Doxycycline plus streptomycin versus ciprofloxacin plus rifampicin in spinal brucellosis [ISRCTN31053647]

Human brucellosis is a worldwide zoonosis and affects over 500 000 people yearly in non-industrialized countries of the world. It is still public health problem in most countries of Mediterranean, including Turkey, Balkan, the Middle East and Central and South America [1].

Brucellosis is a systemic disease and may involve any organ system. However, osteoarticular involvement is the most common complication of brucellosis and the reported incidence is 10–85% in most series [2‐7]. Spinal column is one of the most frequently affected site in osteoarticular brucellosis. However its incidence is highly variable (2–54%) in different studies, depending on the study population, the species of Brucella involved and the difficulty of diagnosis [3‐8]. On the other hand, the optimal treatment regimen and duration of the therapy is still controversial in osteoarticular involvement including spinal brucellosis. The combination of doxycycline with streptomycin is accepted to be the most effective regimen in spinal brucellosis [1]. However, therapeutic failure and relapse are still common. The risk of adverse drug reactions is increased due to the prolonged use of these antibiotics and limited the therapy [4, 9, 10]. Other treatment option, the combination of rifampicin and doxycycline, are associated with worst outcomes [11, 12]. In recent years, fluoroquinolones has been shown to be effective both in vitro and in vivo in the treatment of brucellosis [13‐19]. Furthermore, due to ability of fluoroquinolones to penetrate and achieve increased concentration in the tissues and their proven efficacy in bone and soft tissue infections, the outcome may be better in spinal brucellosis. The regimens including fluoroquinolones may be an alternative therapy to doxycycline and streptomycin.

The aim of this study is to compare the efficacy, adverse drug reactions, complications and cost of ciprofloxacin (C) plus rifampicin (R) versus doxycycline (D) plus streptomycin (S) in the treatment of spinal brucellosis.

This study was planned as an open, controlled and non-randomized. The Ethical Committee of Erciyes University approved the study protocol used in this study. The patients also accepted to take place in this study. The patients diagnosed as spinal brucellosis between January 2002 to December 2004 were enrolled into two study groups consecutively. Age less than 16 years, pregnancy, neurobrucellosis, previous history of brucellosis and antimicrobial therapy and discontinuation of the therapy for any reason (allergy to any of the drugs, death, adverse reactions) were exclusion criteria from the study.

Patients were enrolled into the two antimicrobial therapy groups (DS vs. CR). Patients in DS group received doxcycline, 100 mg twice a day orally for at least 12 weeks, plus streptomycin, 1 g per day intramuscularly for 21 days (1 g every other day for patients more than 60 years of age for 21 days). Patients in CR group received ciprofloxacin, 500 mg twice a day orally, plus rifampicin, 600 mg per day in a single morning dose for at least 12 weeks. The antibiotic therapy was prolonged according to clinical improvement and the resolution of magnetic resonance imaging (MRI) findings. The antibiotic therapy was changed if the adverse effects of the drugs or therapeutic failure were observed. If patient's complaints and findings were partially improved and there was not an increase in MRI findings at the end of 12 weeks, the therapy was prolonged until resolving MRI findings. At least two weeks bed rest was suggested to all patients and lumbosacral orthosis was suggested to patients with pain due to instability. Furthermore, the patients were followed up until 12 month after cessation of the therapy for the evaluation of sequelae and relapse. For the cost analysis of the two regimens, only the cost of antibiotic therapy was analysed for each patient.

Spinal brucellosis is one of the most common complication of human brucellosis in endemic regions [1, 9]. In this study its incidence was 9.4% and it was 12.3% in our previous study [4]. Ingestion of unpasteurized dairy products (raw milk, butter, cheese) is the most common source of infection [1] and 90% of our patients had a history of ingestion of raw milk and products. Furthermore, brucellosis has been accepted an occupational disease and consequently male gender are frequently affected [1]. In our study each sex is affected similarly. High incidence of ingestion of raw milk products can explain this equivalence.

The symptoms of spinal brucellosis may be insidious and non-specific, especially in subacute and chronic forms [1, 9]. In our study, subacute form was more common and back pain, malaise, and arthralgia were the most common complaints. Fever is another common clinical manifestation in spinal brucellosis and undulant fever is described in untreated patients [1, 23]. However, 65% of our patients complained about fever, only four patients had fever on physical examination. Since most of our patients were followed up in outpatient clinic, fever may be overlooked in our study.

The diagnosis of spinal brucellosis is very important for the duration of antibiotics but early diagnosis is difficult. Since the early radiological signs of spinal brucellosis are non-specific and may appear late after the onset of symptoms [24]. The appearance of destructive changes on direct radiographs is about three months and computed tomography may not be always helpful in the differential diagnosis from other degenerative diseases and disc herniation [7, 25]. MRI is the most useful method for diagnosis, assessment and management of spinal brucellosis. Also it has high sensitivity for differential diagnosis and identifying extent of brucellar spondylitis [26]. Consequently, MRI was used for the diagnosis and assessment in this study. Spinal involvement was observed in forty-one levels and lumbar spine was the most common involved site, particularly L4-L5 and L5-S1 in our patients, as reported in previous studies [2‐7]. Moreover, multiple site involvement can be seen and the reported incidence was 9% in the previous studies [2, 7]. In our study, multifocal involvement was 30.3% and, moreover, four patients had non-contiguous involvement at multiple sites. Abscess formation has been rarely reported in the past [7], but in recent years, with highly sensitive diagnostic techniques, it becomes a common finding (21%) [6], as in this study (42%). Sacroiliitis is one of the most common form of osteoarticular involvement [4, 5], however only 9% of our patients had sacroiliitis.

Despite its frequency in developing countries, the optimal treatment regimen is still controversial. The combination of tetracycline/doxycycline and streptomycin is accepted by the World Health Organization (WHO) as the treatment of choice for spinal brucellosis [1]. However, the reported success rate with this classical regimen is 60%–90%. Moreover, therapeutic failure and relapse rates are still common [2, 7]. In our previous report, the relapse rate in osteoarticular involvement was 12.1% and the highest relapse rate was in tetracycline+streptomycin group (14.3%) [4]. Also, the inconvenience of intramuscular streptomycin administration and the toxicity of tetracycline/doxycycline limited the adaptation of patients. The second alternative regimen, the combination of rifampin-doxycycline, has been shown to be less effective than doxcycline-streptomycin [11, 12]. As a result, alternative regimens are investigated in recent years. Quinolones are the most commonly studied antibiotics because of their good bactericidal activity and penetration into leukocytes and macrophages. However, monotherapy with these antibiotics showed high relapse rates [27, 28], combination regimens with rifampin were found as effective as doxcycline plus rifampin [17, 18]. To our knowledge, combination regimens including quinolones were not compared with the classical regimen, doxcycline-streptomycin, in spinal brucellosis. Because of their success in bone infections, they can be alternative as oral regimen to doxcycline-streptomycin therapy in spinal brucellosis. The aim of this study was to compare quinolone-rifampicin efficacy in spinal brucellosis with doxcycline-streptomycin. However, there were differences between the both groups, (patients in CR group were older, the median ESR and CRP were higher at baseline, abscess formation and operation indication before therapy were more prevalent in this group and multifocal involvement was more prevalent in DS group), the outcome was not different in the two groups. Therapeutic failure wasn't observed in the two treatment groups.

The duration of antibiotics in spinal brucellosis is also controversial. As relapse rate and therapeutic failure were significantly higher in shorter durations [1, 9], the duration of antibiotics was at least 12 weeks in our study. It was prolonged according to the resolution of clinical and MRI findings. There was no difference in median duration (12 weeks) of therapy in the both groups. In the comparison of the adverse effects, no adverse effect was observed in DS group and two patients in CR group stopped the treatment because of dizziness. Relapse was not observed in any patient after one year of the therapy. Probably, this is the result of appropriate duration of antibiotics.

The diagnostic or curative surgery can be performed in spinal brucellosis but it is last option in the management of spinal brucellosis. The indications for surgery are; spinal instability, cauda equina syndrome, severe weakness of muscle due to extradural inflammatory mass or progressive collapse [6, 7]. Eight patients in our study underwent operation and the most common indication for surgery was instability. The need for operation after therapy was not different in the both treatment group. Abscess drainage is not always needed in the absence of neurological deterioration [7] and only one of our patients underwent abscess drainage for neurological deficit; others all improved with medical management.

A delayed convalescence after treatment is the well-known clinical sequela in spinal brucellosis. The reason can't be understood, but pscychoneurosis exacerbated by the infection is believed to be a reason for the delay [1]. Mild clinical sequela was observed equally in the both group one year after the therapy. Only one patient in CR group who had operation before and after the therapy had moderate sequale.

The reported disadvantage of quinolone including is the high cost rates [18, 19]. In this study we only calculated the drug cost for the median duration (12 weeks) of the both group and the cost of CR therapy was 1.2 fold higher than the cost of DS therapy.

Ciprofloxacin-rifampicin regimen is as effective as classical regimen. Despite the disadvantages (older age, more prevalent operation and abscess formation before the therapy) of the patients in CR group, the duration of the therapy and clinical response were not different from the classical regimen. So combination of ciprofloxacin and rifampin can be an alternative regimen in spinal brucellosis. However it can't be considered as first-line therapy because of its high cost that is important for developing countries. Classical regimen (doxycycline with streptomycin), with the appropriate duration (at least 12 weeks), is still the first line antibiotics and alternative therapies should be considered when adverse reactions observed.