Research Article
LASSO Model Better Predicted the Prognosis of DLBCL than Random Forest Model: A Retrospective Multicenter Analysis of HHLWG
Table 1
The baseline characteristics between the training cohort and the validation cohort.
| Variables | Training cohort | Validation cohort | | | |
| Gender (%) | | | | Male | 451 (53.2) | 203 (55.9) | 0.416 | Female | 397 (46.8) | 160 (44.1) | | Age (year) | 62.00 (52.00, 70.00) | 62.00 (52.00, 70.00) | 0.903 | TC (mmol/L) | 4.32 (3.68, 4.96) | 4.23 (3.67, 4.96) | 0.730 | ALB (g/L) | 38.80 (34.80, 42.80) | 39.00 (34.40, 43.35) | 0.744 | RBC (1012/L) | 4.10 (3.66, 4.47) | 4.06 (3.73, 4.50) | 0.565 | HB (g/L) | 124.00 (108.00, 135.00) | 123.00 (108.00, 138.00) | 0.442 | PLT (109/L) | 217.00 (165.00, 269.00) | 213.00 (154.00, 272.50) | 0.304 | LDH (U/L) | 236.00 (185.00, 404.25) | 233.00 (181.20, 350.00) | 0.328 | Ki-67 | 0.75 (0.60, 0.80) | 0.70 (0.60, 0.80) | 0.915 | B symptom (%) | | | | Absence | 636 (75.0) | 264 (72.7) | 0.449 | Presence | 212 (25.0) | 99 (27.3) | | CNS involvement (%) | | | | Absence | 761 (89.7) | 333 (91.7) | 0.332 | Presence | 87 (10.3) | 30 (8.3) | | BM involvement (%) | | | | Absence | 776 (91.5) | 333 (91.7) | 0.987 | Presence | 72 (8.5) | 30 (8.3) | | Liver involvement (%) | | | | Absence | 808 (95.3) | 343 (94.5) | 0.661 | Presence | 40 (4.7) | 20 (5.5) | | Ann Arbor stage (%) | | | | I/II | 391 (46.1) | 166 (45.7) | 0.954 | III/IV | 457 (53.9) | 197 (54.3) | | NCCN-IPI (%) | | | | LR/LIR | 477 (56.2) | 193 (53.2) | 0.355 | HIR/HR | 371 (43.8) | 170 (46.8) | | IPI (%) | | | | LR/LIR | 529 (62.4) | 221 (60.9) | 0.743 | HIR/HR | 318 (37.5) | 141 (38.8) | | Bulky (%) | | | | Absence | 799 (94.2) | 343 (94.5) | 0.961 | Presence | 49 (5.8) | 20 (5.5) | |
|
|
Note: TC: total cholesterol; ALB: albumin; RBC: red blood cell count; HB: hemoglobin; PLT: platelet; LDH: lactate dehydrogenase; CNS involvement: central nervous system involvement; BM involvement: bone marrow involvement; IPI: International Prognostic Index.
|