A total of 266 clinically suspected cases of neonatal sepsis due to Candida species were studied from January 2007 to June 2007. Candida species were isolated from 49 (19.14%) patients. Candida glabrata was the commonest isolate in 30 patients (61.22%) followed by Candida parapsilosis in 10 cases (20.40%), Candida albicans in 6 cases (12.24 %), Candida krusei in 2 cases (4.08%), while Candida tropicalis in a single case (2.04 %). Inspite of giving intravenous Amphotericin B, six infants died due to Candida glabrata septicemia. Commonest predisposing factors observed were antibiotic therapy (100%), prematurity (100%), low birth weight (100%), respiratory distress syndrome (66.67%), patients on ventilators (66.67%), patients on central line catheters (33.33%), and urinary catheters (50%). Clinical features were nonspecific, difficult to differentiate from bacterial sepsis.
Keywords: Candida glabrata , Candida septicemia , neonatal intensive care unit.
Over the last two decades, yeasts have become important nosocomial pathogens. Candida species being the most frequent isolate. Importance of Candida species in nursery and Intensive care setup is increasingly being recognized. Candida species account for 9-13% of all blood isolates in neonatal intensive care units (NICU)., Though Candida albicans is the commonest species isolated, recently non-albicans Candida notably , Candida parapsilosis and Candida tropicalis have emerged as important opportunistic pathogens in NICU. ,, Infections due to Candida glabrata are characterized by high-mortality rates and they are difficult to treat due to reduced susceptibility of these species to azoles, specially fluconazole. ,,,,, Multiple risk factors have been identified, e.g. prior antibiotic therapy, presence of central line, endotracheal intubations, low birth weight, prematurity, prior fungal colonization, respiratory distress syndrome, and patients on ventilators. ,,,,,,,, Considering all these facts the present study was conducted to find out the prevalence of fungal septicemia; prevalence of Candida glabrata in septicemia, predisposing factors, and mortality.
A total number of 264 clinically suspected cases of neonatal septicemia were included in the study, from January 2007 to June 2007. Retrospective analysis of prevalence, predisposing factors; clinical outcome and speciation were done. For diagnosis of fungal septicemia, 1-2 ml of blood was collected from peripheral vein under aseptic precautions and inoculated in blood culture bottles containing 10-20 ml of brain heart infusion broth (BHIB). The bottles were inoculated at 37°C and on third, fifth, and seventh day subcultures were done on Sabouraud’s Dextrose agar slants. Speciation of Candida was done by germ tube test, growth patterns on Corn Meal agar, sugar assimilation tests. 
From total number of 264 suspected cases of neonatal sepsis Candida species were isolated in 49 cases (19.14%).
The commonest species identified was Candida glabrata in 30 (61.22%) cases, Candida parapsilosis in 10 cases (20.40%), Candida albicans in 6 cases (12.29%), Candida krusei in 2 cases (4.08%), and Candida tropicalis from a single case (2.04%). In our study, inspite of IV Amphotericin B administration, six infants succumbed to death due to Candida glabrata septicemia.
When clinical diagnosis was made, all the patients were started on Fluconazole, but following growth and speciation all patients were given intravenous Amphotericin B. All the patients were on intravenous Amphotericin B, inspite of that six infants died due to sepsis. From all these six infants, Candida glabrata was isolated from blood culture. Thus, crude mortality among 264 patients was 2.27%.
The signs and symptoms of patients were nonspecific. They include temperature instability (100%), refusal of feeds (100%), and respiratory distress (100%). Others were abdominal distension, lethargy, jaundice, and bradycardia in single case each.
The predisposing factors in patients from which Candida glabrata was isolated were prior antibiotic therapy (100%), low birth weight (100%), prematurity (100%), respiratory distress syndrome in 20 cases (66.67%), ventilation 20 cases (66.67%), patients on central line catheter in 10 cases (33.33%), and urinary catheters in 15 patients (50%). All these patients were on antibiotics since 7 days, presented with sepsis after 7 days (1 week).
The age of neonates at the time of admission varied between just born to 15 days. Male-to-female ratio was 2:1. All attempts were made to find out source of infection. These included fluid used for total parenteral nutrition, swab from ventilator, and swabs from health-care workers. No Candida species was isolated in any case, pointing toward a particular common source of infection.
Out of all patients from which Candida species were isolated, all responded to Amphotericin B except six infants. Out of 30 patients from which Candida glabrata was isolated, 10 patients were on central line catheter which was removed as probable risk factor and the patients survived.
There were no other Candida associated lesions observed in the patients.
In the present study, isolation rate observed was 18.6%. This was comparable with study conducted by Agarwal et al.  showing isolation rate 13.6% and another study conducted by Rani et al. ,  where isolation rate was 11%.
In the present study Candida glabrata was the commonest pathogen (61.22%) followed by Candida parapsilosis (20.40%), Candida albicans (12.24%), and Candida krusei (4.08%) , while in a single case Candida tropicalis was isolated.
In recent years, there is marked shift in isolation rates of non-albicans Candida species compared to Candida albicans in cases of neonatal sepsis. Kossoff et al.  showed significant shift from Candida albicans to non-albicans, i.e. Candida parapsilosis over 15 years. Rani et al.  observed Candida tropicalis as predominant pathogen (92%), followed by Candida albicans and Candida kefyr (4% each). The finding of Candida albicans as predominant pathogen in neonatal sepsis was noted earlier in year 2003 from Mumbai,  in which Candida albicans was isolated in 16 out of 30 isolates (53.3%). All these findings are in contrast to present study which showed that Candida glabrata is emerging as predominant cause of neonatal sepsis. This is in accordance with Karen et al. ,  which showed Candida glabrata as emerging pathogen. Many studies ,,,, show that though Candida albicans is the commonest cause of neonatal sepsis; Candida glabrata is the second most common cause of neonatal sepsis. Malani et al. has reported similar findings as noted in his study conducted at a tertiary care center. Candida glabrata caused 103 (16.91%) of 609 fungemic episodes and predisposing factors were antibiotics (86%), and use of central venous catheters (77%). The overall mortality at day 30 was 64%. The outcome appeared to be unrelated to whether fluconazole or Amphotericin B was administered.
Historically, Candida glabrata has been considered to be relatively nonpathogenic saprophyte of normal flora of healthy individuals rarely causing serious infections. However, following widespread and increased use of immunosuppressive therapy, broad spectrum antibiotic therapy, increased conditions causing compromise of the immune system, the frequency of mucosal as well as systemic infections caused by Candida glabrata has increased significantly. ,,, It is the only species of Candida that does not form pseudohyphae, it is found as blastoconidia (1-4 µm), both as commensal and as pathogenic states. On Sabouraud’s Dextrose agar), forms it forms glistening smooth, cream colored colonies, indistinguishable from Candida albicans but on Cornmeal agar it does not form pseudohyphae. It assimilates only glucose and trehalose. 
Though Candida glabrata has emerged as important nosocomial pathogen, yet little is known about its epidemiology.  Infection with this species is associated with high-mortality rate. ,,, Candida glabrata is of special importance because of its innately increased resistance to antifungal drugs, specially azoles. ,,, The reasons behind emergence of the species as predominant pathogen could be because of selection of lesser susceptible species due to frequent use of fluconazole as prophylaxis. 
Most reports describing the epidemiology, as present study have been retrospective, few studies have evaluated independent risk factors associated with nosocomial Candida glabrata acquisition and subsequent infection. In a recent multivariate prospective case control analysis along with molecular analysis of Candida glabrata demonstrated that patients with new acquisition of Candida glabrata had a longer duration of hospitalization (18.8 and 7.6 days, respectively; P < 0.001) compared to patients from whom Candida species were not recovered during study.  Similar predisposing factors were observed in the present study.
Little is known about the hospital reservoirs of Candida glabrata , but as with Candida albicans probable source include a complex interaction of environmental and human reservoirs. The unique role of hospital environment as a potential reservoir of Candida species is further suggested by findings in a recent study in which identical strains of Candida glabrata were isolated from environment before being newly acquired by patients admitted into a bone marrow transplant unit. Fungal organisms isolated from the inanimate hospital environment were previously considered to contribute little to nosocomial fungal infection. Although infecting strains can be cultured from environmental surfaces, it is believed that the environment becomes possibly contaminated by organisms from patients. Two studies have implicated carriage on the hands of hospital personnel as possible source of an outbreak. Thus, Candida glabrata may be similar to Candida albicans and other nosocomial pathogens that are acquired directly or indirectly from contaminated environmental surfaces. Previous understanding of the pathogenesis of Candida glabrata colonization and infection assumed that organisms responsible for disease were endogenously acquired exclusively from patient’s own flora.  In our study, no source from hospital environment or hospital personnel could be traced.
Probably infections due to Candida species are endogenous. It has been studied that about 10% babies in NICU get colonized in the first week of life and 64% babies get colonized by 4 weeks of hospital stay. The gastrointestinal tract is the first to become colonized though multiple sites may be involved.  There is some evidence showing correlation between fungal colonization and invasive disease in very low birth weight, premature babies.  This endogenous source, together with other predisposing factors as long-term antibiotics, catheterization, patients on ventilator, respiratory distress syndrome lead to candida septicemia in NICU. ,,,,,,,,,,,, Little is known about host defense against Candida glabrata. Recent studies suggest that innate resistance may also be critical for as virulence factors. There have been few clinical observations that provide some indication of what may be important for host defense against mucosal or systemic Candida glabrata mucosal or systemic infections in cancer patients, AIDS patients, transplant recipients, in whom T-cell function is impaired, suggesting that T cell may be important for protection against Candida glabrata infection. As in premature infants, T cells are not working normally, it acts as an important predisposing factor because of immaturity of immune apparatus. According to several workers, the increase in frequency of Candida glabrata infection has paralleled the increased use of fluconazole in hospitals,  as observed in the present study.
Candida may involve any organ, candidemia has diverse clinical picture ranging from low grade fever to fulminant septic shock, indistinguishable from bacterial sepsis. Similarly there is no unique feature associated with Candida glabrata. ,,,,,,,,,,,,,
Karen et al,  compared Candida glabrata sepsis with other. According to this significant study, Candida glabrata sepsis occurred in infants with higher gestational periods (29.7 weeks as against 26.6 weeks in case of Candida albicans ) and birth weight ( Candida glabrata 1442 g, Candida albicans 931 g), but no such difference was observed in the present study. The only finding in that study which matches with the present study is that broad spectrum antibiotics at the time of diagnosis acted as important predisposing factors.
A higher mortality associated with Candida glabrata is probably due to resistance to antifungals agents including Amphotericin B.
However, the relationship between fluconazole and emergence of fluconazole-resistant Candida glabrata remains controversial, as recently reported by Pfaller et al. ,  who suggested that factors other than antifungal usage, such as patient’s age, underlying disease and also geographical location, likely contribute to the emergence of Candida glabrata as a cause of blood stream infections. He for the first time pointed out that geographic location has also an impact on emergence of Candida glabrata infection.
In conclusion, previously ignored, Candida glabrata received little attention; therefore, not surprisingly our knowledge of it is not only incomplete, but also significantly lacking. We now need to have more studies and more tools; specially molecular tools to study the epidemiology of this emerging pathogen. Understanding the mechanisms of innate and acquired resistance may facilitate development of new targets for antifungal agents. More and more comprehensive studies of its epidemiology, pathogenesis, and resistance are needed to control the infections by Candida glabrata properly.
Source of Support: None, Conflict of Interest: None
Clinical trial registration None