrences of pain in the lumbar and right lower limb regions, these episodes were effectively controlled through this optimized, stepwise analgesic regimen. Under this 9-day hospitalization, the patient exhibited a highly favorable clinical course. Continuous biomarker monitoring indicated a progressive resolution of systemic inflammation, with PCT and WBC counts steadily declining to physiological baselines. A follow-up blood culture obtained on Hospital Day 13, prior to discharge, returned negative, documenting microbiological clearance of bloodstream infection. Concurrently, hepatic and renal functions showed significant recovery. Additionally, the patient’s blood pressure spontaneously normalized (e.g., 117/80 mmHg on Hospital Day 10) without the administration of antihypertensive medications, confirming that the initial elevation was a transient stress response. Following a 9-day hospitalization, the patient achieved full clinical stability and was discharged on Hospital Day 14. Review of invasive A. caviae infections A comprehensive literature search of the PubMed, Web of Science, and China National Knowledge Infrastructure (CNKI) databases was conducted for articles published up to December 31, 2025 (Supplementary Fig. 1), using search terms including “Aeromonas caviae” combined with “infection,” “bacteremia,” “sepsis,” and related clinical syndromes. Studies were included if they reported culture-confirmed, clinically significant A. caviae infections from any anatomical site with sufficient clinical detail. This search yielded 43 well-documented cases of invasive infections [9, 12–50] (Supplementary Table 1), encompassing a broad spectrum of clinical syndromes including bacteremia/sepsis, pneumonia, gastrointestinal infection, skin and soft tissue infection, spontaneous bacterial peritonitis, and other invasive presentations. Clinically, the 43 cases were categorized by primary site of infection as follows: (1) Primary Bacteremia/Sepsis without identifiable focal source (8 cases, 18.60%), predominantly occurring in patients with underlying malignancies, hematologic disorders, or decompensated cirrhosis; (2) Pulmonary Infection (8 cases, 18.60%), ranging from community-acquired pneumonia to severe necrotizing pneumonia, often complicating chronic respiratory comorbidities or near-drowning events; (3) Gastrointestinal Infection (8 cases, 18.60%), predominantly acute gastroenteritis but also including chronic diarrhea, colorectal ulcer, hemolytic uremic syndrome, and infectious enteropathy; (4) Skin and Soft Tissue Infection (8 cases, 18.60%), encompassing necrotizing fasciitis (4 cases), post-traumatic wound infection (2 cases), cellulitis (1 case), and burn wound infection (1 case); (5) Spontaneous Bacterial Peritonitis (3 cases, 6.98%), occurring exclusively in patients with decompensated cirrhosis; (6) Urinary Tract Infection (2 cases, 4.65%); (7) Central Nervous System Infection (2 cases, 4.65%), including community-acquired meningitis and post-neurosurgical ventriculoperitoneal shunt infection; (8) Ocular Infection (2 cases, 4.65%), comprising endogenous endophthalmitis and contact lens-associated keratitis; and (9) Hepatobiliary Infection (2 cases, 4.65%), including acute cholecystitis and bile duct infection. Notably, bloodstream infection was documented in 18 cases (41.86%), comprising 8 cases of primary bacteremia and 10 cases of secondary bacteremia complicating focal infections (4 from skin and soft tissue, 2 from gastrointestinal sources, 2 from peritonitis, 1 from pulmonary, and 1 from central nervous system origins). The overall case-fatality rate was 18.60% (8/43), rising to 33.33% (6/18) among patients with bloodstream infection, underscoring the substantial systemic virulence potential of A. caviae. Geographically, these cases exhibited significant diversity, albeit with a marked predominance from Asia (69.77%, 30/43), followed by sporadic reports from Europe and North America. Demographic analysis revealed a broad age distribution spanning from neonates to 93 years (median age, approximately 48 years), with a notable male predominance (26/39, 66.67%). Discussion and conclusions A. caviae is increasingly recognized as a significant opportunistic pathogen capable of causing life-threatening extraintestinal infections [5]. In this report, we present a fulminant case of sepsis and bacteremia in an elderly male, following an agricultural injury in a paddy field. This case is particularly noteworthy due to the rare clinical manifestation of severe, migratory musculoskeletal pain (Supplementary Table 1), and the presence of a MDR genomic profile. To contextualize our findings, our literature review (Supplementary Table 1) reveals that invasive A. caviae infections, particularly bacteremia [12, 15–21, 24, 25, 27–29, 35, 41, 43–45] and spontaneous bacterial peritonitis [19, 30, 41], predominantly afflict severely immunocompromised individuals, such as those with decompensated liver cirrhosis [19, 30, 41, 44], hematologic malignancies [15], or advanced solid tumors [17, 20]. In stark contrast, our patient had no formal immunodeficiency disorder, with no history of cirrhosis or active malignancy. Instead, his susceptibility appears to have been driven by a convergence of host- and exposure-related factors specific to his clinical context. First, the route of infection was percutaneous inoculation: a penetrating knee laceration sustained during agricultural work in a paddy field provided a direct portal of entry, bypassing the intact skin barrier that normally precludes invasion by waterborne Aeromonas species. Paddy fields and freshwater environments are well-recognized ecological reservoirs of A. caviae [51, 52], and minor traumatic injuries in such settings have been increasingly implicated in invasive Aeromonas infections [1, 51, 53]. Notably, this case occurred in summer. The seasonal distribution of Aeromonas infections is well recognized: a multicenter prospective cohort study from Japan demonstrated that the incidence of human Aeromonas infections is significantly higher in months with an average temperature above 20 °C (June to September) [54]. Similarly, a population‑based study from Australia reported the highest rates of Aeromonas bloodstream infections during warmer and wetter months, with a nadir in winter [55]. Therefore, the presentation of this case in summer aligns with the known warm‑season predilection of this pathogen. Second, advanced age and age-associated immunosenescence likely impaired both innate and adaptive immune responses to the inoculated pathogen, particularly the early containment of bacterial dissemination [56, 57]. Third, although the patient had no overt immunosuppressive comorbidity, his history of subtotal gastrectomy and chronic nutritional compromise may have contributed to a generalized state of reduced physiological reserve, potentially attenuating his capacity to mount an effective systemic response to acute bacterial sepsis [58]. Together, these factors—rather than impaired gastric acid barrier function, which is most relevant to enteric routes of infection—provide a more biologically plausible framework for the fulminant clinical course observed in our patient. A striking clinical feature of this case was the paroxysmal, migratory musculoskeletal pain localized to the gluteal and lower limb regions, distant from the healed entry site (left knee). Our systematic literature review indicates that while skin and soft tissue infections (SSTIs) [20, 23, 24, 29, 32–34, 43, 45], including necrotizing fasciitis and cellulitis, have been documented following aquatic or soil exposure, they typically manifest as localized, rapidly progressive necrosis at the primary inoculation site. Notably, our aggregated data demonstrate that patients who develop deep soft tissue involvement concurrent with bacteremia face a markedly elevated mortality rate approaching 45% (4/9 cases) [24, 29, 43, 45]. The clinical phenotype observed in our patient, namely transient migratory myalgia accompanied by extensive pelvic soft tissue edema and bilateral hip joint effusion, remote from the healed knee abrasion, represents an exceptionally rare presentation that is most consistent with hematogenous dissemination of the pathogen to distant musculoskeletal sites rather than contiguous spread from the initial inoculation site. This pathogenic mechanism remains underappreciated in Aeromonas infections [51]. Although no muscle biopsy or aspirate culture was performed to confirm the presence of A. caviae at these distant sites, the clinical and radiological findings strongly support hematogenous seeding. MRI elegantly delineated this early-stage infectious myositis and fasciitis, underscoring the critical imperative for clinicians to maintain a high index of suspicion for deep tissue involvement when managing Aeromonas bacteremia, even in the absence of overt localized abscess formation. The MDR phenotype of this A. caviae isolate poses a significant therapeutic challenge, rooted in a complex and synergistic resistome [5, 51]. Genomic analysis revealed a sophisticated array of β-lactamases, including blaCTX−M−3, blaTEM−1B, and blaOXA−10 [9, 52]. The presence of blaCTX−M−3, a potent plasmid-mediated ESBL, explains the high-level resistance to cefepime and aztreonam [59]. Notably, the co-existence of blaOXA−10, a Class D β-lactamase, is of particular clinical significance. Unlike Class A enzymes, OXA-10 exhibits poor susceptibility to traditional β-lactamase inhibitors such as clavulanate and tazobactam, which likely accounts for the intermediate susceptibility observed for piperacillin/tazobactam (MIC = 64/4 µg/mL) [60]. The aggregation of these genes, potentially localized on a single mobile genetic element, underscores the evolutionary pressure on environmental Aeromonas to accumulate diverse enzymatic resistance mechanisms. Intriguing discrepancies were observed regarding aminoglycoside resistance. While the isolate carried the bifunctional aac(6’)-Ib-cr gene, it remained phenotypically susceptible to amikacin (MIC = 16 µg/mL). This phenomenon is consistent with the enzyme’s substrate specificity. While aac(6’)-Ib-cr efficiently acetylates tobramycin and kanamycin, its catalytic efficiency against amikacin is significantly lower, often resulting in MICs remaining within the susceptible range [52]. Conversely, the isolate displayed high-level resistance to gentamicin (MIC ≥ 16 µg/mL) despite the absence of known gentamicin-modifying enzymes in the ResFinder database. This “genotype-phenotype gap” suggests the involvement of unidentified AMEs, 16 S rRNA methyltransferases not captured by current databases, or the overexpression of specific resistance-nodulation-division (RND) family efflux pumps [61, 62]. Efflux systems like AheB [62] or the recently discovered mobile cluster tmexCD-toprJ have been shown to mediate broad-spectrum resistance in Aeromonas [61], and their role in our isolate warrants further investigation. Furthermore, the susceptibility of the isolate to minocycline despite harboring the tet(A) gene exemplifies the substrate specificity of this resistance determinant. The Tet(A) efflux pump confers resistance to first- and second-generation tetracyclines (e.g., tetracycline and doxycycline) but does not affect third-generation tetracyclines such as minocycline or the glycylcycline tigecycline due to their distinct steric properties [63]. Although this phenotype–genotype discordance suggested that minocycline could theoretically have served as a viable alternative therapeutic option [64], it should be explicitly noted that minocycline was not administered to our patient; definitive therapy was achieved with biapenem based on confirmed carbapenem susceptibility, and the patient demonstrated rapid clinical recovery without the need for tetracycline-class agents. The discussion of minocycline here is intended solely to illustrate the substrate specificity of Tet(A)-mediated resistance and the broader principle of genotype–phenotype discordance in Aeromonas species, rather than to imply its clinical use in this case. From a clinical perspective, the successful transition from ertapenem to biapenem was a pivotal turning point. Although direct MIC testing for ertapenem and biapenem was not performed at our institution—consistent with standard laboratory practice of testing representative carbapenem agents—the isolate’s confirmed susceptibility to both imipenem and meropenem (MIC ≤ 1 µg/mL) strongly indicated preserved carbapenem class activity and the absence of carbapenemase production. The initial suboptimal clinical response to ertapenem at the referring hospital, despite predicted carbapenem class susceptibility, is unlikely to be primarily attributable to enzymatic inactivation, since the same β-lactamase profile (blaCTX−M−3, blaTEM−1B, blaOXA−10) did not preclude in vitro activity of biapenem, imipenem, and meropenem—all of which exhibit significantly greater stability against ESBLs and OXA-type enzymes than ertapenem. Rather, the differential outcome is most plausibly explained by ertapenem-specific pharmacokinetic (PK) and pharmacodynamic (PD) vulnerabilities operative in this clinical context. Ertapenem is highly protein-bound (approximately 85–95%) and administered once daily, properties that render it particularly susceptible to subtherapeutic exposure in critically ill patients with hypoalbuminemia and altered renal function, in whom an increased unbound fraction combined with disturbed clearance frequently compromises the time-above-MIC pharmacodynamic target [65, 66]. Consistent with this mechanism, our patient exhibited severe hypoalbuminemia (serum albumin nadir 29.1 g/L on Hospital Day 9; reference 40.0–55.0 g/L) with concomitant hypoproteinemia (total protein 51.3 g/L) and acute kidney injury during the ertapenem treatment period (serum creatinine 146 µmol/L and eGFR 43 mL/min/1.73 m² on Hospital Day 1, with persistent renal impairment through Hospital Day 5: creatinine 141 µmol/L, eGFR 45 mL/min/1.73 m²) (Table 1). This combination of severe hypoalbuminemia and acute renal dysfunction is precisely the clinical milieu in which ertapenem PK is most disturbed, with both increased volume of distribution of the unbound drug and unpredictable clearance contributing to subtherapeutic free-drug exposure [65, 66]. Although specific ertapenem dosing details and serial drug-level measurements from the referring hospital were not available for retrospective review—a limitation we explicitly acknowledge—the patient’s biochemical profile is highly consistent with the PK vulnerabilities described in critically ill populations [65, 66]. Additionally, ertapenem possesses an intrinsically narrower antimicrobial spectrum than other carbapenems, lacking activity against non-fermenting Gram-negative bacilli, which may be clinically disadvantageous in severe sepsis of uncertain microbiological etiology. In contrast, biapenem—like imipenem and meropenem—exhibits lower plasma protein binding, twice-daily dosing that facilitates more sustained time-above-MIC exposure, and a broader antimicrobial spectrum, all of which may have contributed to its clinical efficacy in this case. While biapenem-specific clinical data in Aeromonas infections remain limited and most published evidence pertains to Enterobacterales [67, 68]—an extrapolation we acknowledge as a limitation—the in vitro carbapenem class susceptibility of our isolate, combined with biapenem’s favourable PK/PD profile in the setting of hypoalbuminemia, provided a rational basis for its selection. The progressive recovery of both renal function (eGFR 69 mL/min/1.73 m² by Hospital Day 9 and 82 mL/min/1.73 m² by Hospital Day 13) and serum albumin (35.3 g/L by Hospital Day 13) following the therapeutic switch, together with rapid clinical and microbiological response, supports the appropriateness of carbapenem class-guided therapy in this case of severe Aeromonas bacteremia, although the precise contribution of PK/PD factors remains inferential in the absence of direct comparative MIC and pharmacokinetic data. In our patient, microbiological clearance was documented by a negative follow-up blood culture (FUBC) on Hospital Day 13. Recent evidence supports a broader role for FUBCs in Gram-negative bloodstream infection (GN-BSI): a 2024 consensus guidance recommended selective use of FUBCs in patients with risk factors for persistent bacteremia [69], and persistent GN-BSI detected by FUBCs has been associated with increased mortality [70]. Reported predictors of persistent bacteremia include multidrug-resistant (MDR) pathogens, inadequate source control, and host immunocompromise [71]. Our patient exhibited several of these features, including an MDR A. caviae isolate harboring blaCTX−M−3 and age-related immune decline; in retrospect, earlier FUBCs at 48–72 h after initiation of targeted therapy would have helped characterize bacteremia kinetics. Early FUBCs were not performed as the patient demonstrated rapid clinical and biochemical improvement, and microbiological reassessment was guided by clinical response. This case highlights that systematic early FUBCs warrant consideration in high-risk GN-BSI, particularly those involving MDR pathogens or immunocompromised hosts. The detection of blaCTX−M−3 within a complex resistance cassette (alongside blaTEM−1B and blaOXA−10) in an environmental A. caviae isolate has epidemiological implications. Aeromonas species are ubiquitous in aquatic environments and function as reservoirs for antimicrobial resistance genes (ARGs), often carrying diverse resistance determinants on mobile genetic elements [61, 72–75]. This case represents an example of spillover of clinically significant resistance determinants from environmental reservoirs into human pathogens. The potential horizontal transfer of these genes to Enterobacterales via plasmids, transposons, or other mobile elements supports the concept of the environment as a ‘mixing vessel’ for resistance, which may facilitate the emergence of multidrug-resistant clones [61, 72, 75, 76]. These observations highlight the need for integrated surveillance within a One Health framework to monitor the flow of ARGs across the environment-human interface and to mitigate the risk of future MDR outbreaks [61, 73, 75]. This case report describes a rare instance of fulminant A. caviae bacteremia caused by a CTX-M-3-producing multidrug-resistant strain in a 71-year-old patient without formal immunodeficiency but with predisposing vulnerabilities, including advanced age and prior subtotal gastrectomy, following percutaneous exposure in a paddy field. The isolate’s complex resistome—encompassing the ESBL genes blaCTX−M−3, blaTEM−1B, and blaOXA−10 alongside aac(6’)-Ib-cr and tet(A)—exemplifies the spillover of clinically significant resistance determinants from environmental reservoirs into human pathogens, while WGS-based phylogenomic analysis combined with ANI determination (98.09%) proved indispensable for accurate species-level identification within the taxonomically complex Aeromonas genus. The clinical failure of initial ertapenem therapy, despite preserved susceptibility to representative carbapenems, highlights the importance of pharmacokinetic/pharmacodynamic considerations in critically ill patients with hypoalbuminemia and acute kidney injury, where biapenem’s lower protein binding, twice-daily dosing, and broader antimicrobial spectrum may have contributed to its clinical efficacy; additionally, the unusual presentation of severe migratory myalgia attributable to occult deep fascial involvement expands the recognized clinical spectrum of invasive Aeromonas disease. Nevertheless, several limitations should be acknowledged: as a single-patient observation, the findings cannot be generalized or used to establish causal inference; direct MIC testing for ertapenem and biapenem was not performed, leaving the proposed PK/PD mechanism of treatment failure hypothesis-generating in the absence of therapeutic drug monitoring; although a negative follow-up blood culture on Hospital Day 13 documented eventual microbiological clearance, earlier follow-up blood cultures at 48–72 h after initiation of targeted therapy were not obtained, as microbiological reassessment was guided by clinical response, precluding characterization of early bloodstream clearance kinetics; transcriptomic and conjugation experiments were not conducted, precluding direct confirmation of the proposed efflux-pump-mediated gentamicin resistance and the horizontal transfer potential of the blaCTX−M−3 platform; environmental sampling of the suspected exposure site was not feasible, precluding direct genomic linkage between the clinical isolate and putative environmental reservoirs; and the literature review may be subject to publication bias toward severe presentations. Overall, this case underscores the emerging clinical relevance of environmentally acquired multidrug-resistant A. caviae and the urgent need for prospective multicentre studies and integrated One Health surveillance to monitor antimicrobial resistance gene flow across the environment–human interface and to optimize therapeutic strategies for invasive infections caused by this emerging waterborne pathogen. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Acknowledgements We thank the Emergency Department of Suqian Hospital Affiliated to Nanjing Drum Tower Hospital Group for providing clinical support. Abbreviations Arterial Blood Gas Alanine Aminotransferase Total protein Albumin estimated Glomerular filtration rate Aminoglycoside-Modifying Enzymes Antimicrobial Resistance Antimicrobial Resistance Genes Average nucleotide identity Antimicrobial Susceptibility Testing Clinical and Laboratory Standards Institute Computed Tomography Extended-Spectrum β-Lactamase High-Sensitivity C-Reactive Protein Intensive Care Unit Interleukin-6 Multidrug-Resistant Minimum Inhibitory Concentration Multiple Organ Dysfunction Syndrome Magnetic Resonance Imaging Procalcitonin Resistance-Nodulation-Division Systemic Inflammatory Response Syndrome Skin and Soft Tissue Infections Total bilirubin Whole-Genome Sequencing Preferred Reporting Items for Systematic Reviews and Meta-Analyses Follow-up blood culture Gram-negative bloodstream infection Author contributions LG and ZZ contributed equally to this work and share first authorship. LG and ZZ were involved in conceptualization, investigation, data curation, and writing the original draft. KL, WH, NM, XJ, BG, QL, CL, and CT performed the investigation and data collection. HZ and JH contributed to conceptualization and writing—review and editing. SG and NL provided supervision, conceptualization, and funding acquisition, and contributed to writing—review and editing. All authors read and approved the final manuscript. Funding This research was supported by the Scientific Research Program of the Affiliated Huai’an No. 1 People’s Hospital of Nanjing Medical University (grant numbers YCT202302 and ZD202502). The funders had no role in the study design, data collection, analysis, interpretation, or manuscript preparation. Data availability The datasets generated and/or analysed during the current study have been deposited in the NCBI GenBank and BioProject repositories. The whole-genome sequencing data are available under NCBI BioProject accession number PRJNA1391908 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1391908), which includes BioSample SAMN54236645 (https://www.ncbi.nlm.nih.gov/biosample/SAMN54236645), and Sequence Read Archive (SRA) run SRR36540707 (https://www.ncbi.nlm.nih.gov/sra/SRR36540707). The assembled whole-genome sequence is deposited under GenBank accession number JBYDFT000000000 (https://www.ncbi.nlm.nih.gov/nuccore/JBYDFT000000000). All other clinical data generated or analysed during this study are included in this published article and its supplementary information files. Declarations Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Review Committee of The Affiliated Huai’an No. 1 People’s Hospital of Nanjing Medical University (Approval No: KY-2024-379-01; date of approval: 13 January 2025). Written informed consent was obtained from the patient for participation in this study and for the publication of this case report and any accompanying images. Consent for publication Written informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Lijing Guo and Ziyu Zhao contributed equally to this work. Contributor Information Sumin Gao, Email: gao666332@sina.com. Ning Lin, Email: linning_paper@163.com. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Data Availability Statement The datasets generated and/or analysed during the current study have been deposited in the NCBI GenBank and BioProject repositories. The whole-genome sequencing data are available under NCBI BioProject accession number PRJNA1391908 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1391908), which includes BioSample SAMN54236645 (https://www.ncbi.nlm.nih.gov/biosample/SAMN54236645), and Sequence Read Archive (SRA) run SRR36540707 (https://www.ncbi.nlm.nih.gov/sra/SRR36540707). The assembled whole-genome sequence is deposited under GenBank accession number JBYDFT000000000 (https://www.ncbi.nlm.nih.gov/nuccore/JBYDFT000000000). All other clinical data generated or analysed during this study are included in this published article and its supplementary information files.