Thrombocytosis as a Marker for Postoperative Complications in Colorectal Surgery.

Background: Blood platelet measurement is a widely available and inexpensive test that is performed routinely. Platelets are thought to act by inducing inflammation and play a role in clotting and antimicrobial defence. A postoperative rise in the platelet count (thrombocytosis) is often dismissed as an incidental finding, but there is growing evidence to suggest that it may act as an indicator to underlying pathology. It correlates with significant pyogenic infections as well as multiple malignancies. In addition to this, recent research indicates that thrombocytosis may be a useful prognostic indicator for postoperative outcomes in patients with malignancies. In patients undergoing surgery for gastric cancer, a combination of platelet count and neutrophil-to-lymphocyte (NLR) ratio collected preoperatively was shown to correlate with postoperative survival.

Objective: To evaluate whether there is a positive correlation between pre- and postoperative thrombocytosis and the risk of complications following colorectal surgery.

Methods: This was a retrospective observational study based in Morriston Hospital, Swansea. Patients undergoing elective colorectal surgery for an 18-month period between 2014 and 2016 were included. Data on patient demographics, pre- and postoperative platelet count, the first date at which the highest platelet count was recorded, length of stay, type of operation, and postoperative complications using the Clavien-Dindo classification was obtained from the theatre booking software (TOMS) and Welsh Clinical Portal. Pearson's chi-square test was used for the analysis of the categorical variables.

Results: Of the 201 patients studied, 75 (37%) had postoperative thrombocytosis (platelets ≥ 500 × 10/L, range 501-1136), 120 (59%) had postoperative normocytosis (platelets < 500 × 10/L, range 107-499), and 6 (2.9%) patients were excluded due to insufficient data. Peak platelet level was seen at a median of 8 days postoperatively but ranged from days 1 to 49. In patients with thrombocytosis, the mean time to peak platelet count was 9.5 days and ranged 1 to 49 days. 101/195 (52%) patients had a Clavien-Dindo III/V postoperative complication: 63% patients with postoperative normocytosis and 24% with postoperative thrombocytosis. In the thrombocytosis group, 16/75 (21%) were found to have postoperative pelvic collections compared to 1/120 (0.8%) of the normocytic patients. The total percentage of medical complications (44% versus 20%, = 0.006) and surgical complications (64% versus 15.8%, = 0.0001) was higher in the thrombocytosis group compared to the normocytosis group.

Conclusion: In this retrospective study, thrombocytosis was shown to have a positive correlation with postoperative medical and surgical complications. An elevated platelet count in the postoperative period should alert the clinician to a developing complication. We recommend that further studies with a larger sample size would test the specific associations with individual complications.