The epidemiology of multidrug-resistant organisms has local, national, and global significance.Reference Bhattacharya ^{1 ,} Reference Bhattacharya, Das, Bhalchandra and Goel ² In this study we describe the epidemiology of antimicrobial resistance from a new oncology and bone marrow transplantation center in eastern India. The method of antimicrobial susceptibility testing was per the Clinical Laboratory Standards Institute guidelines. Stool surveillance culture was performed according to the method described by Landman et al.Reference Landman, Salvani, Bratu and Quale ³ An automated system (Vitek2; bioMérieux) and disc diffusion (Bio-Rad) were employed for antibiotic susceptibility tests. The data refer to the period from April 1, 2012, through March 31, 2013. The data came from 4,723 samples, 1,474 patients (inpatients and outpatients), and 1,965 bacterial and yeast isolates. Gram-positive bacteria were detected in 25% of isolates, gram-negative bacilli in 68%, and yeasts in 7%. Positivity rates for different sample types were blood culture, 15.2%; urine, 33.3%; respiratory samples, 57.9%; pus, 65.8%; and body fluids, 37.0%. Stool samples for surveillance culture of multidrug-resistant organisms were positive in 35.1%. In total 30.6% were positive by culture.

Among patients with various infections antibiotic susceptibility of coliform bacteria (Enterobacteriaceae family) showed a high level of resistance with extended-spectrum beta-lactamase prevalence of 72%, carbapenem resistance in 23%, and resistance to amikacin, gentamicin, piperacillin-tazobactam, and ciprofloxacin to be 26%, 49%, 48%, and 71%, respectively. Nonfermentative gram-negative bacilli (eg, Pseudomonas, Acinetobacter) showed 36% resistance to carbapenems, 35% to piperacillin-tazobactam, 35% to amikacin, 38% to gentamicin, and 43% to ciprofloxacin. Resistance to meropenem and resistance to third-generation cephalosporins such as ceftazidime (marker of extended-spectrum beta-lactamase production) were detected respectively in cultures of 30.6% and 64.2% of blood isolates, 27% and 66.2% of urine samples, 25.7% and 47.8% of respiratory isolates, and 18.1% and 50.0% of pus isolates. Antibiotic resistance in gram-positive bacteria was noted in only 12% of the isolates (eg, methicillin-resistant Staphylococcus aureus), and inducible clindamycin resistance was noted in 23% of isolates. Antifungal susceptibility testing of Candida species (n=123) showed 14% resistance to fluconazole. Among 356 patients with bloodstream infections, 55% were due to gram-negative bacilli, 39% to gram-positive cocci, and 6% to yeasts. Candida albicans was the commonest species of yeast, but C. glabrata, C. haemulonii, C. kefyr, C. norvegensis, C. parapsilosis, and C. tropicalis were also noted.

Surveillance culture for antibiotic-resistant bacteria in stool sample of patients (collected before major therapeutic interventions such as bone marrow transplantation, chemotherapy induction, major gynecologic surgery for cancer) showed meropenem resistance in 17% and extended-spectrum beta-lactamase in 94% among gram-negative bacilli (n=136). In Enterococcus species we found ampicillin resistance in 96.4%, high-level gentamicin resistance in 57.1%, and vancomycin resistance in 18.5% of isolates (Table 1).

Table 1

Surveillance Culture of Stool for Multidrug Resistant Organisms

Resistance to antimicrobial agents is not a new phenomenon.Reference Bhattacharya ^{1 ,} Reference Bhattacharya ^{4 ,} Reference Goel, Hmar, Sarkar, Bhattacharya and Chandy ⁵ However, the degree of antimicrobial resistance in certain settings and the subsequent dwindling of therapeutic options have raised this issue to the level of a national security threat in some countries. What is unusual about the current data from our center is the extent of the problem, especially the severe diminution of therapeutic options posed by the emergence of carbapenem resistance in gram-negative bacterial infections and azole resistance in Candida infections. The current study was done in a new cancer care center (<2 years old) but antimicrobial resistance was noticed from the early days of this hospital, suggesting the possibility of high level of antibiotic resistance already present in the community and the environment. To emphasize this point and also to optimize empirical therapy of infections in neutropenic and other immunocompromised patients, surveillance culture for antibiotic-resistant bacteria was initiated. The results of surveillance culture for antibiotic-resistant bacteria performed predominantly on stool samples before interventions demonstrate the high prevalence of background resistance in the patient population. This could be due to several factors, such as inadequate hygiene and sanitation, improper and inadequate disposal of wastes and sewage, use of antibiotics in food animals (poultry, cattle, fish), unrestricted availability of antibiotics in the chemist shops in developing countries, lack of safe drinking water in large sections of the population, and the entry of antibiotic-resistant bacteria in the food chain. Although it is possible that to some extent resistance could have been acquired during medical interventions, it cannot be ruled out that in a significant proportion of patients it could have been acquired by community exposures of various types. The question about the extent of antibiotic-resistant bacteria colonizing healthy general population needs investigation so that possible sources can be identified and remedial measures taken. Surveillance for multidrug-resistant organisms in stool or rectal swab samples is not routinely performed in most hospitals because of cost and technical difficulties.Reference Bhattacharya ⁴ It may be argued that rapid sensitive and cost-effective tests such as those offered by real-time polymerase chain reaction assays targeting carbapenem-resistant coliforms and vancomycin-resistant enterococci would be useful tools for infection control and rational selection of empirical antimicrobial therapy.

Early diagnosis of drug-resistant pathogens would require new diagnostic and surveillance strategies, including perhaps the wider use of molecular technologies. It is possible that intricate host-pathogen interaction is at play and genetic studies of both bacteria and their victims may reveal new insights into pathogenesis and identify future therapeutic strategies. It is hoped that the creation of a national antibiotic resistance database in India and the close interaction of national, regional, and collaborating centers internationally would lead to more evidence-based prescribing and create greater understanding about the serious threat posed to the population from antimicrobial resistance.