Journal of Infection and Public Health (2009) 2, 147 152 Resistance pattern of breakthrough urinary tract infections in children on antibiotic prophylaxis Ali Reza Nateghian a, Joan L. Robinson b,, Shahab Mohandessi c, Nakysa Hooman d a Pediatric Infectious Diseases Department, Ali Asghar Children s Hospital, Iran University of Medical Sciences, Tehran, Iran b Department of Pediatrics and Stollery Children s Hospital, University of Alberta, Edmonton, Canada c Ali Asghar Children s Hospital, Iran University of Medical Sciences, Tehran, Iran d Pediatric Nephrology Department, Ali Asghar Children s Hospital, Iran University of Medical Sciences, Tehran, Iran Received 27 May 2009; received in revised form 7 August 2009; accepted 11 August 2009 KEYWORDS Antibiotic prophylaxis; Cefixime; Pyelonephritis; Resistance; Antibiotic; Urinary tract infection Summary Prophylactic antibiotics are commonly used for prevention of urinary tract infections (UTIs) in children. It was postulated that the organisms and resistance patterns of breakthrough infections would differ with the choice of antimicrobial prophylaxis. This was a retrospective descriptive study of all breakthroughs UTI from 2000 to 2006 in children over 1 month of age discharged from a referral children s hospital in Tehran, Iran on continuous antibiotic prophylaxis for UTIs. Fifty-seven children discharged on prophylaxis had breakthrough UTIs of which 32 (56%) had a previously diagnosed urinary tract anomaly. Escherichia coli was responsible for the majority of infections irrespective of choice of prophylaxis. Thirty-three of 56 breakthrough UTIs (59%) were with organisms that were resistant to the prophylactic antibiotic. There was an increased incidence of resistance to prophylaxis in children on cefixime (16 of 22; 78%) when compared with children on cephalexin (7 of 19; 37%; p = 0.02) and a trend toward increased resistance when compared with children on trimethoprim sulfamethoxasole (3 of 8; 37%) (p = 0.10). In conclusion, the resistance pattern of organisms causing breakthrough UTIs varies with the choice of prophylaxis which should be taken into consideration in chosing empiric therapy for such infections. 2009 King Saud Bin Abdulaziz University for Health Sciences. Published by Elsevier Ltd. All rights reserved. Corresponding author at: Room 8213, Aberhart Centre One, 11402 University Avenue, Edmonton, AB, Canada T6G 2J3. Tel.: +1 780 407 3666; fax: +1 780 407 7136. E-mail address: jr3@ualberta.ca (J.L. Robinson). 1876-0341/$ see front matter 2009 King Saud Bin Abdulaziz University for Health Sciences. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jiph.2009.08.002
148 A.R. Nateghian et al. Introduction Urinary tract infections (UTI) occur in 1 5% of children with recurrences being common. A study of 134 females and 128 males with a UTI in the first year of life described a 35% recurrence rate in males and 32% in females during the 3-year follow-up period [1]. A significant number of these UTIs result in hospitalization. Local unpublished data showed that of 1159 children hospitalized with a UTI March 2001 to March 2009, 66 children (5.7%) were subsequently readmitted with a recurrence. Continuous prophylaxis with trimethoprim sulfamethoxazole (TMP SMX), nitrofurantoin [2] or cephalosporins [3 5] have been commonly used for children who are prone to UTIs on the basis of voiding dysfunction, obstructive uropathy, high grade vesicoureteral reflux (VUR), or infected renal stones to prevent development of renal scars [6,7]. Of late, there has been less enthusiasm for antibiotic prophylaxis, primarily because studies have demonstrated questionable efficacy in preventing UTIs [7], especially for children with low grade VUR. Nonetheless, prophylaxis is likely to continue to be used for children who are very high-risk for recurrent pyelonephritis. A major concern is the development of UTIs with antibiotic-resistant flora in children on prophylaxis. The goal of this study was to analyze the resistance pattern of UTI isolates from children on prophylaxis and correlate it with the choice of antibiotic, with the hypothesis being that the antibiogram of organisms from breakthrough infections would be influenced by the choice of prophylaxis. Methods Design, setting, and participants This was a retrospective descriptive cross-sectional study of all breakthrough UTIs from January 1, 2000 to December 31, 2006 in Ali Asghar Children s Hospital a referral teaching hospital in Tehran, Iran for children up to age 18 years. All patients over 1 month of age who had been discharged on antibiotic prophylaxis for UTIs and were subsequently admitted up to December 31, 2006 with a breakthrough UTI after being on antibiotic prophylaxis for minimum 1 month were enrolled. The dose used for prophylaxis was one-third of the therapeutic dose of the antibiotic. Study protocol Breakthrough infection was defined as greater than 10 8 colony forming units per liter (10 5 colony forming units per milliliter) in a urine culture with an abnormal urinalysis (defined as greater than 5 white blood cells per high-powered field or a positive nitrite test or both) and a clinical course compatible with a UTI. Suprapubic urines are rarely obtained in our center; this relatively high colony count was chosen to ensure all patients had true UTIs. Patients with mixed growth in the urine were excluded. Results of bagged urines were accepted only if there were two specimens collected with identical results. Data was collected on the presence of urinary tract anomalies, most recent prophylactic antibiotic, and the resistance pattern of the organism isolated from the urine. Children were excluded if it appeared compliance with prophylaxis had been poor and were enrolled only once even if they had more than one breakthrough infection. The standard method for processing urine samples is that the sample is placed in a sterile dish and sent to the laboratory within 20 min. It is then plated on the blood agar medium for gram positive and the MacConkey s agar for Gram-negative bacteria, and subsequently on Muller Hinton agar for antibiogram. Antibiotic discs (Padtan Teb CO. made by Iran) are used with antimicrobial sensitivity interpreted according to standard NCCLS criteria by measuring the halo diameter of each disc after 48 h. Outcomes The types of microorganisms and their resistance patterns were correlated with the most recent prophylactic antibiotic. Ethical issues This research was approved by the Research Committee of the Iran University of Medical Sciences. Results Demographics Breakthrough infection associated with hospital admission was identified in 57 patients who had been discharged with UTI as their final diagnosis, including 35 females and 22 males aged 1.9 months to 18 years (mean 4.2 years). There were no fungal breakthrough infections. The duration of prophylaxis had been 1 2 months for 32 children (56%), 3 6 months for 13 children (23%), 7 12 months for 3 children (6%), 13 24 months for 1
Breakthrough urinary tract infections 149 Fig. 1 Organisms causing breakthrough UTIs in 57 children on antibiotic prophylaxis. child (2%), 25 36 months for 5 children (8%), and over 36 months for 3 children (5%). The indication for prophylaxis was vesicoureteral reflux (N = 14; 25%), neurogenic bladder (N = 9; 16%), ureteropelvic junction obstruction (N = 3; 5%), rectovesical fistula (N = 2; 4%), duplicated ureter (N = 2; 4%), posterior urethral valves (N = 1; 2%), and dysplastic kidney (N = 1; 2%). The other 25 children (44%) had idiopathic recurrent UTIs, renal scars with no recognized predisposing conditions, voiding dysfunction, or were very young at the onset of their first UTI. Forty children (70%) had no previous UTIs with prophylaxis being started because of renal anomalies, 14 (24%) had one previous UTI and 3 (6%) had two previous UTIs. Etiologic organisms causing breakthrough UTIs Etiologic organisms were Escherichia coli (N = 31; 54%), Staphylococcus saprophyticus (N = 10; 17.5%), Pseudomonas species (N = 8; 14%), Enterobacter species (N = 4; 7%), Klebsiella species (N = 3; 5%) and Proteus vulgaris (N = 1; 2%) with organisms appearing to vary by the choice of antibiotic for prophylaxis (Fig. 1). Antibiotic prophylaxis in the 57 patients and their prophylactic agents were as shown in Table 1 with the most common agents being cefixime in 22 cases (39%) and cephalexin in 19 cases (33%). Escherichia coli was responsible for the majority of the breakthrough UTIs irrespective of choice of prophylaxis (Table 1). Thirty-three of 56 breakthrough UTIs (59%) were with organisms that were resistant to the prophylactic antibiotic (Table 1), with susceptibilities not being available for one child on ciprofloxacin prophylaxis. There was a trend towards patients on cefixime being more likely to have breakthrough infections that were resistant to the prophylactic antibiotic than were children on TMP SMX (p = 0.10). Children on Table 1 Type of antibiotic prophylaxis, organisms isolated, and resistance to that prophylactic agent for 57 breakthrough urinary tract infections. Organism causing breakthrough urinary tract infection Total Klebsiella species Proteus vulgaris Enterobacter species Pseudomonas species Prophylaxis Escherichia coli Staphylococcus saprophyticus Cefixime (N = 22) N =13 (R =7;S =6) N =4 (R =4;S =0) N =3 (R =3;S =0) N =1 (R =1;S =0) N =1 (R =1;S =0) N =22 (R = 16;S =6) a Cephalexin (N = 19) N =9 (R =2;S =7) N =5 (R =0;S =5) N =3 (R =3;S =0) N =2 (R =2;S =0) N =19 (R =7;S = 12) TMP SMX (N =8) N =5 (R =1;S =4) N =1 (R =0;S =1) N =2 (R =2;S =0) N =8 (R =3; S =5) N =3 (R =2;S =1) Nitrofurantoin (N =3) N =1 (R =0;S =1) N =1 (R =1;S =0) N =1 (R =1;S =0) Nalidixic acid (N =3) N =1 (R =1;S =0) N =1 (R =1;S =0) N =1 (R =0;S =1) N =3 (R =2; S =1) Ampicillin (N =1) N =1 (R =0;S =1) N =1 (R =0;S =1) Ciprofloxacin (N =1) N =1 (ND=1) b N =1 (ND=1) N =57 (R = 30;S = 26;ND = 1) N =8 (R =8;S =0) N =4 (R =3;S =1) N =3 (R =3;S =0) N =1 (R =1) N =10 (R =4;S =6) Total N =31 (R = 11;S = 19; ND=1) ND, susceptibilities not done; R, resistant to listed antibiotic; S, susceptible to listed antibiotic; TMP SMX, trimethoprim sulfamethoxazole. a p = 0.02 in comparison to cephalexin. b Antimicrobial susceptibility testing is not routinely performed for ciprofloxacin in this hospital.
150 A.R. Nateghian et al. cefixime were more likely to have isolates resistant to their prophylaxis than were children on cephalexin (p = 0.02). The number of children on other antibiotics was too small to compare resistance rates. Table 2 demonstrates a high rate of resistance to all commonly used antibiotics in isolates from breakthrough urinary tract infections. Discussion This study showed that as expected, over half of breakthrough UTIs in children discharged on antibiotic prophylaxis were with organisms resistant to the antibiotic chosen for prophylaxis. However, there was an increased incidence of resistance to the prophylactic antibiotic in children prescribed cefixime when compared with children prescribed cephalexin and a trend toward increased resistance when compared with children prescribed TMP SMX. It is possible this finding is because compliance was better with cefixime, ensuring that UTIs only occurred with resistant organism. However, it is also possible that this result may indicate that cefixime is more likely than other choices of antibiotic to select for resistant bacteria in the bowel flora that then infect the urinary tract. We are not aware of any previously published data on choice of antibiotics or on resistance rates in breakthrough infections in Iran. However, in the neighboring country Turkey, a study showed increasing resistance to ampicillin and TMP SMX with a trend towards increasing resistance to gentamicin and ceftriaxone among isolates from breakthrough infections in children who were taking chemoprophylaxis [8], but this study did not correlate resistance rates with choice of prophylaxis. Recent guidelines published by the National Institute of Health and Clinical Excellence in the United Kingdom recommend abandoning the routine prophylaxis of UTIs in children [9], in part due to a significant increase of antibiotic resistance [10]. However, prophylaxis is likely to continue to be used by many practitioners, especially for children with frequent pyelonephritis. Choice of antibiotic for prophylaxis is complex. In the current study the isolates from breakthrough UTIs in children taking TMP SMX had low resistance rates to the usual empiric UTI therapies (cephalosporins and gentamicin) such that this agent might be still a good candidate for prophylaxis in Iran. Furthermore, a recent study from Taiwan showed that children on TMP SMX as prophylaxis for VUR had a lower risk of having breakthrough UTIs with multi- Table 2 Frequency of in vitro resistance of isolates from breakthrough urinary tract infections to common antimicrobials. TMP/SMX Ampicillin Cephalexin Ceftriaxone Cefixime Ceftazidime Gentamicin Amikacin Nitrofurantoin Nalidixic acid E. coli (N = 31) 31 (100%) 29 (93%) 17 (55%) 17 (55%) 28 (90%) 13 (42%) 9 (29%) 17 (55%) 21 (68%) 30 (97%) Enterobacter species NP 4 (100%) 4 (100%) 4 (100%) 4 (100%) 3 (75%) 3 (75%) 2 (50%) 2 (50%) NP (N =4) Pseudomonas species NP 8 (100%) 6 (75%) 6 (75%) 2 (25%) 6 (75%) 6 (75%) 8 (100%) 8 (100%) 8 (100%) (N =8) Klebsiella sp. (N = 3) 3 (100%) 2 (67%) 0% 0% NP NP 0% NP 2 (67%) 3 (100%) Staphylococcus. 10 (100%) 5 (50%) 10 (100%) 10 (100%) NP 5 (50%) NP NP NP 5 (50%) saprophyticus (N = 10) Proteus vulgaris NP 1 (100%) 0 0 NP 0 0 1 (100%) 1 (100%) 1 (100%) (N =1) Total 44/44 (100%) 49/57 (86%) 37/57 (65%) 37/57 (65%) 34/43 (79%) 27/54 (50%) 18/47 (38%) 28/44 (64%) 34/47 47/53 (89%) (72%) NP, not performed.
Breakthrough urinary tract infections 151 resistant organisms, including extended-spectrum beta-lactamase producing organisms as compared with children on cephalosporins. On the other hand, the same study showed a lower UTI recurrence rate for children on cephalosporins than for children on TMP SMX as prophylaxis [11]. In the Cochrane review, over half of children on cefixime prophylaxis had a possible adverse reaction to the drug [12] but it is not clear if this differs from other cephalosporins or from other classes of antibiotics. Nitrofurantoin is frequently prescribed for UTI prophylaxis, but it does not cover Proteus species and its gastrointestinal side effects and hemolysis in G6PD deficiency (which has high prevalence in Iran) restrict its use. Choice of antibiotics for breakthrough infections is equally complex and is markedly influenced by local resistance patterns. The current study showed a high incidence of resistance to antibiotics that would be commonly chosen such a cefotaxime (65% resistance) and gentamicin (50% resistance). However, it is possible that organism such as S. saprophyticus might be susceptible in vivo despite resistance in vitro so these resistance rates may be falsely high. Over half the breakthrough UTIs occurred within 2 months of the prophylaxis being prescribed. It is possible that in some cases this occurred as the child was discharged on prophyalxis that their original UTI organism was resistant to. However, since 70% of the children had no previous UTI, it seems likely that resistance to prophylaxis sometimes develops quickly. This study had many limitations. Only children admitted to hospital with UTIs could be included, biasing the results by including only children with more severe breakthrough infections. The diagnosis of pyelonephritis (versus cystitis) was presumptive in some cases. The doses of the prophylactic antibiotics were not recorded, although we think all children likely received standard dosing. We could not be certain all parents or patients were compliant with prophylaxis and the fact that 46% of children had UTIs with organisms that were susceptible to their prophylactic agent suggests compliance may have been sub-optimal, presumably related to cost or taste of the drug, adverse events perceived by the parent to be drug-related, inconvenience, or lack of belief that prophylaxis was necessary. However, this makes the study more representative of what happens in practice than is a prospective study where compliance is monitored. Some children might have received antibiotics other than their prophylactic drug for non-urinary tract infections prior to being admitted with a UTI which might have altered the susceptibility pattern of the urinary isolate. We did not have susceptibility data on the isolates from the original UTI in the 30% of children with a previous UTI to be certain the choice of prophylactic agent was always appropriate. The incidence of breakthrough UTIs with different types of prophylaxis could not be calculated as we did not have data on the type of prophylaxis prescribed for the children discharged on prophylaxis who did not have breakthrough UTIs. It is vital to note that results from Iran may not be applicable to other settings with different resistance patterns for uropathogens. Finally, there was not a control group to determine the susceptibility pattern of organisms isolated from subsequent UTIs in children not on antibiotic prophylaxis. In conclusion, the resistance pattern of organisms from breakthrough UTIs is related to the type of antibiotic prophylaxis that has been prescribed. Children prescribed cefixime as prophylaxis for UTIs may be more likely to develop UTIs with organisms that are resistant to their prophylaxis than are children prescribed other antibiotics including cephalexin and possibly TMP/SMX. However, the relative incidence of UTIs in children prescribed different types of antibiotic prophylaxis is of prime importance and is yet to be established. The greatest priority in the field is to establish the efficacy of antibiotic prophylaxis for UTIs in different settings. Then, for settings where it appears to be indicated, there should be large randomized trials of different choices of prophylaxis, collecting data on susceptibilities of breakthrough infections, keeping in mind that the ideal choice of antibiotic will vary depending on the local resistance patterns. Conflict of interest Funding: No funding sources. Competing interests: None declared. Ethical approval: This research was approved by the Research Committee of the Iran University of Medical Sciences. References [1] Nuutinen M, Uhari M. Recurrence and follow-up after urinary tract infection under the age of 1 year. Pediatr Nephrol 2001;16:69 72. [2] Brumfitt W, Hamilton-Miller JM. A comparative trial of low dose cefaclor and macrocrystalline nitrofurantoin in the prevention of recurrent urinary tract infection. Infection 1995;23:98 102. [3] Lettgen B, Tröster K. Prophylaxis of recurrent urinary tract infections in children. Results of an open, con-
152 A.R. Nateghian et al. trolled and randomized study about the efficacy and tolerance of cefixime compared to nitrofurantoin. Clin Padiatr 2002;214:353 8. [4] Seracini D, Materassi M, Danti A. Non-comparative open study of the efficacy and tolerance of cefaclor in the prevention of urinary tract infections in children. Pediatr Med Chir 1996;18:383 5. [5] Stranieri G, Zampogna S, Ielapi V, Defilippo RG, Defilippo V, Cristofaro G, et al. Cefixime for the prophylaxis of urinary tract infections in children with malformative uropathies: an open study. Eur Rev Med Pharmacol Sci 2003;7:57 64. [6] Yokota T. Prophylactic effect of small doses of antibiotics for the prevention of recurrent urinary tract infections. Jpn Pediatr Soc 1992;96:984 93. [7] Roussey-Kesler G, Gadjos V, Idres N, Grellier A, Hazart L, De parscau, et al. Antibiotic prophylaxis for the prevention of recurrent urinary tract infection in children with low grade vesicoureteral reflux: results from a prospective randomized study. J Urol 2008;179: 674 9. [8] Yüksel S, Oztürk B, Kavaz A, Ozçakar ZB, Acar B, Güriz H, et al. Antibiotic resistance of urinary tract pathogens and evaluation of empirical treatment in Turkish children with urinary tract infections. Int J Antimicrob Agents 2006;28:413 6. [9] Mori R, Lakhanpaul M, Verrier-Jones K. Guidelines. Diagnosis and management of urinary tract infection in children: summary of NICE guidance. BMJ 2007;335:395 7. [10] Hajibagheri K, Priesemann M, Morrison I, Parikh A, Lessof N, Price E. NICE guidance on urinary tract infection in children abandons routine antibiotic prophylaxis. Arch Dis Child 2008;93:356. [11] Cheng C, Tsai M, Huang Y, Su L, Tsau Y, Lin C, et al. Antibiotic resistance patterns of community-acquired urinary tract infections in children with vesicoureteral reflux receiving prophylactic antibiotic therapy. Pediatrics 2008;122:1212 7. [12] Williams GJ, Wei L, Lee A, Craig JC. Long-term antibiotics for preventing recurrent urinary tract infection in children. Cochrane Database Syst Rev 2006;3:CD001534. Available online at www.sciencedirect.com