Abstract

Introduction. Achromobacter species (spp.) peritonitis has seldom been identified in medical literature. Scarce cases of Achromobacter peritonitis described previously have been correlated with peritoneal dialysis and more sparingly with spontaneous bacterial peritonitis. Achromobacter exhibits intrinsic and acquired resistance, especially in chronic infections, to most antibiotics. This article conducts a literature review of all previously reported Achromobacter spp. peritonitis and describes the first reported case of Achromobacter peritonitis as a complication of percutaneous endoscopic gastrostomy (PEG) tube placement. Discussion. Achromobacter peritonitis as a complication of PEG-tube placement has not been previously reported. In our patients’ case, the recently placed PEG-tube with ascitic fluid leakage was identified as the most plausible infection source. Although a rare bacterial peritonitis pathogen, Achromobacter may be associated with wide antimicrobial resistance and unfavorable outcomes. Conclusion. No current guidelines provide significant guidance on treatment of PEG-tube peritonitis regardless of microbial etiology. Infectious Disease Society of America identifies various broad-spectrum antibiotics targeting nosocomial intra-abdominal coverage; some of these antimicrobial selections (such as cefepime and metronidazole combination) may yet be inadequate for widely resistant Achromobacter spp. Recognizably, the common antibiotics utilized for spontaneous bacterial peritonitis, i.e., third generation cephalosporins and fluoroquinolones, to which Achromobacter is resistant and variably susceptible, respectively, would be extensively insufficient. Piperacillin/tazobactam (P/T) and carbapenem were identified to provide the most reliable coverage in vitro; clinically, 5 out of the 8 patients who received either P/T or a carbapenem, or both, eventually experienced clinical improvement.

1. Introduction

Achromobacter species (spp.) are nonfermenting, oxidase positive, catalase positive, aerobic, and motile Gram-negative rods [1]. As environmental organisms, Achromobacter spp. are naturally distributed in aqueous surroundings including soil and water [24]. Similarly, they may contaminate fluids in the health care setting, and have been considered infectious etiologies in nosocomial infection outbreaks associated with contaminated fluids [5], pressure transducers [6], incubators [7], and disinfectants [7]. Although opportunistic organisms [8], Achromobacter spp. are deemed to be the etiology for a myriad of infectious diseases with significant morbidity and mortality such as biliary tract sepsis [9], meningitis [10], pneumonia [11], spontaneous bacterial peritonitis [12, 13], peritoneal dialysis associated peritonitis [5, 14], urinary tract infection [11], conjunctivitis [15], osteomyelitis [16], prosthetic knee infection [17], mesh infection [18], necrotizing pancreatitis [19], prosthetic valve endocarditis [20], bacteremia [6, 11, 21], and cystic fibrosis [4, 22], among others. While the majority of these diseases occur as a polymicrobial infection in immunocompromised states such as cancer [21], human immunodeficiency virus/acquired immunodeficiency syndrome [23], chronic renal failure or diabetes mellitus [20], a portion of them occur due to nosocomial outbreaks in immunocompetent hosts [11, 24].

The organism is highly resistant to commonly utilized antibiotics in intra-abdominal infections. Piperacillin/tazobactam (P/T) and carbapenems show activity in vitro and are considered the most active antimicrobials against Achromobacter spp. [2]. On the other hand, sulfamethoxazole/trimethoprim may be a reasonable non beta-lactam alternative [25]. Due to the extensive inherent and acquired resistance to significant number of antimicrobials, providing dual coverage for chronic infections may be prudent [25].

2. Case Summary

A 65-year-old Caucasian male with a history of chronic alcoholism (without cirrhosis or ascites), hypertension and prior tonsillar adenocarcinoma status post chemo-radiation presented to a regional hospital with progressive fatigue and jaundice, epigastric pain, and decreased appetite causing twenty-five-pound weight loss spanning over a month. Leukocytosis along with lactic acidosis was observed. Patient was treated empirically with P/T for biliary pancreatitis secondary to choledocolithiasis. Initial endoscopic retrograde cholangiopancreatography (ERCP) on day (D) 2 of hospitalization identified distal common bile duct stone; sphincterotomy was performed with 7- French 5 cm plastic common bile duct stent placement. Despite the interventions, liver enzymes continued to rise; antibiotic was altered to levofloxacin and metronidazole on D4 to cover intra-abdominal infections and a repeat ERCP was performed. Recently placed CBD stent, which appeared clogged with debris, was removed uneventfully with snare. Appropriate drainage from CBD and cystic duct was confirmed; however, gall bladder could not be visualized with contrast. Due to the development of pneumonia, antibiotics were switched to linezolid and doxycycline on D6 of hospitalization. On D11, due to poor nutritional status and emerging aspiration pneumonia, PEG-tube was inserted. Subsequently, developing and rapidly worsening ascites and ascitic fluid leakage around the PEG-tube insertion site was identified. Paracentesis was performed on D13 of hospitalization without evidence of peritonitis; however, antibiotics were escalated to meropenem and linezolid due to worsening clinical status. Due to thrombocytopenia, linezolid was discontinued, and antimicrobial therapy was narrowed to meropenem on D17. Due to worsening of ascitic fluid leakage around the PEG insertion site, and exacerbation of liver enzymes, the patient was transferred to our academic facility for further management, including possible cholecystectomy on D17.

Upon transfer, the patient was hemodynamically stable with a constellation of physical exam findings significant for leaking ascites, icterus, spider angioma, fine crackles throughout lung fields, and 3+ bilateral pitting edema; no evidence of encephalopathy was noted. The laboratory tests were remarkable for white blood cells 21,600/mcl, serum creatinine 0.55 mg/dl, aspartate aminotransferase 201 U/L, alanine transferase 74 U/L, alkaline phosphatase 458 U/L, total bilirubin 9.0 mg/dl, serum albumin 1.8 g/dl, and INR 1.5. Patient was further diagnosed with alcoholic cirrhosis decompensated by ascites with MELD-Na score of 23, and suspected alcoholic hepatitis with Maddrey’s discriminant function of 28. Antibiotic therapy was held until further work-up due to previously negative cultures and extended duration of therapy prior to transfer. Interventional radiology guided paracentesis was performed on D22 of hospitalization as bedside procedure was unsuccessful. Ascitic fluid analysis revealed albumin of 0.3 gm/dL and polymorphonuclear neutrophil count of 1950 cells/mcl. PEG-tube was retained to allow for maturation of PEG-tube tract; P/T and vancomycin were initiated and continued for two days (D22-D23); therapy was empirically altered to cefepime, metronidazole, and vancomycin to prevent acute kidney injury secondary to concomitant vancomycin and P/T use, which was continued for four days (D24-27). Ascitic fluid cultures revealed P/T susceptible and cefepime resistant Achromobacter xylosoxidans as the infectious etiology (Isolate was identified by MALDI TOF: MALDI Biotyper CA system. Please refer to Table 1 for minimal inhibitory concentration values). Empiric regimen of cefepime, metronidazole, and vancomycin was narrowed to P/T on D27. Bronchoalveolar Lavage on D27 due to worsening respiratory failure also identified A. xylosoxidans. Following five days of treatment with P/T, however, his abdomen demonstrated peritoneal signs with increasing ascites. Repeat paracentesis confirmed worsening bacterial peritonitis, with additional findings of vancomycin resistant Enterococcus and Candida glabrata. Despite broadened antimicrobial coverage with meropenem and vancomycin for four days (D28–D31), the patient deteriorated with acute kidney injury and hypotension, and was deceased within twenty-four hours of intensive care unit transfer.

3. Discussion

3.1. Nomenclature

The taxonomic designation for Achromobacter spp. has been inconsistent due to the limited differentiating capability of previous identification methods [26, 27]. Solely multilocus sequence typing (MLST) and species identification via nrdA sequence analysis for Achromobacter genus have been deemed to be effective in differentiating between various Achromobacter spp. [26, 27]. Prior to utilization of these methods, various alternate Achromobacter species might have been misidentified as Achromobacter xylosoxidans; currently, reclassification of Achromobacter species has resulted from utilization of these techniques [2631]. The ascitic culture in our patient case was assessed by MALDI TOF MS (Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry); thus, exact species identification might not have been accurate. As such, our literature review attempts to include all previously documented Achromobacter species peritonitis cases to ensure an inclusive analysis.

3.2. Antimicrobial Susceptibility Patterns

Achromobacter susceptibility profile identifies the organism as being extensively resistant among various classes of antimicrobials, including fluoroquinolones, aminoglycosides, and the majority of broad-spectrum β-lactam antibiotics [32]. The resistance-nodulation-cell division (RND)-type multidrug efflux pumps, AxyABM [33, 34] and AxyXY-OprZ [35] have been associated with innate resistance. Essentially, resistant to all cephalosporins, except ceftazidime, due to the presence of AxyABM efflux system [29, 33], Achromobacter spp. may also contain AxyXY-OprZ efflux pump associated with aminoglycoside resistance [29, 35]. On the other hand, most isolates are typically susceptible to carbapenems, piperacillin, and P/T [29]. While a few acquired betalactamases have been described, overall acquired resistance mechanisms remain unknown [3640].

3.3. PEG-Tube Associated Complications

This case report documents the first percutaneous endoscopic gastrostomy (PEG) tube-associated Achromobacter peritonitis. PEG-tube is a reliable source of enteral nutrition for patients with the inability to maintain oral nutrition on a long-term basis. Although minimally invasive, gastrostomy tubes have been associated with major complications such as peritonitis, major gastrointestinal bleeds, aspiration pneumonia, gastroenteric fistula, and gastrocolocutaneous fistula; minor complications include cellulitis, leakage around the tube, and granulation tissue at gastrostomy site [4143]. In order to minimize these complications, the European Society for Clinical Nutrition and Metabolism guidelines on artificial enteral nutrition identifies a variety of contraindications to gastrostomy tubes including, serious coagulation disorders, INR greater than 1.5, platelets less than 50,000/mcl, marked peritoneal carcinomatosis, interposed organs (liver and colon), severe ascites, peritonitis, anorexia nervosa, severe psychosis, and discernibly limited life expectancy [44].

Specific risk factors for PEG-tube peritonitis include poor tissue healing, procedural, and technical issues, BMI over 30 kg/m2, albumin less than 2.5 g/dL, and dislodgement or reinsertion of tube [41, 42]. Pseudomonas aeruginosa, Methicillin resistant Staphylococcus aureus, and enteric Gram-negative rods are implicated as the most common infective etiologies of PEG site infections [4547]. Abuksis et al. report 30-day mortality rate of 4.1–26% for stomal infections [48].

Despite the wide array of Achromobacter spp. infections previously reported, Achromobacter peritonitis is extremely rare and has been documented primarily in the setting of peritoneal dialysis. Despite the fact, Achromobacter infections in peritoneal dialysis are also exceptionally infrequent. Moreover, etiologies reported prior to nrdA sequencing might have been inaccurately identified as A. xylosoxidans and may rather represent Achromobacter spp. [11, 2628]. In order to identify all reported and published cases of Achromobacter peritonitis, thorough PubMed (January 1st, 1966–April 7th, 2019) literature searches were conducted from March 20th, 2019 through April 7th, 2019, and October 10th, 2019 through October 21st, 2019 independently by both primary authors. Search terms that were used include “PEG or percutaneous endoscopic gastrostomy tube peritonitis” with no relevant results, “Achromobacter peritonitis” with four relevant results [4952], and “Alcaligenes peritonitis” with additional 8 relevant results [12, 13, 5358]. Five additional cases were identified through previously published case reports, and extensive search in PubMed with search terms “Achromobacter xylosoxidans”, “Achromobacter”, and “Alcaligenes” [5, 8, 14, 59, 60].

In totality, eighteen previous Achromobacter peritonitis cases were encountered; of which, two were diagnosed as spontaneous peritonitis [12, 13] while remaining sixteen cases of peritonitis were associated with peritoneal dialysis [5, 8, 14, 4960]. No previous cases of PEG tube-associated Achromobacter peritonitis were discovered. Tables 1 and 2 summarize, in as much detail as could be obtained, the infectious diagnosis of each patient case, cultures, and sensitivities, antimicrobial therapies selected, and overall outcomes. Likewise, Figure 1 utilizes a bar graph to indicate the overall in vitro susceptibility profiles of the Achromobacter strains that were encountered in all nineteen cases of Achromobacter spp. peritonitis.

While 14 patients ultimately experienced clinical improvement, 4 patients were deceased. Patient responses to each antimicrobial group were assessed. P/T and carbapenem were identified to provide the most reliable coverage in vitro; clinically, 5 out of the 8 patients who received P/T or a carbapenem, or both, eventually experienced clinical improvement. Of the three patients who did not, patient in case nine was administered P/T but patient was deceased very shortly after presentation. Patient likely presented with high disease severity; thus, overall outcome may not reflect utility of the treatment. Likewise, patient in case thirteen was initiated on amikacin upon presentation, which was continued for seven days without confirmed susceptibility; the case report identified susceptible agents, which did not include any aminoglycosides. Following seven days of likely inadequate treatment, therapy was escalated to imipenem and P/T dual-therapy, which was narrowed to imipenem. Despite the eventual initiation of antimicrobials with verified susceptibility, patient failed to respond to therapy despite over 30 days of carbapenem therapy. Such response may have been due to delayed removal of peritoneal dialysis catheter as well as delayed initiation of adequate antimicrobial coverage. Finally, patient described in this case report was initiated on P/T and vancomycin; empirically, therapy was altered to cefepime, metronidazole and vancomycin for four days to prevent acute kidney injury. Unfortunately, the ascitic fluid culture resulted cefepime-resistant but P/T sensitive Achromobacter spp. Despite definitive treatment with P/T for five days and meropenem for four additional days, no recovery was observed. Such outcome may have resulted due to interruption of appropriate treatment as well as retention of PEG-tube despite multiple positive ascitic fluid cultures. No trend in antibiotic selection was observed among nine additional patients with clinical recovery who received agents other than carbapenems and P/T. While the International Society for Peritoneal Dialysis (ISPD) guideline suggests structurally similar Pseudomonas associated peritonitis treatment with dual antibiotic regimen, the guideline makes no specific recommendations for Achromobacter treatment [43]. Per in vitro and in vivo data collected in this article and through numerous studies, P/T and carbapenems yield the most preferable clinical outcomes in Achromobacter peritonitis [2, 24, 25, 34].

PD catheter was removed in 9 out of 14 non-SBP patients with clinical cure. All patients with resolved infection received at least seven days of antimicrobial treatment. Five patients received seven to ten days; four patients received fourteen days; two patients received twenty-one days, and one patient received over 30 days of antibiotics. Glucose containing peritoneal dialysate and aqueous environment foster Achromobacter proliferation and associated infection. Moreover, Achromobacter colonies may have the ability to surround the PD catheter with biofilm [14, 52, 61, 62]. Bacteria in biofilms in vivo are significantly less susceptible to antimicrobials identified to have adequate coverage through laboratory testing in vitro [43]. This fact may further clarify the treatment failures, relapses, and patient deaths resulting in patients who received P/T or carbapenems in these case reports. In addition to appropriate antimicrobial therapy, case seventeen also highlights the importance of timely removal of peritoneal dialysis catheter. Donderski et al. [52] report multiple peritonitis relapse associated readmissions; fortunately, no further episodes were reported following PD catheter removal and a course of adequate antimicrobial regimen. Thus, multiple case studies highlight the importance of timely removal of Achromobacter infected PD catheters; Donderski et al. [52] identify catheter removal as the most preferred treatment method. ISPD Peritonitis Recommendations: 2016 Update on Prevention and Treatment identifies peritonitis as a common and serious complication of PD [43]. Severe or prolonged peritonitis can induce functional and structural deformations resulting in membrane failure. Gram-negative peritonitis presents higher risks of catheter loss and death compared to Gram-positive infections [6367]. ISPD guideline recommends removal of PD catheter in cases of recurrent, relapsed, and refractory (lack of clearance of ascitic fluid despite 5 days of appropriate antimicrobial treatment) peritonitis in order to protect the peritoneal membrane [43]. Timely catheter removal is essential in recurrent episodes. Nonetheless, PD catheter or PEG-tube removal within ten days to four weeks of insertion may not allow respective insertion tracts to mature causing further leakage and complications [44, 6870].

Additionally, ISPD suggests at least 3 weeks of treatment in nonpseudomonas Gram-negative, and polymicrobial PD peritonitis [43]. No current guidelines provide significant direction on treatment of PEG tube peritonitis regardless of microbial etiology. Nevertheless, clinical practice guidelines for antimicrobial prophylaxis in surgery strongly recommends prophylactic antibiotic use to significantly reduce the risk of peristomal wound infection associated with PEG-tube insertion [71]. ASPEN guidelines recommend gastropexy using temporary sutures or T-fasteners to secure the stomach to the abdominal wall [72]. These techniques allow for appropriate attachment of tube, minimizing potential for peritoneal leakage, and reducing difficulty and challenges associated with tube replacement, if indicated [72]. Upon development of peritonitis, tube feeds should be held with initiation of broad-spectrum antibiotics. Evidence of perforation should be assessed via abdominal imaging; potential for wound should be evaluated. Surgical consult to perform laparotomy should be obtained. As misplaced tube may cause peritonitis and death, test must be performed to detect misplacement of tube at exchange. In instances of early inadvertent gastrostomy tube removal, parenteral broad-spectrum antibiotics and monitoring for peritonitis signs are indicated [72]. Although recommendations for total duration of antibiotics for treatment of PEG-tube peritonitis could not be encountered in current literature, in case of suspected peritonitis, clinical knowledge extrapolated from similar gastrointestinal perforations suggests that antibiotics may be discontinued after seven days if peritonitis is not observed [70]. Longer duration of treatment may be required in cases of confirmed PEG-tube peritonitis, particularly in cases without adequate source control [70]. Due to the lack of data to guide treatment of PEG tube-associated Achromobacter peritonitis, we attempted to draw similarities between Achromobacter peritonitis due to PEG-tube and PD. Nonetheless, the authors acknowledge that one of the major differences between the two insertions is the accessibility of the peritoneal cavity; while PD catheter readily allows for diagnosis as well as intraperitoneal treatment of peritonitis, PEG tubes should ideally not provide direct access to the peritoneal cavity.

Multiple risk factors for PEG-tube related complications could be identified in the patient presented in our case study. Undiagnosed advanced liver disease, severe ascites, hypoalbuminemia, prolonged hospitalization, and immunocompromise secondary to liver cirrhosis, all place the patient is at high risk for PEG-tube related complications. Polymicrobial nature of the patient’s peritonitis also portends adverse outcomes; likewise, patient’s ultimate demise indicates the severity of the disease.

4. Conclusion

Achromobacter spp. peritonitis as a complication of PEG-tube placement has not been previously reported. In our patients’ case, the recently placed PEG-tube with ascitic fluid leakage was identified as the most plausible infection source. Although a rare bacterial peritonitis pathogen, Achromobacter spp. may be associated with wide antimicrobial resistance and unfavorable outcomes. Thus, clinicians should consider broad empirical antimicrobial regimen, with Achromobacter coverage, such as P/T or a carbapenem, if any concern for Achromobacter spp. secondary peritonitis associated with PEG placement exists, as they may not always represent clinical sample contamination. While the management of secondary peritonitis involving biliary stent and PEG-tube would generally include surgical intervention for infection source control with broad spectrum antibiotics targeting nosocomial coverage per Infectious Disease Society of America, some of antimicrobial selections (such as cefepime and metronidazole combination) may yet be inadequate for widely resistant Achromobacter species. Recognizably, the common antibiotics for spontaneous bacterial peritonitis, i.e., third generation cephalosporins and fluoroquinolones, to which Achromobacter is resistant and variably susceptible, respectively, would be extensively insufficient. Moreover, repeat ascitic fluid culture may be beneficial in verifying continued susceptibility of selected definitive antibiotics. In addition, as these pathogens are generally identified in polymicrobial infections, their presence may serve as a surrogate marker of unfavorable outcomes. Due to the high disease severity, delay or interruption in efficacious therapy may lead to increased mortality rates.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Nishant Tripathi and Niki Koirala developed the concept and design of the literature review; Hirotaka Kato and Nishant Tripathi developed the concept of case report; Nishant Tripathi, Niki Koirala, Hirotaka Kato, Tushi Singh, Kishore Karri, and Kshitij Thakur performed acquisition of clinical information regarding the patient case. Nishant Tripathi and Niki Koirala each independently conducted PubMed searches to gather relevant articles for the literature review. Nishant Tripathi and Niki Koirala acquired relevant data from the gathered articles, analyzed, and interpreted the data. Nishant Tripathi, Niki Koirala, and Hirotaka Kato drafted the manuscript. All authors revised the manuscript critically, approved the final draft of the manuscript to be published, and agree to be accountable for all aspects of the work.