In this retrospective study, we characterized a cohort of FPF patients in Northern Finland. In addition to clinical information, HRCT scans and histopathology were re-evaluated and classified according to the present international guidelines [22]. Furthermore, the functionality of the survival prediction models was also tested. We found that there was no male or smoking predominance in FPF patients, that both the radiological and histopathological findings were diverse, and that prediction models could be used for FPF patients. To the best of our knowledge, this study of 68 FPF patients is one of the largest published Northern European cohorts to date and included re-evaluated HRCT scans from all the patients.

Our results supported earlier studies showing that disease onset in FPF patients occurred earlier than that usually observed in IPF patients, at least in men [4, 19]. In the study of Krauss et al., the FPF patients were 58.3 years old, and in the study of Bennet et al., the men with familial IPF were 58.5 years old at the time of diagnosis [5, 19]. In our study, the men were significantly younger than the women (61.7 and 68.6 years, respectively; p<0.003), and the mean age was 64.4 years. Similarly, in the study of Froidure et al., the median age of the FPF population was 65 years [27]. In our study, age was not associated with survival, which may be explained by the fact that men had more often UIP pattern although it did not reach significant difference. Similarly to the results of Bennett et al. and Hodgson et al., there was no predominance of males in our study [19, 20]. According to both Krauss et al. and Steele et al., more than fifty percent of the patients were former or current smokers, and Kropski et al. suggested smoking as a risk factor for FPF [2, 5, 28]. In the study of Froidure et al., more than 75% of FPF patients had a smoking history, but they had fewer pack-years than the IPF group [27]. Cutting et al. reported that familial IPF patients had a shorter smoking history than sporadic IPF patients, which is in line with our study in which 54.4% of the patients were nonsmokers [4]. In our study, however, the nonsmokers were predominantly women (67.7% vs 43.2%, p = 0.043).

We were able to find only two previously published European studies on comorbidities in FPF patients [5, 19]. In both of the abovementioned studies, there were fewer patients than in our study (27 and 46 patients, respectively). Our results showed that hypercholesterolemia, CAD, hypertension and gastroesophageal reflux disease were present in at least 20% of the patients, similar to the findings of Bennett and coauthors. Interestingly, nineteen percent of the FPF patients in our study suffered from asthma, which was not observed in previous studies. In the recently published study on IPF, the most common pulmonary comorbidities were COPD (37%) and lung cancer (3%), while the most common nonpulmonary comorbidities were GERD, dyslipidemia and hypertension [15]. We found no COPD patients and only one patient with suspected lung cancer, which is a unique finding and may be due to the predominance of nonsmokers in our study, although the prevalence of nonpulmonary comorbidities in our study was similar to that in Lee’s study [15]. As in our findings, CAD and hypertension were the most common comorbidities in Finnish IPF patients regardless of their smoking status [29].

Earlier studies have shown that although IPF is the most common clinical phenotype in FPF, the radiological patterns are diverse. Lee et al. reported that UIP was the most common HRCT pattern in 22% of the cases [30]. while in the study of Bennet et al. UIP was found in 54% of the patients [19]. Cecchini et al studied 44 FPF patients, including 26 with shortened telomeres and 7 with telomere-related gene variants revealing that probable UIP was the most common pattern in the shortened telomere group, while indeterminate for UIP was the most common pattern in the group with telomeres >10th percentile [10]. In the study of Diaz de Leon et al., 39 CT scans of patients with TERT gene variants with pulmonary fibrosis were evaluated; 74% had a typical UIP pattern, and the other patterns were consistent with UIP without honeycombing and atypical UIP [31]. By using the IPF guideline 2022, we also observed diverse radiological findings although the findings of our study were not straightforwardly comparable to those of other studies due to the different classification criteria. Interestingly, 13.2% of our patients exhibited PPFE in a single or combined pattern, which has not been shown in previous studies.

According to the study of Cecchini et al., IPF was the most common clinical diagnosis, although CHP was more common in patients with telomeres > the 10th percentile [10]. Our clinical diagnoses were similar to those of Cutting et al., who reported IPF as the predominant clinical diagnosis, followed by unclassifiable PF [4]. Six patients in our study with shortened telomeres had UIP or combined UIP with PPFE patterns. Indeterminate for UIP was the most common pattern in histologic samples (8/47.1%), UIP (7/41.2%), and of probable UIP and alternative diagnoses, one of each was found.

Cecchini et al found that patients with shortened telomeres in particular had atypical histopathological findings [10]. On our histological re-evaluation, we found that unusual combinations of histological patterns existed in some patients. Interestingly, there were histological features of two patients with the TERC gene variant, one of which presented with a combination of UIP and constrictive bronchiolitis and the other with a combination of PPFE and granulomatous inflammation with fibrosis. Among 29 TERT gene variant patients, Diaz de Leon et al. reported a histologic UIP pattern in 89% of the patients but also additional histologic features, including chronic inflammation, scattered histiocytes and non-necrotizing granulomas [31]. Cecchini and coauthors analyzed the histological patterns of patients with shortened telomeres and revealed that although the UIP pattern was the most common, other patterns also existed, similar to the results of our study [10]. van Batenburg and coauthors compared the extent of inflammatory cells and fibrosis in the lung tissues of FPF and IPF patients and found no differences [32]. According to our reanalysis of histological features, however, in some cases, the extent of inflammation was high. In our experience, histological diversity may cause diagnostic problems and difficulties in classification, but it can also suggest FPF in an appropriate clinical context.

Cutting et al showed that FPF patients had shorter survival than patients with the sporadic form of the same clinical diagnosis, while familial IPF was associated with the shortest survival [4]. In our study, the median survival time of the deceased or lung-transplanted patients was 39.9 (20.5‒59.3) months, which was nearly the same as that of the IPF group in the studies of Newton et al. (3.75 years), in the TERT gene variant group of Diaz de Leon et al. (3 years, mean) and the TERT or TERC gene variant group of Borie et al. (4.2 years) [31, 33, 34]. The overall estimated survival, 78.3 (71.0‒85.6) months, was slightly worse than the survival reported by Bennett et al. for FPF patients (7.31 years) [19]. Planas-Cerezales et al. investigated the GAP index in patients with shortened telomeres and reported that patients younger than 60 years had the worst survival, although they were categorized into the GAP I stage [11]. In contrast, we found that the GAP I group had better survival than the combined GAP II and III groups. Furthermore, the CPI was significantly related to survival as a median value of 38 was set as the cut-off point.

Limitations of the study include its retrospective nature and small sample size; however, this study is one of the largest FPF studies in Northern Europe. Because the FPF does not have a specific ICD-10 code, the cohort might not be completely comprehensive, and some patients may be missing. Missing genetic testing information for most of the patients is a weakness of our study. Twenty-four patients underwent genetic testing, which was due to the clinical practice of the hospital and the retrospective nature of the study protocol; these patients included FPF patients from the year 2000 onwards. Systematic genetic testing of FPF patients started in our hospital in 2017; therefore, most of the patients were not tested due to the point of time of their illness. A few PFTs were missing, but HRCT was available from every patient. A quarter of the patients underwent histological investigation of lung tissue, which enabled us to perform reanalysis of histological features in a subset of the patients. The strengths of this study were the accurate collection of data and the radiological and histopathological re-evaluations provided by expert physicians.