Table 2.
Hazard ratios for the association of BMI with OS and PFS (along with interactions) and odds ratios for the association of BMI with durable clinical benefit among patients treated with anti-PD-1 based checkpoint inhibition
| Outcome (N = 136)b | Effect Estimate | |||
|---|---|---|---|---|
| Overall Survival (OS) | Unadjusted Hazard Ratio (95% CI) | p-value | Adjusted Hazard Ratioa (95% CI) | p-value |
| Overweight/Class-I Obese vs. Normal Weight (reference) | 0.41 (0.24–0.72) | 0.002 | 0.26 (0.1–0.71) | 0.008 |
| Class-II/III Obese vs. Normal Weight | 0.88 (0.41–1.91) | 0.756 | 0.42 (0.1–1.77) | 0.238 |
| Interaction Modelc 1 (OS) | ||||
| Interaction of BMI (Overweight/Class-I vs. normal weight) with gender (males vs. females) | – | – | 0.19 (0.04–0.95) | 0.044 |
| Interaction Model 1 (OS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in females | – | – | 0.56 (0.16–1.89) | 0.346 |
| Interaction Model 1 (OS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in males | – | – | 0.11 (0.03–0.4) | 0.001 |
| Interaction Model 2 (OS) | ||||
| Interaction of BMI (Overweight/Class-I vs. normal weight) with serum creatinine (> = 0.9 mg/dL vs. < 0.9 mg/dL)c | – | – | 0.11 (0.02–0.7) | 0.020 |
| Interaction Model 2 (OS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine < 0.9 mg/dL | – | – | 0.43 (0.15–1.24) | 0.119 |
| Interaction Model 2 (OS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine > = 0.9 mg/dL | – | – | 0.045 (0.08–0.262) | 0.001 |
| Progression Free Survival (PFS) | ||||
| Overweight/Class-I Obese vs. Normal Weight (reference) | 0.5 (0.31–0.81) | 0.005 | 0.43 (0.19–0.95) | 0.038 |
| Class-II/III Obesity vs. Normal Weight | 1.03 (0.53–2) | 0.932 | 1 (0.34–2.94) | 0.991 |
| Interaction Model 1 (PFS) | ||||
| Interaction of BMI (overweight/Class-I vs. normal weight) with gender (males vs. females) | – | – | 0.29 (0.07–1.18) | 0.084 |
| Interaction Model 1 (PFS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in females | – | – | 0.80 (0.27–2.41) | 0.695 |
| Interaction Model 1 (PFS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in males | – | – | 0.23 (0.08–0.66) | 0.006 |
| Interaction Model 2 (PFS) | ||||
| Interaction of BMI (overweight/Class-I vs. normal weight) with serum creatinine (> = 0.9 mg/dL vs. < 0.9 mg/dL) | – | – | 0.17 (0.04–0.77) | 0.021 |
| Interaction Model 2 (PFS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine < 0.9 mg/dL | – | – | 0.71 (0.29–1.77) | 0.464 |
| Interaction Model 2 (PFS) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine > = 0.9 mg/dL | – | – | 0.12 (0.03–0.45) | 0.002 |
| Durable Clinical Benefit (N = 132)b | Unadjusted Odds Ratio (95% CI) | p-value | Adjusted Odds Ratiod (95% CI) | p-value |
| Overweight/Class-I Obese vs. Normal Weight (reference) | 2.76 (1.18–6.46) | 0.020 | 11.4 (1.65–78.6) | 0.013 |
| Class-II/III Obesity vs. Normal Weight | 0.91 (0.23–3.54) | 0.891 | 3.3 (0.22–50.5) | 0.391 |
| Interaction Model 1 (DCB) | ||||
| Interaction of BMI (overweight/Class-I vs. normal weight) with gender (males vs. females) | – | – | 10.39 (0.17–634.68) | 0.265 |
| Interaction Model 1 (DCB) | ||||
| Overweight/Class-I Obese (vs. normal weight) in females | – | – | 2.91 (0.16–54.05) | 0.473 |
| Interaction Model 1 (DCB) | ||||
| Overweight/Class-I Obese (vs. normal weight) in males | – | – | 30.27 (2.01–455.72) | 0.014 |
| Interaction Model 2 (DCB) | ||||
| Interaction of BMI (overweight/Class-I vs. normal weight) with creatinine (> = 0.9 mg/dL vs. < 0.9 mg/dL) | – | – | 2.82 (0.03–229.67) | 0.644 |
| Interaction Model 2 (DCB) | ||||
| Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine < 0.9 mg/dL | – | – | 8.03 (0.85–76.27) | 0.070 |
| Interaction Model 2 (DCB) | ||||
| Main effect (Serum Creatinine > = 0.9 mg/dL): Overweight/Class-I Obese (vs. normal weight) HR in patients with serum creatinine > = 0.9 mg/dL | – | – | 22.66 (0.51–1013.5) | 0.108 |
a Adjusted for the following covariates: age (<=45, > 45–75 and > 75 years), gender, serum creatinine (< 0.9 and > = 0.9 mg/dL), treatment (monotherapy/combination), current drinker (vs. non-current/never-drinker), smoking history (ever vs. never), KPS (<=70 and > 70), LDH (<=231 vs. > 231 U/L), stage at baseline, Charlson’s score (< 10 vs. > = 10), hemoglobin (< 11.5 vs. > = 11.5 g/dL), ANC (<=8 and > 8 K/uL), ALC (< 3 and > =3 K/uL), albumin (< 3.5 g/dL), autoimmune disease, diabetes, CV disease, CKD, BRAF mutation, NRAS mutation, hypertension, hyperlipidemia, comedications (anti-platelet agents, statins, metformin, ACE/ARB inhibitors), prior treatments (immunotherapy/CTLA-4, radiation, chemotherapy and targeted therapy), fasting glucose (<=110 vs. > 110 mg/dL), type of melanoma,; N = 127 after excluding three patients who were underweight, 8 patients with missing LDH values and 1 patient whose KPS could not assessed (missing)
b Three underweight patients were excluded
c Interactions were studied in separate models along with main effects and adjusted for the same covariates as the main models
d Adjusted for the same covariates listed for OS and PFS (except KPS which was included as a continuous variable); N = 123 (excluding 3 underweight patients; DCB was not assessable/available for 4 patients; patients with missing data for LDH, KPS were not included for the adjusted analysis)