The use of co-trimoxazole in catheter lock therapy. A report on a difficult case

Dear Editor,

Today, the use of long-term catheters to administer therapeutic agents (chemotherapy, antibiotic therapy, or blood or plasma derivatives), as well as the extraction of blood samples, is becoming increasingly common.^1,2 Such devices avoid the need for repeated venipunctures, which are associated with discomfort for patients and derived complications (e.g. phlebitis, etc.). However, it is not uncommon that they lead to infections (bacterial or fungal).¹ The management of catheter-related infections is complex, as, on the one hand, removing the catheter decreases the options for vascular access (which are often limited), while keeping it in place is associated with maintaining the focus of the infection. In some cases, an alternative option is used, consisting of both systemic treatment of the bacteraemia and the local administration of antimicrobial agents, known as “locks”.³ The examples of vancomycin (gram-positive cocci) and amikacin (gram-negative bacilli) are well established,² and there are efficacy data for other antimicrobial agents (e.g. amoxicillin-clavulanic acid or ciprofloxacin).^2,4 Nevertheless, there are some circumstances in which the options are more limited due to the characteristics of the patient or the type of microorganism.

We report the case of a 51-year-old woman diagnosed with seronegative myasthenia gravis (MG) who underwent a thymectomy in February 2003. She subsequently required various hospital admissions due to myasthenic crises, including on one occasion to the ICU. She attended the Accident and Emergency Department due to presenting a fever of 40 °C, primarily in the evenings, accompanied by tremors for one month prior to admission. The patient had had a Port-a-Cath^® (BARD) since 2005 for the administration of immunoglobulins, but did not report other signs of local involvement, such as the existence of pain, erythema or purulent discharge at the insertion site. On admission, and after taking blood cultures peripherally and from the catheter for study using qualitative methods via an automated system, a lock was created with vancomycin 4 mg/ml for 24 h (5 ml) and empirical intravenous treatment was started, avoiding antimicrobial agents contraindicated in patients with MG, with vancomycin (1 g/12 h) and ceftriaxone (1 g/12 h). Given the importance of quickly initiating an effective treatment, and because ceftriaxone is a time-sensitive antimicrobial agent, administration every 12 h was preferred. The fever remitted and the blood cultures revealed Stenotrophomonas maltophilia (S. maltophilia) and Streptococcus agalactiae (S. agalactiae) in the catheter and peripheral blood. In view of the bacterial growth times in the blood cultures, it was very likely that the catheter was the source of the bacteraemia. The treatment was later adjusted based on the antibiogram to co-trimoxazole (800/160 mg/12 h) given orally and vancomycin (1 g/12 h) administered peripherally to treat both microorganisms for 10 days. Due to their coexistence, a lock was created using two different antibiotics. S. agalactiae was susceptible to vancomycin, but for S. maltophilia, based on the patient's antibiogram and due to the restricted antimicrobial options available, only co-trimoxazole could be used for the lock. After performing a search of the literature, only a limited number of articles were found in which such a lock was used, the majority in vitro.^4–7 Once informed consent had been obtained from the patient, the catheter lock was created by alternating 5 ml vancomycin-heparin (4 mg/ml) and 5 ml co-trimoxazole-heparin (10 mg/ml) for 12 h each. The blood cultures obtained from the catheter at 72 h, 7 days and 10 days were negative after 7 days’ incubation. The intention was to substitute the vancomycin with linezolid (600 mg/12 h orally), but the patient reported nausea and it had to be withdrawn. At 30 days from the end of antibiotic treatment, blood cultures were taken and were negative.

The case presented illustrates a complex situation as S. maltophilia is not a microorganism normally encountered in bacteraemia and several of the drugs to which it is susceptible are contraindicated in MG. In this case, due to the patient's underlying illness, it was necessary to use an antibiotic lock for which there is practically no published experience. After reviewing the stability of the antimicrobial agent in the summary of product characteristics and always with the patient's consent, the decision was made to attempt its use before removing the catheter, with the result that we were able to avoid the latter. We believe that catheter locks, including potential antibiotics other than those traditionally used, should always be taken into account together with systemic treatment.