25 mm Hg versus 35 mm Hg elastic compression stockings to prevent post-thrombotic syndrome after deep vein thrombosis (CELEST): a randomised, double-blind, non-inferiority trial.

Background: The optimal strength of compression needed to prevent post-thrombotic syndrome (PTS) after a proximal deep vein thrombosis (DVT) is debated. We aimed to assess whether 25 mm Hg elastic compression stockings (ECS) are non-inferior to 35 mm Hg ECS in preventing PTS after a DVT.
Methods: In this multicentre, double-blind, non-inferiority, randomised controlled trial, we enrolled adults (≥18 years) with a first ipsilateral proximal DVT attending 46 French vascular medicine hospital departments or private practices. Participants were randomly allocated (1:1, stratified by centre, age, and sex; with varying block sizes of two and four) to wear 25 mm Hg or 35 mm Hg ECS for 2 years. The primary outcome was the cumulative rate of PTS 2 years after inclusion, defined by a Villalta scale (≥5). Efficacy was assessed by intention-to-treat and in eligible participants who had complete primary outcome data. A per-protocol analysis was also conducted among compliant patients as a secondary outcome measure. Safety was assessed in all participants who used ECS at least once, and for which we have at least some tolerance information during follow-up. The margin for non-inferiority was 12·5%. This study is registered with ClinicalTrials.gov, NCT01578122, and has been completed.
Findings: Between June 28, 2012, and July 21, 2017, we enrolled 341 eligible participants who consented to randomisation. 233 (68%) were men and median age was 59 years (IQR 45-70). Collection of ethnicity and race as a routine research variable is not authorised in France. Median follow-up was 735 days (IQR 721-760). 249 (73%) had complete data at 2 years. For the primary analysis, 40 (31%) of 129 participants with complete data in the 25 mm Hg ECS group and 40 (33%) of 120 in the 35 mm Hg group had PTS (absolute difference -2·3% [90% CI -12·1 to 7·4], p _{non-inferiority} =0·0062; relative risk 0·93, 95% CI 0·65 to 1·33). Results remained similar after imputation of missing data in patients we were authorised to do so: the cumulative proportion of PTS was 45 (29%) of 154 in the 25 mm Hg ECS group versus 52 (35%) of 148 in the 35 mm Hg ECS group (relative risk 0·83, 95% CI 0·60 to 1·16). Absolute difference was -5·9%, (90% CI -14·7 to 2·9), p=0·0003 for non-inferiority. Adherence was optimal (>80% and modified GIRERD score of 0-2) for 75 (51%) of 146 patients assigned to 25 mm Hg ECS and for 56 (42%) of 134 patients assigned to 35 mm Hg ECS (p=0·11). Regarding major adverse events related to ECS, there were no between-group differences in rates of deep vein thrombosis (0 vs 1 [0·6%]), ipsilateral leg ulcer (0 vs 1 [0·6%]), infection (0 vs 0), or death (0 vs 0) between the 169 patients evaluated in the 25 mm Hg ECS group and the 159 patients in the 35 mm Hg ECS group. Two (1%) of 328 patients who ever wore ESC developed ECS-related serious adverse events, one distal DVT and one leg ulcer (both in the 35 mm Hg ECS group). In the 25 mm Hg group, 6 patients died, 14 had a venous thromboembolic recurrence (proximal DVT or pulmonary embolism), and 7 had a major bleed. In the 35 mm Hg group, 5 patients died, 10 had a venous thromboembolic recurrence (proximal DVT or pulmonary embolism), and 6 had a major bleed.
Interpretation: Although we did not reach the prespecified sample size, our results suggest that 25 mm Hg ECS are non-inferior to 35 mm Hg ECS in preventing PTS. Larger more powerful studies are needed.
Funding: Laboratoires Innothera, France.
Competing Interests: Declaration of interests All authors declare financial support from Laboratoires Innothera for the submitted work through their institutions. VJ and FV are employees of Laboratoires Innothera but were not involved in the analysis and interpretation of data, or in writing the Article. AB-R declares consultant fees and travel grants from Bristol-Myers Squibb Pfizer and Bayer. SB declares honoraria from Urgo and Thuasne.
(Copyright © 2022 Elsevier Ltd. All rights reserved.)