Predictive factors of outcome following staphylococcal

Clinical Nephrology, Vol. 64 – No. nn/2005 (nnn-nnn)

Predictive factors of outcome following staphylococcal peritonitis in continuous ambulatory peritoneal dialysis Original ©2005 Dustri-Verlag Dr. K. Feistle ISSN 0301-0430

M. de Lourdes Ribeiro de Souza da Cunha1, A.C. Montelli2, A.M. Fioravante3, J.E. Neves Batalha1,2, J.C. Teixeira Caramori2 and P. Barretti2 1Department

of Microbiology and Immunology, Bioscience Institute, of Internal Medicine of Botucatu Medical School, and 3Department of Biostatistics of Bioscience Institute, UNESP, Botucatu, SP, Brazil 2Department

Predictive factors of outcome following staphylococcal peritonitis

Key words peritonitis – staphylococci – CAPD – oxacillin resistance – predictive factors

Received June 23, 2004; accepted in revised form April 19, 2005 Correspondence to Dr. P. Barretti Departamento de Clínica Médica, Faculdade de Medicina de Botucatu, UNESP, Rubião Júnior, Botucatu, SP, Brasil, Caixa Postal 584 CEP 18618-000 [email protected]

Abstract. Background and aims: Staphylococcus epidermidis and other coagulasenegative staphylococci (CoNS) are the most common agents of continuous ambulatory peritoneal dialysis (CAPD) peritonitis. Episodes caused by Staphylococcus aureus evolve with a high method failure rate while CoNS peritonitis is generally benign. The purpose of this study was to compare episodes of peritonitis caused by CoNS species and S. aureus to evaluate the microbiological and host factors that affect outcome. Material and methods: Microbiological and clinical data were retrospectively studied from 86 new episodes of peritonitis caused by staphylococci species between January 1996 and December 2000 in a university dialysis center. The influence of microbiological and host factors (age, sex, diabetes, use of vancomycin, exchange system and treatment time on CAPD) was analyzed by logistic regression model. The clinical outcome was classified into two results (resolution and non-resolution). Results: The odds of peritonitis resolution were not influenced by host factors. Oxacillin susceptibility was present in 30 of 35 S. aureus lineages and 22 of 51 CoNS (p = 0.001). There were 32 of 52 (61.5%) episodes caused by oxacillin-susceptible and 20 of 34 (58.8%) by oxacillin-resistant lineages resolved (p = 0.9713). Of the 35 cases caused by S. aureus, 17 (48.6%) resolved and among 51 CoNS episodes 40 (78.4%) resolved. Resolution odds were 7.1 times higher for S. epidermidis than S. aureus (p = 0.0278), while other CoNS had 7.6 times higher odds resolution than S. epidermidis cases (p = 0.052). Episodes caused by S. haemolyticus had similar resolution odds to S. epidermidis (p = 0.859). Conclusions: S. aureus etiology is an independent factor associated with peritonitis non-resolution in CAPD, while S. epidermidis and S. haemolyticus have a lower resolution rate than other CoNS. Possibly the ag-

gressive nature of these agents, particularly S. aureus, can be explained by their recognized pathogenic factors, more than antibiotic resistance.

Introduction In spite of technological advances and significant reductions in peritonitis, it is still the most important complication in continuous ambulatory peritoneal dialysis (CAPD). It is also the most frequent cause of therapy failure [Kavanagh et al. 2004] and strongly impacts patient’s mortality [Fried et al. 1996]. Staphylococcus epidermidis and other coagulase-negative staphylococci (CoNS) are considered the most common etiological agents in CAPD peritonitis. They have been identified in approximately one third of all episodes worldwide [Kavanagh et al. 2004, Keane et al. 1996, Nakamoto et al. 2004]. CAPD peritonitis shows clinical improvement in the first 96 hours and resolves without complications in 54 – 65% of cases [Troidle et al. 1998]. However, clinical outcome is strongly influenced by the microbiological characteristics of the etiological agent. Episodes caused by Staphylococcus aureus and most cases caused by Gram-negative bacteria, particularly Pseudomonas aeruginosa, result in persistent infection and a high dialysis modality failure rate [Bunke et al. 1997, Szeto et al. 2001]. On the other hand, CoNS peritonitis is generally benign with few complications [Bunke et al. 1997]. However, there are cases of CoNS peritoneal infections with characteristics similar to

CN-4793 / 23.06.2005

de Lourdes Ribeiro de Souza da Cunha, Montelli, Fioravante et al.

S. aureus. Schnitzler et al. [1998] reported a severe case of peritonitis due to Staphylococcus lugdunensis, similar to S. aureus infections. Gruer et al. [1984] studied 43 lineages of CoNS isolated during 41 CAPD episodes of peritonitis and observed that multiple bacterial resistances were common. Holley et al. [1990], comparing in vitro susceptibility of CoNS lineages isolated in 57 episodes between 1984 and 1986 and 32 diagnosed in 1987 and 1988, reported that methicillinresistant lineage rates had increased from 5 – 28%. Differences between species and drug susceptibility profiles can explain the different clinical outcome characteristics in infections caused by these microorganisms. The purpose of this study was to compare episodes of peritonitis caused by CoNS species and S. aureus to evaluate the microbiological and host factors that affect outcome. Literature is scarce on this type of analysis.

Material and methods Microbiological and clinical data were retrospectively studied from all episodes of peritonitis caused by staphylococci species between January 1996 and December 2000 in a university dialysis center. We included only new episodes of peritonitis. These were the patients’ first episodes or episodes diagnosed at least 14 days after completion of the last peritonitis treatment [Bunke et al. 1997, Peacock et al. 2000]. Cases associated with exit site or tunnel infections were excluded. Peritonitis was defined as the presence of cloudy peritoneal fluid associated with one or more of the following symptoms: abdominal pain, fever, nausea, and vomiting. Occurrence of only one of these was considered peritonitis if the dialysate white cell count was greater than 100/ml, with at least 50% polymorph nuclear cells. Resolution was defined as disappearance of signs and symptoms within 96 hours of initiating antibiotic therapy, with negative results from peritoneal fluid culture at least 14 days after treatment completion. Relapse was defined as peritonitis recurrence with the same organism and susceptibility profile or no organism within 14 days of completing antibiotics [Bunke et al. 1997, Peacock et al. 2000]. Non-resolution

2

was the generic term used for cases with initial non-resolution, relapse, peritoneal catheter removal or death. For each case, information was collected on: – episode: date, clinical findings, treatment, outcome (resolution, relapse, catheter removal or death), – presence of diabetes mellitus, – demographics data: age, sex, and race (Caucasian, non-Caucasian), treatment time on dialysis, and – exchange system (standard or double bags). All episodes were treated according to a local protocol adapted from the third report of The Ad Hoc Advisory Committee on Peritonitis Management [Keane et al. 1996] for staphylococcal episodes. All patients were treated within 24 hours of first clinical signs or symptoms. Antibiotic therapy was started with 500 or 750 mg/l i.p. cefazolin (patients with urinary volume more than 500 ml/24 h) and 250 mg/l i.p. amikacin as loading dose, followed by 500 or 750 mg/l cefazolin and 2 mg/kg amikacin each day in the last PD bag. As soon as culture results were available, therapy was evaluated. According to the National Committee for Clinical Laboratory Standards (NCCLS) [2002], for oxacillin-resistant S. aureus and CoNS, all penicillins, cephems (cephalosporins) and other b-lactams may appear active in vitro but are not effective clinically. Thus, for oxacillin-susceptible cocci the cefazolin was maintained, while for oxacillin-resistant cocci the cefazolin was replaced by 1 g/l i.p. vancomycin, repeated every five (in patients with urinary volume more than 500 ml/24 h) or seven days. In both situations, amikacin was discontinued. Vancomycin was also used for patients with no clinical improvement within the first 96 hours of antibiotic treatment, although microbiological tests revealed oxacillin susceptibility. Antibiotic therapy duration was 14 days for CoNS and 21 days for S. aureus episodes.

Microbiological methods Microbiological materials were stored in a collection of cultures. Clinical specimen

CN-4793 / 23.06.2005


Table 1. Breakpoints (mg/l) for susceptibility and resistance of Staphylococcus aureus and coagulase-negative staphylococci (CoNS) as per NCCLS [2002]. Antimicrobial agent

Penicillin Oxacillin for S. aureus Oxacillin for CNS Cephalothin Levofloxacin

Interpretative standards Susceptible Resistant £ 0.12 £ 2.0 £ 0.25 £ 8.0 £ 2.0

³ 0.25 ³ 4.0 ³ 0.5 ³ 32 ³ 8.0

Table 2. Characteristics of 63 CAPD patients at first peritonitis episode by staphylococci. Characteristic

Number (%)

Age (years) Birth – 20 21 – 40 41 – 59 60+ Sex (female) Race (Caucasian) Diabetics Treatment time Less than 1 year More than 1 year Exchange system Standard Double bag

4 (6.3) 12 (19.0) 22 (34.9) 25 (39.7) 43 (68.3) 39 (61.9) 28 (44.4) 28 (44.4) 35 (55.6) 19 (30.2) 44 (69.8)

Table 3. Causative microorganisms in 86 new episodes of peritonitis by S. aureus and coagulase-negative staphylococci (CoNS). Microorganism

Lineages n (%)

S. aureus CoNS S. epidermidis S. haemolyticus S. warneri S. hominis S. xylosus S. cohnii S. simulans S. lugdunensis

35 51 (100) 24 (47) 11 (21.5) 5 (9.8) 5 (9.8) 2 (3.9) 2 (3.9) 1 (1.9) 1 (1.9)

3

CoNS were identified by using the simplified biochemical test scheme proposed by Kloos and Schleifer [1975] and Kloos and Bannerman [1995]. In vitro susceptibility was determined by minimal inhibited concentration (MIC) by the E-test (AB Biodisk, Solna, Sweden), a new quantitative method that uses a transparent strip of inert plastic with drug concentrations from 0.002 – 256 mg/ml. Sample susceptibility proportion to each drug was defined as per NCCLS breakpoints [2002] (Table 1). Samples with intermediate values were considered resistant.

Statistical analysis In this observational study, several factors could have influenced the outcome: staphylococcus species, age, sex, race, presence of diabetes mellitus, exchange system, treatment time on CAPD, antibiotic susceptibility and use of vancomycin. The oxacillin susceptibility was not included in the model due to its association with the species (CoNS). Also, the susceptibility to cephalothin, penicillin, levofloxacin and ofloxacin was not included due to the association with oxacillin susceptibility. We, therefore, adopted a regression model that incorporates all factors, of interest and control, affecting infection outcome. Logistic regression model outcome was classified into two results, exhausted and mutually exclusive (resolution or non-resolution). The c2-test was used for frequency comparisons, p < 0.05 was considered statistically significant.

Results

isolates were seeded on blood agar and Gramstained to confirm purity and determine morphology and specific color. They were then submitted to catalase and coagulase tests.

There were 86 new peritonitis episodes diagnosed in 63 patients on regular CAPD. Demographic data at first episode are shown in Table 2. There were 35 episodes caused by S. aureus and 51 by CoNS. From CoNS, S. epidermidis was the most frequent species (24 cases, 47% of all CoNS episodes), followed by S. haemolyticus (11 cases, 21.6%). Other species were present in the remaining 16 cases (Table 3). Oxacillin susceptibility was observed in 30 of 35 (85.7%) episodes due to S. aureus and in 22 of 51 (43.1%) due to CoNS (p = 0.001).

CN-4793 / 23.06.2005

4

de Lourdes Ribeiro de Souza da Cunha, Montelli, Fioravante et al.

Table 4. Susceptibility profile for antibiotics in Staphylococci species isolated in 86 new episodes of peritonitis in CAPD. Susceptible n (%)

S. aureus n = 35

S. epidermidis n = 24

S. haemolyticus n = 11

Other CoNS n = 16

p#

Oxacillin Levofloxacin Cephalothin Penicillin

30 (85.7) 31 (88.6) 30 (85.7) 14 (40.0)

10 (41.7) 20 (87.0) 24 (100.0) 5 (20.8)

6 (54.5) 9 (81.8) 9 (81.8) 3 (27.3)

6 (37.5) 13 (92.9) 15 (93.7) 9 (56.2)

0.001 0.892 0.125 0.124

# refers to comparison between S. aureus versus all CoNS together

Table 5. Comparison of odds by regression logistic analysis of resolution of peritonitis. Factor

Log (Odds)

p

Age (years) (birth to 20/60+) Age (years) (21 – 40/60+) Age (years) (41 – 59/60+) Sex (male/female) Race (Caucasian/non-Caucasian) Diabetes mellitus (no/yes) System (standard/double bags)

0.5150 1.4076 0.7775 0.2022 0.4956 –0.5555 0.5956

0.7450 0.1516 0.3701 0.7921 0.5600 0.3951 0.4861

Treatment time on CAPD Less than one year More than one year

2.0445 0.9547

0.2729 0.2065

–0.2115

0.8839

–2.2462 –1.9629 2.0272 0.1730

0.0051 0.0278 0.0526 0.8592

Treatment with vancomycin (no/yes) Etiological agent All CoNS/S. aureus S. epidermidis/S. aureus Other CoNS/S. epidermidis S. haemolyticus/S. epidermidis

Odds ratio

9.4 7.1

Table 4 shows the antibiotic susceptibility profile in S. aureus, S. epidermidis, S. haemolyticus and other CoNS. The percentage of susceptible lineages was similar between CoNS species. Overall, 57 episodes (66.3%) were resolved, 14 (16.3%) relapsed, 12 (13.9%) needed catheter removal and 3 (3.5%) died. Of the 35 S. aureus cases, 17 (48.6%) resolved, 8 (22.8%) relapsed, 7 (20%) needed catheter removal, and 3 (8.6%) resulted in death. For CoNS episodes, 40 (78.4%) resolved, 6 (11.8%) relapsed, and 5 (9.8%) needed catheter removal. Of the 11 non-resolved episodes, 7 were due to S. epidermidis,

2 to S. haemolyticus and 2 to other CoNS. There were significantly more CoNS cases resolved than S. aureus episodes (p < 0.001). There were 52 episodes involving oxacillin-susceptible lineages, of which 32 (61.5%) resolved, 10 (19.2%) relapsed and 10 (19.2%) had catheter losses; 34 infections were caused by oxacillin-resistant lineages, of which 20 (58.8%) resolved, 9 (26.5%) relapsed, and 5 (14.7%) needed catheter removal. The percentage of resolved cases was similar between susceptible and resistant oxacillin lineages (p = 0.9713); 2 of the 3 deaths associated with S. aureus were due to oxacillin-susceptible and one to oxacillin-resistant lineages. Vancomycin was used in 56 episodes, 18 caused by S. aureus and 38 by CoNS. For S. aureus infections, vancomycin was prescribed for bacterial resistance in six cases, no improvement in clinical signs and symptoms in eight, and other unknown causes in two. For CoNS vancomycin was prescribed for bacterial resistance in 22 cases, no clinical improvement in 12, and other unknown causes in four. No cefazolin allergy was diagnosed. The time between diagnosis and the first vancomycin dose was 3.7 ± 1.7 days for S. aureus episodes and 3.9 ± 1.2 days for CoNS (p = 0.793). Table 5 shows results adjusted to the regression model. Controlling for covariables, resolution odds were not influenced by host factors such as age, sex, diabetes, exchange system, treatment time on CAPD or vancomycin use. Resolution odds were 7.1 times higher for S. epidermidis than for S. aureus (p = 0.0278), while other CoNS had similar odds

CN-4793 / 23.06.2005


resolution to S. epidermidis (p = 0.0526) and to S. haemolyticus cases (p = 0.859). For all CoNS episodes together, the resolution odds were 9.4 times higher than for S. aureus (p = 0.0051).

Discussion The incidence of peritonitis due to CoNS has decreased significantly in recent years [Nakamoto et al. 2004]. However, CoNS still remains the most frequent etiologic group in CAPD peritonitis. In our study, S. epidermidis was the most frequently isolated lineage. This finding corresponds to other reports [Bunke et al. 1997, Kavanagh et al. 2004). In addition to S. epidermidis, seven other CoNS species were identified, predominantly S. haemolyticus (11 lineages). This result is important, since there is little literature on other CoNS, except when peritonitis is caused by uncommon agents. The frequency of oxacillin-resistant agents was greater in CoNS than in S. aureus lineages. This agrees with other authors who have reported an increased resistance to methicillin in CoNS over the last two decades [Holley et al. 1990, Kim et al. 2004, Nakamoto et al. 2004]. Although some authors have shown the influence of demographic factors, it is only recently that the impact of clinical and laboratory parameters on peritonitis outcome was first analyzed. Krishnan et al. [2002] showed that CAPD duration and the number of days effluent cell count remained > 100/ml were the only factors that independently predicted episode outcome. Caucasians seemed to have a lower resolution rate than blacks, while other variables did not affect outcome. In our study, variables such as age, race, sex, treatment time on CAPD, diabetes and exchange system did not influence clinical outcome. These results were similar to Krishnan et al. [2002] except for race and treatment time on dialysis. Troidle et al. [1999] also did not find differences in peritonitis outcome associated with dialysis therapy duration. Peritonitis outcome from S. aureus was worse than from CoNS. Bunke et al. [1997] and Peacock et al. [2000] in prospective studies made similar observations. Factors related to species and antimicrobial resistance could

5

explain this different outcome. Since oxacillin resistance rate was higher in CoNS than S. aureus, the contribution from drug resistance is inconsistent. Furthermore, there was no difference in episode resolution rate between oxacillin-susceptible and oxacillin-resistant lineages. We can conclude that S. aureus etiology is an independent factor associated with peritonitis episode non-resolution in CAPD. Possibly, the aggressive nature of this agent can be explained by its recognized pathogenic factors, more than for antibiotic resistance.

Acknowledgment We thank Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) for financial support.

References Bunke MC, Brier ME, Golper TA, Academic Subcommittee of the Network 9 Peritonitis Study 1997 Outcome of single organism peritonitis in peritoneal dialysis: Gram-negatives versus Gram-positives in Network 9 Peritonitis Study. Kidney Int 52: 524-529 Fried LF, Bernardino J, Johnson JR, Piano B 1996 Peritonitis influences mortality in peritoneal dialysis patients. J Am Soc Nephrology 7: 2176-2182 Gruer LD, Bartlett R, Ayliffe GA 1984 Species identification and antibiotic sensitivity of coagulase-negative staphylococci from CAPD peritonitis. J Antimicrob Chemother 3: 577-583 Holley JL, Bernardino J, Johnston JR, Piraino B 1990 Methicillin-resistant staphylococcal infections in an outpatient peritoneal dialysis program. Am J Kidney Dis 16: 142-146 Kavanagh D, Prescott GJ, Mactier RA 2004 Peritoneal dialysis-associated peritonitis in Scotland (1999 – 2002). Nephrol Dial Transplant 19: 2584-2591 Keane WF, Alexander SR, Baillie GR, Boeschoten E, Gokal R, Golper TA, Holmes CJ, Huang CC, Kawaguchi Y, Piano B, Riella M, Schefer F, Vas S 1996 Peritoneal dialysis-related peritonitis treatment recommendations: 1996 update. Perit Dial Int 16: 557-573 Kim DK, Yoo TH, Ryu DR, Xu ZG, Kim HJ, Choi KH, Lee HY, Han DS, Kang SW 2004 Changes in causative organisms and their antimicrobial susceptibilities in CAPD peritonitis: a single center’s experience over one decade. Perit Dial Int 24: 424-432 Kloos WE, Schleifer KH 1975 Simplified scheme for routine identification of human Staphylococcus species. J Clin Microbiol 1: 82-88 Kloos WE, Bannerman TL 1995 Staphylococcus and micrococcus. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH (eds) Manual of Clinical Microbiology (6th ed). American Society Microbiology, Washington, pp 282-298

CN-4793 / 23.06.2005

de Lourdes Ribeiro de Souza da Cunha, Montelli, Fioravante et al. Krishnan M, Thodis E, Ikonomopoulos D, Vidgen E, Chu M, Bargman JM, Vas SI, Oreopoulos DG 2002 Predictors of outcome following bacterial peritonitis in peritoneal dialysis. Perit Dial Int 22: 573-581 Nakamoto H, Hashikita Y, Itabashi A, Kobayashi T, Suzuki H 2004 Changes in the organisms of resistant peritonitis in patients on continuous ambulatory peritoneal dialysis. Adv Perit Dial 20: 52-57 National Committee for Clinical Laboratory Standards 2002 Performance Standards for Antimicrobial Susceptibility testing: Twelfth Informational Supplement. NCCLS document M100-S12, Pennsylvania, USA Peacock SJ, Howe PA, Day NP, Crook DW, Winearls CG, Berendt AR 2000 Outcome following staphylococcal peritonitis. Perit Dial Int 20: 215-219 Schnitzler N, Meilicke R, Conrads G, Frank D, Haase G 1998 Staphylococcus lugdunensis: report of a case of peritonitis and an easy-to-perform screening strategy. J Clin Microbiol 36: 812-813 Szeto CC, Chow KM, Leung CB, Wong TY, Wu AK, Wang AY, Lui SF, Li PK 2001 Clinical course of peritonitis due to Pseudomonas species complicating peritoneal dialysis: a review of 104 cases. Kidney Int 59: 2309-2315 Troidle L, Gorban-Brennan N, Kliger A, Finkelstein F 1998 Differing outcomes of Gram-positive and Gramnegative peritonitis. Am J Kidney Dis 32: 623-628 Troidle L, Gorban-Brennan N, Kliger AS, Finkelstein F 1999 Effect of duration of chronic peritoneal dialysis therapy on the development of peritonitis. Perit Dial Int 19: 376-379

CN-4793 / 23.06.2005

6