Laparoscopic splenectomy for immune thrombocytopenia (ITP): long-term outcomes of a modern cohort.

Background: The advent of newer second-line medical therapies (SLMT) for immune thrombocytopenia (ITP) has contributed to decreased rates of splenectomy, following a trend to avoid or delay surgery. We aimed to characterize the long-term outcomes of laparoscopic splenectomy (LS) for ITP at our institution, examining differences in LS efficiency when performed before or after SLMTs.

Methods: Adults with primary ITP who underwent LS between 2002 and 2016 were identified. Retrospective review of electronic medical records was supplemented with telephone interviews. Treatment response was defined according to current guidelines as complete responders (CR), responders (R), and non-responders (NR). Kaplan-Meier estimates assessed relapse-free rates, and predictors of long-term response were investigated using logistic regression.

Results: 109 patients met inclusion criteria, from which 42% were treated with an SLMT before referral to LS. LS was completed in all cases, with no conversions or intraoperative complications. The perioperative morbidity was 7.3%, including 3 deep vein and 2 portal vein thrombosis, one reoperation for bleeding, and no mortalities. Splenectomy was initially effective in 99 patients (CR + R = 90.8%), and 10 patients were NR. At a median 62-month follow-up, 25 patients relapsed, resulting in a 68% CR + R rate. Proportion of CR + R was similar in patients who previously received SLMT and those who did not (61 vs. 76.7%, p = 0.08). CR + R patients were younger (45 vs. 53, p = 0.03), had higher preoperative platelet counts (36 vs. 19, p = 0.01), and experienced a higher increment in platelet counts during hospital stay (117 vs. 38, p < 0.001) as well as 30-days postoperatively (329 vs. 124, p < 0.001). Only a robust response in platelet count at 30-days postoperatively was independently associated with long-term response (OR 1.005, p = 0.006).

Conclusion: LS was curative in 68% of patients, with no statistically significant difference when performed before or after SLMTs. Outcomes remain challenging to predict preoperatively, with only a robust increase in platelet counts on short term being associated with long-term response.