GPC: Gram-Positive Cocci

AOMI: Obliterative Arteriopathy of the Lower Limbs

CAPD: Continuous Ambulatory Peritoneal Dialysis

APD: Automated Peritoneal Dialysis

IP: Intraperitoneal

NPC:Non-Polymorphonuclear Cells

CRP: Non-Polymorphonuclear Cells

RDPLF: French Language Peritoneal Dialysis Registry

Peritonitis currently represents 14% of the cases necessitating a transition from peritoneal dialysis to hemodialysis in France and accounts for 3% of all mortality rates associated with dialysis [1]. Innovations in both the technical and medical spheres, including the implementation of the «flush before fill» technique, enhanced antisepsis protocols, and the application of mupirocin, in conjunction with the establishment of therapeutic patient education initiatives, have contributed to a decrease in the incidence of peritonitis. This reduction is particularly notable in the case of gram-positive cocci (GPC) infections, which are prevalent among diabetic individuals.

The capacity to identify causative germs is not consistently achievable, with the prevalence of culture-negative peritonitis exhibiting significant variation, ranging from 10% to 50% across different facilities [2]. The process for germ identification necessitates a prompt analysis of the effluent dialysate, which is incubated in blood culture bottles (e.g., BACTEC, Kent, UK) following a minimum dwell time of two hours. Employing an optimal dialysate culture methodology involving centrifugation of the dialysate followed by resuspension of the supernatant in a culture medium should curtail the incidence of undocumented peritonitis to below 10%. Typically, a phenotypic analysis utilizing mass spectrometry (bioMérieux VITEK MS) is conducted initially, with subsequent identification of bacteria through automated biochemical galleries (bioMérieux VITEK 2) in cases where spectrometry is inconclusive.

Staphylococci, encompassing coagulase-positive staphylococci, such as Staphylococcus aureus, and coagulase-negative staphylococci, predominantly account for GPC-related peritonitis cases. In both micro and macroscopic evaluations of cultures, it is common to encounter difficulties distinguishing coagulase-negative staphylococci from other GPC species, such as Kocuria.

Species within the Micrococcus and Kocuria genera, members of the Micrococcaceae family, are characterized as cluster or tetrad-forming GPCs. These species are distinguishable from staphylococci by their larger size and more pronounced yellow pigmentation under light microscopy. Although typically saprophytic on human skin, mucous membranes, and oropharynx, they can become pathogenic in the context of immunodepression. Concerning peritoneal dialysis, these bacteria have been implicated in cases of peritonitis.

The International Society for Peritoneal Dialysis has issued guidelines for the probabilistic antibiotic treatment of peritonitis, which were recently summarized by Taghavi and Dratwa [3]. These guidelines primarily focus on GPCs and gram-negative bacilli; however, they do not explicitly address peritonitis caused by Micrococcus and Kocuria species.

In this report, we present two instances of Kocuria rhizophila peritonitis identified at our facility within 3 months in 2022, both of which necessitated the removal of the peritoneal dialysis catheter. The objective of this study is to delineate the clinico-biological profiles and outcomes of these K. rhizophila peritonitis cases and compare the findings with data recorded in the French Language Peritoneal Dialysis Registry (RDPLF) as well as existing literature on the subject.

Case 1 involves a 70-year-old male, presenting with chronic renal failure attributed to probable nephroangiosclerosis. His medical history included active smoking, exogenous factors, and endovascular revascularization for stage 3 obliterative arteriopathy of the lower limbs (AOMI). He had been undergoing continuous ambulatory peritoneal dialysis (CAPD) for 5 years, with three daily exchanges. Two years prior, he experienced Streptococcus salivarius and Streptococcus vestibularis peritonitis, which responded favorably to intraperitoneal (IP) cefazolin.

Upon first evaluation, the patient presented with a compromised infusion line, necessitating its replacement, yet there were no clinical or biological indicators initially suggesting peritonitis. Subsequent routine analysis of his dialysate identified the presence of Micrococcus luteus. This bacterium was interpreted as a contaminant due to its non-pathogenic nature and the absence of increased cellularity in the dialysate, indicating a lack of inflammatory response typically associated with infection.

Despite developing severe abdominal pain, widespread defensiveness, and functional ileus 3 weeks later, the patient remained apyretic and showed no signs of hemodynamic compromise. The effluent appeared turbid, and analysis confirmed peritonitis with a dialysate hypercellularity of 9440 elements/mm3, of which 80% were non-polymorphonuclear cells (NPCs).

The patient’s blood analysis indicated a biological inflammatory syndrome with a C-reactive protein (CRP) level of 125 mg/L, notably without hyperleukocytosis. Plasma albumin concentration was measured at 30 g/L. In accordance with local protocols and International Society for Peritoneal Dialysis recommendations, empiric treatment commenced with IP cefazolin and ceftazidime. Subsequent direct examination detected the presence of GPCs, leading to the discontinuation of ceftazidime and the continuation of cefazolin as monotherapy.

The patient’s condition rapidly improved, as evidenced by diminished pain following lavage and the placement of a nasogastric tube for functional ileus. However, cultures of the dialysate remained positive for K. rhizophila for 6 days, prompting the substitution of cefazolin with IP vancomycin, pending the results of susceptibility testing. The susceptibility testing revealed a strain responsive to multiple antibiotics. Cultures became sterile after 48 hours of vancomycin therapy, which was extended for a total duration of 21 days. After treatment, the dialysate analysis showed no microbial growth and normalization of peritoneal cellularity to 17 elements/mm3, albeit with a predominance of 68% NPCs. The CRP level remained mildly elevated at 40 mg/L.

Ten days post-antibiotic therapy, the patient experienced severe abdominal pain and diffuse tenderness. The dialysate appeared cloudy, and analysis confirmed a recurrence of peritonitis, with 5760 nucleated cells/mm3, including 89% NPCs. Blood tests indicated an elevated inflammatory response, with a CRP level of 80 mg/L; however, no hyperleukocytosis was observed. Treatment was reinitiated with IP vancomycin, ceftazidime, and amikacin.

Cultures of the dialysate once again isolated K. rhizophila, without detection of mycelial filaments. Considering the recurrence of K. rhizophila peritonitis, potentially due to catheter colonization, the management strategy included a one-stage catheter removal and replacement under vancomycin coverage, which resulted in an improvement in the infection.

This case highlights the unique pathogenic potential of K. rhizophila, an organism typically considered low-virulence, and its ability to form a biofilm on the catheter, necessitating its removal. Notably, catheter cultures remained sterile, likely due to the antibiotic regimen.

The case study underscores the necessity of recognizing Kocuria and M. luteus not merely as contaminants but as potential pathogens in dialysate cultures, emphasizing their clinical significance in the context of peritoneal dialysis-associated peritonitis.

Case 2 involves a 77-year-old female with chronic renal failure secondary to acute tubular necrosis during hemorrhagic shock. Her medical history includes endovascularly revascularized AOMI, arterial hypertension, and an anaphylactic reaction to cephalosporins. Notably, the patient had experienced 12 prior episodes of peritonitis, attributed to exposure to cats at home, leading to multiple instances of Pasteurella-induced peritonitis.

Following a failed kidney transplant, she had been receiving automated peritoneal dialysis (APD) for 7 years. The patient was admitted to the hospital presenting with cloudy dialysate but no abdominal pain or fever. Dialysate analysis confirmed peritonitis, with 800 nucleated elements/mm3, 80% of which were NPCs. The biological workup did not indicate an inflammatory syndrome. Due to a documented allergy to ceftazidime and cefazolin, probabilistic treatment was initiated with aztreonam IV and vancomycin IP. Subsequent direct examination revealed GPCs, leading to the discontinuation of aztreonam. Culture results yielded a multi-susceptible K. rhizophila strain. The patient responded well to IP vancomycin therapy, and after 4 weeks of treatment, one-stage catheter removal and reinstallation were performed due to recurrent peritonitis.

This second case, occurring 3 months after the first, prompts consideration of the transmission mode and the potential for nosocomial transmission of Kocuria. In both instances, the Kocuria strains were multi-susceptible, with phenotypic analysis alone not providing sufficient evidence to conclude the identity between the strains. Pulsed-field gel electrophoresis is currently being conducted at Nantes University Hospital to compare the strains and investigate the possibility of nosocomial transmission.

K. rhizophila is an environmental commensal GPC. It is a member of the Micrococcaceae family, comparable to staphylococci and Micrococcus, characterized by the formation of tetrads and being catalase positive as well as coagulase negative (Figure 1.).

Distinguishing between Staphylococcus and Kocuria based on direct, micro, and macroscopic examination of cultures presents challenges due to the similar appearance of their colonies. Nonetheless, spectrometric analysis of the bacterial proteins facilitates the identification of the species with notable specificity. This method is employed at our center for the accurate identification of K. rhizophila. To date, 19 species of Kocuria have been identified (Figure 2.), with five species, including K. kristinae, K. varians, K. marina, K. rosea, and K. rhizophila, being documented as pathogens in human infections.

K. rhizophila is ubiquitously found in soil, dust, water, and air, forming part of the natural flora on mammalian skin. This bacterium is also capable of colonizing the oral cavity and mucous membranes, particularly within the oropharynx and upper respiratory tract in humans. Transmission can occur through contact with contaminated objects or surfaces or through droplet transmission and inhalation of contaminated aerosols. Moreover, K. rhizophila has been isolated from oxalic acid-treated chicken, indicating that contact with contaminated meat could serve as a potential source of exposure. A recent study has underscored the bacterium’s propensity for adhering to silicone tubing, which may contribute to failures in antibiotic treatments. While biofilm production has been posited for Kocuria species, it has not been conclusively demonstrated across all strains [7].

Given the diverse environments K. rhizophila inhabits and its potential for biofilm formation, managing infections poses unique challenges, particularly in the context of antibiotic treatment. In the absence of species-specific guidelines, the antibiotic susceptibility thresholds applied to Kocuria are aligned with those established for staphylococci. To our knowledge, there has been no documented antibiotic resistance profile specific to K. rhizophila in the existing literature.