Hydrochloric Acid Lock Therapy for Central Line-associated Bloodstream Infections

BACKGROUND

Patients with cancer or hematologic diseases and a long-term central venous access device (CVAD), including central venous catheters and intravenous ports, are at risk of developing central line-associated bloodstream infections (CLABSI). A patient with CLABSI is treated with systemic antibiotic therapy. Additionally, prompt removal of the CVAD because of CLABSI is often recommended. However, removal and replacement of a long-term CVAD is an invasive procedure and not always possible. Therefore, various types of lock therapies have been tried in attempt to cleanse and preserve the CVAD.

At more hospitals in Denmark, hydrochloric acid lock therapy (HALT) has been used routinely for patients with cancer and hematologic diseases and a CLABSI. However, HALT is not recommended outside Denmark due to limited evidence of its use.

PATIENT AND METHODS

Eligible patients with CLABSI will be enrolled during antibiotic therapy and a maximum of 5 days after the detection of a positive blood culture/CLABSI. The patients will be randomized to receive blinded lock therapy with hydrochloric acid (2 M) or placebo (normal saline). HALT and placebo lock therapy is carried out by trained oncologic nurses in identical fashion: Hydrochloric acid or normal saline in a volume corresponding to dead space (0.3-1.8ml) will be instilled and aspirated 10 min later; the instillation is completed 3 times. In addition to the study intervention, patients in both arms will receive standard systemic antibiotic therapy.

The patient is followed 6 weeks from instillation with HALT/placebo. The patient can be re-included if the patient has 1) replacement of the CVAD, or 2) a new CLABSI later than 6 weeks from a previous CLABSI. In case of re-inclusion the patient, the patient will be randomized again.

SAMPLE SIZE CALCULATION

The primary objective is to compare the risk of treatment failure between the two treatment groups.

A recent randomized, controlled trial showed a treatment failure rate of 33% within 42 days for children with CLABSI treated with placebo lock therapy (heparin in normal saline) (Wolf et al, 2018). External data from 2015-2020 of patients admitted with CLABSI and receiving HALT at Copenhagen University Hospital, Denmark, led to an estimated treatment failure rate of 17% (unpublished data).

Based on the above-estimated treatment failure rate in each group as well as feasibility constraints (incl. timeline and accrual rate), a sample size of 125 patients in each group was chosen. This choice was estimated to provide approximately 80% power to show a difference in risk between the two groups. This power was computed for a two-sided test at the type-I error level 5%, assuming a 2% dropout rate, using standard asymptotic normal approximations.

CHANGES IN THE PROTOCOL SINCE THE FIRST EDITION OF THE PROTOCOL

The first version of this protocol was published in June 2022. In October 2023, after inclusion of 61 episodes of CLABSI in paediatric patients the investigators made the following changes:

1. The titles: The word 'children' has been omitted and replaced with 'patients with cancer and hematologic diseases'.
2. Sample size: Increased from 100 to 250 patients.
3. Patient population: Changed from age 0-17 years to age 0-100 years and the inclusion of non-malignant hematologic diseases.
4. Re-inclusion of patients: Changed from 'cross-over' to new blinded randomization.
5. Time to enrolment: Changed from 48 hours to 5 days after detection of CLABSI.
6. Definition of persistent infection: Changed from "persistent positive blood cultures for more than 48 hours" to "persistent positive blood cultures 24 hours after HALT/placebo".

Reasons for changes

1. The titles: The titles now reflect inclusion of adults and patients with of non-malignant hematologic diseases.
2. Sample size: In the first power calculation, the investigators estimated a treatment failure rate in patients receiving HALT to 10%. This estimation was based on two older Danish articles, including patients in 1999-2005 (Larsen et al, 2011, Madsen et al, 2013), and data from 2010-2015 of children with CLABSI receiving HALT in Denmark (unpublished data). Recently, external data from 2015-2020 suggested a higher treatment failure rate of 17% following HALT (unpublished data). An updated power calculation using a treatment failure of 17% following HALT versus 33% following placebo suggested to increase the sample size to 250 (i.e. 125 in each group, see above).
3. The patient population: To meet the new sample size requirement and also address the research question in adults, the investigators will include adults from the Department of Haematology, Copenhagen University Hospital, Denmark, from October 17th, 2023. This choice was thought reasonable because: 1) External data suggest that the risk of bacteraemia in adults, and its recurrence, is similar to that observed in the paediatric population (unpublished data). 2) It is logistically feasible, as the placement and supervision of the CVADs is handled by the same anaesthesia personnel. A subgroup analysis of paediatric and adult patients will be made to complement the main analysis that will now include both paediatric and adult patients.
4. Changing of cross over in case of re-inclusion: In the first version of the protocol from June 2022, re-inclusion of patients was performed with cross-over design. The decision of cross-over design was mainly motivated by power considerations; it was expected to increase the power of the study. However, after re-inclusion of 10 episodes of CLABSI, the investigators realized that it led to logistical challenges and ethical issues, as it required some unblinding for the re-inclusions. The investigators have therefore changed the setup to a new randomization. The impact of this decision on the power was estimated to be very minor and therefore it was thought that such power considerations could not outweigh the importance of blinding.
5. Time to enrolment: The investigators have expanded the time window for including patients in the study from 48 hours after the detection of a positive blood culture to 5 days. This is because 1) new external data has shown that the risk of relapse during the first 5 days, when the patients receive antibiotic therapy, is only approximately 1% (unpublished data), 2) The expected effect of HALT is to reduce the occurrence of relapse of infection or new infection after the cessation of antibiotic treatment and 3) there is a risk of exclusion of otherwise eligible patients with a time-limit of 48 hours due to practical circumstances, e.g., during weekends.
6. Definition of persistent infection:

Definition of persistent infection is changed from "persistent positive blood cultures for more than 48 hours" to "persistent positive blood cultures 24 hours after HALT/placebo" since the installation of HALT/placebo is not always carried out immediately after the detection of a positive blood culture.