Utility of C-reactive protein in assessing the disease severity and complications of community-acquired pneumonia.

Previous studies on the usefulness of C-reactive protein (CRP) in patients with community-acquired pneumonia (CAP) have yielded somewhat inconsistent results. Our aim was to assess the value of CRP in estimating the severity and complications of CAP. CRP levels during the first 5 days of hospitalization were measured in 384 adult patients with CAP, and the data were evaluated using comprehensive statistical analyses. Significantly higher CRP levels on admission were detected in Pneumonia Severity Index (PSI) classes III-V than in classes I and II (p <0.001). An increment of 50 mg/L CRP on admission was associated with a 1.22-fold odds for a patient to be in PSI classes III-V as compared with classes I and II (OR 1.22, 95% CI 1.11-1.34; p <0.001). CRP levels were significantly higher in bacteraemic pneumonia than in non-bacteraemic pneumonia (p <0.001). An increment of 50 mg/L CRP was associated with a 1.67-fold odds for a patient to be bacteraemic (OR 1.67, 95% CI 1.46-1.92; p <0.001). CRP levels >100 mg/L on day 4 after the admission were significantly associated with complications (p <0.01). There was a trend for an association between the level of CRP on admission and the time to reach clinical stability (p <0.01). In conclusion, CRP may be valuable for revealing the development of complications in CAP. It may also be useful to assess the disease severity, thus being complementary to the assessment of the PSI. In our patients, high CRP levels were associated with a failure to reach clinical stability.

Duration of antibiotic treatment in disseminated Lyme borreliosis: a double-blind, randomized, placebo-controlled, multicenter clinical study.

Despite rather strict recommendations for antibiotic treatment of disseminated Lyme borreliosis (LB), evidence-based studies on the duration of antibiotic treatment are scarce. The aim of this multicenter study was to determine whether initial treatment with intravenous ceftriaxone (CRO) for 3 weeks should be extended with a period of adjunct oral antibiotic therapy. A total of 152 consecutive patients with LB were randomized in a double-blind fashion to receive either amoxicillin (AMOX) 1 g or placebo (PBO) twice daily for 100 days. Both groups received an initial treatment of intravenous CRO 2 g daily for 3 weeks, followed by the randomized drug or PBO. The outcome was evaluated using the visual analogue scale at the follow-up visits. The final analysis included 145 patients, of whom 73 received AMOX and 72 PBO. Diagnoses of LB were categorized as either definite or possible, on the basis of symptoms, signs, and laboratory results. The diagnosis was definite in 52 of the 73 (71.2%) AMOX-treated patients and in 54 of the 72 (75%) PBO patients. Of the patients with definite diagnoses, 62 had neuroborreliosis, 45 arthritis or other musculoskeletal manifestations, and 4 other manifestations of LB. As judged by the visual analogue scale and patient records, the outcome after a 1-year follow-up period was excellent or good in 114 (78.6%) patients, controversial in 14 (9.7%) patients, and poor in 17 (11.7%) patients. In patients with definite LB, the outcome was excellent or good in 49 (92.5%) AMOX-treated patients and 47 (87.0%) PBO patients and poor in 3 (5.7%) AMOX-treated patients and 6 (11.1%) PBO patients (difference nonsignificant, p = 0.49). Twelve months after the end of intravenous antibiotic therapy, the levels of antibodies against Borrelia burgdorferi were markedly decreased in 50% of the patients with definite LB in both groups. The results indicate that oral adjunct antibiotics are not justified in the treatment of patients with disseminated LB who initially receive intravenous CRO for 3 weeks. The clinical outcome cannot be evaluated at the completion of intravenous antibiotic treatment but rather 6-12 months afterwards. In patients with chronic post-treatment symptoms, persistent positive levels of antibodies do not seem to provide any useful information for further care of the patient.

A retrospective nationwide case study on the use of a new antifungal agent: patients treated with caspofungin during 2001-2004 in Finland.

The aim of this study was to evaluate the efficacy and safety of caspofungin in patients treated in Finland during the period 2001-2004. The medical records of 78 adult patients treated with caspofungin in five major hospitals were reviewed retrospectively. Fifty-nine (76%) patients had proven invasive fungal infection, of whom 22 (28%) had aspergillosis and 37 (47%) had candidiasis. Nineteen (24%) patients were treated empirically; only 13 (17%) patients received caspofungin as primary therapy. A favourable response was achieved in 52 (67%) patients. The response rate was 78% in patients with candidiasis, and 50% in patients with aspergillosis. At the end of the study period, 40 (51%) patients remained alive; of the 38 deaths, nine (24%) were caused by fungal infection. The response rates were lower, although not significantly, for patients with high (>20) vs. low (< or =20) Acute Physiology and Chronic Health Evaluation (APACHE II) scores (response rate 50% vs. 68%, p 0.48, respectively), and were also lower in patients with long-term (>20 days) vs. shorter duration (< or =20 days) neutropenia (55% vs. 73%, p 0.32, respectively), and in those with an underlying haematological malignancy vs. patients with other diseases (59% vs. 73%, p 0.2, respectively). In five (6%) patients, caspofungin therapy was discontinued prematurely because of adverse drug reactions (ADRs) (elevated liver enzyme values in three patients, neuropathic pain in one, and skin rash in one). Serious ADRs occurred in two (3%) patients (severe hepatic insufficiency with consequent death, and eosinophilia with elevated alkaline phosphatase levels), and laboratory abnormalities, mostly mild and reversible, in 24 (31%) patients. In this unselected patient population, caspofungin was safe, well-tolerated, and had an efficacy comparable to that in previous reports from prospective trials.

Prospective open-label study of pharmacokinetics, efficacy and safety of a new 10% liquid intravenous immunoglobulin in patients with hypo- or agammaglobulinemia.

BACKGROUND AND OBJECTIVES: The aim of this study was to evaluate the pharmacokinetics, efficacy and safety of a newly developed 10% liquid immunoglobulin preparation in patients with primary immunodeficiency diseases. This new preparation for intravenous use includes three dedicated virus clearance steps in its manufacturing process to ensure a high margin of viral safety. MATERIALS AND METHODS: This was a prospective, open-label, non-controlled, multicentre study. Twenty-two subjects with primary immunodeficiency were treated initially with three infusions of a licensed intravenous immunoglobulin to standardize the immunoglobulin G (IgG) replacement therapy of all subjects to the same intravenous product. A total of nine infusions of the new 10% liquid preparation were subsequently administered. RESULTS: The median terminal half-life of total IgG following administration of the new preparation was 30.1 days. Median terminal half-lives for IgG subclasses IgG(1), IgG(2), IgG(3) and IgG(4) were 28.3, 31.3, 20.9 and 24.2 days, respectively. The median total serum IgG steady-state trough level was 8.51 g/l. No severe infection episodes started after initiation of treatment with the new preparation. The median rate of mild or moderate infection episodes was 0.48 per month. A total of 194 infusions with the new 10% liquid immunoglobulin preparation were administered. The mean dose per infusion was 0.41 g/kg body weight and the maximum infusion rates recorded were 8 ml/kg/h. Adverse experiences were mostly mild and unrelated to the study drugs. Only 4% of infusions with the new product were followed by one or more related adverse experiences. CONCLUSION: The new 10% liquid immunoglobulin preparation was well tolerated and shown to have an excellent pharmacokinetic, efficacy and safety profile. The liquid formulation provides convenience to patients and healthcare professionals.

Infective endocarditis in a Finnish teaching hospital: a study on 326 episodes treated during 1980-2004.

OBJECTIVES: To evaluate potential changes of infective endocarditis (IE) in patients treated in a Finnish teaching hospital during the past 25 years. PATIENTS: 326 episodes of IE in 303 patients treated during 1980-2004 were evaluated for clinical characteristics and their changes over time. RESULTS: The mean age of the patients increased with time (from 47.2 to 54.5 years, p = 0.003). Twenty-five (7.7%) episodes were associated with intravenous drug use (IVDU), with a significant increase of these episodes after 1996 (from 0 to 19 (20%), p < 0.001). Viridans streptococci were the most common causative agents of IE during 1980-1994, but after that Staphylococcus aureus was the most common pathogen (p = 0.015). The proportion of IE of the aortic valve decreased during the study (from 30 (49%) to 26 (27%), whereas the proportions of mitral (11 (18%) to 33 (35%) and tricuspid valve IE (0 to 13 (14%) increased correspondingly (p = 0.001). This was mainly due to more patients with IVDU. Chronic dialysis for renal failure as an underlying condition increased over time (from 0 to 7 (7.4%), p = 0.015) but no other predisposing conditions changed. Complications such as neurological manifestations and heart failure did not change in frequency, but the incidence of lung emboli increased (from 0% to 10.5%, p < 0.001); 83% of these emboli occurred in patients with IVDU. The proportion of patients requiring surgical treatment and mortality due to IE did not change. CONCLUSIONS: During these 25 years, the causative agents, affected valves and complications of IE changed to some degree. These changes were mainly attributed to the increase of IVDU-associated IE. Except for the increase in age, the clinical presentation and outcome in non-addicts remained substantially unchanged.

Levofloxacin does not decrease mortality in Staphylococcus aureus bacteraemia when added to the standard treatment: a prospective and randomized clinical trial of 381 patients.

OBJECTIVES: To study whether levofloxacin, added to standard treatment, could reduce the high mortality and complication rates in Staphylococcus aureus bacteraemia. DESIGN: A prospective randomized multicentre trial from January 2000 to August 2002. SETTING: Thirteen tertiary care or university hospitals in Finland. SUBJECTS: Three hundred and eighty-one adult patients with S. aureus bacteraemia. Patients with meningitis, and those with fluoroquinolone- or methicillin-resistant S. aureus were excluded. INTERVENTIONS: Standard treatment (mostly semisynthetic penicillin) (n = 190) or that combined with levofloxacin (n = 191). Supplementary rifampicin was recommended if deep infection was suspected. MAIN OUTCOME MEASURES: Primary end-points were mortality at 28 days and at 3 months. Clinical and laboratory parameters were analysed as secondary end-points. RESULTS: Adding levofloxacin to the standard treatment offered no survival benefit. Case fatality rates were 14% in both groups at 28 days, and 21% in the standard treatment and 18% in the levofloxacin group at 3 months. Levofloxacin combination did not differ from the standard treatment in the number of complications, time to defervescence, decrease in serum C-reactive protein concentration or length of antibiotic treatment. Deep infection was found in 84% of patients within 1 week following randomization with no difference between the treatment groups. At 3 months, the case fatality rate for patients with deep infection was 17% amongst those who received rifampicin versus 38% for those without rifampicin (P < 0.001, odds ratio = 3.06, 95% confidence intervals = 1.69-5.54). CONCLUSIONS: Levofloxacin combined with standard treatment in S. aureus bacteraemia did not decrease mortality or the incidence of deep infections, nor did it speed up recovery. Interestingly, deep infections in S. aureus bacteraemia appeared to be more common than previously reported.

Bronchoalveolar lavage in immunocompromised patients with haematological malignancy--value of new microbiological methods.

PURPOSE: To assess the usefulness of new culture-independent microbiological methods to analyse bronchoalveolar lavage (BAL) fluid from haematological patients with clinical pneumonia. PATIENTS AND METHODS: Results of 135 BALs from 122 disease episodes in 99 patients treated between 1996 and 2002 were retrospectively analysed. Forty-three patients had undergone haematopoietic stem cell transplantation and 56 patients had been treated with conventional chemotherapy for haematological malignancy. In addition to conventional microbiological methods, polymerase chain reaction (PCR) tests for Pneumocystis carinii, cytomegalovirus (CMV), Legionella sp., mycobacterium, Mycoplasma pneumoniae, and Chlamydia pneumoniae and the Aspergillus antigen test were performed. RESULTS: Three (2.2%) quantitative and four (3.0%) special bacterial cultures gave an aetiological diagnosis. A respiratory virus was isolated in 10 episodes (8.2%). The diagnostic yield increased to 35.6% (48 of 135) by other methods. The P. carinii PCR test was positive in 21 of 24 patients with P. carinii pneumonia, being the only microbiological indication of P. carinii in four cases. The CMV PCR test was positive in 18 patients, but in 14 patients the clinical significance of the finding remained unproven. The Aspergillus antigen test was positive in seven of nine patients with aspergillosis, being the only microbiological indication of Aspergillus in three cases. The result of BAL indicated commencement of specific antimicrobial treatment in 27 episodes (22.1%). CONCLUSION: The contribution of new culture-independent methods to the total diagnostic yield was of note. Among these methods, the P. carinii PCR and Aspergillus antigen tests proved the most valuable, while the CMV PCR test was not clinically useful.

Measurement of the D-arabinitol/L-arabinitol ratio in urine of neutropenic patients treated empirically with amphotericin B.

The aim of the present study was to evaluate the diagnostic significance of the D-arabinitol/L-arabinitol ratio in urine of neutropenic patients with suspected fungal infection. D-arabinitol/L-arabinitol ratios were determined in 373 serial urine samples of 104 patients with haematological malignancies receiving empirical amphotericin B treatment for suspected invasive fungal infection. Twenty-eight (8%) urine samples obtained from 17 (16%) patients were positive (ratio > or =4). Eight (47%) patients had positive urine samples at the initiation of empirical amphotericin B treatment and the rest from 7 to 30 days after empirical therapy was started. Several urine samples were positive in six patients. Only one of the five patients with candidemia had elevated D-arabinitol/L-arabinitol ratios (persistent Candida krusei fungaemia). Four patients with transient candidemia and seven patients with invasive mould infections were negative. Patients who died during the study period had significantly higher D-arabinitol/L-arabinitol ratios than patients who survived (P=0.0002). Pneumonia was the most common manifestation of infection (53% of patients with elevated D-arabinitol/L-arabinitol ratios) and was associated with an especially high mortality (67%). The present study shows that elevated urine D-arabinitol/L-arabinitol ratios are common in febrile, neutropenic patients. However, the urine arabinitol test did not detect transient candidemia at elevated levels during the course of infection. Furthermore, D-arabinitol/L-arabinitol ratios were often elevated in the late phase of infection only. This contests the use of this test in guiding the initiation of antifungal therapy. The detection of elevated arabinitol levels in neutropenic patients during empirical amphotericin B treatment is associated with poor prognosis.

Diagnostic findings in 95 Finnish patients with common variable immunodeficiency.

Common variable immunodeficiency is the most frequent of the primary hypogammaglobulinemias. It is manifested by a wide variety of clinical signs and symptoms. In this retrospective, nationwide survey data were collected on all patients with common variable immunodeficiency who were receiving immunoglobulin replacement therapy in Finland to study the prediagnostic clinical picture, diagnostic delay, and diagnostic findings. Ninety-five patients were identified. The median age of the patients was 33 years. Sixteen of the patients were children. Sinopulmonary infections were the most common prediagnostic signs and symptoms; 66% had suffered from recurrent pneumonia, 60% from recurrent maxillary sinusitis, and 45% from recurrent bronchitis. There was a considerable delay in diagnosis. The mean delay was 8 years. At the time of diagnosis chronic pulmonary complications had already developed in 17% of the patients. The diagnosis was based on low serum immunoglobulin concentrations. This study showed that the awareness of common variable immunodeficiency is low. To improve the recognition of hypogammaglobulinemia, it should be suspected in every patient with recurrent bacterial infections. In addition to a low serum IgG concentration, measurement of specific antibody production is recommended to establish the diagnosis before initiation of a life-long and costly replacement therapy.

Parainfluenza virus type 3 infections in a hematology unit.

Parainfluenza virus type 3 (PIV3) is associated with a high mortality rate in BMT recipients with lower respiratory tract infections. We describe nine patients with hematological malignancies (five having undergone either allogeneic or autologous stem cell transplantation) identified as having PIV3 infection during a 2-month period in a Hematology Unit. Four patients with infiltrates on chest radiograph received intravenous ribavirin therapy; all survived. The infection was community-acquired in two patients, while nosocomial origin of the disease was evident, or presumed, in the remaining seven. The policy implemented to control the spread of PIV3 was as follows: (1) nasopharyngeal samples for antigen detection were obtained from all patients presenting with respiratory symptoms; (2) all diagnosed (or suspected) PIV3-positive hematological patients were nursed following contact isolation precautions, preferably in the Infectious Diseases Unit; and (3) staff were given further education on hospital hygiene. Our experience shows that it may be possible to avoid mortality for PIV3 lower respiratory tract infection in immunocompromised patients by early commencement of intravenous ribavirin. It is also possible, even without closing the ward, to contain nosocomial spread of PIV3 by implementing systematic nasopharyngeal sampling for rapid diagnostics, and by strict adherence to cohorting and contact isolation precautions.

Early identification of bacteremia by biochemical markers of systemic inflammation.

Group II phospholipase A2 (PLA2-II), procalcitonin (PCT) and C-reactive protein (CRP) are useful indicators of the severity of inflammation in various infections. To compare their discriminatory abilities at an early phase of bacteremia, PLA2-II, PCT and CRP were measured upon admission and 24-48 h thereafter in 29 patients with bacteremia, non-bacteremic bacterial or viral infections. The levels of PLA2-II and PCT were higher in bacteremia than in non-bacteremic bacterial or viral infections. PCT was highest upon admission, PLA2-II peaked at 12-24h, whereas CRP peaked one day later. At < or =24h, the AUC(ROC)s of PLA2-II and PCT were superior to those of CRP. Thereafter, the AUC(ROC)s of PLA2-II and PCT decreased and those of CRP increased. PLA2-II at cut-off level of 150 microg/L and PCT at 2-6 microg/L showed high sensitivity and specificity for bacteremia within the first 24h. In conclusion, PLA2-II and PCT are useful markers for early diagnosis of bacteremia. Devising analytical methods suitable for point-of-care testing would further enhance the clinical utility of the measurement of serum PLA2-II and PCT.

Neurologic manifestations of infective endocarditis: a 17-year experience in a teaching hospital in Finland.

BACKGROUND: Many previous studies have endeavored to find appropriate means to reduce the occurrence of neurologic manifestations in patients with infective endocarditis (IE). We evaluated patients with IE-associated neurologic complications and compared them with patients with IE who did not have neurologic symptoms. Particular attention was focused on assessing the impact of cardiac surgery and the presence of potential risk factors for complications on the outcome of the patients. METHODS: A total of 218 episodes designated as definite or possible IE according to Duke criteria and treated during the years 1980 through 1996 in a Finnish teaching hospital were retrospectively evaluated for neurologic manifestations. RESULTS: Neurologic complications were identified in 55 episodes (25%), with an embolic event as the most frequent manifestation (23/55; 42%). In the majority (76%) of episodes, the neurologic manifestation was evident before antimicrobial treatment was started, being the first sign of IE in 47% of episodes. Only 1 recurrent cerebral embolization was observed. Neurologic complications were significantly associated with Staphylococcus aureus infection (29% vs 10%; P =.001) and with IE affecting both the aortic and the mitral valves (56% vs 23%; P<.01), but not with echocardiographic detection of vegetations or anticoagulant therapy. Death during the acute phase of IE occurred in 13 episodes (24%) with neurologic complications and in 17 episodes (10%) without neurologic complications (P<.03). In episodes with neurologic complications, the IE-associated mortality rate was 25% (10/40) in the medical treatment group and 20% (3/15) in the surgical group. No neurologic deterioration was observed in these surgically treated patients postoperatively. CONCLUSIONS: Our results reinforce the belief that rapid diagnosis and initiation of antimicrobial therapy may still be the most effective means to prevent neurologic complications. These data underscore the importance of diagnostic alertness to the prognosis of patients with IE.

Aspergillus antigen in serum, urine and bronchoalveolar lavage specimens of neutropenic patients in relation to clinical outcome.

We have used a new, commercial enzyme-linked immunosorbent assay (ELISA, Platelia Aspergillus) to detect Aspergillus antigen in serum, urine and bronchoalveolar lavage (BAL) samples of 105 haematological patients who received empirical amphotericin B treatment for suspected fungal infection. 14% (60/419) of serum and 5% (18/373) of urine samples were positive. Ten-fold concentration of urine increased the number of positive samples to 31 (8%). The antigen was detected in 5 of 20 BAL samples, but fungal culture was negative in all of them. 22 patients had positive antigen test. Serum was positive in 17, urine in 7 and concentrated urine in 12 patients. Six patients had confirmed invasive aspergillosis. In all these patients, antigen was detected in serum, but urine was positive in only 2 patients. Patients whose antigen test turned negative during the amphotericin B treatment had significantly lower mortality than patients with persistently positive antigen test (2/10 vs. 8/8, p = 0.002). We conclude that Aspergillus galactomannan can be detected by ELISA in serum, urine and BAL samples of haematological patients, but the higher sensitivity of serum testing makes it preferable for screening. Disappearance of the antigen during antifungal therapy seems to correlate with good, and persistence with poor, clinical outcome.

Fungal colonization of haematological patients receiving cytotoxic chemotherapy: emergence of azole-resistant Saccharomyces cerevisiae.

Fungal colonization during cytotoxic chemotherapy was studied in 42 patients with a recent diagnosis of a haematological malignancy. In total, 2759 surveillance cultures were taken from the nostrils, throat, urine, stool and perineal region. Seven hundred and ninety-six positive surveillance cultures (28.9%) yielded 968 fungal isolates. The rate of fungal colonization did not differ between patients with acute leukaemia, patients with other haematological malignancies and control patients in the same ward at admission (71% vs. 67% vs. 80%). Patients with acute leukaemia were colonized at a significantly lower rate in samples from the throat (32%), urine (10%), stool (45%) and perineum (29%) taken during hospitalization when compared with other haematological patients (respective values 58%, 21%, 67% and 45%; P-values 0.001). This could be attributed to differences in the use of antifungal drugs. Although 21/42 (50%) of our patients had multiple-site fungal colonization at the end of follow-up, only one systemic Candida infection was diagnosed. Extensive use of antifungal treatment may have influenced the low incidence of systemic fungal infections during the follow-up. In addition to Candida species, Malassezia furfur, Geotrichum candidum and Saccharomyces cerevisiae were frequently isolated. The rate of S. cerevisiae isolation increased significantly over time after admission (1%, vs. 18% of isolates, P<0.001), suggesting hospital-acquired transmission. These isolates were highly resistant to azole antifungals (MIC90 128 microg/mL for fluconazole and 16 microg/ml, for itraconazole), and caused persistent multiple site colonization in 12 patients. Extensive use of antifungal agents in a haematological ward may keep the incidence of invasive fungal infections low in spite of heavy fungal colonization. However, there may be a risk of emergence of resistant fungal strains.

Eikenella corrodens prosthetic valve endocarditis in a patient with ulcerative colitis.

A 33-y-old male with ulcerative colitis developed prosthetic valve endocarditis (PVE) caused by Eikenella corrodens. The outcome of conservative treatment was successful. Only 2 cases of E. corrodens PVE were found in a survey of the English-language medical literature. In contrast to previous data indicating that eikenella infections usually derive from the oral cavity, our patient most likely acquired the infection by colonoscopy and mucosal biopsies, which were performed a few days before onset of the disease.

Pulmonary abnormalities in patients with primary hypogammaglobulinemia.

BACKGROUND: Pulmonary complications are common in patients with primary hypogammaglobulinemia. Intravenous immunoglobulin replacement therapy has been thought to reduce the occurrence of pulmonary complications, yet they do occur. OBJECTIVE: The purpose of this study was to evaluate pulmonary abnormalities in 22 patients with primary hypogammaglobulinemia (18 with common variable immunodeficiency, 4 with X-linked agammaglobulinemia) and to conduct a prospective 3-year follow-up study to assess the possible progression of pulmonary abnormalities. METHODS: Pulmonary changes were evaluated with use of pulmonary imaging (chest radiographs, high-resolution computed tomography), and pulmonary function testing. RESULTS: High-resolution computed tomography revealed pulmonary abnormalities in 21 patients. Bronchiectasis was present in 16 patients, whereas chest radiographs revealed bronchiectasis in only 3 patients. Pulmonary function testing showed obstruction in 5 patients. A prospective 3-year follow-up was conducted in 14 patients. It showed silent progression of bronchiectasis in 5 of the 14 patients, all of whom were receiving intravenous immunoglobulin replacement therapy and had preinfusion serum IgG concentrations of 5 g/L or more. CONCLUSIONS: Pulmonary abnormalities develop in most patients with primary hypogammaglobulinemia. A new finding is that silent and asymptomatic progression of pulmonary changes may occur in patients despite an adequate immunoglobulin replacement therapy. High-resolution computed tomography is the method of choice in monitoring pulmonary changes.

Borrelia burgdorferi detected by culture and PCR in clinical relapse of disseminated Lyme borreliosis.

A total of 165 patients with disseminated Lyme borreliosis (diagnosed in 1990-94, all seropositive except one culture-positive patient) were followed after antibiotic treatment, and 32 of them were regarded as having a clinically defined treatment failure. Of the 165 patients, 136 were tested by polymerase chain reaction (PCR) during the follow-up. PCR was positive from the plasma of 14 patients 0-30 months after discontinuation of the treatment, and 12 of these patients had a clinical relapse. In addition, Borrelia burgdorferi was cultured from the blood of three patients during the follow-up. All three patients belonged to the group with relapse, and two of them were also PCR positive. This report focuses on the 13 patients with clinical relapse and culture or PCR positivity. Eight of the patients had culture or PCR-proven initial diagnosis, the diagnosis of the remaining five patients was based on positive serology only. All 13 patients were primarily treated for more than 3 months with intravenous and/or oral antibiotics (11 of them received intravenous ceftriaxone, nine for 2 weeks, one for 3 weeks and one for 7 weeks, followed by oral antibiotics). The treatment caused only temporary relief in the symptoms of the patients. All but one of them had negative PCR results immediately after the first treatment. The patients were retreated usually with intravenous ceftriaxone for 4-6 weeks. None of them was PCR positive after the retreatment. The response to retreatment was considered good in nine patients. We conclude that the treatment of Lyme borreliosis with appropriate antibiotics for even more than 3 months may not always eradicate the spirochete. By using PCR, it is possible to avoid unnecessary retreatment of patients with 'post-Lyme syndrome' and those with 'serological scars' remaining detectable for months or years after infection.