Fig 1.
Flowchart of the participant selection process.
In total, 196 IPF patients who had been treated with nintedanib (NT) were enrolled from April 2017 to March 2022. Patients with missing data at follow-up (n = 9), the absence of pulmonary function test results at baseline (n = 8), the presence of complications of lung cancer at the initial visit (n = 6), and an observation period of less than six months (n = 3) were excluded. Of the 170 patients, 123 were classified as continuing NT for more than 12 months and 47 were classified as discontinuing NT within 12 months. During the observation period, 130 patients discontinued NT as a first-line treatment. Their reasons for discontinuing the treatment were as follows: Disease progression (n = 36), adverse events (n = 53), death (n = 30), self-interrupted (n = 4), lung transplantation (n = 3), and others (n = 4). In addition, 28 of 36 patients were switched from NT to pirfenidone due to disease progression and 25 of 53 due to adverse effects.
Table 1.
Comparison of the characteristics of IPF patients continuing and discontinuing NT.
Table 2.
Treatment status of NT.
Table 3.
Comparison of the tolerability of IPF patients to NT treatment between elderly and non-elderly groups.
Fig 2.
The predicted baseline forced vital capacity % of patients continuing nintedanib during 12 months (stratified by FVC % predicted).
The stratified analysis for %FVC of patients continuing NT for 12 months demonstrated that 2 (18.2%) of them had a baseline predicted FVC <50%, 9 (56.2%) had a predicted FVC between ≥50% and <60%, 19 (73.0%) had a predicted FVC between ≥60% and <70%, 31 (72.1%) had a predicted FVC between ≥70% and <80%, and 62 (83.7%) had a predicted FVC of ≥80%.
Table 4.
Risk factors for discontinuing NT within 12 months.
Fig 3.
Change in FVC before and after treatment with switching anti-fibrotic drugs.
Each line represents one patient. The bold red line represents the mean value. The mean FVC value was 2.46 ± 0.52 L under nintedanib treatment six months prior to switch to pirfenidone, 2.16 ± 0.50 L at the time of the switch, and 2.08 ± 0.55 L at six months after pirfenidone initiation. The decline in FVC was suppressed after switching to pirfenidone using the one-way repeated-measures ANOVA with Bonferroni’s multiple comparison (P = 0.0001, P = 0.3).
Fig 4.
Swimmer plots showing the time course for switching from nintedanib to pirfenidone.
The mean (median) periods for NT and PFD administration were 356 ± 242 (324) days, and 705 ± 305 (673) days, respectively. Thirteen of the 28 patients discontinued the treatment after the switch. The reasons for discontinuation were as follows: [Died: 11 (cases no. 1, 2, 3, 5, 8, 14, 19, 20, 21, 23, 24), Lung transplantation: 1 (case no. 18), Liver injury (as an adverse effect): 1 (case no. 13)]. The causes of death were respiratory failure due to the worsening of the present illness in five patients, AE in four patients, pneumonia in one patient, and lung cancer in one patient.
Table 5.
Logistic regression analysis for verifying predictors of deterioration after switching from NT to PFD.