Abstract
Abstract
Background:
Prosthetic joint infection from Listeria monocytogenes is rare. We decided to shed light on this illness and review the reported cases to better understand its characteristics.
Patients and Methods:
We conducted a comprehensive review of the English literature using PubMed. We also included one case that we had managed.
Results:
We found 25 cases of prosthetic joint infection from L. monocytogenes reported individually and a retrospective study of 43 cases of joint and bone listerial infection, including 34 with prosthetic joint infection, conducted in France. We have described their clinical and para-clinical features and tried to elaborate on the pathophysiology, treatment, and prevention.
Conclusion:
Prosthetic joint infection from L. monocytogenes is mainly late. Systemic inflammation may be absent. Although rare, it must be suspected in patients at high risk for both prosthetic joint and listerial infections. In addition, those patients must be instructed on appropriate preventive measures.
L
Listeria monocytogenes is widely spread in the nature. It is found in soil, sewage, surface water, vegetation, slaughterhouse waste, insects, amphibians, fish, poultry, animal feed, dairy products, and even dust. It is also normal flora of many mammals. In fact, 1%–5% of asymptomatic adult human beings are carriers. Although 13 serovars of L. monocytogenes have been identified, serovars 1/2a, 1/2b, 1/2c, and 4b account for 98% of infections with Listeria cases in human beings [1].
Infection is the most serious complication after prosthetic joint replacement. The infection rate is less than 2% in knee prostheses and less than 1% in hip prostheses in the first 2 years [2]. The infection rate after revision arthroplasty is significantly higher. The most frequently isolated bacteria from infected prosthetic joints are coagulase-negative staphyloocci, Staphylococcus aureus, streptococci, enterococci, gram negative bacilli, and anaerobes. In approximately 10% of cases, the infection is polymicrobial [3]. Because of the extreme rarity of the illness, we decided to conduct a comprehensive review of cases of prosthetic joint infection (PJI) from L. monocytogenes reported in the English literature.
Patients and Methods
We searched the English literature via PubMed and found 25 cases of PJI from L. monocytogenes that are, to the best of our knowledge, all the cases reported in the English literature. In addition, we found one retrospective study conducted in France that analyzed cases reported to the French National Reference Center for Listeria between 1992 and 2010. We are also including a case of PJI from L. monocytogenes that we have encountered recently.
Results
All the characteristics of the 25 cases including the age, gender, affected joint, laboratory data, co-morbidities, Gram stain and cultures results, interval from the index operation to time of infection, prosthesis management, antibiotic therapy and outcome are depicted in Tables 1 and 2. We have also included a description of the case in which we managed the infection.
ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; WBC = white blood cell; F = female; NA = not available; M = male; RA = rheumatoid arthritis; DM = diabetes mellitus; CC = colon cancer; GI = gastrointestinal; nl = normal; AB = antibiotic; OC = our case. WBC × 1,000/mm3.
PJI = prosthetic joint infection; ex = exchange; TMP-SMX = trimethoprim-sulfamethoxazole; IV = intravenous; PO = per ora (by mouth).
Note that the prosthesis was retained in 15 patients and replaced in 8. Five patients underwent two-stage exchange and three patients one-stage exchange. In addition, the prosthesis was removed in three patients, but the subsequent management was not clear. Most antibiotic agents used were ampicillin, trimethoprim-sulfamethoxazole, and gentamicin.
Cefamandole and cefotaxime were used in one patient with poor response. In fact, despite intrinsic resistance of Listeria to cephalosporins, they were inappropriately used in that patient based on an incorrect report of sensitivity. Ciprofloxacin was the choice for another patient because of an allergy to penicillin and sulfamethoxazole interaction with the patient's medications with subsequent treatment failure. A combination of teicoplanin and rifampicin was used as well in one patient based on the sensitivity report with an unclear result.
Finally, levofloxacin and co-trimoxazole combination were chosen for one patient based on the sensitivity report and ampicillin allergy. Recurrent infection was subsequently managed with a combination of linezolid, co-trimoxazole, and rifampicin with good response. The duration of the treatment was variable. Overall, the prognosis was good.
The study conducted in France analyzed all cases reported to the French National Reference Center for Listeria between 1992 and 2010 [4]. Forty-three patients had joint and bone listerial infection, including 34 with PJI; 88% of the patients were older than 60 years. Male gender was predominant, with a male to female ratio of 0.63. All four serotypes were reported, with serotypes 4b and 2a being the most predominant. The hip was the most commonly involved joint (60%) followed by the knee (21%).
Most patients had leukocytosis with a mean of 10,134/mm3 and elevated C-reactive protein (CRP) level. Underlying conditions included rheumatoid arthritis, neoplasms, diabetes mellitus, chronic kidney disease, kidney transplantation, human immunodeficiency virus infection, alcoholism, and immunosuppressive treatment. Gram stain of the sample was positive in just 21%, but culture results were positive in all cases. Bacteremia was present in only 16% of patients. Approximately 40% of the patients had fever.
The median time of infection after prosthesis insertion was 9 years. In 34% of the patients, the prosthesis was retained; 41% of patients underwent one-stage exchange and 7% two-stage exchange. In 17% of patients, the prosthesis was removed without replacement. The most commonly used antibiotics were amoxicillin, aminoglycosides, and co-trimoxazole. Other antibiotics used included rifampicin, doxycycline, pristinamycin, and vancomycin. No infection-related death was reported. The findings in the French study were fairly similar to those we obtained in analyzing the 26 individual cases.
The patient we treated was a 57-year-old male with diabetes mellitus who had presented to the emergency department in September 2015 for right knee pain. He had a history of right total knee replacement performed in 2010 with subsequent revision in 2014. He had started to have fever, nausea, and vomiting a few days before presentation. He had eaten soft cheese at a Mexican restaurant on the day his symptoms had started. On presentation, his temperature was 103°F. His knee was swollen and tender to palpation.
The erythrocyte sedimentation rate was 45, CRP 523, and white blood cell count 11,100/mm3. Arthrocentesis drained 40 mL of purulent fluid. Gram stain results showed gram positive bacilli, and fluid culture grew L. monocytogenes. Blood culture results remained negative. Antibiotic therapy based on ampicillin and gentamicin was started. The patient then underwent debridement with implant removal and placement of a spacer loaded with gentamicin. He was discharged to a rehabilitation facility with the plan to continue ampicillin for a total course of 6 weeks with subsequent prosthetic knee implantation. The case was reported to the public health department.
The patient finished antibiotic therapy with no complications. Two months after the initial operation, the spacer was removed, and definitive implants were successfully fixed in place with excellent range of motion. Tissue specimens taken intra-operatively before implantation showed less than five polymorphonuclear cells per high power field. Cultures performed at re-implantation also had negative results. He progressively regained normal activity with physical therapy. There was no evidence of infection on subsequent regular clinic visits. He was last seen 5 months after reimplantation. He was doing well and was considering going back to work.
Discussion
Listeria monocytogenes is primarily a food pathogen. It is resistant to heat and may survive sub-optimal pasteurization. In addition, it may multiply at low temperatures and reach high concentrations even in refrigerated food [5]. The main portal of entry is the gastrointestinal tract. After ingestion of contaminated food, listeriosis manifests as acute gastroenteritis. A person can also remain totally asymptomatic [6]. On average, the incubation period is 24 hours, and the symptoms last 48 hours.
Listeriosis, however, may be invasive, mainly in immunocompromised patients; the bacteria will invade the mesenteric lymph nodes and reach the blood stream [7]. Because L. monocytogenes organisms are facultative intracellular bacteria, resistance to the infection relies mainly on cell-mediated immunity. Any condition altering the cell-mediated immunity will predispose a person to listeriosis. The main reported underlying conditions associated with listeriosis were immunosuppression, malignant disease, old age, pregnancy, diabetes mellitus, cirrhosis, chronic kidney disease, collagen-vascular diseases, mainly rheumatoid arthritis, and alcoholism.
Listeriosis is observed most commonly as meningitis or primary bacteremia, less often as localized infection such as arthritis, osteomyelitis, endocarditis, pericarditis, conjunctivitis, endophthalmitis, gastroenteritis, cholecystitis, liver or spleen abscess, peritonitis, pneumonitis, urethritis, cervical adenitis, non-meningitic central nervous system infection, skin infection, and mononucleosis-like syndrome. In pregnant females, it can cause premature delivery, peri-natal sepsis, and even fetal death.
In 2013, the incidence of listerial infections was 2.6 cases per million in the United States [8]. The mortality rate in patients infected with Listeria is estimated at 15.9% [9]. Listeriosis was added to the list of nationally notifiable diseases in 2001. In 2014, 675 cases of listeriosis were reported to the department of public health. Note that the incidence of listeriosis in Europe is 4.7 cases per million [10], which is almost twice that in the United States.
On the basis of the onset of symptoms after implantation, PJI can be classified as early (occurring in fewer than 3 mo), delayed (occurring in 3–24 mo), and late (occurring more than 24 mo) after arthroplasty [3,11]. On the basis of our review, most cases of PJI from L. monocytogenes are late. Early and delayed infections are mainly acquired peri-operatively. Late infections, however, are predominantly caused by hematogenous spread.
Patients with early infections present with arthralgia, edema, erythema, and fever. Delayed infections manifest as persisting or worsening pain as well as early loosening. Signs of infection may be absent on physical examination. Late infections most commonly occur after missed bacteremia. They also may be associated with systemic symptoms [3].
The pathogenesis of PJI differs from that of a native joint. The infection relies mainly on biofilm formation. Listeria monocytogenes has a high affinity for foreign bodies. Surprisingly, we found two cases of listerial arthritis in which the prosthetic joint was not affected despite involvement of a native joint [12,13]. The infection was even poly-articular in one of them. Listeria group IIa isolates were found to have a higher capability of biofilm formation. Interestingly, that was not correlated with worse prognosis [4].
The white blood cell count is not helpful in predicting the presence or absence of PJI [14]. The CRP level is useful if the value is compared with a previous measurement, because it may take 7–10 days to normalize after the surgical procedure. Synovial fluid leukocyte count and differential seem to be more reliable in predicting infection. Microbiologic phenotypic features are the standard tools to identify the causative pathogens. Listeria monocytogenes can be isolated easily on routine media. In case of contaminated specimens, selective enrichment media stimulate growth.
Usually, cultures grow in 1–2 days. Importantly, a negative culture result does not rule out Listeria infection if the latter is strongly suspected clinically [15]. That is particularly true in case of ongoing antibiotic treatment or aberrant phenotype. Molecular methods based on sequencing 16S ribosomal ribonucleic acid seem to solve this problem. A case of Listeria PJI with a negative culture result but positive molecular study was reported [16]. The patient was receiving antibiotic treatment when the specimen was collected. A sophisticated method called Matrix-assisted Laser Desorption Ionization Time-of-Flight Mass Spectrometry can also be used to identify L. monocytogenes [17]. Unfortunately, those techniques are not available readily.
Management of PJI was nicely depicted in an algorithm by Trampuz and Zimmerli [3]. The treatment depends on duration of symptoms, condition of the implant and surrounding soft tissue, pathogen involved, and the patient's performance. Listeria monocytogenes was reported to have in vitro sensitivity to many antibiotics. The drugs of choice for listerial infections, however, are ampicillin or penicillin G. Because of delayed bactericidal activity of those antibiotics [18], gentamicin is added for synergy [19,20], mainly in neonates or immunocompromised patients as well as in cases of central nervous system infections and endocarditis. Patients who are allergic to penicillin can be treated with trimethoprim-sulfamethoxazole. Other antibiotics such as imipenem, meropenem, linezolid, and vancomycin have been also used with variable outcomes.
In addition to general measures recommended to avoid foodborne infections, specific recommendations apply for persons with higher risk for listeriosis, mainly immunosuppressed and old persons as well as pregnant females. For those categories, the Centers for Disease Control and Prevention recommends the avoidance of soft cheeses and seafood consumption as well as hot dogs, luncheon meats, and other deli meats unless specific conditions are met [21]. Soft cheeses made with pasteurized milk are safe to consume with the exception of Mexican-style cheeses that have caused listerial infections probably because of contamination during cheese making.
There is no sufficient evidence to suggest prophylactic antibiotic therapy for patients with prosthetic joints before invasive procedures. Prophylaxis is controversial in immunocompromised patients undergoing procedures with secondary high risk of bacteremia. Although the mortality rate in listerial infections was estimated close to 16%, the prognosis of listerial PJI was surprisingly good.
We were surprised to find that 34 cases were reported in France over 18 years, although we had just found 25 cases reported in the English literature. The obvious factors that may explain that difference are the higher incidence of listeriosis in Europe and the mandatory reporting of cases since 1999. Note again that listeriosis became a nationally notifiable disease in the United States in 2001. Another plausible explanation may be the increasing prevalence of risk factors for PJI and listeriosis in France. Further studies need to be conducted to confirm or deny that possibility. That is beyond the scope of our current work.
Conclusion
Listerial PJI is rare. It occurs mainly late, and systemic inflammation may be absent. We believe that PJI incidence will keep increasing because of the growing number of implanted prostheses and the aging population, which implies prolonged prosthesis lifetime and subsequently increased risk of infection. That will certainly result in a huge financial burden because of the high cost of implants replacement and long-term antibiotic therapy.
Physicians must instruct patients at risk for PJI and listeriosis on appropriate preventive measures. As mentioned, the patients at highest risk for listeriosis are adults 65 years and older (58% of Listeria infections), pregnant females (14% of Listeria infections), and persons with weakened immunity [22]. These high-risk patients will definitely benefit from counseling about appropriate prevention. In addition, listerial PJI must be suspected in high-risk patients, even if culture results are negative, if the clinical suspicion is high. Molecular tests increase the diagnostic power but unfortunately are not widely available.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
