Between January 2004 and November 2016, 24 suspected patients with clinico-pathologic-radiologic features of BHD syndrome were screened at Asan Medical Center, Seoul, South Korea. The diagnosis of BHD syndrome was based on the criteria proposed by the European BHD consortium [8]. The diagnosis was made when patients fulfilled at least one major or two minor criteria. The major criteria were as follows: (1) at least five fibrofolliculomas or trichodiscomas (at least one histologically confirmed, of adult onset); (2) pathogenic FLCN germline mutation. The minor criteria were as follows: (1) multiple lung cysts (bilateral basally located lung cysts with no other apparent cause, with or without spontaneous primary pneumothorax); (2) renal cancer (early onset before age 50 or multifocal or bilateral, or mixed chromophobe and oncocytic histology); (3) a first-degree relative with BHD [8]. Based on these criteria, 12 patients were diagnosed with BHD syndrome (10 confirmed by direct sequencing of the FLCN gene and two confirmed by clinical diagnosis) (Fig. 1). Other multiple cystic lung disorders, such as lymphangioleiomyomatosis (LAM), were excluded by characteristic high-resolution computed tomography (HRCT) and/or histologic findings. For example, the diffuse thin walled cysts with 2 to 5 mm-sized, round or ovoid shape and surrounded by normal lung without regional sparing on HRCT images, and smooth-muscle cell proliferation in the pulmonary interstitium affecting vessels, airways, lymphatics, alveolar septa, and pleura on the histology supported the diagnosis of LAM [9].

Board of Asan Medical Center (approval number: 2016-0744) and informed consent was waived because the data were analyzed anonymously and retrospectively.

Clinical and survival data for all patients were obtained from medical records, telephone interviews, and/or the records of the National Health Insurance of Korea. All clinical parameters and lung function data were obtained within 1 month after the initial visit.

Spirometry, total lung capacity by plethysmography, and diffusing capacity of the lung for carbon monoxide were measured according to the recommendation [10-12] and the results were expressed as percentages of normal predicted values. All patients underwent HRCT scans at their first visit. Abdominopelvic computed tomography (CT) or kidney ultrasonography was per formed in patients who were first diagnosed with BHD to evaluate renal lesions. Patients who had skin lesions were referred to dermatologists and underwent punch biopsy.

Chest CT images were reviewed by one chest radiologist (E.J.C.). The number, distribution, size and shape of pulmonary cysts were assessed on both axial and coronal reformatted images. The number of pulmonary cysts was classified into three groups: less than 50, 50 to 100, and more than 100. The distribution of pulmonary cysts was classified as upper dominant, lower dominant, or diffuse which was divided by the level of carina. The distribution was also classified as central dominant, peripheral dominant, or diffuse which was divided by the midline between hilar structure and costal pleural surface. The relation of the cysts to visceral pleura and interlobular fissures was also analyzed. Cyst size was measured along the long axis of the cysts. Cyst shape was classified as round, oval, lentiform, or irregular; round cysts were defined as having a smooth surface and a diameter ratio of long to short axes ≤ 1.1, oval as having a diameter ratio of long to short axes > 1.1, lentiform as being convex on both sides meeting at an acute angle, and all others as being irregular-shaped.

Genomic DNA was extracted from peripheral blood leukocytes. Eleven coding exons (exons 4 to 14) and exon-intron boundaries of the FLCN gene were amplified by polymerase chain reaction (PCR) using the primers designed by Primer3 (Supplementary Table 1) [13]. After PCR amplification, the PCR products were purified and directly sequenced using an ABI3130x1 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). DNA sequences were analyzed according to GenBank reference DNA sequences NG_008001.2 (genomic DNA), NM_144997.5 (mRNA), and NP_659434.2 (protein) (http://www.ncbi.nlm.nih.gov).

All values were given as the mean ± standard deviation for continuous variables or as percentages for categorical variables. Student’s t test or Mann-Whitney U test was used for continuous data, and Pearson’s chi-square test or Fisher exact test was used for categorical data. All statistical analyses were performed using SPSS version 21.0 (IBM Co., Armonk, NY, USA). All reported p values are two-sided, and p < 0.05 was considered as significant.

The median follow-up period was 52 months (range, 1 to 149). The mean age was 41.3 years, 66.7% were female, and 58.3% were ever-smokers (Table 1). Pneumothorax was the most common initial manifestation in seven patients (58.3%), followed by asymptomatic screening (41.7%).

A family of seven patients (63.6%, among total 11 non-family patients) had suspected clinical features of BHD syndrome (Table 1). Among total 12 patients, two (patient no. 2 and 9) were sisters; the older sister was diagnosed after the second event of pneumothorax and the younger sister underwent diagnostic evaluation by genetic counseling. Their other sister and father, not included in this study, also had histories of pneumothorax; however, they did not undergo diagnostic evaluation. The other patient (no. 4) had five family members with a history of pneumothorax. Four patients (36.4%) had family histories of colon cancer.

Various comorbidities were found in these patients (Table 2). Thyroid nodules were detected in three patients (25%), one of whom (no. 10) underwent a right thyroid lobectomy that revealed nodular hyperplasia. Extrapulmonary cystic lesions (liver, ovary, breast, and spleen) were also found in three patients (25%), and arteriovenous malformation in brain, early gastric cancer, colon polyps (tubular adenoma and hamartoma), and a gall bladder polyp were found in one (no. 4).

Five different mutations of the FLCN gene were confirmed by genetic testing in 10 patients (Table 3). The most frequent mutation (40%) was the duplication of cytosine in the C8 tract of exon 11 (c.1285dupC), as was reported in previous studies [2,14-17]. One patient (no. 8) harbored a novel heterozygous sequence variant of c.31T>C (p.C11R) in exon 4 that replaces a cysteine with an arginine at codon 11.This mutation was not found in the gnomAD browser (http://gnomad.broadinstitute.org/) and was predicted to be probably damaged by Poly-Phen-2 (http://genetics.bwh.harvard.edu/pph2/).

All 12 patients had multiple pulmonary cysts (Table 4). The total number of pulmonary cysts varied, and a number of 50 to 100 cysts per patient was the most common (50%). Cyst size varied from 3 to 70 mm. The most dominant cyst shape was irregular in 10 patients (83.3%), followed by lentiform (n = 1, 8.3%) and round (n = 1, 8.3%), respectively. The cysts were distributed in the predominantly lower and peripheral regions of the lungs (Fig. 2). Most of the patients had fissural (91.7%) and subpleural (100%) cysts. Pneumothorax developed in eight patients (66.7%) and 75% were recurrent (up to five times during follow-up). The mean age of the initial episodes of pneumothorax was 29.8 ± 8.4 years. The number, maximal size, and distribution of cysts, along with patient age and smoking status were not associated with pneumothorax.

Most of the patients had normal lung function (58.3%); however, mild restrictive pattern was also detected in 33.3% (Table 5). Bronchoalveolar lavage (BAL) was performed in two patients, and BAL fluid analysis showed nonspecific results. Wedge resection was performed in four patients due to pneumothorax. The specimen showed multiple small intraparenchymal cysts rimmed by thin, fibrous walls or normal pulmonary parenchyma (Fig. 2).

Simple cysts were detected in two patients (16.7%) and renal cancers in three (25%), two of whom carried an FLCN mutation in exon 11. The mean age at the initial diagnosis of renal cancer was 54 ± 4 years. All three patients underwent partial nephrectomy. The histologic type was chromophobe in two patients (no. 6 and 7) and oncocytic in one (no. 4). Patient no. 4 had a 0.9-cm renal nodule that was difficult to differentiate at the time of diagnosis. Ten years later, the size of the nodule had increased from 0.9 to 1.2 cm, and oncocytoma was confirmed by surgery. In patient no. 6, a 1.2 cm-sized renal cancer developed 4 years after the diagnosis.

Skin lesions were detected in six patients. Most of the lesions presented as small, dome-shaped, whitish papules visualized on the face, neck, and/or upper trunk. Skin biopsies were performed in five patients. Fibrofolliculoma was confirmed in two patients (16.7%), angiofibroma in one (8.3%), and mucinosis and milia, which do not commonly present in BHD syndrome, in two (16.7%).

During follow-up, lung function remained almost unchanged (Fig. 3). Pneumothorax recurred in two patients (16.7%; no. 1 and 2). One patient (no. 1) had repeated recurrence of pneumothorax even though he had undergone bullectomy twice. Another patient (no. 2) had already had a history of pneumothorax twice; however, he experienced three more pneumothoraces during follow-up. One patient (no. 8) developed breast cancer 3 years after the diagnosis of BHD syndrome and underwent breast conserving surgery, followed by adjuvant chemotherapy and radiotherapy. No one died during the follow-up.