Case report of peritoneal dialysis-associated peritonitis caused by Mycobacterium abscessus in a child, successfully treated with dual β-lactam combination therapy

ABSTRACT

Background

Peritoneal dialysis-associated peritonitis caused by Mycobacterium abscessus is a rare and difficult-to-treat infection that frequently results in peritoneal dialysis failure. Since M. abscessus is intrinsically resistant to many antibiotics, therapeutic decisions are challenging. While data are limited, a prolonged course of antibiotics with at least two or three agents is recommended, guided by susceptibility testing. Many regimens use amikacin, which can worsen renal function and cause deafness. There is limited safety and efficacy data on newer antimycobacterial medications in children. Safe and well-tolerated options are needed for the treatment of infections caused by M. abscessus.

Case Summary

Herein, we present the case of a toddler with dysplastic congenital renal disease, who developed peritoneal dialysis catheter-related M. abscessus peritonitis and was treated with a multidrug regimen including two β-lactam antibiotics. This resulted in clinical and microbiologic cure, despite peritoneal failure and transition to hemodialysis. Based on predicted synergy testing, the patient was treated with a combination that included meropenem and ceftaroline for the initial intensified intravenous phase, followed by a regimen that included amoxicillin combined with cefdinir for step-down therapy. This allowed the option of enteral therapy and limited the use of more toxic medications or agents with inadequate information for use in children.

Conclusion

This case highlights the potential benefit of dual β-lactam therapy for the treatment of M. abscessus infection as a well-tolerated regimen for a difficult-to-treat infection. To our knowledge, this is the first report using this approach for the treatment of M. abscessus peritonitis.

INTRODUCTION

Nontuberculous mycobacteria (NTM) peritoneal dialysis (PD)-associated peritonitis is a rare and difficult-to-treat infection (1, 2). Complete cure is low, and mortality can be high (3, 4). Less than 20% of patients are able to resume PD following NTM infection (1). Due to biofilm formation, source control with removal of the infected catheter is recommended (1, 2). Multidrug regimens are used to prevent the emergence of resistance. A prolonged multidrug course is often recommended (5); however, drug toxicity can lead to intolerance or poor adherence, and the optimal combinations and duration of therapy have not been established (6). Consultation with an expert in infectious diseases is recommended to guide treatment for NTM infections (1, 2).

NTM peritonitis with Mycobacterium abscessus can be particularly challenging to treat due to additional antimicrobial resistance concerns and limited therapeutic options. While no specific guidelines are established for the treatment of peritonitis, society guidelines for the treatment of NTM pulmonary disease classify M. abscessus infection based on macrolide susceptibility and suggest a multidrug regimen that includes three or more active drugs guided by in vitro testing in the initial phase of treatment (5). There are three genomically characterized subspecies, of which M. abscessus subsp. abscessus can be particularly challenging to treat due to intrinsic and acquired resistance (1, 2). Alarmingly, treatment outcomes for drug-susceptible M. abscessus subsp. abscessus are often worse than those for multidrug-resistant tuberculosis (MDR-TB) and may approximate those of extensively drug-resistant tuberculosis (XDR-TB) (7). Elevated minimum inhibitory concentrations to backbone β-lactam treatments (imipenem-cilastatin, cefoxitin) in M. abscessus subsp. abscessus make achieving pharmacodynamic targets difficult, and the optimal combination regimens and durations are not established (8). While most M. abscessus organisms are susceptible to amikacin, the toxicity profile of aminoglycosides limits their use over extended periods (9, 10). Hence, safe and effective treatment options are needed (6). Dual β-lactam therapy is emerging as a promising, well-tolerated synergistic combination for M. abscessus infection to restore susceptibility to β-lactam agents (11). We present the case of a child with M. abscessus subsp. abscessus PD-associated peritonitis successfully treated with dual β-lactam combination therapy.

CASE PRESENTATION

A 2-year-old toddler with end-stage renal disease secondary to congenital renal dysplasia, managed with nighttime continuous cycler peritoneal dialysis (PD), was admitted to the hospital for treatment of Pseudomonas aeruginosa catheter-related exit-site infection, after presenting with one week of thick green mucous discharge at the catheter site. On arrival to the emergency room, vital signs were within normal range, and the patient was well-appearing despite thick purulent discharge emerging from the PD exit site. The abdomen was soft and non-tender to palpation. Initial laboratory findings revealed a white cell count of 13.8 × 10⁹/L (reference range 6.0–17.5 × 10⁹/L), hemoglobin of 9.5 g/dL (reference range 10.5–13.5 g/dL), platelet count of 595 × 10⁹/L (reference range 150–450 × 10⁹/L), and C-reactive protein of 5.48 mg/L (reference range 0–5.00 mg/L). Analysis of the peritoneal fluid from admission did not reveal evidence of peritonitis by cell count (27 total nucleated cells/µL, 18% neutrophils) or Gram stain, and PD fluid culture was negative. Empiric therapy was initiated with intravenous ceftazidime 50 mg/kg (adjusted for renal impairment), intraperitoneal cefepime 125 mg/L, and topical gentamicin 0.1% three times daily over the exit site.

On the night of admission, she developed hypotension and leukocytosis of 22.4 × 10⁹/L (reference range 6.0–17.5 × 10⁹/L) with left shift (69% neutrophils, 18% bands) and an increase of the C-reactive protein to 175 mg/L (reference range 0–5.00 mg/L). Antibiotic therapy was broadened to meropenem, initially dosed at 20 mg/kg (adjusted for renal impairment). Blood culture was negative, leukocytosis improved, and blood pressure stabilized, but she developed a fever of 39.4°C on the fifth day. Evaluation for a new source of fever indicated development of peritonitis with an increase in peritoneal fluid cell count (11,625 total nuclear cells/µL, 77% neutrophils) on day 8 of admission. Enteral linezolid 10 mg/kg every eight hours, intraperitoneal vancomycin 25 mg/L, and intraperitoneal fluconazole 40 mg/L were added to the antibiotic regimen.

Repeat bacterial cultures collected from the peritoneal fluid on day 8 of hospitalization detected 1 + acid-fast bacilli five days later, on day 13 of hospitalization. In the setting of a prior negative routine screening for interferon gamma release assay and lack of tuberculosis risk factors, treatment was empirically initiated for rapidly growing nontuberculous mycobacterium (NTM), with the addition of intravenous amikacin 5 mg/kg (re-dosed by drug levels) and azithromycin 10 mg/kg every 24 hours (Fig. 1). The PD catheter was removed on day 15, and the patient was transitioned to thrice weekly hemodialysis. Mycobacterium abscessus subsp. abscessus was identified via HAIN line-probe assay and isolated from multiple peritoneal fluid cultures prior to catheter removal (Fig. 1). All blood cultures, including bacterial and mycolytic cultures, were negative. The fever resolved following PD catheter removal, and inflammatory markers improved. CT scan of the abdomen and pelvis demonstrated multiple intraperitoneal fluid collections concerning for abscess formation, with the most well-defined collection located in the right anterior hemipelvis measuring 3.4 × 1.9 × 3.2 cm. The pediatric surgical team was consulted, but no surgical intervention was initially advised to limit the risk of creating draining fistulas.

Fig 1.

Antimycobacterial treatment timeline. Timeline is not scaled. ¹Antibiotics administered after hemodialysis. ²Cefdinir administered with amoxicillin to optimize synergy. GT, gastric tube; h, hours; IV, intravenous; kg, kilograms; mg, milligrams; PD, peritoneal dialysis; q, every; TDM, therapeutic drug monitoring with redosing based on trough levels. ⁺Mycobacterium abscessus grew from all bacterial cultures collected from the peritoneal fluid on day 8, 9, 10, 12, 13, and 14 of admission. The peritoneal dialysis catheter was removed on day 15 of admission.

Following the results of susceptibility testing, therapy was optimized to meropenem 40 mg/kg every 24 hours (adjusted for renal impairment), cefoxitin 500 mg/kg every 24 hours (adjusted for renal impairment), azithromycin 10 mg/kg every 24 hours, and amikacin 5 mg/kg (re-dosed by drug levels) on day 36, utilizing dual β-lactam therapy to optimize synergy between cefoxitin and meropenem (Table 1). Additional dual β-lactam synergy testing further suggested enhanced synergy with the combination of imipenem or meropenem and ceftaroline (Table 2); hence, cefoxitin was switched to intravenous ceftaroline 5 mg/kg every 24 hours (adjusted for renal impairment) on day 47. The largest fluid collection was ultimately drained by interventional radiology on day 49, and the fluid culture was sterile. Amikacin was discontinued after six weeks of therapy on day 55, due to decreasing otoacoustic emissions.

TABLE 1.

Mycobacterium abscessus subsp. abscessus susceptibility testing^a

Antibiotic	Minimal inhibitory concn (µg/mL)	Interpretationb
Amikacin	8	Susceptible
Cefoxitin	32	Intermediate
Ciprofloxacin	4	Resistant
Clarithromycin	0.25	Susceptible
Clofazimine	0.25	No CLSI interpretation available
Doxycycline	>8	Resistant
Imipenem	16	Intermediate
Linezolid	16	Intermediate
Moxifloxacin	>4	Resistant
Tigecycline	0.25	No CLSI interpretation available
Trimethoprim/sulfamethoxazole	>4	Resistant

^a Macrolide screen: macrolide susceptibility is predicted, no detectable mutation in rrl gene. Aminoglycoside screen: aminoglycoside susceptibility is predicted, no detectable mutation in rrs gene. Inducible macrolide resistance: inducible macrolide resistance is not predicted; a cytosine has been detected at position 28 (C28) of the erm(41) gene.

^b Breakpoints from the CLSI M24S, 2nd ed. (12). Organism identification performed via HAIN line-probe assay. Macrolide and aminoglycoside resistance screens were performed using the GenoType NTM-DR molecular assay. Performed at the University of Florida Health Pathology Laboratories Nontuberculosis Mycobacterial Lab.

TABLE 2.

Mycobacterium abscessus subsp. abscessus dual-β lactam synergy testing^a

Antibiotic (µg/mL)	Test method	Minimal inhibitory concn (µg/mL)
Imipenem	Alone	16
Cefuroxime	Alone	256
Cefdinir	Alone	128
Ceftaroline	Alone	64
Imipenem + cefuroxime	Cefuroxime fixed at 4	0.5
Imipenem + cefdinir	Cefdinir fixed at 4	0.5
Imipenem + ceftaroline	Ceftaroline fixed at 4	0.25
Imipenem + amoxicillin + relebactam	Amoxicillin fixed at 8, relebactam at 4	≤0.25
Cefdinir + amoxicillin	Amoxicillin fixed at 8	1

^a MICs of ceftaroline, imipenem, cefuroxime, and cefdinir were determined using broth microdilution. Approximately 5 × 10⁵ CFU/ml were inoculated into Middlebrook 7H9 broth supplemented with 10% (vol/vol) oleic albumin dextrose catalase and 0.05% (vol/vol) Tween 80. When more than two drugs were combined, the second drug was added at a fixed concentration of 4 μg/mL or 8 μg/mL to serial dilutions of imipenem, meropenem, or cefdinir. Isolates were incubated with test agents at 30°C for 3 to 7 days, and MIC was defined as the lowest antibiotic concentration that prevented visible bacterial growth. Active moiety compounds, and not pro-drug, e.g., ceftaroline (unphosphorylated) and not ceftaroline fosamil, were used. Performed at Case Western Reserve University at the Cleveland VA Medical Center.

Abdominal imaging findings continued to improve on combination therapy, showing healing of the fluid collections and peritoneal inflammation with calcification. Inflammatory markers declined further, and C-reactive protein normalized by day 167 of admission. Based on the expanded synergy testing (Table 2), the patient was transitioned to an all-enteral step-down regimen with azithromycin 10 mg/kg every 24 hours, amoxicillin 20 mg/kg every 24 hours (adjusted for renal impairment), and cefdinir 7 mg/kg every 24 hours (adjusted for renal impairment). The patient tolerated the therapy well with good adherence. Following prolonged hospitalization, the patient was discharged home after 210 days. Additional complications that arose during the hospital course included ventricular fibrillation from hyperkalemia requiring defibrillation and the development of pulmonary hypertension requiring oxygen supplementation and pulmonary vasodilator therapy. The patient completed a total of 12 months of antimycobacterial therapy from the day of catheter removal. Repeat ultrasound at the end of therapy and two months after showed resolution of all fluid collections.

DISCUSSION

In this report, we describe the use of dual β-lactam synergistic combination therapy as part of a multidrug regimen to successfully treat a child with Mycobacterium abscessus subsp. abscessus PD-associated peritonitis.

M. abscessus is considered among the most pathogenic and drug-resistant of the rapidly growing NTM (13). In our case, evidence of inducible or mutational macrolide resistance was not shown; hence, we were fortunately able to treat with a macrolide-containing regimen for the entire course. Most M. abscessus isolates are also susceptible to amikacin, but development of side effects, such as otovestibular toxicity, often precludes prolonged use, as noted in our case. Susceptibility to all other tested antibiotics was reduced (Table 1); hence, we sought a novel option that could be safely administered to children over an extended period.

Our therapeutic rationale for using combination synergistic dual β-lactam therapy with ceftaroline and meropenem during the initial intensified treatment phase is derived from accumulating clinical reports and encouraging observations using this approach (11, 14–17). This includes a single pediatric report of a 3-year-old child with bronchiectasis and M. abscessus pulmonary infection treated with a multidrug regimen that contained ceftaroline to enhance the activity of meropenem, resulting in both clinical and microbiologic cure (14). M. abscessus is known to produce β-lactamases that lead to variable β-lactam susceptibility (18). In vitro studies demonstrate that imipenem combined with ceftaroline significantly lowers the imipenem minimum inhibitory concentration of clinical isolates. Both drugs target the same peptidoglycan synthesis enzymes. A plausible mechanism for synergy may be explained by either binding of multiple targets with differential inhibition of enzymes involved in peptidoglycan synthesis or blocking of β-lactam hydrolysis to restore susceptibility (11, 19–21). Though not tested in our case, meropenem was preferentially used given its lower risk of inducing seizures in children (22). Our decision to use amoxicillin with cefdinir for step-down enteral therapy, after infection was deemed well-controlled, was based on in vitro testing (Table 2) and reports showing synergy with amoxicillin and other β-lactam combinations, such as imipenem (17, 23). The synergy testing uses varying concentrations in doubling dilutions of one β-lactam with a fixed dose of another β-lactam agent. The minimal inhibitory concentration (MIC) of monotherapy is compared to the MIC in the presence of a fixed-dose second agent to determine whether there is an appreciable drop in MIC. This was the case for both combination regimens utilized, where the imipenem MIC dropped from 16 mcg/mL when tested alone to an MIC of 0.5 mcg/mL when tested with a fixed concentration of 4 mcg/mL of cefdinir. The cephem concentration was fixed at 4 μg/mL because this is the level where synergy is observed (19, 20). Similarly, the cefdinir concentration dropped from 128 mcg/mL to 1 mcg/mL when tested with a fixed concentration of 8 mcg/mL amoxicillin.

Antimycobacterial treatment dosages for children are not well established. When available, guideline-based (24) pediatric dosages were renally adjusted and increased for growth throughout the treatment course. There is a paucity of data on the pharmacokinetic profiles of these medications in patients receiving renal replacement therapies. Aggressive dosing using maximum tolerable doses should be considered when using oral β-lactams for synergy, while carefully considering dose-limiting toxicities that can occur with many of the β-lactams (25, 26).

The optimal duration of therapy is also unknown, but often exceeds one year (5, 24). We opted to treat for this duration to decrease the risk of relapse of infection and increase the chance of a successful outcome as a renal transplant candidate. Alternative approaches were considered. Newer antimycobacterial agents include the tetracycline derivatives eravacycline and omadacycline, though safety is not established in children. Similarly, there is limited data on using the diarylquinoline antibiotic bedaquiline, which is associated with cardiotoxicity and is difficult to access. The riminophenazine antibiotic, clofazimine, and bacteriophage therapy are investigational therapies confined to expanded access pathways. In contrast, β-lactam antibiotics are widely available and are among the best tolerated antibiotics with an excellent established profile.

In summary, we report the successful treatment of M. abscessus subsp. abscessus PD-associated peritonitis complicated by multiple intra-abdominal abscesses using dual β-lactam synergistic therapy as part of a multidrug approach. A similar strategy has been used to treat other challenging cases of M. abscessus infection. To our knowledge, this is the first report using dual β-lactam therapy to treat peritonitis and provides further support for use in hard-to-treat infections based on the safety of this approach. Larger studies and clinical investigations are needed to determine the best combinations, dosages, and durations of treatment for this promising antimycobacterial synergism.