SSAT - the Validation of C-Reactive Protein and Albumin As Predictors of Post-Operative Infective Complications and Their Clinical Utility in Patients With Colorectal Cancer

Back to 2015 Annual Meeting Program

the Validation of C-Reactive Protein and Albumin As Predictors of Post-Operative Infective Complications and Their Clinical Utility in Patients With Colorectal Cancer
David G. Watt^*, Stephen T. McSorley, James H. Park, Paul G. Horgan, Donald C. McMillan

Academic Unit of Colorectal Surgery, Glasgow Royal Infirmary, Glasgow, United Kingdom

Background
Infective post-operative complication rates remain a significant problem following surgery for colorectal cancer and have been associated with poorer long term outcome. It has previously been reported that, in patients undergoing resection for colorectal cancer, a CRP of >170 mg/L on post-operative day 3 and a CRP of > 145 mg/L on post-operative day 4 and hypoalbuminaemia (<25 g/L) on post-operative days 3 and 4 were associated with the development of infective complications. The aim of the present study was to validate these post-operative thresholds and ascertain whether a post-operative scoring approach, using these thresholds, was capable of predicting the development of infective complications.
Methods
Patients with histologically proven colorectal cancer who were considered to have undergone potentially curative resection were included in the validation study (n = 410) between 2008 and 2013. Patients were assessed daily for post-operative complications. Diagnostic accuracy of CRP and albumin were analysed using receiver operating characteristics curves and threshold values derived appropriately.
Results
156 patients (38%) developed a complication with 106 (68%) of these infective. CRP was significantly higher and albumin significantly lower in those who developed infective complications compared to no complications (Table 1). The previously identified thresholds for CRP on post-operative days 3 & 4 were validated with day 3 CRP optimal threshold of 160-180 mg/L and day 4 CRP optimal threshold approximately 130 -150 mg/L. The previously identified thresholds for albumin on post-operative days 3 & 4 were validated with day 3 and 4 albumin optimal thresholds of 23-27 g/L. A post-operative systemic inflammation scoring system based on the validated thresholds on post-operative days 3 and 4 was constructed. Patients with a score of 0 on either day 3 (CRP ≤170 mg/L and albumin < 25 OR ≥ 25) or day 4 (CRP ≤ 145 mg/L and albumin < 25 OR ≥ 25) had infective complication rates of 17% and 20% respectively. Patients with a score of 2 on either day 3 (CRP > 170 mg/L and albumin < 25 g/L) or day 4 (CRP > 145 mg/L and albumin < 25 g/L) had infective complication rates of 52% and 63% respectively (p < 0.001).
Conclusions
The use of CRP and albumin as predictors of post-operative infective complications was validated in the present study. The magnitude of the post-operative systemic inflammatory response, evidenced by CRP thresholds of 170 mg/L on day 3 and 145 mg/L on day 4 and albumin thresholds of 25 g/L on days 3 and 4 (termed poGPS), were associated with the development of infective complications. This simple, objective post-operative scoring approach may be clinically useful in predicting which patients are likely to develop infective complications.

Table 1. Relationship between serial post-operative values and development of post-operative comlpications (n = 410)

Characteristic No Complications (n = 254) Infective Complications (n = 106) p-value ^a
Median (range) Median (range)
Pre-operative CRP (mg/L) 4.8 (0.3-208) 5.9 (0.4-212) 0.234
Day 1 101 (2-286) 106 (8-264) <0.001
Day 2 192 (25-417) 194 (60-387) <0.001
Day 3 175 (26-374) 216 (40-430) <0.001
Day 4 134 (22-388) 182 (27-425) <0.001
Day 5 103 (17-399) 185 (22-425) <0.001
Day 6 83 (12-362) 179 (15-431) <0.001
Day 7 76 (6-360) 146 (12-381) <0.001
Day 3 CRP (<170 mg/L/ >170 mg/L) ^b 156 (65)/ 83 (33) ^c 33 (33)/ 68 (67) ^c <0.001
Day 4 CRP (<145 mg/L/ > 145 mg/L) ^b 141 (69)/ 64 (31) ^c 36 (35)/ 66 (65) ^c <0.001
Pre-operative albumin (g/L) 36 (15-46) 37 (17-44) 0.594
Day 1 26 (9-35) 25 (11-34) 0.071
Day 2 25 (10-33) 25 (12-37) 0.003
Day 3 24 (11-32) 23 (12-31) <0.001
Day 4 25 (12-33) 23 (10-31) <0.001
Day 5 25 (13-38) 23 (14-32) <0.001
Day 6 25 (11-36) 23 (6-30) <0.001
Day 7 26 (12-38) 23 (5-31) <0.001
Day 3 Albumin (>25 g/L/ < 25 g/L) ^b 142 (60)/ 96 (40) ^c 42 (42)/ 59 (58) ^c 0.002
Day 4 Albumin (> 25 g/L/ <25 g/L) ^b 130 (63)/ 75 (37) ^c 37 (37)/ 64 (63) ^c <0.001

a compared to no complications
b thresholds have been previously described for prediction of infective post-operative complcations
c n (%)

Table 2. Relationship between day 3 and day 4 post-operative GPS (poGPS) and development of infective complications

Complications
No Complications
n (%) Infective Complications
n (%) p value
Day 3 poGPS 0 156 (83) 33 (17)
Day 3 poGPS 1 41 (63) 24 (37)
Day 3 poGPS 2 40 (48) 44 (52)
<0.001
Day 4 poGPS 0 141 (80) 36 (20)
Day 4 poGPS 1 36 (66) 19 (34)
Day 4 poGPS 2 27 (37) 46 (63)
<0.001

Back to 2015 Annual Meeting Program