1. Background
Sphingomonas paucimobilis (formerly Pseudomonas paucimobilis) is an aerobic, non-fermentative, Gram-negative bacillus characterised by production of a yellow pigment known as nostoxanthin, slow motility with a single polar flagellum, and positive oxidase and catalase reactions (1). Sphingomonas paucimobilis is distributed widely in nature, especially in water and soil (2). It is an oligotrophic bacterium and can be isolated from low-nutrient environments, such as drinking water distribution systems, water treatment plants, tap water, and water demineralisation filters and biofilms collected from water system (3-6). This bacterium was even isolated from water supply and humidity condensate samples gathered from the international space station (7).
In hospital settings, S. paucimobilis is seen in sporadic cases of various infections, such as catheter-related sepsis, primary bacteraemia, pneumonia, meningitis, peritonitis, septic arthritis, and urinary tract infections, and small outbreaks caused by contaminated hospital devices and contaminated intravenous fluids (8-17). The presence of indwelling devices, an impaired immune system, and co-morbidities such as malignancy and diabetes mellitus are reported risk factors for S. paucimobilis infection (14, 15, 18, 19). Biofilm production is a well-known bacterial virulence factor and causes treatment failure, particularly in patients with indwelling devices (20). Biofilm production by S. paucimobilis has been investigated extensively in non-clinical, but not clinical, isolates (4, 6, 21, 22).
2. Objectives
This is the first study to assess biofilm production by S. paucimobilis clinical isolates, together with patient demographics, clinical aspects, risk factors, and outcomes.
3. Methods
3.1. Patients
The study was conducted at in a 750-bed Training and Research hospital in Sakarya/Turkey from January 2010 to December 2014 with the ethical approval number 71422473/050.01.04/54. Forty-three non-duplicate S. paucimobilis isolates were identified from clinical samples, but ten of the samples were excluded from the study because they were deemed either colonisation or contamination based on clinical signs and laboratory and radiographic findings. Consequently, 33 isolates from patients clinically determined to have S. paucimobilis infections were enrolled in the study. Clinical and demographic data for these patients were collected retrospectively from the hospital records.
3.2. Bacteriological Studies
Conventional cultivation methods were used to isolate S. paucimobilis from clinical specimens. Identification and antimicrobial susceptibility tests were performed using a VITEK II Automated System (bioMérieux, Marcy L’Eoile, France). The EUCAST 2014 criteria were used to assess antimicrobial resistance. The strains were stored at -80°C as stock cultures until biofilm production was evaluated.
3.3. Detection of Biofilm Formation With Tube Method
The tube method described by Christensen et al. in 1982 (23) and 1985 (24) was used to assess biofilm formation qualitatively, with some modifications as proposed by Stepanovic et al. (25). Sphingomonas paucimobilis isolates were recultivated from stock cultures and incubated for 24 hours at 37°C. A loopful of freshly grown colonies was inoculated into glass tubes (13 × 10 cm) containing 3-mL tryptic soy broth (Merck, Darmstadt, Germany) with 2.5% glucose (Merck, Darmstadt, Germany). The tubes were incubated overnight at 37°C under aerobic conditions. The content of each tube was removed carefully using an automatic pipette. Safranin solution (2 mL, 0.25%) from the Gram staining set was then added. After staining for 1 minute, the safranin was removed using a pipette. The tubes were placed upside down for 24 hours at room temperature. No wash steps were performed during the procedure. A non-inoculated tube containing tryptic soy broth was used as the negative control. The well-known biofilm producer Escherichia coli ATCC 25922 was used as the positive control (20). Biofilm formation was accepted as positive when adherent film material was attached to the inner wall and bottom of the tube. Colourful ring formation at the air-liquid interface was not considered. According to the intensity of the adherent dye, the tubes were scored as 0 (absent), +1 (low), +2 (moderate), and +3 (high). Experiments were repeated three times for each isolate.
3.4. Statistical Analyses
Groups were compared using the chi-square or Fisher’s exact test for categorical variables and Student’s t-test for continuous variables. A P < 0.05 was considered to indicate significance. Statistical analyses were performed using SPSS ver. 21.0 (SPSS, Chicago, IL, USA).
4. Results
Sphingomonas paucimobilis infection was identified in 33 patients (17 females, 16 males). The mean age of the patients was 61.2 years, ranging from a 1-year-old male to a 78-year-old female. The infections included 27 nosocomial and six community-acquired infections. Sphingomonas paucimobilis was isolated most commonly from blood samples (n = 16) as primary bacteraemia. In addition, seven cases of pneumonia, six cases of wound infection, and four cases of urinary tract infection were detected. Twenty-four of the patients were followed in the intensive care unit. Twenty-two patients had indwelling catheters. Malignancy (n = 11) and diabetes mellitus (n = 10) were the most common concomitant diseases. Biofilm formation by all isolates was determined and the degree of biofilm formation and other data are summarised in Table 1.
Table 1. Demographic Data and Clinical Characteristics of the S. paucimobilis Infections
| Age | Sex | Nosocomial | Sample | Biofilm Formation | ICU Stay | Indwelling Catheter | Surgery | Steroid Use | Concominant Disease | |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 69 | F | - | Urine | +++ | - | + | - | - | - |
| 2 | 64 | M | + | Blood | +++ | + | + | - | - | MA |
| 3 | 53 | M | - | Wound | ++ | - | - | - | - | - |
| 4 | 86 | M | + | Blood | +++ | + | + | + | - | DM , ARF |
| 5 | 68 | F | + | Blood | ++ | + | + | - | + | MA |
| 6 | 64 | F | + | Tr. Asp. | +++ | + | + | - | - | CVD |
| 7 | 64 | M | + | Blood | + | - | + | - | - | - |
| 8 | 77 | M | - | Wound | + | - | - | + | - | - |
| 9 | 71 | F | + | Blood | ++ | + | - | - | - | HT,DM, CVD |
| 10 | 62 | F | + | Urine | + | + | + | - | - | HT ,DM |
| 11 | 72 | F | + | Tr. Asp. | +++ | + | + | - | - | CAD, DM |
| 12 | 47 | M | - | Wound | ++ | - | - | - | - | MA |
| 13 | 59 | M | - | Wound | + | + | - | - | - | MA, DM |
| 14 | 54 | F | - | Urine | ++ | - | + | - | - | CRI |
| 15 | 63 | F | + | Tr. Asp. | ++ | + | - | - | - | MA |
| 16 | 49 | M | + | Blood | +++ | + | + | - | - | MA |
| 17 | 78 | M | - | Wound | + | - | + | + | - | - |
| 18 | 49 | M | + | Blood | +++ | + | + | - | - | MA |
| 19 | 59 | M | + | Blood | ++ | - | + | - | - | MA , DM |
| 20 | 15 | F | + | Urine | +++ | - | - | - | - | - |
| 21 | 79 | F | + | Blood | +++ | + | + | - | + | MA |
| 22 | 77 | F | + | Blood | ++ | - | + | - | - | DM, HT |
| 23 | 78 | F | + | Tr. Asp. | +++ | + | - | - | - | - |
| 24 | 79 | M | + | Blood | +++ | + | + | + | - | - |
| 25 | 67 | M | + | Tr. Asp. | ++ | + | - | - | - | HT, DM, CAD |
| 26 | 69 | F | + | Wound | + | - | + | - | - | - |
| 27 | 49 | F | + | Tr. Asp. | ++ | + | - | - | - | - |
| 28 | 59 | M | + | Blood | +++ | + | + | - | + | MA |
| 29 | 59 | F | + | Tr. Asp. | ++ | + | + | - | - | - |
| 30 | 1 | M | + | Blood | +++ | + | + | - | - | - |
| 31 | 54 | F | + | Blood | ++ | + | + | - | - | DM |
| 32 | 52 | F | + | Blood | +++ | + | - | - | - | COLD, HT |
| 33 | 75 | M | + | Blood | +++ | + | + | + | - | MA, DM |
Abbreviations: F, female; M, male; Tr. Asp., trakeal aspiration; MA, malignancy; DM, diabetes mellitus; ARF, acute renal failure; CRI, chronic renal failure; HT, hypertension; COLD, chronic obstructive lung disease; CVD, cerebrovascular disease; CAD, coronary artery disease.
Table 2 shows the antimicrobial susceptibility test results. Tigecycline (84.85%) was the antimicrobial agent to which the isolates were most frequently susceptible, together with carbapenems and aminoglycosides, whereas fewer isolates were susceptible to ampicillin (21.21%), ampicillin/sulbactam (24.24%), cefuroxime (24.24%), and cefuroxime axetil (24.24%).
Table 2. Antimicrobial Susceptibilities of the S. paucimobilis Isolates
| Antimicrobial Agent | S. paucimobilis ,n = 33 | |||||
|---|---|---|---|---|---|---|
| Resistant | Intermadiate | Sensitive | ||||
| Amikacin | 8 | 24.24 | 2 | 6.06 | 23 | 69.70 |
| Amoxicillin-clavulanate | 23 | 69.70 | 0 | 0.00 | 10 | 30.30 |
| Ampicillin | 25 | 75.76 | 1 | 3.03 | 7 | 21.21 |
| Ampicillin-sulbactam | 23 | 69.70 | 1 | 3.03 | 9 | 27.27 |
| Cefazolin | 23 | 69,70 | 2 | 6.06 | 8 | 24,24 |
| Cefepime | 19 | 57.58 | 0 | 0.00 | 14 | 42.42 |
| Cefoperazone-sulbactam | 21 | 63.64 | 2 | 6.06 | 10 | 30.30 |
| Ceftazidime | 22 | 66.67 | 0 | 0.00 | 11 | 33.33 |
| Cefuroxime | 25 | 75.76 | 0 | 0.00 | 8 | 24.24 |
| Cefuroxime-axetil | 25 | 75.76 | 0 | 0.00 | 8 | 24.24 |
| Ciprofloxacin | 22 | 66.67 | 0 | 0.00 | 11 | 33.33 |
| Ertapenem | 10 | 30.30 | 0 | 0.00 | 23 | 69.70 |
| Gentamicin | 10 | 30.30 | 2 | 6.06 | 21 | 63.64 |
| Imipenem | 8 | 24.24 | 0 | 0.00 | 25 | 75.76 |
| Levofloxacin | 18 | 54.55 | 0 | 0.00 | 15 | 45.45 |
| Meropenem | 8 | 24.24 | 0 | 0.00 | 25 | 75.76 |
| Netilmicin | 6 | 18.18 | 1 | 3.03 | 26 | 78.79 |
| Piperacillin | 24 | 72.73 | 0 | 0.00 | 9 | 27.27 |
| Piperacillin-tazobactam | 18 | 54.55 | 3 | 9.09 | 12 | 36.36 |
| Tigecycline | 5 | 15.15 | 0 | 0.00 | 28 | 84.85 |
| Trimethoprim-sulfamethoxazole | 20 | 60.61 | 0 | 0.00 | 13 | 39.39 |
All isolates produced biofilms. Table 3 shows the correlations of biofilm formation intensity with demographic data and clinical variables. The intensity of biofilm formation was correlated only with blood samples (P = 0.003) in the sample types group and a stay in the intensive care unit (P = 0.002) in the risk factors group. No other variable was correlated with biofilm formation intensity (P > 0.005).
Table 3. Correlation of Intensity of Biofilm Formation With Patient Characteristicsa
| Patient Characteristics | S. paucimobilis Infection | Biofilm Formation | P Value | ||
|---|---|---|---|---|---|
| +1 | +2 | +3 | |||
| Age, y | |||||
| 1 - 50 | 6 (18.18) | 0 | 3 (50.0) | 3 (50.0) | 0.098 |
| 50 - 70 | 17 (51.52) | 4 (23.53) | 9 (52.94) | 4 (23.53) | |
| over 70 | 10 (30.30) | 2 (20.00) | 1 (10.00) | 7 (70.00) | |
| Sex | |||||
| Female | 17 (51.52) | 2 (11.76) | 9 (52.94) | 6 (35.29) | 0.241 |
| Male | 16 (48.48) | 4 (25.00) | 4 (25.00) | 8 (50.00) | |
| Source of infection | |||||
| Nosocomial | 26 (78.79) | 3 (11.54) | 9 (34.62) | 14 (53.85) | 0.024 |
| Community | 7 (21.21) | 3 (42.86) | 4 (57.14) | 0 | |
| Sample | |||||
| Blood | 16 (48.48) | 1 (6.25) | 4 (25.00) | 11 (68.75) | 0.003 |
| Tracheal aspiration | 7 (21.21) | 0 | 4 (57.14) | 3 (42.86) | |
| Wound | 6 (18.18) | 4 (66.67) | 2 (33.33) | 0 | |
| Urine | 4 (12.12) | 1 (25.00) | 3 (75.00) | 0 | |
| Risk factors | |||||
| Indwelling catheter | 24 (72.73) | 4 (16.67) | 8 (33.33) | 12 | 0.346 |
| ICU stay | 22 (66.67) | 2 (9.09) | 6 (27.27) | 14 | 0.002 |
| Surgery | 5 (15.15) | 2 (40.00) | 0 | 3 | 0.262 |
| Steroid use | 3 (9.09) | 0 | 1 (33.33) | 2 | 0.792 |
| Concominant disease | |||||
| Malignancy | 11 (33.33) | 6 (54.55) | 4 (36.36) | 1 (9.09) | 0.507 |
| Diabetes mellitus | 10 (30.30) | 2 (20.00) | 4 (40.00) | 4 (40.00) | 0.977 |
| Hypertension | 5 (15.15) | 1 (20.00) | 2 (40.00) | 2 (40.00) | 0.990 |
| Chronic renal insufficiency | 3 (9.09) | 1 (33.33) | 1 (33.33) | 1 (33.33) | 0.774 |
| Coronary artery disease | 2 (6.06) | 0 | 1 (50.00) | 1 (50.00) | 0.778 |
| Cerebrovascular disease | 2 (6.06) | 0 | 0 | 2 (100.00) | 0.236 |
| Chronic obstructive lung disease | 1 (3.03) | 0 | 0 | 1 (100.00) | 0.497 |
aValues are expressed as No. (%).
5. Discussion
Sphingomonas paucimobilis was first isolated from the leg ulcer of a Japanese seaman in 1979 (26). Since then, many case reports, case series, and outbreaks have been presented. Clinical infections related to S. paucimobilis include bacteraemia, peritonitis, wound infections, adenitis, diarrhoeal disease, sepsis arthritis, osteomyelitis, and meningitis (8, 13, 18). Outbreaks have also been attributed to contaminated water sources or contaminated intravenous fluids (10, 16, 17, 27). Bacteraemia is the most common clinical infection caused by S. paucimobilis (8, 15, 18). In our series, 48.48% of the cases were identified as bacteraemia, followed by pneumonia, wound infections, and urinary tract infections. Most of the cases (78.79%) were nosocomial in origin. According to the reviews of Lin et al., Huesh et al., and Cheong et al. (8, 15, 19), most S. paucimobilis infections are healthcare associated. However, Toh et al. investigated 55 cases of S. paucimobilis infection and Bayram et al. evaluated 24 paediatric patients; both reported that the incidence of community-acquired infections was higher than that of nosocomial cases (14, 28). Since these studies assessed relatively few cases, it is impossible to reach a reliable conclusion regarding the source of such infections.
In many case reports, indwelling catheters and an immunosuppressed host were identified as risk factors for S. paucimobilis infections (8, 9, 18, 19, 28-30). Community-acquired infection, diabetes mellitus, and alcoholism were determined as risk factors for primary bacteraemia in the multivariate logistic regression presented by Toh et al. (14). Comorbidities such as malignancy and diabetes mellitus were also reported to be risk factors (18, 19). In our study, the most common concomitant diseases were malignancy (33.33%) and diabetes mellitus (30.30%). The presence of an indwelling catheter (72.73%) and hospitalisation in the intensive care unit (66.67%) were the most common risk factors in our study.
The antimicrobial susceptibility patterns of S. paucimobilis differ among studies. Tigecycline (84.85%) was the agent to which S. paucimobilis was most frequently susceptible in this study, but there are no data in the literature for comparison. Sphingomonas paucimobilis was also found to be sensitive to imipenem (75.76%) and meropenem (75.76%), netilmicin (78.79%), and amikacin (69.70%). Toh et al. and Bayram et al. (14, 28) reported that fluoroquinolones, carbapenems, and trimethoprim/sulfamethoxazole were the most effective antibiotics. Lin et al. stated that S. paucimobilis was frequently sensitive to fluoroquinolones, carbapenems, and beta-lactam/beta-lactamase combinations (19). Conversely, in our study, cephalosporins and beta-lactam/beta-lactamase were found to be less effective. In our study, the antimicrobials most commonly used for S. paucimobilis infections were carbapenems and third-generation cephalosporin/aminoglycoside combinations. With the exception of two patients who died of clinical syndromes other than infections, the remaining 31 patients were treated successfully. Some researchers suggest third-generation cephalosporin/aminoglycoside combinations for the treatment of this type of infection (8, 18, 19). However, standardised therapies cannot be established at present, because of the different antimicrobial susceptibility patterns among studies. Relying on antimicrobial susceptibility testing results is the most logical and appropriate approach, as always.
Demographic data showed no correlation with the degree of biofilm formation. However, biofilm formation intensity was significantly higher in primary bacteraemia (P < 0.005). Intensive care unit stay was also positively correlated with a high degree of biofilm production. Higher biofilm production is expected to correlate with the presence of an indwelling catheter, but that was not the case in our study (P = 0.346). This situation might be due to the relatively small number of isolates. There was no correlation between the degree of biofilm formation and presence of concomitant diseases. Studies including more patients and isolates are needed to explain the correlation of biofilm production with such variables.
Previously, this bacterium was regarded as being of low virulence (18, 19, 28, 31). In a recent study, however, the S. paucimobilis virulence factors were reported to resemble those of Pseudomonas spp. These factors included proteases (alkaline protease, LasA, LasB), adherence factors (lipopolysaccharide, type IV pili), iron uptake (pyoverdin, pyochelin), a biosurfactant (rhamnolipid), and antiphagocytosis factors (alginate) (32). Type IV pili and alginate production are also involved in biofilm formation. We determined biofilm production in all clinical isolates of S. paucimobilis included in this study. Therefore, biofilm production by S. paucimobilis can be accepted as an important virulence factor, especially in patients in intensive care units and those with suspected primary bacteraemia.
5.1. Conclusion
Our study is the first research article to demonstrate biofilm production by clinical S. paucimobilis isolates. Sphingomonas paucimobilis infections are not uncommon as has been thought, and can cause serious infections, especially in immunocompromised patients with determined risk factors, due to the effects of multiple virulence factors, together with biofilm formation.
