Adrenocortical Carcinoma as a Manifestation of Birt-Hogg Dubé Syndrome

Abstract

Birt Hogg Dubé syndrome (BHDS) is characterized by fibrofolliculomas, trichodiscomas, renal tumors, lung cysts, and spontaneous pneumothoraces. Other than renal cell carcinoma, no cancers have been confirmed as part of the BHDS tumor spectrum. Here, we set out to uncover if adrenocortical carcinoma (ACC) is associated with BHDS. A retrospective review at a single tertiary care center between 2010–2024 was performed using search terms describing the clinical manifestations of BHDS, adrenal tumors, and iterations thereof. Initial review identified 1010 patients with ACC. Of these patients, 466 were evaluated by a genetic counselor, and 346 underwent genetic testing – seven of which were found to have either characteristic clinical features of BHDS or a confirmed diagnosis. Literature review over the same time period demonstrated five additional cases of malignant adrenal tumors in patients with BHDS. We observed BHDS, confirmed by a pathogenic FLCN variant in 0.7% of all ACC patients, or 1.5% of patients evaluated by a genetic counselor and 2.0% of all patients who underwent genetic testing. Testing for FLCN pathogenic variants should be considered during routine genetic evaluation of patients with ACC and special attention should be paid to adrenal tumors in BHDS patients during surveillance for renal neoplasms as more than 50% of patients in this series developed metastatic disease.

Introduction

In 1977, a family with multiple fibrofolliculomas, trichodiscomas and acrochordons were described in what would become the first report of Birt-Hogg-Dubé Syndrome (BHDS) (1). Later BHDS was defined as an autosomal dominant tumor predisposition syndrome caused by germline pathogenic variants in the folliculin gene (FLCN), resulting in the distinctive skin findings along with pulmonary cysts, spontaneous pneumothoraces, and renal tumors. While identification of a heterozygous pathogenic variant in FLCN provides a definitive diagnosis, the diagnosis can be made clinically and is established with one major criterion (>5 fibrofolliculomas or trichodiscomas; >1 must be proven on biopsy) or two minor criteria (multiple lung cysts, with or without spontaneous pneumothorax, renal cell cancer: diagnosed <50 years, multifocal or bilateral, mixed chromophobe or oncocytic histology, first-degree relative BHDS) (2). Currently, there are no other recognized phenotypic features of BHDS, although multiple reports have suggested colon cancer, lipomas, parathyroid adenomas, thyroid nodules, and parotid oncocytomas as potential manifestations of this syndrome (3).

While the majority of adult-onset adrenocortical carcinomas (ACC) are apparently sporadic, approximately 5–10% are associated with familial cancer syndromes, including Li-Fraumeni syndrome (LFS), Lynch syndrome (LS), Multiple Endocrine Neoplasia Type I (MEN1), and rarely, Familial Adenomatous Polyposis (FAP) (4, 5). There have also been case reports of patients with neurofibromatosis type 1 and Carney complex with ACC. Given the high prevalence of germline pathogenic germline variants in pediatric ACC (50–80% LFS and rarely Beckwith-Widemann) it has long been recommended that all children with ACC be at minimum screened for TP53 germline variants. (5, 6). To further our understanding of the prevalence of ACC in patients with BHDS, we reviewed patients with ACC at our institution along with the current literature.

Materials and Methods

Patients were identified using search terms describing the clinical manifestations of BHDS, adrenal tumors, as well as iterations thereof using EMERSE (7). Results were filtered for the main providers caring for patients with ACC. The total number of patients with ACC was estimated using provider names and search terms describing ACC. The number of patients with ACC or pancreatic cancer evaluated by a genetic counselor and the total number tested were derived from a local cancer genetics registry (HUM00226722). NM_114997.7 was used as a reference transcript and ClinVar was accessed for review of variant interpretation and the Genome Aggregation Database (gnomAD v4.1.0) was used as a reference population database. Studies were approved by the local institutional review board (HUM00091004, HUM00226722). For statistical analysis of prevalences, Fisher’s exact test was used.

Results

Review of electronic medical records between January 2010 and June 2025 using the above-described parameters identified 1010 patients with ACC. A genetic counselor met with 466 of these patients, of which only 346 underwent genetic testing. While originally pursued via single gene testing or limited panels, standard of care are now large next generation sequencing panels; however, FLCN was only included for 68 patients.

Ten patients were found to have characteristic clinical features of BHDS or a confirmed pathogenic variant in FLCN. Three were excluded – the first had a personal history of renal and thyroid cysts along with spontaneous pneumothoraces and a pathology-confirmed ACC but opted not to pursue genetic testing, the second underwent somatic tumoral testing that demonstrated loss of exons 10–14 FLCN but without confirmation of a germline alteration, and the third had a personal history of pulmonary and renal cysts with a pathology-confirmed ACC but did not elect genetic testing.

Of the remaining seven patients, two were previously published in isolated case reports with confirmed FLCN pathogenic variants and adrenocortical tumors (8–10). This left five new patients with adrenal tumors and germline FLCN mutations (Table 1).

Table 1:

Clinical Characteristics of Adrenal Tumors in Patients with Birt-Hogg-Dubé Syndrome

	AD	Sex	Race	BC*	Imaging	BHDS Features/FH	FLCN Variant	gnomAD	Histology	Outcome
1	50	F	C	NF	>15 cm L adrenal mass	Personal fibrofolliculomas	c.490del (p.Arg164fs) Pathogenic	Not Reported	ACC, Sarcamoid High grade by MI (160 mitoses/HPF)	Surgery, radiation. Maintained on mitotane and chemotherapy (EDP); wide-spread metastatic disease
2	42	M	C	NF	>12 cm R adrenal mass	None	c.113del (p.Ser38fs) Pathogenic	AC: 1 AF: 6.20e-7 AN: 1614000	ACC, Clear Cell High grade by Ki-67 (40%)	Surgery, radiation. Maintained on mitotane and chemotherapy (EDP); abdominopelvic metastatic disease
3	37	F	C	NF	>10 cm L adrenal mass	Personal fibrofolliculomas FH: Spontaneous PNX (mother)	c.296del (p.Asp99fs) Pathogenic	Not Reported	ACC, Sarcamoid/Mixed Low grade by Ki-67 (6.5%)	Surgery. Local recurrence (surgery, radiation, chemotherapy (EDP)). Maintained on mitotane; no metastatic disease
4 #	62	F	C	NF	>6 cm R adrenal mass	Personal trichodiscoma, pulmonary cyst	c.1252del (p.Leu418fs) Pathogenic	Not Reported	Uncertain malignant potential+ Low grade by Ki-67 (5%)	Surgery, radiation; no metastatic disease
5 ^	38	M	C	NF	>20 cm L adrenal mass	Personal pulmonary cysts	c.1285dup (p.His429fs) Pathogenic	AC: 111 AF: 6.88e-5 AN: 1612356	ACC, NOS High grade by Ki-67 (40%)	Surgery. Mitotane and chemotherapy (EDP) transitioned Lutetium 177 Dotatate; widespread metastatic disease
6	31	M	C	NF	>3 cm R adrenal mass	Personal fibrofolliculomas, spontaneous PNX	c.1539-?_c.1740+?del (EX11_3’UTRdel) Likely Pathogenic	Not Reported	ACC, NOS Low grade by MI (<20/HPF)	Surgery; no metastatic disease
7	65	F	C	NF	>15 cm L adrenal mass	Personal hepatic cysts	c.889_890delGA (p.E297Kfs*6) Pathogenic	Not Reported	ACC, Oncocytic Low grade& by Ki-67 (10–20%)	Chest wall and abdominal LN metastases on presentation. Chemotherapy (EDP)

Abbreviations: ACC (adrenocortical carcinoma), AC (allele copy), AD (age of diagnosis), AF (allele frequency), ANF (female), BHDS (Birt-Hogg-Dubé Syndrome), C (Caucasian, European heritage), EDP (etoposide, doxorubicin, cisplatin), FH (family history), HPF (high-powered field), L (left), LN (lymph node), M (male), MI (mitotic index), NF (non-functional), NOS (not otherwise specified) PNX (pneumothorax), R (right)

^#Patient first reported in 2014 by Raymond et al.

^{^}Patient first reported in 2022 by Kim et al.

^*Four patients had confirmed non-functional tumors prior to initial resection; however, the remaining two patients did not undergo a preoperative evaluation, but denied signs or symptoms of hormone excess and are therefore presumed to be non-functional

⁺Tumor not felt to be a typical adenoma or possess enough histologic features to be diagnosed as an outright ACC. Observed labeling index borderline between adenoma and carcinoma, consistent with a tumor of uncertain malignant potential. At worst, felt to be a low-grade ACC with oncocytic features

^&Insufficient tumor tissue to do a formal mitotic grading, but available data suggests low grade carcinoma

Assuming a maximum prevalence of pathogenic FLCN variants of one in 3234 in a presumably unaffected population, the expected number of patients with a variant in 1010 consecutively evaluated ACC patients is one (11). Therefore, the prevalence of pathogenic FLCN variants in the ACC cohort is significantly higher than expected (p=0.0003, Fisher’s exact test). We also reviewed the genetic testing results of 271 patients diagnosed with pancreatic cancer who were tested using a panel that includes FLCN as a comparison cohort. No pathogenic FLCN variants were observed. Finally, we accessed data from the Michigan Oncology Sequencing Center (MI-ONCOSEQ), which performs high-throughput DNA and RNA sequencing for tumor analysis, including germline sequencing, and found that of the 4,885 patients (without adrenal or renal carcinoma), there were four pathogenic FLCN variants for a frequency of 0.08%.

Discussion

In addition to the above-described cases, literature review demonstrated isolated reports from North America (8–10, 12) and Europe (13–16) detailing seven patients with adrenocortical tumors and pathogenic variants in FLCN (Table 2). Two additional cases were identified but omitted due to incomplete evaluation of an adrenal lesion (17) and further review identifying the adrenal lesion as likely ectopic renal tissue (18).

Table 2:

Previously Reported Adrenal Tumors⁺ in Patients with Birt-Hogg-Dubé Syndrome^*

	Originally Published	AD	Sex	FLCN Variant	Biochemical Profile	Tumor Characteristics
1	Juszczak et al., 2011	26	F	c.1285 dup C p.His429fs	Non-Functional	Adrenocortical Carcinoma
2	Rato et al., 2017	39	M	c.1015C > T p.Gln339*	Non-Functional	Malignant perivascular epithelioid cell tumor
3	Ramsingh & Watson, 2018	32	M	c.1285 dup C	Non-Functional	Oncocytoma
4	Kalloniati, 2021	30	F	c.941_942del;p.P314fs	Non-Functional	Adrenocortical Carcinoma
5	Hofstedter et al., 2023	50	F	c.694C>T (p.Gln232Ter)	Non-Functional	Adrenocortical Carcinoma

Abbreviations: AD (age of diagnosis) dup (duplication), F (female), FLCN (folliculin), fs (frameshift), M (male)

⁺Kunogi et al., 2009 described a 33-year-old male with a non-functioning adrenal adenoma without further confirmatory work-up. Reese et al., 2009 reported a 32-year-old male with a large adrenal tumor, although it was felt the mass arose from ectopic renal tissue

^*Raymond et al., 2014 is omitted, but included in Table 1 and denoted with a () as they were seen in our institution

^{^}Kim et al., 2022 and Murvelashavili et al., 2022 (detail the same patient) and are omitted, but included in Table 1 and denoted with a (^{^}) as they were seen in our institution

With two of the above reports detailing a patient included in our current study, a total of twelve confirmed cases of malignant adrenocortical tumors in patients with BHDS have been reported, suggesting that ACC is a manifestation of BHDS.

The function of folliculin is largely unknown, but it likely serves as a tumor suppressor (19). Yet the inactivating mechanisms are debated, with somatic mutations, hypermethylation, and loss of heterozygosity (LOH) hypothesized as contributors (19, 20). In renal tumors associated with BHDS, multiple reports did not identify LOH as the etiology of inactivation and instead found somatic mutations (19) and promotor methylation (20). With respect to cutaneous manifestations, LOH was not discovered in several fibrofolliculomas, suggesting that haplo-insufficiency might be sufficient in initiating uncontrolled growth of skin lesions (21).

From our cohort, tumor samples are not currently available to confirm LOH and thus a second hit to support causality; however, reports have confirmed loss of the wild-type FLCN allele in a patient’s tumor, supporting its pathogenic role (12). However, larger numbers of ACC in BHDS are needed to prove LOH as a tumorigenic mechanism as the FLCN locus on chromosome 17 is commonly lost in sporadic ACC (22).

This study has several limitations. Our clinical cohort were patients with ACC found to have BHDS; however, it is important to think about the estimates of ACC in patients with germline pathogenic FLCN variants as well. In the future, it may be worthwhile to review ACC incidence in BHDS-specific databases if/when they emerge to estimate relative risk increase. Furthermore, this is a single institution study of two rare disorders. Future directions include a multi-institution, international collaboration to assess a larger population data set.

In our single institutional study, we observed ACC and a pathologic FLCN variant in 0.7% of patients seen in our clinic (1.5% of patients evaluated by a genetic counselor, 2.0% who underwent genetic testing, and 10% of those included FLCN). However, this was a select population of patients with ACC – less than 50% (466/1010) met with genetic counselors and another 25% (120/466) opted not to pursue genetic testing. While we must be cautious with applicability for clinical practice given this select cohort, the prevalence of BHDS amongst ACC patients far exceeds the conservatively estimated prevalence of BHDS (1–3/million), and even recent higher estimates of up to 1 in 3234 individuals (0.03%) (11). Furthermore, given that (1) no pathogenic variants in FLCN were identified in a comparison cohort of patients with pancreatic cancer that underwent genetic counseling and testing at the same institution and (2) 0.08% of patients with non-adrenal or renal cancers were found to have a pathogenic FLCN variant in a larger population of more than 4,500 patients, a true, specific association of BHDS and ACC is supported over a chance event.

While we demonstrate a higher-than-expected prevalence of adrenal tumors in BHDS, the overall incidence remains low. Regardless, adrenal lesions identified in BHDS patients during routine imaging surveillance for renal neoplasms should be evaluated in detail. Prompt assessment of this population is vital as more than 50% of patients developed metastatic disease. Clinicians specializing in ACC should perform a dedicated skin exam for the characteristic lesions associated with BHDS and review imaging for renal and/or pulmonary cysts. Current recommendations for genetic evaluation focus on LFS, LS, MEN1, and FAP (23). FLCN pathogenic variants should be considered during the genetic evaluation of patients with ACC given the implication for future tumor surveillance and the basis for family cascade testing to identify other at-risk family members that may benefit from tumor surveillance.