Table 1.

Comparison of baseline variables for the matched versus unmatched patients

ParametersTotal (n = 180)Matched (n = 87)Unmatched (n = 93)Pa
Age at diagnosis (Median, 95% CI)52 (50–55)52 (50–55)54.2 (48–58)0.556
Gender (n; %)<0.001
 Women121 (67.2%)70 (80.5%)51 (55%)
 Men59 (32.8%)17 (19.5)42 (45%)
Tumor type (n; %)
 Breast60 (33.3%)45 (52%)15 (16%)<0.001
 Brain15 (8.3%)9 (10%)6 (6.5%)0.423
 Gastrointestinal35 (19.4%)8 (9%)27 (29%)0.001
 Genitourinary22 (12.2%)10 (12%)12 (13%)0.823
 Head and neck16 (8.9%)6 (7%)10 (11%)0.437
 Lung16 (8.9%)7 (8%)9 (9.7%)0.796
 Skin/melanoma11 (6.1%)2 (2%)9 (9.7%)0.059
Total n of alterations (Median, 95% CI)4 (3–5)5 (4–6)3 (3–4)0.017
Alterations, n (%)
CDKN2A34 (18.9%)17 (19.5%)17 (18%)0.851
TP5387 (48.3%)44 (50.6%)43 (46.2%)0.655
PTEN19 (10.6%)10 (11.5%)9 (9.7%)0.810
Metastasis at time of biopsy, n (%)138 (76.7%)71 (82%)67 (72%)0.159
Metastasis at time of diagnosis, n (%)39 (21.7%)16 (18%)23 (25%)0.366
Therapy was 1st line, n (%)43 (23.9%)14 (16%)29 (31%)0.023
  • aFisher's exact tests were used for categorical variables. The Mann–Whitney U test was used and medians (95% CI) were reported for continuous variables [“age at diagnostic” and “total number (n) of alterations”]. Variables with P values (P < 0.1; bolded in the table) were included in a logistic regression to compute the propensity score (propensity for a patient to be matched vs. unmatched); as the “gender (women),” P = 2.6E−4 and “breast cancers,” P = 6E−7 strongly correlated, and to avoid collinearity in the model, only “breast cancers” were included. Five patients had “other” tumor types and were not included.