Table 1.

Univariate and multivariate analyses of clinical findings between concordant and discordant cases in the overall cohort.

Univariate analysisMultivariate analysis
Categorical variablesaOR (95% CI)PbOR (95% CI)Pb
Age1.0 (0.6–1.7)0.869NANA
SexMale/female1.4 (0.6–3.3)0.383NANA
Previous chemotherapyYes/no2.4 (1.1–5.2)0.0320.7 (0.2–2.0)0.500
Location of primary tumorcLeft/right2.9 (1.0–8.6)0.0592.0 (0.6–6.8)0.261
Metastatic siteLiver/other0.3 (0.1–0.6)<0.0010.9 (0.2–3.9)0.897
Peritoneum/other0.8 (0.2–2.7)0.663NANA
Lung/other5.2 (2.3–11.4)<0.0011.8 (0.4–8.1)0.443
Collection interval (tissue vs. plasma)2.2 (1.5–3.3)<0.0011.9 (1.0–3.3)0.036
Source of tissue samplesMetastasis/primary1.9 (0.7–5.7)0.237NANA
Baseline longest diameter and number of lesions<20 mm and <10/≥20 mm or ≥105.0 (2.3–10.9)<0.0013.8 (1.5–9.5)0.004
Resection of primary tumorYes/no3.8 (1.1–13.1)0.0321.4 (0.3–5.6)0.633
CEA<5 ng/mL/≥5 ng/mL1.6 (0.7–3.6)0.296NANA
  • Abbreviations: CEA, carcinoembryonic antigen; NA, not available; OR, odds ratio.

  • aThe second category for each variable was used as reference in the model to calculate the OR.

  • bP values were derived using Fisher's exact test for categorical variables and the Kruskal–Wallis test for continuous variables.

  • cLeft includes splenic flexure to rectum; right includes caecum to transverse colon.