Patient Mortality Risk

HOW CAN MORTALITY RISK BE ASSESSED IN PATIENTS WITH IPF?

Several years ago, a baseline-risk prediction model for IPF was developed and validated.¹ The "GAP model" (so called for the measures it uses, Gender, Age, and Physiology) includes variables – patient gender, age, forced vital capacity (FVC), percent predicted, and diffusing capacity for carbon monoxide (DL_CO), percent predicted – that most clinicians can easily obtain.

However, the GAP model does not take into account other aspects of the disease, specifically longitudinal disease behavior, which may be informative regarding risk prediction.² Brett Ley, Williamson Bradford, Derek Weycher, Eric Vittinghoff, et al sought to validate the GAP baseline model in clinical trial cohorts and to assess GAP-based models incorporating additional baseline and longitudinal variables in terms of predictive value.²

WHAT THEY DID

The authors utilized patients with IPF from 3 clinical trials (a clinical trial of interferon γ1b, and 2 clinical trials of pirfenidone) as the source population.² The study cohort included patients in the source population who had a 24-week trial visit (N=1109).

First, the predictive performance (discrimination and calibration) of the original GAP model was evaluated.

Then, all potential models that included the individual GAP predictors plus 1 to 4 additional variables were ranked using a screening procedure. Models with GAP plus novel predictors were compared to the GAP model and a novel longitudinal GAP model was selected, balancing optimum discrimination with practical considerations.

WHAT THEY FOUND

In this cohort, the GAP model had discriminative performance similar to that found in the previous clinical cohorts.²

Significant associations with mortality were found for the following novel predictor variables: respiratory hospitalization in the prior 24 weeks, dyspnea severity as measured by the baseline University of California San Diego Shortness of Breath Questionnaire (UCSD SOBQ) and 24-week change in UCSD SOBQ, baseline 6-minute walking distance (6MWD) and 24-week change in 6MWD, 6-minute walk test desaturation, 24-week change in forced vital capacity (FVC), and 24-week change in diffusing capacity for carbon monoxide.

Multiple models constructed from the GAP model plus multiple additional variables were evaluated for discriminative performance. The simplest model (termed the Longitudinal GAP model) that resulted in the largest improvement in discrimination consisted of the GAP model plus respiratory hospitalization and 24-week change in FVC. Similar to the original GAP model,¹ points are assigned for each variable of the scoring system to obtain a total point score (Table 1).² The total point score is then used to estimate mortality risk for that particular patient (Table 2).

Table 1.

The longitudinal GAP categorical and point index models. Adapted from Ley B, Bradford WZ, Weycker D, et al. Unified baseline and longitudinal mortality prediction in idiopathic pulmonary fibrosis. Eur Respir J. 2015;45(5):1374-1381. doi: 10.1183/09031936.00146314.

DL_CO, diffusing capacity for carbon monoxide; FVC, forced vital capacity; GAP, gender, age, and lung physiology variables (FVC and DL_CO); HR, hazard ratio.

Table 2.

The longitudinal GAP point index models: predicted 1- and 2-year mortality risks. Adapted from Ley B, Bradford WZ, Weycker D, et al. Unified baseline and longitudinal mortality prediction in idiopathic pulmonary fibrosis. Eur Respir J. 2015;45(5):1374-1381. doi: 10.1183/09031936.00146314.

WHAT IT MEANS

The GAP models provide clinicians with a 2-tiered approach to risk prediction in IPF patients: initial assessment of mortality risk using the GAP model and later assessment using the Longitudinal GAP model.

The use of these tools in clinical practice may help guide discussions with patients about life planning, facilitate treatment decisions, and provide a framework for timely referral to lung transplant centers.

REFERENCES