1. Background
Ranking as the fourth most common cause of nosocomial bloodstream infections (BSIs) (1), candidemia has been surging as a daunting challenge to both clinicians and infection control practitioners. In the neonatal intensive care unit (NICU), candidemia has been the third most common cause of late-onset sepsis, accounting for 9% - 13% of the neonatal BSIs (2).
Because of their immature immune system, neonates are particularly vulnerable to invasive candidemia. Incriminated factors include prolonged endotracheal intubation, total parenteral nutrition (TPN), indwelling venous catheterization, and broad-spectrum antibiotics (3). While Candida albicans has long been accountable for the majority of candidemia cases, non-albicans Candida spp. (NAC) are now soaring on a global scale (4).
Early initiation of antifungal therapy would have a huge impact on the disease prognosis. However, due to the lack of specific signs and symptoms, clinching a definite diagnosis of candidemia is a tedious process. This, in turn, has led to a substantial surge in morbidity, mortality, and health-care expenditures (5).
2. Objectives
In this study, we aimed to evaluate the risk factors predisposing to neonatal candidemia with special emphasis on non-albicans candidemia. Hence, we endeavored to investigate the incidence, risk factors, antifungal susceptibility, and the case fatality rate of candidemia patients in the NICU of a tertiary care hospital.
3. Methods
3.1. Ethical Consideration
Before commencement of the study, the protocol was approved by the Ethics Committee of the Pediatrics and Neonatology Department, Cairo University, Cairo, Egypt.
3.2. Study Population
This cross-sectional analytical study was conducted from October 2015 throughout October 2017 in the Departments of Pediatrics and Neonatology, Clinical Pathology, and Microbiology at Kasralainy Tertiary Care Hospital, Cairo, Egypt. Out of 1296 neonates admitted to the NICU at Al-Mounira Pediatric University Hospital with a clinical suspicion of septicemia, 214 neonates had positive blood culture results and were included in this study. Prior to enrollment, informed consents were obtained from the parents or guardians. All participating neonates were subjected to history taking and clinical examination.
The following data were gathered: Admission age; gestational age; mode of delivery; premature rupture of membranes (PROM); gender; birth weight; underlying maternal illness; hypoglycemia; hyperglycemia; metabolic acidosis; seizures; poor feeding; skin rash; hepatomegaly; invasive and non-invasive ventilation; pneumonia; TPN; necrotizing enterocolitis (NEC); intracranial hemorrhage (ICH); central venous catheterization; chest intubation; steroid administration; bronchopulmonary dysplasia (BPD); duration of hospitalization; use of H2 blockers; use of carbapenems, vancomycin, fluoroquinolones, and other antibiotics; antifungal drug prophylaxis; and the outcome of candidemia.
3.3. Microbiological Methods
From each of the 1296 neonates, two blood samples were collected into pediatric Bactec blood culture bottles (BD Diagnostic Systems). The bottles were transported to the laboratory and immediately incubated in the blood culture systems (BD BACTEC™ 9050). Any growth was subcultured onto 5% sheep blood agar, MacConkey’s agar, and Sabouraud’s dextrose agar (SDA) with chloramphenicol (0.05%) (Oxoid Co. England) and then incubated at 37°C. The diagnosis of candidemia was settled by the presence of at least one positive blood culture showing pure growth of Candida spp. together with supportive clinical findings (6). The isolated Candida spp. were identified by colony morphology on SDA, germ tube test (7), and chromogenic media (CHROMagar™ Candida).
The minimum inhibitory concentrations (MICs) of fluconazole (FLU), itraconazole (ITR), voriconazole (VOR), amphotericin B (AMB), and caspofungin (CSP) were determined using the E-test (bioMérieux-SA, France). The guidelines for the in vitro susceptibility of Candida spp. to FLU, ITR, and VOR were adopted from the M27-A3 document of the Clinical and Laboratory Standards Institute (8), while literature guidelines for AMB were adopted from Sanita et al. (9) and Negri et al. (10). The MIC results for CSP were interpreted using revised clinical breakpoints for echinocandins (11, 12). Quality control strains (Candida albicans ATCC® 10231, C. albicans ATCC 90028, C. parapsilosis ATCC 22019, and C. krusei ATCC 6258) were supplied by Microbiologics 200 Cooper Avenue North, St. Cloud, MN 56303, United States.
3.4. Statistical Analysis
Data management and statistical analysis were performed using the statistical package for social sciences (SPSS) version 21. Numerical data were summarized using means and standard deviations or medians and ranges. Categorical data were summarized as percentages. Comparisons between groups with respect to normally distributed numeric variables were made using the t-test. Non-normally distributed numeric variables were compared by the Mann-Whitney test. For categorical variables, differences were analyzed with the Chi-square test and Fisher’s exact test. Comparisons between the different prognostic factors were made using the Logrank test. All significant variables were entered into the Cox Proportional Hazards Model to detect the most important factors that independently affected the outcome. All P values were two-sided. A P value < 0.05 was considered significant.
4. Results
Out of 214 neonates with positive blood cultures, 139 (65%) patients were females. The male to female ratio was 1:1.8.
Upon cultivation, pure Candida growth was retrieved from 32/214 cases (15%) while 182/214 (85%) neonates had a bacterial BSI. In the candidemia group, C. albicans was isolated from only 25% of the cases while NAC spp. were responsible for 75% of the cases, with C. tropicalis being the most frequently retrieved isolate (43.8%). Meanwhile, C. krusei was isolated from 6.2% of the cases and other Candida spp. from 25% of the cases.
The MICs of AMB, ITR, FLU, and VOR, as well as CSP, were determined using the E-test (Table 1).
Table 1.
Antifungal Drug Susceptibility Results as Revealed by the E-test
| Drug | MIC range, µg/mL | Sensitivea | Intermediatea | Resistanta |
|---|---|---|---|---|
| AMB | 0.023 - 32 | 28 (87.5) | - | 4 (12.5) |
| ITR | 0.016 - 32 | 20 (62.5) | 2 (6.25) | 10 (10.25) |
| FLU | 0.25 - 256 | 26 (81.25) | 2 (6.25) | 4 (12.5) |
| VOR | 0.016 - 32 | 24 (75) | 2 (6.25) | 6 (18.75) |
| CSP | 0.003 - 32 | 22 (68.75) | 4 (12.5) | 6 (18.75) |
Abbreviations: AMB, amphotericin B; CSP, caspofungin; FLU, fluconazole; ITR, itraconazole; MIC, minimum inhibitory concentration; VOR, voriconazole.
aValues are expressed as No. (%).
In the candidemia group, the mean gestational age was 34 weeks while the mean birth weight was 2.25 kg. The demographics of albicans candidemia patients are compared with those of NAC patients in Table 2. Of note, a statistically significant difference was found between the two groups concerning the age at admission, as well as the birth weight.
Table 2.
Demographics of the Enrolled Neonatesa
| Variables | Bacteremia Group | Candidemia Group | P Value | Albicans Candidemia Subgroup | NAC Subgroup | P Value |
|---|---|---|---|---|---|---|
| Age at admission, d | 4 (1 - 14) | 4 (1 - 14) | 0.717 | 7 (1 - 14) | 3 (1 - 12) | 0.021 |
| Length of hospital stay | 36 (2 - 88) | 40 (2 - 88) | 0.347 | 32 (6 - 82) | 43 (2 - 88) | 0.383 |
| Gestational age, wk | 34 ± 4 | 34 ± 3 | 0.370 | 37 (36 - 38) | 35 (27 - 38) | 0.063 |
| Birth weight, kg | 2.11 ± 0.82 | 2.25 ± 0.9 | 0.397 | 2.85 (1.9 - 3.7) | 1.93 (0.9 - 3.9) | 0.023 |
| Male gender | 62 (82.7) | 13 (17.3) | 0.473 | 4 (30.8) | 9 (69.2) | 0.684 |
aValues are expressed as median (range), mean ± SD and No. (%).
In the meantime, the potential predisposing factors and clinical characteristics were compared between the candidemia and bacteremia groups (Table 3). Risk factors incriminated in candidemia were antifungal (FLU) prophylaxis and antibiotic use (fluoroquinolones).
Table 3.
Potential Risk Factors in Neonates with Candidemia Compared to Those with Bacterial BSIa
| Variables | Candidemia Group | Bacteremia Group | P Value |
|---|---|---|---|
| Mode of delivery | |||
| NVD | 18 (15.9) | 95 (84.1) | 0.672 |
| CS | 14 (13.9) | 87 (86.1) | |
| PROM | 6 (11.8) | 45 (88.2) | 0.464 |
| Maternal illness | 16 (15.0) | 91 (85.0) | 1.0 |
| Signs of sepsis | |||
| RD | 22 (14.9) | 126 (85.1) | 0.957 |
| Hypotension | 26 (14.9) | 148 (85.1) | 0.993 |
| Poor capillary perfusion | 26 (14.9) | 148 (85.1) | 0.993 |
| Edema | 8 (14.3) | 48 (85.7) | 0.870 |
| Hypoglycemia | 14 (15.1) | 79 (84.9) | 0.971 |
| Hyperglycemia | 22 (15.1) | 124 (84.9) | 0.945 |
| Metabolic acidosis | 26 (14.9) | 148 (85.1) | 0.993 |
| Seizures | 10 (15.2) | 56 (84.8) | 0.957 |
| Poor feeding | 22 (14.9) | 26 (85.1) | 0.957 |
| Hepatomegaly | 8 (14.5) | 47 (85.5) | 0.922 |
| Jaundice | 18 (15.1) | 101 (84.9) | 0.937 |
| Pneumonia | 26 (15.3) | 144 (84.7) | 0.783 |
| Invasive ventilation | 24 (14.9) | 137 (85.1) | 0.947 |
| TPN | 23 (17.3) | 110 (82.7) | 0.219 |
| Intercostal tube insertion | 18 (15.3) | 100 (84.7) | 0.942 |
| NEC | 18 (14.9) | 103 (85.1) | 0.971 |
| ICH | 12 (17.6) | 56 (82.4) | 0.451 |
| BPD | 14 (15.1) | 79 (84.9) | 0.971 |
| Antenatal steroids | 14 (14.9) | 80 (85.1) | 0.983 |
| Antimicrobial therapy before culture results | |||
| Aminoglycoside use | 2 (20) | 8 (80) | 0.647 |
| Amikacin use | 14 (14.6) | 82 (85.4) | 0.891 |
| Fluoroquinolone use | 15 (10.8) | 124 (89.2) | 0.020 |
| Carbapenem use | 26 (14.9) | 148 (85) | 0.993 |
| Vancomycin use | 26 (14.8) | 150 (85.2) | 0.873 |
| Fluconazole prophylaxis | 24 (55.8) | 19 (44.2) | <0.001 |
| Central venous line | 20 (15.0) | 113 (85) | 0.965 |
| Central venous line duration, median (range) | 10 (5 - 32) | 10 (5 - 32) | 0.944 |
Abbreviations: BPD, bronchopulmonary dysplasia; CS, cesarean section; ICH, intracranial hemorrhage; NEC, necrotizing enterocolitis; NVD, normal vaginal delivery; PROM, premature rupture of membranes; RD, respiratory distress; TPN, total parenteral nutrition.
aValues are expressed as No. (%) unless otherwise indicated.
On the other hand, NAC was more likely to develop in neonates subjected to invasive ventilation, as well as neonates with prolonged venous catheterization (Table 4).
Table 4.
Risk Factors and Clinical Characteristics of albicans Candidemia Compared to NACa
| Variables | Albicans Candidemia | NAC | P Value |
|---|---|---|---|
| Mode of delivery | |||
| NVD | 2 (11.1) | 16 (88.9) | 0.096 |
| CS | 6 (42.9) | 8 (57.1) | |
| PROM | 0 (0) | 6 (100) | 0.296 |
| Maternal illness | 4 (25) | 12 (75) | 1 |
| Signs of sepsis | |||
| RD | 4 (18.2) | 18 (81.8) | 0.186 |
| Hypotension | 4 (15.4) | 22 (84.6) | 0.023 |
| Poor capillary perfusion | 6 (23.1) | 20 (76.9) | 0.625 |
| Edema | 0 (0) | 8 (100) | 0.059 |
| Hypoglycemia | 4 (28.6) | 10 (71.4) | 0.073 |
| Hyperglycemia | 6 (27.3) | 6 (72.7) | 0.66 |
| Metabolic acidosis | 6 (23.1) | 20 (76.9) | 0.625 |
| Seizures | 2 (20) | 8 (80) | 0.66 |
| Poor feeding | 4 (18.2) | 18 (81.8) | 0.186 |
| Hepatomegaly | 0 (0) | 8 (100) | 0.056 |
| Jaundice | 4 (22.2) | 14 (77.8) | 0.703 |
| Pneumonia | 4 (15.4) | 22 (84.6) | 0.023 |
| Invasive ventilation | 4 (16.7) | 20 (83.3) | 0.059 |
| TPN | 6 (26.1) | 17 (73.9) | 0.82 |
| Intercostal tube insertion | 4 (22.2) | 14 (77.8) | 0.703 |
| NEC | 2 (11.1) | 16 (88.9) | 0.96 |
| ICH | 0 (0) | 12 (100) | 0.011 |
| BPD | 2 (14.3) | 12 (85.7) | 0.412 |
| ntenatal steroids | 2 (14.3) | 12 (85.7) | 0.412 |
| Antimicrobial therapy before culture results | |||
| Aminoglycoside use | 0 (0) | 2 (100) | 1 |
| Amikacin use | 2 (14.3) | 12 (85.7) | 0.421 |
| Fluoroquinolone use | 4 (26.7) | 11 (73.3) | 1 |
| Carbapenem use | 6 (23.1) | 20 (76.9) | 0.625 |
| Vancomycin use | 6 (23.1) | 20 (76.9) | 0.625 |
| Fluconazole prophylaxis | 6 (25) | 18 (75) | 1 |
| Central venous line | 6 (30) | 14 (70) | 0.399 |
| Central venous line duration, median (range) | 7 (5 - 10) | 15 (7 - 32) | 0.007 |
Abbreviations: BPD, bronchopulmonary dysplasia; CS, cesarean section; ICH, intracranial hemorrhage; NEC, necrotizing enterocolitis; NVD, normal vaginal delivery; PROM, premature rupture of membranes; RD, respiratory distress; TPN, total parenteral nutrition.
aValues are expressed as No. (%) unless otherwise indicated.
In the candidemia group, the case fatality rate was 56.2%, which was quite close to that of the bacteremia group (56%). Meanwhile, the case fatality rate was slightly higher in NAC (58.3%, 14/24) than in albicans candidemia (50%).
5. Discussion
Candida spp. have been emerging as leading pathogens in neonatal BSIs, pertaining to substantial morbidity and mortality (13). Therefore, it is crucial for both physicians and microbiologists to promote the prevention, early detection, and prudent management of candidemia in high-risk neonates.
In this study, 65% of the patients were females, with a male to female ratio of 1:1.8. Consistent with our finding, Juyal et al. (6) reported that 60.4% of their patients were females, with a male to female ratio of 1:1.5.
In this study, candidemia accounted for 15% of the neonatal BSIs cases. Corroborating our finding, Jain et al. (14) reported that candidemia accounted for 15.8% of the neonatal BSIs. Nonetheless, a higher incidence of 20.4% was reported by Rao et al. (15).
Out of the 32 retrieved Candida isolates, 75% were NAC. In line with our result, Juyal et al. (16) reported NAC as responsible for 80.6% of neonatal candidemia.
Meanwhile, C. tropicalis was the most commonly encountered Candida spp. (43.8%), followed by C. albicans (25%). This was in accordance with Yadav et al. (17) who stated that C. tropicalis was the most frequent Candida spp. (26.9%), followed by C. albicans (11.5%). On the other hand, Jain et al. (14) reported that C. albicans was the most common Candida spp. (37.6%).
Among the 32 Candida isolates, 87.5% were susceptible to AMB, 62.5% to ITR, 81.25% to FLU, 75% to VOR, and 68.75% to CSP. That was lower than the results of earlier studies (4, 17), which revealed that 100% of Candida isolates were sensitive to FLU, VOR, ITR, and AMB.
Notably, numerous risk factors including broad-spectrum antibiotics have been reported as predisposing factors for candidemia (18). The widespread use of antimicrobial agents can suppress the bacterial microbiota, paving the way for the dominance of Candida colonization (19). In this study, fluoroquinolone empiric use and fluconazole prophylaxis proved to be significant risk factors; however, none of the other potential risk factors proved significant.
In an earlier study (3), central venous catheterization and TPN were predisposing factors for candidemia. On the other hand, Yadav et al. (17) stated that prolonged antibiotic use, very low birth weight, and prematurity were significant risk factors. This myriad of the findings is attributable in part to variations in host factors and infection control practices, and they notably reflect the epidemiological shift of Candida spp. (20).
In the meantime, factors contributing to the development of NAC were low birth weight, the age at admission, invasive ventilation, and the duration of central venous catheterization. In accordance with that, Garzillo et al. (21) stated that NAC was linked to birth weight, gestational age, and the duration of central venous catheterization, but not to the length of hospital stay. These results were in line with those of Juyal et al. (6) who reported low birth weight and indwelling catheters as contributory factors to NAC; yet, they added broad-spectrum antibiotics as a risk factor.
Indeed, several studies (7, 16) have drawn attention to the tendency of Candida spp. to adhere to foreign materials. Some studies have proven that non-albicans Candida have even a higher capability of such adherence (22). Once they adhere to fibrinogen or platelets on a venous catheter, they can form a biofilm. On the one hand, this biofilm protects the fungus against immune responses and antifungal agents and on the other hand, it acts as a reservoir for systemic spread (19).
The case fatality rate in the candidemia group was 56.2%, while in the bacteremia group, it was 56%. This was close to the fatality rate of 54% reported by Hammoud et al. (23). Meanwhile, the case fatality rate in the albicans candidemia group was 50%, while in the NAC group, it was 58.3%. This was in accordance with the results of Al Thaqafi et al. (24) who reported the mortality rate in albicans candidemia as 50% and in NAC as 57.8%.
It has been observed that drug resistance in Candida spp. is rare and thus an unlikely cause of treatment failure. This is exemplified by the absence of a clear correlation between susceptibility patterns and patients’ outcomes (25). Hence, studies have reiterated the importance of host factors in the failure of antifungal therapy. One of these factors is the immaturity of the neonates’ immune system that may render them unable to eliminate a fungus despite that its growth has been inhibited (26). The higher fatality in NAC patients may be attributable to the underlying medical conditions of the neonates (21); however, further studies are warranted to investigate the mortality risk factors.
Worth mentioning, the current study is limited by having a single center design with a relatively small sample size; hence, it might be difficult to generalize the findings to various settings. Nonetheless, the study emphasizes the need for vigilant surveillance and monitoring of neonates admitted to the ICU. In order to attain timely and actionable results, judicious antibiotic stewardship and early antifungal treatment should essentially be implemented in all NICUs.
