1. Baseline characteristics of patients
As presented in Table 1, a total of 184 patients with IPAF were enrolled from January 2014 to January 2019. Among them, 81 (44.0%) patients were in the pirfenidone group and 103 (56.0%) patients were in the control group. There were no significant differences in gender, BMI, smoking history, arterial oxygen partial pressure and oxygen saturation. Both FVC% and DLCO% in the pirfenidone group were lower than those in the control group (P < 0.001 and P = 0.003). 151 (82.1%) patients were treated with oral glucocorticoid, and 13 (7.1%) patients were treated with immunosuppressants. There were no differences in the treatment of glucocorticoid and immunosuppressants between pirfenidone and control group. The mean duration of follow-up was 15.0 months, and no significant difference was observed between the pirfenidone group (14.6 months) and the control group (15.4 months). The range of pirfenidone dose was 600-1800 mg/day, with a median dose of 1492 mg/day. The duration of pirfenidone treatment was 3-44 months, with a mean duration of 14.4 months.
2. Diagnostic characteristics of IPAF patients
The diagnostic characteristics of all the IPAF patients were shown in Table 2. Overall, 66(35.9%) patients met IPAF criteria through a combination of serological and morphological domains, 53 (28.8%) by clinical and morphological domains, 34 (18.5%) by clinical and serological domains and 26 (14.1%) by all the three diagnostic domains.
A breakdown of features within each IPAF domain showed that the most common clinical findings were Raynaund’s phenomenon (49, 26.6%) and inflammatory arthritis or polyarticular morning joint stiffness lasting >60min (45, 24.5%). 131 patients had positive serum autoantibody (71.2%), of which, 51 cases had two or more positive antibodies. An ANA ≥ 1:320 (or nucleolar or centromere pattern of any titer) was the most common serological finding (81, 44.0%). Morphological features were all obtained from HRCT. Within the morphological domain (150 81.5%), NSIP pattern by HRCT was found in 62.0% (114), while organizing pneumonia (OP) pattern was found in 14.1% (26). There were no significant differences in any of the diagnostic characteristics between the pirfenidone group and control group.
3. Arterial oxygen saturation and CT scan of chest
The differences after the treatment in chest CT scan and arterial oxygen saturation (SaO2% Figure 2) were analyzed. Although SaO2% of pirfenidone group tend to improve in 12 months (month 0 (95.48%±2.43%) vs. month 12 (96.51%±1.49%), P=0.098), no statistical significance was observed (Figure 2C). The SaO2% at other follow-up time points had no significant differences, when compared with the baseline SaO2% (m0). In addition, the changed SaO2% between two groups had no differences (Figure 2D).
According to the evaluation criteria, the changes in CT scan of chest were divided into four groups: complete remission, partial remission, stabilization, and progression. The representative CT images are shown in Figure 3A. There were no significant differences in changes of chest CT scan between the two groups (Figure 3B).
4. Changes in pulmonary function
The differences in FVC% (Figure 4A) and DLCO% (Figure 4B) between the two groups were compared at the time of 3, 6, 12, 18, and 24 months with or without pirfenidone treatment. We found that after 12 months of treatment, FVC% in pirfenidone group was improved by 10.44%±18.83%, while that was decreased in the control group by 1.18%±6.96%, and the difference was statistically significant (P=0.016). In addition, FVC% was significantly increased in pirfenidone group after 6 (P=0.047) and 24 months (P=0.040) of treatment, compared with that of control group. The improvement of DLCO% in pirfenidone group was higher than that in the control group at 6 months (P=0.043).
Considering the bias in the retrospective research, we used mixed-effect model to estimates the differences of FVC% (Figure 4C) and DLCO% (Figure 4D) in the two groups. After adjusted by the sex, age, baseline FVC% and DLCO%, we found that compared with the control group, changed FVC% in pirfenidone group tend to improve by 1.49% (95% CI (0.14%, 2.84%)) within 40 months (χ2 (1) =4.59, P=0.032), however there was no difference in the change of DLCO% (χ2 (1) =0.49, P=0.48).
5. Subgroup analysis of pulmonary function
Patients in the two groups were subdivided according to different conditions. The average annual changes in absolute value of FVC within 40 months are presented in Table 4. The volume of FVC (liters) in the pirfenidone group increased by 0.0390 L/year on average, while that decreased by 0.0769 L/year on average in the control group (P=0.038). In patients with FVC < 70%, the improvement of FVC in the pirfenidone group significantly increased, compared with the control group (P=0.021). In patients with FVC > 70%, there was no significant difference in changes of FVC between the two groups. Patients in the pirfenidone group were subdivided according to the dosage and length of pirfenidone treatment. The results showed that the improvement of FVC significantly increased in patients with dose of pirfenidone > 600 mg/day (P=0.010) and total medication time > 12 months (P=0.007) compared with patients in control group. In addition, for patients with dose of prednisone > 20 mg combined with the medication of pirfenidone, the improvement of FVC was higher than that of the control group (P=0.031). According to the diagnostic characteristics, the changes of FVC increased in the pirfenidone group, who were diagnosed on the basis of morphological and serological domains (P=0.033), whereas there were no significant differences among the other three groups analysis in terms of the diagnostic characteristics.
6. Administration of pirfenidone with IPAF patients can reduce the dose of prednisone after 12 months.
We compared total dose (Figure 5A) and average daily dose (Figure 5B) of prednisone in the two groups. There was no significant difference of total dose and daily dose in two groups during the whole follow-up of 40 months. But we found that both total dose (P=0.012) and daily dose (P=0.032) of prednisone were less in pirfenidone group than those of control group after 12 months of treatment. Then, mixed-effect model was used to analyze differences of the prednisone dose. We found that during the whole follow-up of 40 months (χ2 (1) =0.7318, P=0.392, pirfenidone n=53, control n= 47) and initial 12 months (χ2 (1) =0.1396, P=0.709, pirfenidone n=53, control n= 47), there was no significant difference in dose of prednisone (Figure 5C&D) between two groups. However, prednisone dose was reduced by 6.27 mg per day in pirfenidone group in 12-40 months (Figure 5E), compared with control group (χ2 (1) =9.8385, P=0.002, pirfenidone n=34, control n= 27).
7. Side effects of pirfenidone
Analysis of side effects showed that 17 (19.32%) patients had side effects after taking pirfenidone (Figure 6A). Among them, 7 (7.95%) patients stopped medication due to side effects and 1 patient had anaphylactic shock. Skin rash (10.23%) and liver damage (5.68%) were the most common side effects, which were similar to those of IPF patients who were treated with pirfenidone [12]. Additionally, 14 (14/17, 82.35%) patients experienced side effects at the initial dose (600 mg Figure 6B), and 3 (17.65%) patients experienced side effects during the increasing dose of pirfenidone (Figure 6B).