Clin Microbiol Infect 2012; 18: E431–E434
We report a case of viral peritonitis caused by coxsackievirus B1 in a 79-year-old male undergoing continuous ambulatory peritoneal dialysis (CAPD), and review the English language literature. Clinicians should be aware of viral peritonitis in patients on CAPD presenting with a viral syndrome and mononuclear peritoneal dialysis effluent. Currently, viral diagnostic tests are available to confirm the diagnosis and avoid unnecessary treatment with antibiotics.
Peritonitis is one of the major complications of continuous ambulatory peritoneal dialysis (CAPD). Bacteria cause at least 70–80% of CAPD-associated peritonitis episodes, and bacterial cultures remain negative in 20% .
Only a few cases of viral CAPD-associated peritonitis have been reported [2–8]. It is unclear whether viral CAPD-associated peritonitis is truly infrequent or remains largely unrecognized because laboratory tests for the detection of viruses in peritoneal dialysis (PD) effluent or viral antibodies are not often performed.
We hereby report on an episode of CAPD-associated peritonitis during the course of a coxsackievirus B1 infection in a 79-year-old male patient. Previously reported cases of viral CAPD-associated peritonitis were searched for through MEDLINE, and are briefly summarized.
A 79-year-old man was admitted with respiratory-induced pain in the left hemithorax (pleurodynia), fever, and malaise (day 0). Clinical examination revealed a temperature of 38°C and a pre-cordial pleural rub. The patient was treated with CAPD for 4 months because of end-stage ischaemic vascular kidney disease. Further medical history was insignificant, apart from a multinodular goitre and an inguinal herniorraphy.
The peripheral white blood cell count was 4000/mm3, with 87.8% neutrophils and a C-reactive protein (CRP) level of 38.8 mg/L. The PD effluent contained 154 leukocytes/mm3, with 98.5% monocytes. Owing to this unusual finding, and because the patient mentioned that his toddler granddaughter had symptoms of a viral infection, i.e. fever, myalgia, and vesicular lesions of the tongue, the clinician also requested an analysis of the PD effluent for viral pathogens. For viral culture, samples were inoculated on RD and MRC5 cell lines. Cultured monolayers showing a typical cytopathic effect for enteroviruses were examined for the presence of enteroviruses by means of indirect immunofluorescence (Monoclonal Mouse anti-Enterovirus Antibody, clone 5-D8/1; Dako Diagnostics, Glostrup, Denmark). A chest X-ray was normal, but echocardiography revealed a mild pericardial effusion. A presumed diagnosis of pericarditis was made, and acetylsalicylic acid 500 mg three times daily was started orally. Empirical cefazolin and amikacin were started for presumed CAPD peritonitis. The antibiotic therapy was discontinued on day 3, when the PD effluent culture showed no bacterial growth. On day 4, the patient had recovered and was discharged. The CRP level had decreased to 29.8 mg/L. Bacterial, mycobacterial and viral cultures of PD effluent remained negative.
On day 8, the patient was re-admitted with renewed complaints of malaise and fever. His temperature was 38.6°C. Abdominal examination findings were unremarkable. The PD effluent was cloudy, and contained 4590 leukocytes/mm3, with 91.5% monocytes. The peripheral white blood cell count was 3400/mm3, with 80% neutrophils, and the CRP level was 66.8 mg/L. Antibiotics (cefazolin and amikacin) were restarted. On day 9, the laboratory reported the isolation of an enterovirus from the PD effluent sample of day 0. The antibiotic therapy was stopped. Another PD effluent sample was obtained for viral culture and RT-PCR for enteroviruses. The in-house PCR technique was as follows: after RNA extraction with the QIAMP viral RNA MINI Kit (Qiagen, Venlo, The Netherlands) and a reverse transcriptase step with Superscript III reverse transcriptase (Invitrogen, Merelbeke, Belgium), real-time PCR was performed on a Rotor-Gene 3000 (Qiagen), amplifying a highly conserved region of the 5′-untranslated region of the enterovirus genome . The presence of enterovirus was confirmed by PCR in this second PD sample, but at this time the culture remained negative. Two days later (day 10), a blood sample was also positive for enterovirus by PCR. A positive sample was sent to the reference centre (Laboratory of Clinical Virology, Rega Institute, Catholic University Leuven) for confirmation and typing of the virus strain by means of VP1 sequencing. The viral strain was identified as coxsackievirus B1 by the reference laboratory. Serological testing for enterovirus antibodies (complement fixation test; Serion Immundiagnostica, Würzburg, Germany) showed a two-fold rise in titre from day 1 to day 10. The anti-inflammatory therapy was continued. The CRP level and number of PD effluent leukocytes gradually decreased. On day 13, the patient was discharged without complications.
The majority of peritonitis episodes in PD can be ascribed to Gram-positive bacteria of the skin and, to a minor degree, to Gram-negative bacteria presumably originating from the enteric flora. About 20% of cultures remain negative. These culture-negative cases can be ascribed to inadequate culture techniques, prior antimicrobial therapy, fastidious organisms, or chemical substances . However, as diagnostic tests for viral pathogens are not routinely performed on PD effluent, the role of viruses in CAPD peritonitis might be underestimated. On the other hand, Pomeranz et al.  found a non-interferon anti-viral factor in PD effluents that may explain, in part, the rarity of clinical viral peritonitis. HIV, hepatitis C virus or cytomegalovirus DNA has been detected in PD effluent of chronically infected CAPD patients with and without peritonitis [11–13]. DNA/RNA contamination from plasma should be considered in cases of inflammation, particularly for viruses with a high viral load, such as hepatitis C virus.
Our literature search revealed seven cases of viral peritonitis in patients undergoing CAPD [2–8]. The characteristics of the cases, including our case, are given in Table 1. Some common features can be seen. The clinical course in our patient (fever, malaise, and clearance without antimicrobial treatment) and a number of laboratory features (repeated negative bacterial and mycobacterial cultures, slight leukopenia, PD effluent monocytosis (or absence of polymorphonuclear leukocytes, the hallmark of bacterial peritonitis), and a rise in enterovirus antibody titre) pointed to a viral infection. Coxsackievirus infection was demonstrated by PCR and viral culture of PD effluent. As in the case reported by Struijk et al. , the clinical picture and the history of viral infection of the granddaughter triggered the physician to test the PD effluent for viruses. Transmission of enteroviruses is predominantly faeco-oral. Longitudinal studies have demonstrated that secondary infections occur in >50% of susceptible household contacts . During the course of a systemic viral illness, enterovirus spreads by the haematogeneous route throughout the body. This often results in a biphasic illness , as demonstrated in our patient and the case with coxsackievirus pancreatitis . Close contact and faeco-oral transmission were very plausible, but, as no samples from the child were available, this hypothesis could not be proven.
|Case, year (reference)||Patient age (years)/sex||Symptoms at presentation||Trigger for viral testing||Diagnostic laboratory methods||Viral species/result|
|Struijk, 1986 ||42/female||Abdominal pain |
Cloudy PD effluent
|Son had fever and headache 1 week before |
Atypical bodies in leukocytes of PD effluent
|Viral culture (effluent and faeces) |
Complement fixation test (serum)
|Echovirus type 11 |
Enterovirus four-fold antibody titre rise at −1 week and +2 weeks
|Lal, 1988 ||57/male||Cough, dyspnoea, blood-streaked sputum |
Left-sided pleuritic pain, pleural and pericardial friction rub
Five days later, abdominal pain, acute pancreatitis
Cloudy PD effluent
|Acute pancreatitis||Complement fixation test (serum)||Coxsackievirus B1 and B6 four-fold and eight-fold antibody titre rise at +4 days and +35 days|
|Abraham, 1989 ||50/female||Mononucleosis-like illness. Abdominal pain |
Cloudy PD effluent
>100 leukocytes/mm3, lymphocytosis
|Mononucleosis-like illness; lymphocytosis in effluent (predominantly atypical)||CMV IgM and IgG negative; herpes simplex virus type 2-positive titre (serum)||Viral aetiology was presumed|
|Lewis, 1991 ||45/female||Hepatitis |
Cloudy PD effluents
20 leukocytes/mm3, 75% lymphocytes, many atypical
|Atypical lymphocytes in blood and PD effluent after blood transfusion||Viral culture (effluent) |
Indirect fluorescence antibodies (serum)
CMV IgG antibody titre rise from <1 : 10 to 1 : 128
|Braun, 1991 ||6/male||Abdominal pain, fever, chills, generalized vesicular rash |
Cloudy PD effluent
120 leukocytes/mm3, 85% monocytes
|Typical varicella rash||Varicella IgG (serum)||Herpes zoster virus antibody titre rise from undetectable (−2 weeks) to 1 : 1000 (+4 weeks)|
|Yakulis, 1999 ||60/female||Abdominal pain, diarrhoea, nausea, chills |
Cloudy PD effluent
100 leukocytes/mm3, 64% lymphocytes, 32% monocytes
|Probable viral gastroenteritis with prolonged course and severity of symptoms||Histology, immunohistochemistry, electron microscopy (omental biopsy)||HSV (types I and II) |
Intranuclear viral particles
|Liesker, 2006 ||30/female||Abdominal pain |
Cloudy PD effluent
7500 leukocytes/mm3, 79% monocytes
|Small herpetiform genital ulcers||Viral culture (effluent and ulcer) |
Enzyme immunoassay and complement fixation (serum)
|HSV 2 |
No antibodies against HSV-2 6 months before
|Pauwels, 2011 (present case)||79/male||Pleurodynia, pericardial effusion, fever, malaise |
Cloudy PD effluent
154 leukocytes/mm3, 98.5% monocytes
|Granddaughter had symptoms of hand, mouth and foot disease||Viral culture (effluent) and PCR (effluent) |
Complement fixation (serum)
|Coxsackievirus B1 |
Enterovirus two-fold antibody titre rise (day 0 to day 10)
This case demonstrates that, when a viral aetiology of CAPD-associated peritonitis is suspected, viral culture and, preferably, PCR of the PD effluent leads to the diagnosis. PCR is a more rapid method, and has shown superior sensitivity . Culture and PCR of stool and respiratory specimens are considered to be more sensitive. As the initial suspicion of viral peritonitis was rather low, these specimens were only tested later (after the active infection period (day 13)). Cultures remained negative, and PCR on the respiratory sample showed inhibition. Serological testing for enterovirus antibodies should be performed on baseline and convalescent samples, and in order for an acute infection to be documented, a significant rise in antibody titre (four-fold) followed by a decrease in the convalescent phase should be demonstrated . Unfortunately, we were not able to demonstrate this in our case, possibly because of the short time period between the paired samples (10 days). PD effluent leukocytosis, predominantly lymphocytic or monocytic, was also reported in five of seven other published cases of viral CAPD peritonitis [2–8].
In conclusion, although a viral aetiology in CAPD-related infections is rare , viral peritonitis can occur in CAPD patients during the course of an acute systemic viral illness. Currently, virology tests are available to confirm the diagnosis and avoid unnecessary days of antibiotic therapy. Intravenous immunoglobulin therapy has been used for serious enterovirus infections, but the evidence is limited. There could be a beneficial effect for neonates and children with coxsackievirus myocarditis .
All authors declare no potential conflicts of interest.