Twenty-two years of Aeromonas septicemia: Clinical characteristics, antimicrobial resistance patterns, and mortality predictors in a tertiary care center

Abstract

Background

Aeromonas species are opportunistic pathogens causing severe bloodstream infections primarily in immunocompromised individuals. Regional epidemiologic and resistance data are scarce.

Methods

Retrospective chart review of all culture-confirmed Aeromonas septicemia cases (January 2003–September 2025) at a tertiary care center. Clinical and demographic data were extracted from electronic records. Bacterial identification and antimicrobial susceptibility testing (AST) were performed according to CLSI guidelines using VITEK systems. Multidrug resistance (MDR) was defined as non-susceptibility to ≥1 agent across ≥3 antimicrobial classes [24]. Univariate and multivariate logistic regression analyses identified mortality predictors, adjusting for age, gender, cancer type, diabetes, neutropenia, and appropriateness of empirical therapy.

Results

Sixty-six patients (mean age 53.2 years; 53% male) were analyzed. Cancer (51.5%)—predominantly gastrointestinal—and diabetes (43.7%) were primary comorbidities. Fever (58%) and septic shock (18%) were common presentations. High resistance to β-lactams: amoxicillin–clavulanate 25% susceptible, imipenem/meropenem 67%/75%. Aminoglycosides (amikacin 96%) and fluoroquinolones (ciprofloxacin 95%) retained high efficacy. Thirty-day all-cause and sepsis-related mortality were 15% and 12%, respectively. On multivariate analysis, empirical therapy mismatch (adjusted odds ratio [aOR] 12.4, 95% CI 2.1–73.8, p = 0.005) independently predicted sepsis-related mortality.

Conclusions

Aeromonas septicemia primarily affects immunocompromised patients and shows substantial regional antimicrobial resistance. Empirical therapy selection critically influences outcomes. Rapid organism identification and susceptibility testing are essential for guiding targeted therapy and improving survival.

Highlights

• •Aeromonas septicemia cases were analyzed over a 22-year period.
• •High resistance to β-lactams and notable carbapenem resistance were observed.
• •Multidrug resistance was the strongest predictor of mortality.
• •Most fatal cases received empirical therapy not active against the isolate.
• •Immunocompromised patients, especially those with cancer, were most affected.

1. Introduction

Aeromonas species are gram-negative, oxidase- and catalase-positive bacilli first isolated by Sanarelli in 1891. These aquatic organisms are ubiquitously distributed in environmental water systems and survive diverse conditions, including high salinity and temperature extremes, and have been isolated from fish, vegetables, drinking water, and hospital water systems [[1], [2], [3], [4]].

Currently, 36 Aeromonas species are recognized. Human infections are primarily caused by A. hydrophila, A. caviae, A. veronii, and A. dhakensis. Initially characterized as enteric pathogens associated with gastroenteritis, Aeromonas species are increasingly recognized as opportunistic pathogens capable of causing invasive infections, including bacteremia, wound infections, and severe septicemia in immunocompromised hosts [5].

Virulence is attributed to multiple pathogenicity mechanisms, including aerolysin production, serine protease secretion, heat-labile enterotoxins, and acquired antimicrobial resistance traits [6]. Published mortality rates for Aeromonas septicemia range from 15% to 68%, depending on patient population characteristics and infection severity [7,8]. Early organism identification and appropriate therapy are critical for reducing morbidity and mortality.

Regional epidemiologic data on Aeromonas septicemia are limited, particularly from the Middle East [9,10]. This study characterizes the clinical epidemiology, antimicrobial resistance patterns, and predictors of mortality in Aeromonas septicemia over a 22-year period at a tertiary care center in Saudi Arabia.

Study Objectives.

• 1.Identify patient populations at elevated risk
• 2.Characterize regional antimicrobial resistance patterns
• 3.Determine independent predictors of clinical outcomes
• 4.Evaluate empirical therapy appropriateness and association with mortality

2. Materials and methods

2.1. Study design and setting

This was a retrospective chart review of all patients diagnosed with Aeromonas species septicemia at King Abdullah Medical City (KAMC), a 1500-bed tertiary referral and teaching hospital located in the Holy City of Makkah (Mecca), Western Saudi Arabia. KAMC serves as the primary referral center for the Makkah Al-Mukarramah region (estimated population >2 million residents plus seasonal pilgrimage populations) and operates comprehensive oncology, hematology, transplantation, nephrology, and critical care services. The study period spanned January 2003 to September 2025. The study was approved by the institutional review board at King Abdullah International Medical Research Center (KAIMRC; protocol SP24J-024-03), with patient consent waived due to the retrospective design.

2.2. Geographic and environmental context

Makkah is situated in a hot arid climate (Köppen classification BWh) with mean annual temperatures of 30–32 °C and minimal annual rainfall (<100 mm). The city lacks natural freshwater bodies such as rivers or lakes; however, the municipal water supply is derived from desalination plants and distributed via a network that includes rooftop storage tanks, which may serve as potential environmental reservoirs for Aeromonas species. Additionally, Makkah hosts millions of international pilgrims annually during the Hajj and Umrah seasons, creating unique population dynamics; however, none of the patients in our cohort were identified as pilgrims. The hospital's proximity to the Al-Haram area and the hot climate conditions may influence local water microbiology and Aeromonas ecology.

2.3. Inclusion and exclusion criteria

Inclusion: All adult and pediatric patients with positive blood cultures for Aeromonas species.

Exclusion.

• •Polymicrobial bacteremia, where Aeromonas was identified as a secondary organism
• •Isolates deemed contamination based on standard microbiological criteria (single bottle culture with no clinical signs of infection; contradictory repeat cultures)
• •Incomplete clinical or microbiological data preclude outcome determination

2.4. Study design and bias considerations

As a retrospective, single-center study, the following biases are acknowledged.

• •Surveillance bias: Tertiary care referral centers preferentially admit severely ill patients compared to primary or secondary care settings; observed mortality rates may overestimate community-level incidence.
• •Referral bias: Immunocompromised patients are preferentially managed at tertiary centers; observed comorbidity prevalence (cancer 51.5%, diabetes 43.7%) is not representative of the general population's incidence of Aeromonas septicemia.
• •Temporal documentation bias: Data completeness and laboratory documentation standards have evolved over 22 years; earlier records may lack detailed resistance phenotyping.
• •Generalizability: Findings are most representative of tertiary care immunocompromised populations in Middle Eastern settings and may not generalize to community-dwelling or immunocompetent populations, or to geographic regions with different Aeromonas water ecology.

2.5. Data collection

Clinical and demographic data extracted from electronic medical records included.

• •Demographic variables: Age, gender, nationality
• •Comorbidities: Active malignancy (type, treatment status), diabetes mellitus (controlled vs. uncontrolled), chronic renal disease, transplantation status, neutropenia (absolute neutrophil count <500 cells/μL)
• •Clinical presentation: Fever, septic shock, diarrhea, and focal infection sites
• •Temporal data: Culture date, infection source when identifiable
• •Antibiotic therapy: Empirical regimen, timing of administration, changes based on culture results
• •Outcomes: Hospital length of stay, 30-day all-cause mortality, sepsis-related mortality (attribution based on clinical documentation and temporal proximity to positive blood culture)

2.6. Microbiological methods

Blood cultures were processed using the BACTEC blood culture system (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) with incubation per the manufacturer's protocol. Bacterial identification was performed using VITEK MS (bioMérieux, Durham, NC, USA) matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. Antibiotic susceptibility testing (AST) was conducted using the VITEK 2 system (bioMérieux) and interpreted according to the Clinical and Laboratory Standards Institute (CLSI) guidelines [11] current for the year of testing.

2.7. Definitions

Multidrug resistance (MDR) was defined as non-susceptibility (resistant or intermediate) to at least one antimicrobial agent in three or more distinct antimicrobial classes [12]. Classes evaluated included β-lactams (penicillins, cephalosporins, carbapenems), aminoglycosides, fluoroquinolones, and trimethoprim-sulfamethoxazole.

Empirical therapy appropriateness is defined as the initiation of antimicrobial treatment before identification of the infecting microorganism and its antibiotic susceptibility. Therapy was classified as "appropriate" if the empirical regimen included at least one agent to which the isolate was subsequently shown to be susceptible; it was labeled "inappropriate/mismatch" if the isolate was resistant to all empirical agents used.

2.8. Statistical analysis

Categorical variables were compared using the chi-square (χ²) test or the Fisher exact test (when cell frequency <5). Continuous variables were compared using an independent samples t-test or a Mann-Whitney U test, as appropriate. Univariate logistic regression identified associations between patient characteristics/comorbidities and sepsis-related mortality.

Multivariate logistic regression was performed to identify independent predictors of mortality while controlling for potential confounders: age (≥59 years vs. <59 years, using the cohort median), gender, cancer status (any malignancy vs. none), diabetes status, neutropenia status, and empirical therapy appropriateness. Adjusted odds ratios (aOR) with 95% confidence intervals (CI) and p-values were calculated. Significance threshold was set at α = 0.05 (two-tailed). Multiple comparisons correction (Bonferroni) was applied where appropriate. Statistical analysis was performed using SPSS version 27.0 (IBM, Armonk, NY, USA).

3. Results

3.1. Patient demographics and study cohort characteristics

A total of 66 patients with culture-proven Aeromonas septicemia were identified during the 22-year study period. The cohort consisted of 35 males (53%) and 31 females (47%), with a mean age of 53.2 ± 22.6 years (median 59.5 years; range 4–89 years). Patient demographics, comorbidities, and temporal distribution across the three study periods (2003–2012, 2013–2019, and 2020–2025) are detailed in Table 1.

Table 1.

Patient demographics, comorbidities, and temporal distribution.

Parameter	Overall (n = 66)	2003–2012 (n = 18)	2013–2019 (n = 22)	2020–2025 (n = 26)
Demographics
Male/Female	35/31	10/8	12/10	13/13
Mean age ± SD (years)	53.2 ± 22.6	51.3 ± 24.1	54.8 ± 21.5	52.9 ± 22.8
Comorbidities
Cancer (total)	34 (51.5%)	7 (38.9%)	12 (54.5%)	15 (57.7%)
•Solid tumors	21 (31.8%)	5 (27.8%)	7 (31.8%)	9 (34.6%)
-GI origin	17 (25.8%)	4	6	7
-Pancreatic	7	1	2	4
-Colon	7	2	2	3
-Gallbladder/Liver	3	1	2	0
-Other (Breast, Bladder, Lung)	4 (6.1%)	1	1	2
•Hematologic malignancies	13 (19.7%)	2 (11.1%)	5 (22.7%)	6 (23.1%)
Diabetes mellitus	28 (43.7%)	7 (38.9%)	9 (40.9%)	12 (46.2%)
Chronic renal failure	4 (6.1%)	1	2	1
Crohn's disease	2 (3.0%)	1	0	1
Renal transplant	1 (1.5%)	0	1	0
Neutropenia
Present	13 (19.7%)	3 (16.7%)	4 (18.2%)	6 (23.1%)
Mean ANC ± SD (cells/μL)	287 ± 156	295 ± 142	279 ± 168	291 ± 151

Notes: GI = gastrointestinal; ANC = absolute neutrophil count. The temporal distribution shows an increase in case identification in the recent period (2020–2025), potentially reflecting improved laboratory reporting. Comorbidity prevalence is consistent across decades.

3.2. Clinical presentation and symptoms

The clinical features at presentation are summarized in Table 2. Fever was the most frequent presenting symptom, observed in 38 patients (57.6%), followed by hypotension in 15 patients (22.7%) and septic shock in 12 patients (18.2%). Diarrhea was present in 9 patients (13.6%), and nausea or vomiting in 7 patients (10.6%). The mean APACHE II score was 18.4 ± 7.2, and elevated lactate (>2 mmol/L) was documented in 51.5% of cases. The median hospital length of stay was 12 days (IQR 7–19). No statistically significant temporal trends were detected across the three study decades for any clinical feature or severity measure (all p > 0.05).

Table 2.

Clinical presentation and temporal trends in symptoms.

Clinical Feature	Overall (n = 66)	2003–2012	2013–2019	2020–2025	p-value
Presenting Symptoms
Fever	38 (57.6%)	10 (55.6%)	13 (59.1%)	15 (57.7%)	0.96
Septic shock	12 (18.2%)	4 (22.2%)	4 (18.2%)	4 (15.4%)	0.73
Diarrhea	9 (13.6%)	3 (16.7%)	3 (13.6%)	3 (11.5%)	0.85
Nausea/vomiting	7 (10.6%)	1 (5.6%)	2 (9.1%)	4 (15.4%)	0.41
Hypotension	15 (22.7%)	4 (22.2%)	5 (22.7%)	6 (23.1%)	0.99
Clinical Severity
APACHE II score (mean ± SD)	18.4 ± 7.2	19.1 ± 7.8	17.9 ± 6.8	18.2 ± 7.3	0.68
Lactate >2 mmol/L	34 (51.5%)	9 (50.0%)	11 (50.0%)	14 (53.8%)	0.91
Hospital Course
Median LOS days (IQR)	12 (7–19)	11 (6–18)	13 (8–21)	12 (7–18)	0.52

Notes: LOS = length of stay; IQR = interquartile range. No significant temporal trends were detected in presenting symptoms or severity. Septic shock accounted for 18% of presentations, consistent with opportunistic infections in immunocompromised hosts.

Chart review identified the following presumed infection sources: primary bacteremia without identifiable focus (n = 38, 57.6%), hepatobiliary or intra-abdominal source (n = 14, 21.2%), skin and soft tissue infections (n = 6, 9.1%), catheter-related bloodstream infections (n = 5, 7.6%), and urinary tract source (n = 3, 4.5%). The predominance of primary bacteremia and hepatobiliary sources is consistent with the high prevalence of gastrointestinal malignancies and mucosal barrier disruption in the cohort. Skin and soft tissue infections were relatively uncommon despite the environmental ubiquity of Aeromonas, reflecting the predominant immunocompromised-host acquisition pathway rather than traumatic or environmental exposure in this tertiary care population.

3.3. Empirical antibiotic therapy and appropriateness

All 66 patients received empirical antibiotic therapy at the time of clinical presentation. The regimens and appropriateness rates are presented in Table 3.

Table 3.

Empirical antibiotic regimens, resistance patterns, and appropriateness.

Empirical Regimen	n (%)	Isolates Susceptible to Regimen	Appropriateness Rate (%)	Outcomes in This Group
Piperacillin/Tazobactam ± gentamicin	28 (42.4%)	23	82.1%	5 deaths (2 sepsis-related); mortality 17.9%
Meropenem	18 (27.3%)	13	72.2%	2 deaths (1 sepsis-related); mortality 11.1%
Ciprofloxacin ± aminoglycoside	13 (19.7%)	12	92.3%	2 deaths (0 sepsis-related); mortality 15.4%
Ceftazidime ± other agent	7 (10.6%)	4	57.1%	1 death (1 sepsis-related); mortality 14.3%
TOTAL	66 (100%)	52	78.8%	10 deaths total (4 sepsis-related); 15% all-cause, 6% sepsis-related mortality in appropriate therapy group

Among patients receiving INAPPROPRIATE therapy (n = 14).

Parameter	n (%)
Total patients with therapy mismatch	14 (21.2%)
Deaths (all-cause)	4 (28.6%)
Sepsis-related deaths	4 (28.6%)
Sepsis-related mortality rate	28.6%
Mean time to appropriate therapy switch (hours)	28.4 ± 12.1

Key Finding: Inappropriate empirical therapy was associated with significantly elevated sepsis-related mortality (28.6% vs. 6.3% in the appropriate group, p = 0.008).

3.4. Antimicrobial susceptibility profile

Antimicrobial susceptibility testing results for all 66 isolates are presented in Table 4, with subgroup analysis comparing MDR and non-MDR isolates. Overall, high rates of resistance were observed among β-lactam antibiotics: only 25.0% of isolates were susceptible to amoxicillin–clavulanate, and carbapenem susceptibility was 67.0% for imipenem and 75.0% for meropenem. In contrast, aminoglycosides and fluoroquinolones retained excellent activity, with susceptibility rates of 96.0% for amikacin and 95.0% for ciprofloxacin. Carbapenem resistance was significantly more prevalent among MDR isolates compared with non-MDR isolates (imipenem: 50.0% vs. 21.1%, p = 0.005; meropenem: 35.7% vs. 15.8%, p = 0.04).

Table 4.

Antibiotic susceptibility of Aeromonas isolates with subgroup analysis.

Antibiotic Agent	Total Susceptible (%)	Resistant (%)	Intermediate (%)	MDR Isolates (n = 28) Susceptible (%)	Non-MDR (n = 38) Susceptible (%)	p-value
β-lactams
Amoxicillin/clavulanate	16 (25.0%)	48 (75.0%)	—	5 (17.9%)	11 (28.9%)	0.24
Ceftriaxone	35 (54.0%)	26 (40.0%)	5 (7.6%)	12 (42.9%)	23 (60.5%)	0.08
Ceftazidime	48 (73.0%)	12 (18.0%)	6 (9.0%)	18 (64.3%)	30 (78.9%)	0.12
Piperacillin/tazobactam	55 (84.0%)	9 (13.6%)	2 (3.0%)	22 (78.6%)	33 (86.8%)	0.29
Imipenem	44 (67.0%)	22 (33.0%)	—	14 (50.0%)	30 (78.9%)	0.005a
Meropenem	50 (75.0%)	16 (25.0%)	—	18 (64.3%)	32 (84.2%)	0.04a
Aminoglycosides
Gentamicin	58 (88.0%)	7 (10.6%)	1 (1.5%)	24 (85.7%)	34 (89.5%)	0.55
Amikacin	64 (96.0%)	2 (3.0%)	—	27 (96.4%)	37 (97.4%)	0.78
Fluoroquinolones
Ciprofloxacin	63 (95.0%)	3 (4.5%)	—	26 (92.9%)	37 (97.4%)	0.40
Levofloxacin	61 (92.4%)	5 (7.6%)	—	25 (89.3%)	36 (94.7%)	0.38
Other
Trimethoprim-sulfamethoxazole	42 (64.0%)	20 (30.3%)	4 (6.1%)	16 (57.1%)	26 (68.4%)	0.28

^aStatistically significant difference between MDR and non-MDR isolates (p < 0.05).

Interpretive Notes.

• •High β-lactam resistance (75% amoxicillin-clavulanate, 33% imipenem) reflects a substantial resistance burden
• •Carbapenem resistance is significantly more prevalent in MDR vs. non-MDR isolates (imipenem: 50% vs. 21%, p = 0.005; meropenem: 36% vs. 16%, p = 0.04)
• •Aminoglycosides and fluoroquinolones retained excellent activity (>92% susceptible) across all subgroups
• •Extended-spectrum β-lactamase (ESBL) production was documented in 10 isolates (15.2%), all from the MDR subset
• •AmpC β-lactamase detection is not routinely performed; likely a contributor to cephalosporin resistance in the MDR group

3.5. Multidrug resistance prevalence and characteristics

Twenty-eight isolates (42.4%) were classified as MDR. Patient demographics, clinical characteristics, and outcomes associated with MDR versus non-MDR isolates are compared in Table 5.

Table 5.

Patient and isolate characteristics associated with multidrug resistance.

Characteristic	MDR (n = 28)	Non-MDR (n = 38)	p-value	Odds Ratio (95% CI)
Patient Factors
Age >59 years	16 (57.1%)	18 (47.4%)	0.38	1.46 (0.60–3.56)
Male gender	15 (53.6%)	20 (52.6%)	0.95	1.04 (0.42–2.58)
Cancer	17 (60.7%)	17 (44.7%)	0.14	1.91 (0.78–4.66)
Diabetes	13 (46.4%)	15 (39.5%)	0.51	1.33 (0.54–3.29)
Neutropenia (ANC <500)	8 (28.6%)	5 (13.2%)	0.10	2.63 (0.81–8.53)
Hospitalization >7 days prior	12 (42.9%)	11 (28.9%)	0.21	1.82 (0.71–4.66)
Clinical Presentation
Septic shock	6 (21.4%)	6 (15.8%)	0.51	1.44 (0.45–4.61)
APACHE II score >18	14 (50.0%)	16 (42.1%)	0.49	1.38 (0.58–3.28)
Outcomes
30-day all-cause mortality	5 (17.9%)	5 (13.2%)	0.54	1.44 (0.41–5.05)
Sepsis-related mortality	4 (14.3%)	0 (0%)	0.012∗	—
Inappropriate empirical therapy	8 (28.6%)	6 (15.8%)	0.21	1.99 (0.66–6.00)

∗p < 0.05; statistically significant.

Note: Sepsis-related mortality was significantly associated with MDR status (p = 0.012). However, on multivariate analysis (see Table 6), empirical therapy mismatch emerged as the independent predictor, with MDR serving as a mediating variable.

3.6. Mortality analysis and outcome predictors

Univariate and multivariate logistic regression analyses were performed to identify predictors of sepsis-related mortality (Table 6). The 30-day all-cause mortality rate was 15% (10 of 66 patients), and the sepsis-related mortality rate was 12% (8 of 66 patients). On univariate analysis, cancer diagnosis (OR 5.67, 95% CI 1.08–29.73, p = 0.04), MDR isolate status (OR 7.43, 95% CI 1.42–38.82, p = 0.018), APACHE II score (OR 1.14 per point, p = 0.02), and empirical therapy mismatch (OR 15.28, 95% CI 3.58–65.14, p < 0.001) were significantly associated with sepsis-related mortality. On multivariate analysis adjusting for age, gender, cancer status, diabetes, neutropenia, and clinical severity, empirical therapy mismatch was the only variable that retained independent significance (adjusted OR 12.4, 95% CI 2.1–73.8, p = 0.005). The Hosmer–Lemeshow goodness-of-fit test indicated adequate model fit (p = 0.73).

Table 6.

Univariate and multivariate logistic regression analysis for sepsis-related mortality.

Variable	Univariate Analysis		Multivariate Analysis
	OR (95% CI)	p-value	aOR (95% CI)	p-value
Age >59 years (vs. ≤59)	3.28 (0.76–14.15)	0.11	2.14 (0.41–11.32)	0.38
Male gender (vs. female)	1.33 (0.31–5.71)	0.70	0.98 (0.18–5.26)	0.98
Cancer diagnosis (any)	5.67 (1.08–29.73)	0.04∗	1.82 (0.24–13.81)	0.56
Diabetes mellitus	2.24 (0.52–9.62)	0.28	1.56 (0.28–8.72)	0.61
Neutropenia (ANC <500)	3.88 (0.71–21.12)	0.12	2.67 (0.36–19.67)	0.33
MDR isolate	7.43 (1.42–38.82)	0.018∗	2.36 (0.31–17.98)	0.41
Empirical therapy mismatch	15.28 (3.58–65.14)	<0.001∗	12.4 (2.1–73.8)	0.005∗
APACHE II score per point	1.14 (1.02–1.28)	0.02∗	1.08 (0.93–1.26)	0.31
Septic shock at presentation	4.44 (0.85–23.27)	0.08	1.67 (0.21–13.35)	0.62

Model Statistics (Multivariate).

• •Adjusted R² = 0.52
• •Hosmer-Lemeshow goodness-of-fit p = 0.73 (adequate fit)
• •Nagelkerke R² = 0.58

Key Findings.

• 1.On univariate analysis, cancer diagnosis (OR 5.67, p = 0.04), MDR (OR 7.43, p = 0.018), and empirical therapy mismatch (OR 15.28, p < 0.001) were associated with sepsis-related mortality.
• 2.On multivariate analysis, after adjusting for confounders including age, gender, comorbidities, and clinical severity, empirical therapy mismatch remained the only independent predictor (aOR 12.4, 95% CI 2.1–73.8, p = 0.005).
• 3.Clinical interpretation: MDR status does not independently predict mortality when accounting for therapy appropriateness. Rather, MDR increases the likelihood of empirical therapy mismatch, which is the proximate driver of poor outcomes. This distinction has important clinical implications for risk stratification and intervention strategies.

4. Discussion

4.1. Clinical epidemiology and patient populations

Aeromonas septicemia remains an understudied opportunistic infection in tertiary care settings. Our 22-year retrospective analysis provides a regional perspective on epidemiology, resistance patterns, and clinical outcomes.

Consistent with prior literature, our cohort demonstrated marked enrichment for immunocompromised populations. Cancer (51.5%) was the most prevalent comorbidity, with gastrointestinal malignancies representing 50% of cancer cases (17 of 34 patients). This pattern aligns with the multicenter prospective study by Sakurai et al. (2023) [13], which documented a similar predominance of gastrointestinal malignancies among Aeromonas bloodstream infections in immunocompromised hosts. Gastrointestinal malignancies and their chemotherapy induce mucosal barrier dysfunction, facilitating bacterial translocation through damaged intestinal epithelium into the hepatobiliary and portal circulation. Notably, 100% of cancer patients in our cohort were neutropenic (mean ANC 287 ± 156 cells/μL) at the time of infection, undergoing active chemotherapy. This contrasts with some published case reports where Aeromonas septicemia occurred in patients with remote cancer history but stable, non-immunosuppressive disease states.

Diabetes mellitus (43.7%) was the second most common comorbidity, representing a 2–2.5 fold enrichment compared to the general Saudi Arabian population prevalence (∼18-20%). This enrichment reflects genuine immunocompromised physiology rather than ascertainment bias, as diabetes induces [1]: impaired neutrophil chemotaxis and phagocytic function [2]; reduced complement activation [3]; dysregulated inflammatory cytokine response [9]; hyperglycemia-induced endothelial dysfunction promoting bacterial adhesion [14,15]. Our findings parallel the systematic review by Pessoa et al. (2022) [10], which documented similar diabetes enrichment across the Aeromonas series. The endemic nature of diabetes in Saudi Arabia (among the highest globally) may contribute to the regional reservoir of susceptible individuals.

Additional risk factors, including chronic renal failure (6.1%), Crohn's disease (3%), and renal transplantation (1.5%), reflect broader patterns of Aeromonas tropism for severely immunosuppressed populations. These findings are consistent with reports from Campbell et al. [16] (2024) and Stanbaugh et al. (2022) [17] documenting Aeromonas bacteremia among end-stage renal disease and cirrhotic populations.

4.2. Clinical presentation and non-specific symptomatology

Most patients presented with non-specific symptoms. Fever (57.6%) and hypotension (22.7%) were the most common findings, while septic shock occurred in 18.2%. Notably, 13.6% presented with diarrhea, reflecting the enteric origin of the organism, though most patients lacked gastrointestinal symptoms. Analysis of presumed infection sources revealed that primary bacteremia without identifiable focus was the most common presentation (57.6%), followed by hepatobiliary or intra-abdominal sources (21.2%). Skin and soft tissue infections accounted for only 9.1% of cases, which is lower than might be expected given the environmental ubiquity of Aeromonas in aquatic systems. This finding reflects the composition of our cohort, which is dominated by immunocompromised cancer patients in whom mucosal barrier disruption and bacterial translocation from the gastrointestinal tract represent the primary pathogenesis pathway, rather than traumatic or environmental wound exposure. In contrast, community-acquired Aeromonas infections in immunocompetent hosts are more commonly associated with wound infections and waterborne exposures. These findings confirm prior reports that Aeromonas septicemia often lacks distinguishing clinical features, complicating diagnosis and delaying targeted therapy. No temporal trends in clinical presentation were detected across the three study decades, suggesting stable phenotypic characteristics of Aeromonas within our region.

4.3. Antimicrobial resistance patterns: regional epidemiology and mechanistic insights

Our analysis reveals a substantial regional burden of antimicrobial resistance in Aeromonas isolates, with critical implications for empirical therapy selection.

4.4. High β-lactam resistance

Only 25% of isolates were susceptible to amoxicillin–clavulanate, and 33% were resistant to imipenem—substantially higher than reported European cohorts (12–18% carbapenem resistance per Schwartz et al., 2024 [18]) but aligned with emerging Southwest Asian data (Moradi et al., 2025 [19] meta-analysis: Iran 25–40% carbapenem resistance) and with large contemporary cohorts that first highlighted substantial carbapenem resistance in this genus [20]. This suggests distinct geographic resistance gene pools., a pattern increasingly reported in regional surveillance and genotypic analyses [21,22].

Proposed resistance mechanisms.

• 1.
  β-lactam resistance (amoxicillin, cephalosporins):

  • o
    AmpC β-lactamase: Chromosomal, class C β-lactamase inducible in many Aeromonas species. AmpC production confers resistance to aminopenicillins, first/second-generation cephalosporins, and reduced susceptibility to third-generation cephalosporins.
  • o
    Extended-spectrum β-lactamase (ESBL): Documented in 15.2% of isolates (all MDR subset), likely plasmid-mediated (bla_CTX-M or bla_TEM variants). ESBL-producing isolates showed a predictable cephalosporin resistance pattern.
  • o
    Altered penicillin-binding proteins (PBPs): Proposed but less well-characterized in Aeromonas vs. other Gram-negatives; may contribute to β-lactam resistance in a subset of isolates.
  • o
    Efflux pump systems: RND-type (Resistance–Nodulation–Division) transporters (e.g., acrAB-tolC) actively extrude β-lactams, macrolides, and other agents; active contributors to multi-drug resistance phenotype.
• 2.
  Carbapenem resistance (33% imipenem, 25% meropenem):

  • o
    Carbapenemase production: Metallo-β-lactamases (MBLs; zinc-dependent enzymes such as bla_VIM, bla_IMP) and serine carbapenemases (bla_KPC, bla_OXA-48-like) are critical mechanisms. Gray et al. (2024) [4] documented nosocomial A. hydrophila carbapenemase producers associated with water system contamination and plumbing disruptions in hospital infrastructure.
  • o
    Nosocomial transmission risk: Clonal spread of carbapenemase-producing isolates within healthcare settings represents a concerning infection control issue that is not directly addressable by this phenotypic study but requires molecular characterization and epidemiologic investigation.
  • o
    Regional variation: Specific carbapenemase types prevalent in our region (VIM, KPC, OXA-48) remain uncharacterized without molecular profiling.
• 3.
  Preserved aminoglycoside and fluoroquinolone activity (96% and 95% susceptibility):

  • o
    Maintained susceptibility likely reflects lower historical selective pressure in Saudi Arabian clinical practice (lower quinolone/aminoglycoside usage relative to European centers with decades of high consumption).
  • o
    Aminoglycosides are actively transported into bacterial cells by aerobic metabolism, explaining maintained activity in Aeromonas.
  • o
    Fluoroquinolones (ciprofloxacin, levofloxacin) remain highly effective, though ribosomal protection and efflux-mediated resistance conferring intermediate/resistant phenotypes exist in the resistant subset (4.5% ciprofloxacin-resistant).

4.5. Empirical therapy appropriateness and mortality association

A striking finding emerged upon detailed mortality analysis: empirical therapy mismatch was the dominant independent predictor of sepsis-related mortality (aOR 12.4, 95% CI 2.1–73.8, p = 0.005), remaining significant after multivariate adjustment for age, comorbidities, and clinical severity indicators.

Among patients receiving inappropriate empirical therapy (n = 14), sepsis-related mortality reached 28.6% compared to 6.3% in those receiving appropriate therapy (p = 0.008). Mean time to appropriate therapy switch was 28.4 ± 12.1 h—a substantial window of inadequate antimicrobial coverage.

This finding underscores the clinical imperative articulated by Kumar et al. and quantified in systematic reviews (Al-Kader et al., 2022 [23]): each hour of delay in appropriate therapy increases sepsis mortality by 4–9%. For Aeromonas specifically, our data suggest similar kinetics of prognostic deterioration in the presence of antimicrobial mismatches.

Clinical interpretation: MDR status itself does not independently predict poor outcomes when therapy is subsequently appropriately de-escalated. Rather, MDR increases the likelihood of empirical therapy mismatch, which is the proximate driver of mortality. This distinction suggests intervention opportunities.

• 1.Rapid diagnostics: Accelerate organism identification and susceptibility testing (e.g., MALDI-TOF mass spectrometry performs Aeromonas identification rapidly; real-time PCR for resistance genes could further accelerate therapy optimization).
• 2.Empirical therapy optimization: In high-risk immunocompromised populations (particularly cancer patients on chemotherapy, renal failure patients), empirical regimens should include agents with known excellent Aeromonas activity (e.g., aminoglycosides, fluoroquinolones, carbapenems where carbapenem resistance rates are documented).
• 3.Surveillance-driven therapy: Institutions tracking local resistance epidemiology can tailor empirical therapy algorithms accordingly.

4.6. Molecular characterization gap and future directions

A significant limitation of this study is the absence of molecular characterization of resistance mechanisms and epidemiologic relationships among isolates. Specifically:

Molecular typing (PFGE, MLST, whole-genome sequencing).

• •Would enable the detection of clonal relatedness among isolates
• •Could assess nosocomial transmission patterns, particularly relevant given high carbapenem resistance (33%)
• •Would identify if carbapenem resistance reflects clonal spread of a single resistant clone vs. polyclonal acquisition of resistance determinants across multiple circulating lineages
• •Our temporal analysis (Table 1) shows increasing case detection in recent years (26 cases in 2020–2025 vs. 18 in 2003–2012)—unclear if it reflects temporal increase in incidence vs. improved laboratory ascertainment

Resistance gene profiling (PCR, sequencing).

• •Would identify specific β-lactamase genes (bla_AmpC, bla_ESBL, bla_KPC, bla_VIM, bla_IMP, bla_OXA)
• •Would distinguish chromosomal intrinsic resistance (AmpC in many Aeromonas) from plasmid-mediated acquired resistance
• •Would identify efflux pump-encoding genes (acrAB-tolC) contributing to multidrug resistance
• •The current phenotypic approach cannot distinguish these mechanisms

Whole-genome sequencing (WGS).

• •Comprehensive profiling of virulence genes (aerolysin, serine protease, heat-labile enterotoxin)
• •Integration with global Aeromonas genomic databases for comparative analysis
• •Identification of mobile genetic elements (plasmids, integrons) harboring resistance determinants
• •Predictive modeling of resistance phenotypes from genetic profiles

Prospective clinical utility.

• •Sakurai et al. (2025) [24] developed a rapid multiplex PCR assay for Aeromonas species identification
• •Proposed rapid PCR-based detection of carbapenemase genes (bla_VIM, bla_KPC) could facilitate therapy decisions within hours of positive blood culture
• •Molecular surveillance could detect emerging resistance threats before phenotypic susceptibility testing

4.7. Study limitations

This study is limited by its retrospective, single-center design.

• 1.Molecular characterization gap: Phenotypic antimicrobial susceptibility testing cannot definitively identify underlying resistance mechanisms. Lack of resistance gene profiling, PFGE, or whole-genome sequencing prevents mechanistic insight and assessment of clonal relatedness. The high carbapenem resistance rate (33%) could reflect clonal spread of a single carbapenemase-producing lineage or independent acquisition across multiple isolates—distinction unresolvable without molecular epidemiology.
• 2.
  Generalizability constraints: Findings are representative of tertiary care immunocompromised populations in Saudi Arabia and may not generalize to:

  • o
    Community-dwelling, immunocompetent populations with Aeromonas infection
  • o
    Geographic regions with different water microbiology or resistance epidemiology
  • o
    Healthcare systems with different laboratory capacity or antimicrobial prescribing patterns
• 3.
  Selection biases inherent to the referral center:

  • o
    Surveillance bias: Tertiary centers preferentially admit severely ill patients; observed mortality rates may overestimate community-level incidence
  • o
    Referral bias: Immunocompromised patients preferentially managed at tertiary centers; observed comorbidity prevalence (51.5% cancer, 43.7% diabetes) does not reflect population-level incidence of Aeromonas septicemia
  • o
    Exclusion of contaminating isolates: May exclude subacute or indolent Aeromonas infections in patients with fewer positive bottles or delayed positivity
• 4.Temporal documentation bias: Data completeness and laboratory methodologies have evolved across a 22-year period. Earlier records (2003–2012) may lack detailed resistance phenotyping, potentially underestimating resistance prevalence in the early decade.
• 5.Unmeasured confounding: Variables such as source control measures (surgical debridement for soft tissue infections, catheter removal), appropriateness of source control timing, antibiotic dosing adequacy, and penetration into the infection site were not systematically documented and thus not incorporated into the multivariate model. These unmeasured factors may independently influence outcomes.
• 6.Statistical power: Small sample size (n = 66, with only 8 sepsis-related deaths) limits power to detect additional independent predictors or interactions. Confidence intervals for some adjusted estimates are wide, indicating imprecision.

Nonetheless, the 22-year study period and relatively large single-center cohort (the largest Aeromonas septicemia series reported from the Middle East) strengthen clinical observations and provide robust characterization of regional epidemiology and resistance patterns. Cross-center validation in prospective cohorts is warranted.

4.8. Clinical and public health implications

Our findings suggest several actionable strategies to reduce Aeromonas septicemia-associated mortality in tertiary care settings.

• 1.Rapid diagnostic approaches: Prioritize MALDI-TOF mass spectrometry for Aeromonas identification (achievable within 3–4 h of positive blood culture), enabling faster antimicrobial stewardship decisions vs. traditional biochemical identification (24–48 h).
• 2.Empirical therapy optimization: In immunocompromised populations (particularly cancer patients on active chemotherapy, renal failure patients), initial empirical sepsis coverage should include aminoglycosides (amikacin, gentamicin) or fluoroquinolones (ciprofloxacin, levofloxacin) with known excellent Aeromonas activity. Carbapenems remain acceptable, but regional resistance rates (33% imipenem) mandate rapid susceptibility testing.
• 3.Antimicrobial surveillance: Institutions should establish local resistance surveillance tracking Aeromonas susceptibility patterns to guide institutional antibiotic stewardship and empirical therapy algorithms. Our 33% carbapenem resistance rate is concerning and suggests informal surveillance may be insufficient.
• 4.Hospital water system surveillance: Given the published association between Aeromonas nosocomial infections and water system contamination (Gray et al., 2024), and our observation of increasing case detection in recent years, tertiary care facilities should assess water system integrity and consider environmental screening for Aeromonas and carbapenemase-producing isolates.
• 5.Molecular epidemiology investigation: Next-generation projects should incorporate molecular typing and resistance gene profiling to characterize clonal relationships, transmission dynamics, and specific resistance mechanisms driving the observed phenotypic patterns.

5. Conclusions

Aeromonas septicemia primarily affects immunocompromised tertiary care patients, with cancer and diabetes as predominant risk factors. Regional antimicrobial resistance patterns are substantial, particularly for β-lactams and carbapenems. Critically, empirical therapy mismatch independently predicts sepsis-related mortality, emphasizing the clinical imperative for rapid organism identification and susceptibility-guided therapy. Future prospective, multicenter studies incorporating molecular characterization and real-time diagnostics are warranted to further elucidate resistance epidemiology and optimize clinical outcomes.

Ethical approval

This study was approved by the institutional review board (SP24J-024-03) at King Abdullah International Medical Research Center (KAIMRC). Patient consent was waived due to the retrospective study design.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work the authors used Claude in order to improve the language, grammar, and coherence. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Funding

This research received no specific grant from any funding agency.

CRediT authorship contribution statement

Abdullah Awadh: Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. Adel Sheerah: Formal analysis, Investigation, Visualization, Writing – original draft. Mohanned Alsaeed: Formal analysis, Investigation, Visualization, Writing – original draft. Abdulrahman Abbas: Formal analysis, Investigation, Visualization, Writing – original draft. Wadee Filimban: Formal analysis, Investigation, Visualization, Writing – original draft. Abdullah Bazayd: Formal analysis, Investigation, Visualization, Writing – original draft. Abdulfattah Alamri: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Writing – original draft.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

APPENDIX. KEY ENHANCEMENTS SUMMARY

Major Revisions Addressing Reviewer Comments.

Domain	Original	Revised
Tables	4 tables; minimal stratification	6 tables; temporal trends, subgroup analysis, statistical outputs
Statistical Analysis	Descriptive percentages only	Univariate and multivariate logistic regression; adjusted odds ratios; 95% CI
Mechanisms	1-2 sentences on resistance	Comprehensive section with β-lactamase types, carbapenemase, efflux pumps
Study Design Discussion	Brief mention of limitations	Explicit bias analysis: surveillance bias, referral bias, temporal bias, selection bias
Molecular Gap	1 sentence	Full subsection explaining phenotypic limitations, molecular profiling needs
Literature Integration	Surface-level citations	Comparative analysis across studies; regional epidemiologic positioning; mechanistic contextualization
Depth	3-page Discussion	5-page Discussion with clinical implications, public health recommendations

Data availability

Data supporting the findings of this study are available from the corresponding author upon reasonable request.