Mycobacterium fortuitum peritoneal dialysis-related peritonitis in a child: A case report and review of the literature

Abstract

BACKGROUND

Non-tuberculous mycobacteria (NTM) are an uncommon but serious cause of peritoneal dialysis (PD)–related infections. NTM peritonitis typically necessitates PD catheter removal, PD withdrawal, and aggressive, prolonged antimicrobial treatment. Few reported cases of NTM peritonitis in the pediatric population exist.

METHODS

We describe a case of a 9-year-old boy on PD after kidney allograft failure who developed Mycobacterium fortuitum peritonitis, and we summarize the available literature on M. fortuitum peritonitis in pediatric patients receiving PD.

RESULTS AND CONCLUSION

Therapeutic options were limited by adverse medication effects and risk of drug–drug interactions in a patient with complex mental health comorbidities. Clofazimine presented an acceptable oral treatment option for long-term therapy in combination with ciprofloxacin and was well tolerated by this patient. Prompt PD catheter removal followed by 6 months of dual antimicrobial therapy resulted in a full recovery and successful re-transplantation with no infection relapse.

Case Presentation

A 9-year-old boy returned to peritoneal dialysis (PD) 2 years after kidney transplant as a result of allograft failure from T-cell–mediated rejection and BK virus nephropathy. Five months into his PD course, he developed abdominal pain and emesis, cloudy peritoneal dialysis effluent, and intermittent fevers not resolving with home intraperitoneal (IP) cefazolin, which is local empiric gram-positive coverage for patients with suspected peritonitis and no history of methicillin-resistant Staphylococcus aureus. He presented to hospital, and a Psychrobacter spp (a gram-negative organism related to Moraxella) was isolated from his PD fluid, at which point his therapy was changed to IP ceftazidime. After symptomatic improvement, he was discharged home to complete a 3-week course of IP ceftazidime with oral fluconazole added for fungal prophylaxis. Six days later, he re-presented with ongoing symptoms again suggestive of peritonitis, including cloudy effluent, abdominal pain, and fever. His immunosuppressive medication at the time included low-dose tacrolimus and prednisone for prevention of donor-specific antibody development. He had also received anti-thymocyte globulin (ATG) 9 months earlier, resulting in ongoing lymphopenia. His pharmacological management for behaviour and anxiety included methylphenidate, aripiprazole, and sertraline. In addition to immunosuppression, his risk factors for peritonitis included behavioural issues leading to frequent PD catheter disconnections and touch contamination. Before his first kidney transplant, he had an episode of PD peritonitis due to Acinetobacter pittii that was successfully treated.

On physical exam, he was hemodynamically stable with a diffusely tender abdomen but no evidence of PD catheter exit-site infection. Laboratory investigations were notable for ongoing lymphopenia (0.81 × 10⁹/L) post-ATG treatment, with a normal white blood cell count of 5.5 × 10⁹/L and normal neutrophils of 3.95 × 10⁹/L. CRP was elevated at 134 mg/L. Cell count in the peritoneal fluid at this time measured more than 17,000 × 10⁶/L with 77% neutrophils. Peritoneal fluid microscopy and cultures for typical bacteria, mycobacteria, and fungi were initially negative. Peritoneal fluid culture subsequently showed faint gram-positive or variable bacilli from the enrichment broth only. They failed to grow on subculture and were sent to the provincial reference laboratory for further speciation. The patient was discharged home to continue IP ceftazidime, with IP cefazolin added for additional gram-positive coverage.

Nine days after discharge, he was re-admitted with persistent abdominal discomfort and ongoing low-grade fever. Physical exam findings were unchanged. Vancomycin IP was added for further coverage of gram-positive organisms. Shortly thereafter, PD culture results from previous admission reported as gram-positive bacilli in enrichment broth were revised by the reference laboratory and reported as acid-fast bacilli, most likely a rapid-growing non-tuberculous Mycobacterium spp (NTM). Given the patient’s stable condition and possible presence of other colonizing bacteria, IP ceftazidime and IP cefazolin were continued while awaiting NTM speciation. IP vancomycin was stopped. Sequencing of hsp65 gene fragment of the isolate confirmed the organism as Mycobacterium fortuitum on day 3 post-admission. The molecular identification was consistent with rapid organism growth on subculture (<7 d) and nonchromogenic colonies. Workup for disseminated infection included abdominal computed tomography, which did not show any discrete intraperitoneal collections. Mycobacterial blood culture was sent and subsequently reported as negative. There was no new evidence of exit-site or tunnel infection. The patient’s peritoneal catheter was promptly removed for source control, and he was transitioned to hemodialysis. He was empirically treated with intravenous (IV) imipenem, IV amikacin, and oral moxifloxacin, based on susceptibility patterns demonstrated in case reports. At this point, his peritonitis symptoms had resolved, and he did not have any further fever. Baseline audiology assessment at the time IV amikacin was started showed mild sensorineural hearing loss, which was unchanged from a previous exam. Therapeutic drug monitoring was conducted using peaks and troughs to ensure appropriate dosing of amikacin. Empiric IV imipenem was subsequently changed to oral sulfamethoxazole–trimethoprim because most local M. fortuitum isolates possess only intermediate susceptibility to imipenem and because of difficulty administering two IV agents to a child with significant behavioural challenges.

M. fortuitum susceptibility results were returned from the National Molecular Laboratory 2 weeks later and are presented in Table 1. The patient’s sulfamethoxazole–trimethoprim was stopped, and oral ciprofloxacin was initiated in place of moxifloxacin to reduce risk of QT prolongation in combination with his other home medications, including tacrolimus and sertraline. An electrocardiogram (ECG) at this time showed borderline prolonged QT interval, which returned to normal after discontinuation of moxifloxacin. Because of concerns of ototoxicity with long-term use, amikacin was stopped after 2 weeks and was replaced with oral linezolid.

Table 1:

Susceptibility testing results for Mycobacterium fortuitum isolate

Drug	MIC interpretation	MIC (mg/mL)
Amikacin	S	≤1
Cefoxitin	I	64
Ciprofloxacin	S	≤0.12
Clarithromycin	R	16
Clofazimine	*	0.12
Doxycycline	R	16
Imipenem	I	8
Linezolid	S	8
Moxifloxacin	S	≤0.25
Sulfamethoxazole–trimethoprim	R	>8/152

*No validated MIC breakpoint for clofazimine susceptibility

MIC = Minimum inhibitory concentration; S = Sensitive; R = Resistant; I = Intermediate

Concerns about myelosuppression and neuropathies with long-term linezolid use, as well as a potential risk of serotonin syndrome with the combination of linezolid, sertraline, and methylphenidate, prompted a request for access to clofazimine through the Health Canada Special Access Program. Although the patient did not develop any linezolid-associated adverse effects, clofazimine was started 6 weeks later after drug access was secured. Clofazimine was dosed at 50 mg every other day. This regimen was selected on the basis of established dosing for other NTM indications, and the pharmacokinetic properties of the drug made its removal by dialysis unlikely. Clofazimine drug levels were not obtained as a result of limited access to testing facilities. Because of the long half-life of clofazimine (approximately 25 d), linezolid was continued for 4 weeks after the initiation of clofazimine, at which point steady-state achievement was assumed. Monthly ECGs were performed while the patient was receiving clofazimine because of its known QT prolongation risk and ongoing therapy with other potentially QT-prolonging medications. The patient experienced intermittent nausea and decreased appetite while on clofazimine but was able to tolerate the 15-week course. He did not experience other adverse effects related to clofazimine such as QT prolongation, mood changes, or skin discoloration.

In total, the patient completed a 6-month course of antimicrobial therapy for M. fortuitum after PD catheter removal, receiving two effective agents in conjunction at all times (Figure 1). Testing for residual M. fortuitum in the peritoneal fluid was not possible because a PD catheter was no longer present and there was no underlying fluid collection to sample. Complete resolution of the infection was presumed. The patient remained on hemodialysis until time of re-transplantation 2 months after completion of antimicrobial therapy. He has had no febrile events or episodes of abdominal pain since. Given that his was an isolated case of NTM peritonitis at our centre, no investigation of potential hospital-acquired sources of M. fortuitum occurred. The presumed source of M. fortuitum was environmental contamination.

Figure 1:

Timeline of antimicrobial therapy

Discussion

Bacterial peritonitis is a significant complication of PD and the most common reason for cessation of PD. As many as 40% of peritonitis cases are culture negative (1). A culture-negative, relapsing clinical course is typical of NTM peritonitis. NTM accounts for approximately 3% of PD-related peritonitis (1), of which M. fortuitum, a rapid growing NTM, is the most commonly reported species (2,3). In a recently published systematic review, Washida and Itoh described 58 cases of PD-related NTM peritonitis, with M. fortuitum isolated in 37 cases (3). These were primarily in adults ranging in age from 15 to 86 years. M. fortuitum peritonitis has previously been reported in only two pediatric patients, male adolescents aged 15 years and 16 years (4,5). Key clinical features of these patient cases and our own are given in Table 2.

Table 2:

Clinical features of pediatric patients with PD-related Mycobacterium fortuitum peritonitis

Case reference	Age (y), sex	Duration of PD at time of infection	PD catheter removal	Antibiotic therapy	Treatment duration	Adverse medication effects	Outcome
Current case	9, male	5 mo	Yes	IV amikacin (2 wk) PO linezolid PO ciprofloxacin PO clofazimine	6 mo	Decreased appetite, gastrointestinal upset with clofazimine	Recovery, re-transplantation
(4)	15, male	NR	Yes	IV amikacin IV cefoxitin	NR	Desquamation of hands and feet attributed to cefoxitin	NR
(5)	16, male	36 mo	Yes	PO ciprofloxacin PO sulfamethoxazole–trimethoprim	≥9 mo	None reported	Recovery

PD = Peritoneal dialysis; IV = Intravenous; PO = By mouth; NR = Not reported

NTM species are widely present in soil and water sources, are resistant to commonly used disinfectants, and have been associated with health care mycobacterial outbreaks, although cases are typically isolated (6). Diagnosis of NTM peritonitis may be delayed by the initially misleading Gram staining, the failure to prolong culture incubation beyond 2–3 days, and the possibility of mycobacterial culture–negative disease, which may occur in one-third of cases (2). Diagnosis may also be delayed as a result of significant clinical presentation overlap with bacterial peritonitis and lack of clinician familiarity with NTM as a peritonitis pathogen. Prompt identification of NTM species is important because of substantial differences in antimicrobial susceptibilities (2). M. fortuitum, although usually susceptible to multiple oral antimicrobial agents (4), typically necessitates prolonged treatment with two or more antimicrobials in conjunction with source control (2,3). Because of M. fortuitum’s and other NTM species’ hardiness and tendency for growth in biofilms, removal of foreign bodies such as PD catheters is essential to recovery (6).

The optimal treatment duration and approach for NTM peritonitis is unknown. In a systematic review of NTM peritonitis, Song et al described 57 cases in which mean treatment duration was 4.6 months but ranged from 2 days to 24 months (2). Treatment duration was described in only one of the previously reported pediatric cases, in which antimicrobial therapy continued for more than 9 months after PD catheter removal (5). In a review by Washida and Itoh (3), patients received between one and four antimicrobial agents for treatment of NTM peritonitis, with the majority receiving two or more antibiotics. Although no standard treatment approach exists for M. fortuitum peritonitis, the use of at least two agents with activity against the isolate is recommended for both M. fortuitum lung and skin and soft tissue disease (6). Lengthy therapy with multiple antimicrobial agents that are effective against NTM, however, poses significant risk of adverse effects and may be further complicated by a high burden of possible drug–drug interactions. Use of linezolid beyond 28 days of therapy, for example, is cautioned because of the adverse effects of myelosuppression and peripheral and optic neuropathies (7). In our case, long-term use of linezolid was further complicated by potential drug–drug interactions with the patient’s medications for mood and behaviour, including sertraline, aripiprazole, and methylphenidate. Serotonin syndrome has been reported when linezolid is used in combination with selective serotonin reuptake inhibitors and serotonergic psychiatric medications such as aripiprazole (8), and worsening hypertension has been reported when it is combined with methylphenidate (9).

Clofazimine is a riminophenazine antimicrobial approved by the US Food and Drug Administration for treatment of Mycobacterium leprae (10). Clofazimine has been used successfully in the treatment of M. fortuitum and other NTM infections (11–15). Its main adverse effects include gastrointestinal upset and skin discoloration (12). Despite its not infrequent use for the treatment of NTMs, clinical breakpoints have not yet been established by the Clinical and Laboratory Standards Institute. Many studies, however, do report low clofazimine minimum inhibitory concentrations in their isolates (11). Experience with clofazimine in the pediatric population is limited, but the drug was well tolerated in a recent case series of 27 children with extra-pulmonary Mycobacterium abscessus infection (12) and in a recent report of a 6-year-old with M. abscessus peritonitis (15). For our case, clofazimine presented an appealing alternative therapy because of the minimal risk of serious drug–drug interactions, acceptable adverse effects, oral administration, and long half-life. Our patient reported mild gastrointestinal discomfort, decreased appetite, and intermittent diarrhea while on clofazimine, but he was able to tolerate therapy for the planned duration. He did not develop QT prolongation while on clofazimine. Our case is the first reported in which M. fortuitum peritonitis in a pediatric patient has been successfully treated using clofazimine as part of long-term dual therapy.

Although rare, NTM peritonitis is associated with a high mortality rate. In the review by Washida and Itoh (3), despite most patients receiving antimicrobial therapy, 10 of the 58 (17.2%) cases presented died. Serious complications, including bacteremia, abdominal abscesses and adhesions, bowel perforation, and resistant infection, were reported, and very few patients returned to peritoneal dialysis therapy. In our case, early diagnosis and prompt source control with PD catheter removal aided in rapid resolution of peritonitis symptoms. Although adverse effects and drug–drug interactions complicated his treatment options, access to clofazimine allowed this patient to complete the lengthy antimicrobial course with two effective oral agents, without disruption to his stable home medications for behaviour and anxiety and without any relapse of peritonitis symptoms.

Conclusion

Suspicion of NTM infection should remain high in cases of culture-negative peritonitis in which response to empiric antibiotic therapy is poor. This rare cause of peritonitis is associated with high mortality and severe complications and typically necessitates PD catheter removal and prolonged therapy with multiple antibiotic agents to eradicate the bacteria. Clofazimine may be considered as a therapeutic option for children for the treatment of NTM peritonitis, including M. fortuitum, because of its favourable adverse effect profile for long-term use.

Ethics Approval:

N/A

Informed Consent:

Written consent for writing and publication of this case reported was obtained from the patient’s caregivers.

Registry and the Registration No. of the Study/Trial:

N/A

Funding:

No funding was received for this work.

Disclosures:

The authors have nothing to disclose.

Peer Review:

This manuscript has been peer reviewed.

Animal Studies:

N/A