Mycobacterium chelonae Peritonitis in a Patient on Peritoneal Dialysis

Abstract

Non-tuberculous mycobacteria peritonitis is uncommon with the majority of cases reported in patients on peritoneal dialysis with diagnostic and therapeutic challenges. Here we present a case of Mycobacterium chelonae peritonitis and review other cases in the literature to discuss the clinical spectrum, diagnostics, regimens and duration of treatment, and outcome.

Keywords:

• Mycobacterium chelonae
• non-tuberculous mycobacterium
• peritonitis
• peritoneal dialysis

INTRODUCTION

Peritoneal dialysis (PD)-associated peritonitis caused by Mycobacterium species is well documented but uncommon.[1] The majority of cases have been caused by M. tuberculosis and reported in Asia. However, cases caused by non-tuberculous mycobacteria (NTM) also occur and may pose diagnostic and therapeutic challenges.[1–3] We report a case of PD-associated peritonitis caused by M. chelonae.

CASE REPORT

A 45-year-old woman was diagnosed with end-stage renal disease caused by lupus nephritis in 1991. She had been on hemodialysis for seven years but was switched to continuous ambulatory PD because of arteriovenous shunt failure. Her systemic lupus erythematosus was adequately controlled with prednisolone 10 mg per day and hydroxychloroquine 200 mg twice a day.

She developed purulent discharge at the exit site of the Tenckhoff tube, without erythema or granuloma. The peritoneal dialysate fluid was grossly clear. There was no peritoneal tenderness, fever, or other systemic signs. An initial culture of the pus grew methicillin-resistant Staphylococcus aureus, which was treated with intravenous vancomycin 1 gm per week. The discharge persisted for a month, however, so oral rifampicin 300 mg per day was added to the vancomycin. Repeat cultures had no growth the first two times and then Corynebacterium on a third set. Three months after the discharge began, with no improvement on antibiotics, the Tenckhoff tube was removed and a new one implanted. Cefazolin was given post-operatively.

Two weeks after the operation, the patient again had a purulent discharge accompanied by cloudy dialysate fluid. She had a fever of 38.3°C and abdominal tenderness. The peritoneal fluid contained 875/mm³ white blood cells with 85% neutrophils. Her peripheral blood leukocyte count was 12,400 cells/μL and C-reactive protein 1.8 mg/dL. Cefazolin and ceftazidime was administered intravenously because she was unable to tolerate intraperitoneal antibiotics. Her leukocytosis persisted (19,500 cells/μL). One week later, culture of the peritoneal fluid yielded a rapidly growing mycobacterium subsequently identified as M. chelonae. It was sensitive to amikacin, clarithromycin, cefoxitin, and imipenem but resistant to tobramycin, ciprofloxacin, doxycycline, and trimethoprim. The patient was treated with imipenem 250 mg intravenously every 12 hours, amikacin 250 mg intravenously every 3 days, and clarithromycin 250 mg orally every 12 hours. The Tenckhoff tube was also removed because of persistent abdominal pain. Dialysis was performed thrice weekly via an external jugular venous catheter.

The patient had an episode of adhesion ileus and required total parenteral nutrition for nearly one month. Finally, after six weeks of triple antibiotic therapy, the patient's condition improved and the infection resolved. She was discharged on oral clarithromycin 250 mg every 12 hours for another 3 months. There were no further complications.

DISCUSSION

NTM are defined as mycobacterial species other than M. tuberculosis and M. leprae. NTM infections are less virulent than M. tuberculosis and are usually acquired from environmental sources. Abdominal infection caused by NTM is very uncommon, with the majority of cases reported in patients on PD. There are also a few reported cases in patients with other serious underlying disease who presumably were immunocompromised.[3] Our patient thus had two apparent risk factors for the infection, her PD and the steroids for her lupus.

Although more 50 species of NTM have been reported, most NTM infections in PD have been caused by rapidly growing NTM. These organisms, M. fortuitum, M. chelonae, and M. abscessus,[4] become apparent on culture within 5 to 10 days, much earlier than other species of NTM. They have been implicated in surgical site, catheter-related, and superficial skin infections.[5] M. fortuitum PD-related infection is most commonly reported, followed by M. chelonae. M. abscessus is more likely to cause pulmonary infection.[6]

PD-related infection can occur at the catheter exit site, in the skin tunnel, or intraperitoneally, each of which has been reported involving M. chelonae. In 1981, Band et al. first reported 17 cases of PD peritonitis caused by M. chelonae-like organisms, a number of which involving contaminated equipment.[7] Since 1987, only seven similar cases have been reported in the English literature, even though PD is commonly used for renal replacement.[2],[8–13] The clinical characteristics of these seven patients plus ours are summarized in Table 1. In half the cases, including ours, other organisms were also cultured, including Staphylococcus, Streptococcus, and Corynebacterium. One patient had M. chelonae peritonitis caused by hematogenous dissemination from the remains of a shunt.[2] Three of the patients had diabetes, which might have contributed to some degree of immunocompromise.

Table 1 Reports of Mycobacterium chelonae infection in patients on peritoneal dialysis

Case	Age, gender	Cause of renal failure	Site	Other pathogen	Treatment	Duration of antibiotics (months)	Catheter removal	Complication	Outcome	Reference
1	35, F	Agenesis	P	—	Kanamycin Erythromycin	2	Yes	Abscess, adhesions	(NM)	[2]
2	35, M	Diabetes	P	α-hemolytic streptococcus	Erythromycin	2	Yes	—	HD	[8]
3	63, M	Unknown	P	—	Doxycycline Amikacin	6	Yes	—	TP	[9]
4	40, M	ADPKD	E	CNS	Ciprofloxacin^†	2	Yes	—	TP	[10]
5	61, F	Diabetes	E,P	Corynebacterium S. epidermidis	Ciprofloxacin Calrithromycin Amikacin	2	Yes	—	HD	[11]
6	28, F	(NM)	E	CNS	Clarithromycin TMP/SFX Ofloxacin Amikacin	4	Yes	—	HD	[12]
7	60, M	Diabetes, hypertension	E	—	Amikacin Clarithromycin	3.5	No	—	PD	[13]
8	45, F	SLE	E,P	MRSACorynebacterium	Imipenem, Tobramycin, Clarithromicin	4.5	Yes	Adhesion ileus	HD	Our case

*Other pathogens grew before or simultaneously with Mycobacterium chelonae.

^†No antibiotics were given after catheter removal.

Abbreviations: ADPKD = autosomal dominant polycystic kidney disease, CNS = coagulase-negative staphylococci, E = exit-site or tunnel infection, HD = hemodialysis, MRSA = methicillin-resistant Staphylococcus aureus, (NM) = not mentioned, P = peritonitis, PD = peritoneal dialysis, SLE = systemic lupus erythematosus, TMP/SFX = trimethoprim/sulfamethoxazole, TP = transplantation

NTM-related PD infection often mimics the symptoms and signs of typical bacterial peritonitis, including abdominal pain, fever, and turbid fluid, rather than the more insidious presentation of tuberculous peritonitis. The peritoneal fluid cytology is generally neutrophil-predominant.[1],[14] For this reason, it is difficult early on to distinguish NTM infection from the more common bacterial peritonitis. Atypical pathogens should be considered when the symptoms are persistent or recurrent and unresponsive to empirical antibiotics. On Gram stain of the discharge or dialysate, M. chelonae appear fragmented and beaded and may be mistaken for gram-positive rods, diphtheroids, or debris.[9] An acid-fast stain would be helpful, but in fact the correct diagnosis is usually not made until culture results are reported. Rapidly-growing NTM grow on nonselective bacteriologic media, but usually need at least five days before colonies are apparent. If the laboratory policy is to discard cultures with no growth after three days, the pathogen may be missed.[6] It is important, therefore, that the laboratory be notified if an atypical organism is suspected.

Standard anti-tuberculous therapy is usually ineffective in M. chelonae infections.[5] Agents to which it may be sensitive include clarithromycin, amikacin, cefoxitin, imipenem, doxycycline, and fluoroquinolones, but it does tend to have more drug resistance than M. fortuitum.[9],[10] There are no clear guidelines for treating NTM PD-related infections, but it has been recommended that at least two antibiotics to which the organism is sensitive be given for a prolonged period.[1] The mean treatment duration in the cases we reviewed was 3.3 ± 1.5 months (see Table 1). In all those cases reports except for one, the catheter was also removed. In case 4, antibiotics failed to resolve the infection, but it quickly resolved without any further antimicrobial treatment after the catheter was removed.[10] Siu et al. reported the one patient whose catheter was not removed, but the infection in that case was limited to the exit site and tunnel, with no evidence of peritonitis. In addition to giving antibiotics, they deroofed the tunnel and shaved the catheter outer cuff (case 7).[13] If signs of peritonitis are present, however, the catheter most likely should be removed.

Although NTM abdominal infections not associated with PD have a high mortality, this apparently is not true of NTM PD infections.[3],[6] Complications, of course, may develop despite a good response to antibiotics and tube removal. Adhesions, abscess, and ileus have been reported;[2],[6] our patient apparently had adhesive ileus. Some patients have been switched to hemodialysis after an NTM infection. In two of the reported cases (numbers 3 and 4), subsequent renal transplantation was successful.[9],[10]

In conclusion, peritoneal infection caused by NTM is an unusual complication of PD. It should be suspected when a patient with apparent bacterial peritoneal infection does not respond to standard treatment. Recovery can be expected if appropriate antibiotics are given and the catheter removed if indicated.