Pathogenic variants of FLCN and kidney cancer

Birt-Hogg-Dubé Syndrome (BHD) is an inherited autosomal dominant disorder caused by mutations in the folliculin (FLCN) gene. About 200 pathogenic variants in FLCN have been identified to date. The three classic symptoms of BHD are spontaneous pneumothoraces, characteristic skin lesions associated with hair follicles (the most common of which are fibrofolliculomas), and kidney cancer (also known as renal cell carcinoma, RCC). Not all BHD patients have all symptoms. In particular, only about 1 in 3 patients develop RCC. It has been suggested, based on very limited data, that some FLCN variants may not give rise to RCC. If it were possible to identify variants that do not increase the risk of RCC, expensive and uncomfortable imaging (typically CT or MRI every three year) could be avoidable for people with those variants.

Matsumoto et al. undertook an extensive literature search to try to identify pneumothorax only pathogenic variants(POPVs) (1) . They found 158 articles, which described 1059 individuals from 575 families. 194 unique pathogenic FLCN variants were identified, with patient details provided in most reports. Patients were evenly split between East Asia (Japan, South Korea, China and Taiwan) (33.3%), North America (32.7%), and Europe (31.7%). They broke down the data to make several interesting observations.

Of the documented patients, 91.9% were shown to have pulmonary cysts. The prevalence of pneumothorax, characteristic skin lesions, and RCC in BHDS were 50.9%, 47.9%, and 22.5% respectively. The median age at first pneumothorax was 34 years (range 10–78 years). The median age at which skin changes were first noted was 38 years (range 20–65 years). The median age at first diagnosis of RCC was 47 years (range 14–83 years).

In previous reports there appeared to be a higher percentage of pneumothorax and fewer dermatological findings and RCCs among patients from East Asia. The current study bears this out and provides a more detailed overview. The prevalence of pneumothorax in East Asians was 74%, vs 45% in Europe and 35% in North America. It's not clear whether this might be explained by genetic and physiological or environmental differences, but there does seem to be some sampling bias due to differences in diagnostic pathways. Pneumothorax recurred in 2 out of 3 patients (geographic differences were not mentioned).

It typically took 6 years for patients to be diagnosed with BHD following a pneumothorax, compared with 0 years for diagnoses following RCC or skin changes. Lack of awareness by clinicians can delay life-saving screening for RCC.

Out of 991 patients identified, only one had RCC below age 20 (at 14 years of age), while 14 had a pneumothorax before age 20. This bears on the minimum age at which genetic testing should be performed. Ethical considerations are discussed by Borry et al, (2); generally, testing is discouraged before the age at which diagnosis might influence management of the condition.

Matsumoto et al. discuss some of the unavoidable limitations and potential biases of a study based on a review of the literature.

FLCN variants were analyzed according to the type of mutation, because different mutations behave differently. 68% were nonsense and frameshift variants, 16% were intronic variants, 6% were missense variants/in-frame deletions, 7% were large deletions/duplications, and 4% were variants affecting transcription initiation. The class of large deletions is significant because these variants are not readily identified by DNA sequencing. These are identifiable by copy number analysis, a technique that has only become common in recent years. Patients with large deletion variants who were tested more than a few years ago have often been told they were negative for BHDS.

Ideally, there would be enough data to pinpoint which variants are associated with RCC and which are not. But BHD is a rare disease, and there simply aren’t enough documented patients. Many variants have only been documented a few times; not enough to say "this variant, although known to have caused pneumothoraces, will never give rise to RCC." So, in order to get a little traction, we can try to focus on the variants that seem least likely to cause major problems.

For most of these classes of variants, it is thought that the mutation either causes the variant protein to be destabilized and rapidly degraded, or causes much less protein to be made because the mRNA is either rapidly degraded or just not functional. Missense variants, small in-frame deletions, and a small percentage of nonsense / frameshift variants (those near the last exon) are the only ones with much hope of making a normal amount of (possibly only slightly defective) protein. It is assumed that in general, inadequate amounts of protein being made will absolutely prevent normal function of FLCN, while normal amounts of (slightly defective) protein might have selective effects on function, perhaps leading to a milder condition that does not progress to RCC.

Of the 194 pathogenic FLCN variants (after eliminating variants known to be associated with RCC), 76 appeared, given the limited data, to be POPV. The authors looked for differences between POPVs and non-POPVS by assessing variant type, location within the gene, age of patients, and number of patients with that variant. There were no statistical differences in type or distribution. Almost 90% of POPVs were found in only three or fewer individuals; clearly there is a decent chance that for many of these variants, RCCs will be found if more patients are tested. There was a tendency for patients with POPVs to be younger than those with non-POPVs; if these patients are observed for a longer time, RCCs may develop in them.

Given the above, of the 76 apparent POPVs, it seems unlikely that the majority of these variants are only linked to pneumothorax. However, it cannot yet be ruled out that some of these may be genuine POPVs. The ability to follow BHD patients with these variants over time and assess whether they develop further BHD-associated manifestations would be greatly beneficial in furthering our understanding of whether certain variants are linked to particular manifestations.

Based on this conclusion, the authors recommend lifelong monitoring for RCC (by CT or MRI) of all BHD patients, since pneumothorax-only FLCN variants are likely to be rare.

References

1. Matsumoto K, Lim D, Pharoah PD, Maher ER, Marciniak SJ. A systematic review assessing the existence of pneumothorax-only variants of FLCN. Implications for lifelong surveillance of renal tumours. Eur J Hum Genet (2021). https://doi.org/10.1038/s41431-021-00921-x Online ahead of print.
2. Borry P, Evers-Kiebooms G, Cornel M, et al. Genetic testing in asymptomatic minors. Eur J Hum Genet 17, 711–719 (2009). https://doi.org/10.1038/ejhg.2009.25