Fig 1.
Schematic diagram of the workflow followed on the modeling exercise.
Table 1.
Response model and relation with TNM stage during neoadjuvant treatment.
Table 2.
Patients’ characteristics.
Table 3.
Population parameter estimates of response and survival models.
Fig 2.
Results of the visual predictive checks of the model after 500 simulated profiles for the probability of response.
Partial response in green (A), stable disease in blue (B), or progressive disease in red (C). Solid lines represent the observed raw data. Dotted lines correspond to the median of the simulated profiles.
Fig 3.
Results of the visual predictive check as a Kaplan-Meier curve after 500 simulations.
Orange solid lines represent the observed probability of OS. Orange dotted lines represent the median simulated OS probability. Orange shaded areas represent the 95% prediction intervals of 500 simulated datasets.
Fig 4.
Results of the visual predictive checks as Kaplan-Meier curves after 500 simulations showing the probability of survival.
The probability of OS (A) is depicted depending on the appearance (red) or absence (grey) of metastasis. The TNM stage at diagnosis (B) is categorized as <3 (blue) or ≥3 (yellow) to show the effect on OS probability. The probability of OS (C) is shown for patients with linitis (red) or without (green). Solid lines represent the observed OS probability; dotted lines represent the median simulated OS probability and shaded areas represent the 95% confidence intervals of 500 simulated datasets.