Effect of Prophylactic Embolization on Patients With Blunt Trauma at High Risk of Splenectomy: A Randomized Clinical Trial

Abstract

Importance: Splenic arterial embolization (SAE) improves the rate of spleen rescue, yet the advantage of prophylactic SAE (pSAE) compared with surveillance and then embolization only if necessary (SURV) for patients at high risk of spleen rupture remains controversial.

Objective: To determine whether the 1-month spleen salvage rate is better after pSAE or SURV.

Design, setting, and participants: In this randomized clinical trial conducted between February 6, 2014, and September 1, 2017, at 16 institutions in France, 133 patients with splenic trauma at high risk of rupture were randomized to undergo pSAE or SURV. All analyses were performed on a per-protocol basis, as well as an intention-to-treat analysis for specific events.

Interventions: Prophylactic SAE, preferably using an arterial approach via the femoral artery, or SURV.

Main outcomes and measures: The primary end point was an intact spleen or a spleen with at least 50% vascularized parenchyma detected on an arterial computed tomography scan at 1 month after trauma, assessed by senior radiologists masked to the treatment group. Secondary end points included splenectomy and pseudoaneurysm, secondary SAE after inclusion, complications, length of hospital stay, quality-of-life score, and length of time off work or studies during the 6-month follow-up.

Results: A total of 140 patients were randomized, and 133 (105 men [78.9%]; median age, 30 years [interquartile range, 23-47 years]) were retained in the study. For the primary end point, data from 117 patients (57 who underwent pSAE and 60 who underwent SURV) could be analyzed. The number of patients with at least a 50% viable spleen detected on a computed tomography scan at month 1 was not significantly different between the pSAE and SURV groups (56 of 57 [98.2%] vs 56 of 60 [93.3%]; difference, 4.9%; 95% CI, -2.4% to 12.1%; P = .37). By the day 5 visit, there were significantly fewer splenic pseudoaneurysms among patients in the pSAE group than in the SURV group (1 of 65 [1.5%] vs 8 of 65 [12.3%]; difference, -10.8%; 95% CI, -19.3% to -2.1%; P = .03), significantly fewer secondary embolizations among patients in the pSAE group than in the SURV group (1 of 65 [1.5%] vs 19 of 65 [29.2%]; difference, -27.7%; 95% CI, -41.0% to -15.9%; P < .001), and no difference in the overall complication rate between the pSAE and SURV groups (19 of 65 [29.2%] vs 27 of 65 [41.5%]; difference, -12.3%; 95% CI, -28.3% to 4.4%; P = .14). Between the day 5 and month 1 visits, the overall complication rate was not significantly different between the pSAE and SURV groups (11 of 59 [18.6%] vs 12 of 63 [19.0%]; difference, -0.4%; 95% CI, -14.4% to 13.6%; P = .96). The median length of hospitalization was significantly shorter for patients in the pSAE group than for those in the SURV group (9 days [interquartile range, 6-14 days] vs 13 days [interquartile range, 9-17 days]; P = .002).

Conclusions and relevance: Among patients with splenic trauma at high risk of rupture, the 1-month spleen salvage rate was not statistically different between patients undergoing pSAE compared with those receiving SURV. In view of the high proportion of patients in the SURV group needing SAE, both strategies appear defendable.

Trial registration: ClinicalTrials.gov Identifier: NCT02021396.

Figure 1.. Study Flowchart

CT indicates computed tomography; OIS, Organ Injury Scale; pSAE, prophylactic splenic arterial embolization; and SURV, surveillance with embolization only if necessary. ^aSee eTable 1 in Supplement 2 for reasons for ineligibility. ^bDay 5 visit occurred between day 4 and day 8 after inclusion.

Figure 2.. Comparison of the Total Length of Hospitalization

For prophylactic splenic arterial embolization (pSAE), the median length of hospitalization was 9 days (interquartile range, 6-14 days). For surveillance and then embolization only if necessary (SURV), the median length of hospitalization was 13 days (interquartile range, 9-17 days) (P = .002). The horizontal lines indicate the median. The vertical lines indicate the 2 groups.