Combined pulmonary fibrosis and emphysema (CPFE) is a syndrome consisting of two distinct disease entities: fibrosis predominantly in the lower zones and emphysema in the upper zones [1–3]. Although CPFE includes idiopathic pulmonary fibrosis (IPF) and emphysema, it is considered a separate disease entity with distinct clinical and functional characteristics and different prognoses [3]. CPFE is prevalent in males (mean age, 65 years) with a history of significant tobacco exposure (≥40 pack-years) [4]. It is characterized by progressive worsening of respiratory symptoms with high mortality [1,5]. CPFE is more closely related to lung cancer and pulmonary hypertension (PH) than IPF alone [2,3,6,7]. Pulmonary function tests (PFTs) in CPFE reveal preserved lung volumes but a severe reduction in the diffusing capacity of the lung for carbon monoxide (DLCO) [8]. Moreover, the annual decline rates in DLCO and forced vital capacity (FVC) were significantly lower in the CPFE group [9].

Several large, prospective, randomized controlled trials of anti-fibrotic agents have been conducted in patients with IPF [10–14]. Pirfenidone, an oral anti-fibrotic agent for IPF, has demonstrated efficacy in attenuating the decline in FVC and improving survival in these patients [10]. However, despite a CPFE prevalence of up to 30% [5], the efficacy of anti-fibrotic therapy, including pirfenidone, remains unknown in this population. This knowledge gap exists because milestone studies on anti-fibrotic agents excluded patients with a forced expiratory volume in 1 second (FEV1)/FVC ratio below 0.8, a history of chronic obstructive pulmonary disease (COPD), or a predominance of emphysema over fibrotic changes on high-resolution computed tomography (HRCT) scans, thereby excluding CPFE patients [10–13].

Therefore, we aimed to determine whether antifibrotic therapy improves survival and slows disease progression in patients with CPFE by assessing lung function preservation, similar to its established effects in IPF.

Our study evaluated the effectiveness of antifibrotics in patients with CPFE, a group generally excluded from the landmark clinical trial of antifibrotics. There was no significant survival benefit in patients with CPFE prescribed with antifibrotics. There was no significant transplant-free survival benefit in patients with CPFE prescribed with antifibrotics within a 100-month truncation horizon (RMST difference: −6.87 months, p=0.264), which is consistent with previous study among US-Based cohort study [24]. However, a highly significant physiological benefit was achieved in lung function preservation, markedly slowing the continuous annual rate of DLCO decline (from −4.19±0.37%/year pre-treatment to −2.97±0.30%/year post-treatment, p<0.001).

At 5 years post-diagnosis, the No AFD group exhibited a higher transplant-free survival rate than the AFD group (62.5% vs. 50.4%), whereas a distinct survival reversal was achieved at the ultra-late 10-year milestone, with the AFD group outperforming the control group (31.4% vs. 27.1%). This prominent survival crossover may suggest a potential “late catch-up effect,” which could indicate that the cumulative physiological benefit of early antifibrotic exposure gradually unfolds during extended long-term observation by effectively applying a brake to the underlying disease progression. In the ultra-late phase, while the untreated control group experiences a steeper survival decline, the multi-year stabilizing trends in the antifibrotic group might reflect a deceleration of the long-term attrition rate. However, given that the RMST analysis showed no statistically significant survival difference between the groups (p=0.264), this possible trend requires cautious interpretation and warrants further confirmation in prospective, larger-scale studies.

This potential delayed trend might help counterbalance the high-risk baseline phenotype that inherently characterizes this severe CPFE population. Consequently, our findings underscore that intermediate observations may fail to capture the true long-term utility of these agents. Future prospective trials with extended chronological follow-up are highly warranted to definitively determine whether this observed stabilizing trend eventually translates into a statistically significant survival advantage beyond a decade.

Overall, it can be inferred that DLCO could be interpreted as one essential component of a multidimensional assessment for assessing the preservation of lung function with antifibrotics in CPFE and that FVC is inappropriate for monitoring disease progression consistent with findings from previous studies [25]. However, further studies are needed to find other appropriate surrogates for evaluating the effect of antifibrotics on lung function preservation in CPFE because DLCO is largely affected by the patient's condition and intercenter measurement variability [26].

Several limitations of the present study warrant consideration.

First, Pulmonary hypertension (PH), a critical prognostic determinant in CPFE, could not be directly evaluated within our survival analysis framework.

PH is conventionally defined by a mean pulmonary artery pressure (mPAP) greater than 25mmHg via invasive right heart catheterization (RHC). However, due to its invasive nature, RHC is not routinely performed in clinical practice, particularly for patients with IPF; consequently, exact mPAP data could not be acquired in our cohort. To circumvent this clinical barrier, expert consensus guidelines propose an estimated right ventricular systolic pressure (RVSP) greater than 40mmHg via echocardiography as indicative of “suspected PH” [27]. Nevertheless, in this study, valid longitudinal RVSP assessments were available for only 264 patients, rendering the missing data rate too extensive to construct a robust, direct PH-stratified analysis.

To address this limitation, we evaluated alternative validated physiological surrogates. Previous literature has established that longitudinal deteriorations in DLCO and the FVC/DLCO percent predicted ratio possess powerful direct correlations with secondary PH development and progression in chronic lung diseases [23,28–30] Crucially, our piecewise LMM analysis successfully demonstrated a substantial deceleration and flattening in the continuous annual decline kinetics of both gas exchange parameters following early antifibrotic therapy. This physiological stabilization strongly implies that early antifibrotic exposure may exert a protective hemodynamic brake, potentially translating into a positive long-term impact on secondary PH-related prognostic outcomes. Future large-scale randomized controlled trials (RCTs) incorporating standardized, systematic longitudinal assessments of mPAP or comprehensive echocardiographic protocols are highly warranted to definitively validate these continuous hemodynamic preservation trends.

Second, due to its retrospective nonrandomized design, various inherent biases (such as selection bias and immortal bias) and confounding factors are inevitably present. To mitigate immortal time bias, we performed a landmark analysis alongside a time-dependent Cox proportional hazards analysis. We set a predefined landmark time point at 6 months from the index date to ensure that only patients surviving up to this point were evaluated, thereby preventing an artificial overestimation of survival benefits for the treated group. Furthermore, to mathematically control for the varying timing of treatment initiation, antifibrotic use was treated as a time-dependent covariate in the time-dependent Cox proportional hazards analysis. This approach ensures that a patient's person-time is appropriately classified as “untreated” until the actual date of drug initiation and “treated” thereafter, minimizing potential bias in the estimate of the true treatment effect on survival. Nevertheless, despite these rigorous statistical adjustments and the application of propensity score matching, the inherent retrospective and observational nature of this study implies that residual confounding cannot be completely excluded. Unmeasured or unrecorded clinical variables, which could not be accounted for in our models, may still have influenced the survival outcomes.

Third, utilizing a ≥5% emphysema threshold for the research definition of CPFE may introduce certain limitations. As noted in the consensus statement guidelines, an emphysema extent of less than 5% is unlikely to significantly impact patient physiology or clinical outcomes. Furthermore, such a low threshold is inherently more susceptible to interobserver disagreement during visual quantification on HRCT, which could potentially lead to the misclassification of patients. Consequently, this might have contributed to the heterogeneity of our study population. Future studies incorporating stricter clinical criteria (e.g., ≥15% emphysema) or automated quantitative CT software are warranted to validate our findings.

In conclusion, our study provides robust, bias-controlled insights into the long-term therapeutic utility of early antifibrotic therapy in patients with CPFE within a rigorous 6-month landmark milestone network. Although early antifibrotic intervention does not translate into a statistically significant overall survival duration advantage over a decade due to an explicit late-phase survival cross-over, our findings suggest it may help attenuate the continuous, longitudinal physiological decline of lung function, most notably by preserving DLCO. Nevertheless, it must be emphasized that no statistically significant long-term survival benefit was demonstrated after our adjusted analyses. However, we acknowledge that serial changes in DLco can be highly variable and influenced by confounding factors such as pulmonary vasculopathy, hemoglobin fluctuations, and measurement variations. Consequently, we propose that DLco could be utilized as one of the components of a multi-dimensional assessment—integrated alongside longitudinal chest CT findings, exercise tolerance tests, and subjective clinical symptoms—to comprehensively evaluate disease trajectory and therapeutic responses in patients with CPFE in clinical practice. Ultimately, future prospective multi-center randomized controlled trials utilizing standardized direct hemodynamic assessments are highly warranted to definitively validate these long-term survival and preservation trends.

The findings reported in this manuscript have not been previously presented at any conferences and are not part of a PhD thesis.

This study was approved by the Institutional Review Board of the Severance Hospital (approval number: 4-2019-0031).

There is nothing to disclose.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

There were no conflicts of interest.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.