Tuberous sclerosis complex (TSC) has an incidence of up to 1 in 5800 live births.1 While formal diagnosis is still based upon clinical criteria, two genes have been identified in association with the disease: the TSC1 gene on chromosome 9q34, and the TSC2 gene on chromosome 16p13.3.2^–4 The gene products of TSC1 and TSC2 are the proteins hamartin and tuberin, respectively, and are thought to act together in a tumour suppression pathway.5 Additionally, a subset of the TSC population exists for whom no disease-causing mutation can be identified (no mutation identified, NMI).

Lymphangioleiomyomatosis (LAM) is a cystic disease of the lung with a characteristic chest computed tomography (CT) appearance of bilateral, thin walled round cysts measuring between 2–60 mm in diameter.6 7 LAM can occur either sporadically or as a sequela of TSC, in which it has been reported to affect 34–39% of asymptomatic women.8 9 LAM has also been confirmed in some TSC men.10^–12 While biopsy is necessary for a definitive diagnosis, the characteristic CT presentation is adequate for a presumptive diagnosis in the setting of TSC. Among patients with LAM and TSC, one study found that of the 24 disease-causing mutations reported in the literature, 22 (92%) were TSC2 and only two (8%) were TSC1. The same study found that of the seven somatic mutations reported in patients with sporadic LAM, all were TSC2.13

The present study was designed to compare the presence and severity of LAM in TSC patients with a TSC1 mutation, TSC2 mutation, or NMI. We retrospectively assessed the chest CTs of 65 TSC patients with genetic mutational testing available for the presence and severity of cystic lung changes consistent with LAM.

METHODS

Patients

Two hundred and ninety-three patients meeting clinical criteria for TSC were seen in the TSC clinic of a single neurologist (EAT) between October 1996 and June 2008.2 Seventy-three of these patients had at least one chest CT scan at the Massachusetts General Hospital (MGH) over this span and were either adolescents or older. Sixty-five of these 73 also had genetic mutational testing available for review. These 65 patients were ultimately included in the study. Genetic testing of the TSC1 and TSC2 genes, including detection of large DNA deletions and rearrangements of the TSC2 gene, was performed at Athena Diagnostics (Worcester, Massachusetts, USA) or the MGH Neurogenetic Diagnostic Laboratory (Boston, Massachusetts, USA). A single author (DAM) retrospectively reviewed all available medical records, and patients were split into groups of TSC1, TSC2, and NMI. One individual who expressed mosaicism for a TSC2 mutation was included in the TSC2 group. This study was approved by the institutional review board of MGH.

Chest CTs

Chest scans were performed on a range of CT scanners varying between 4 and 16 slice detectors, with either 2.5 or 5 mm sections. All studies were performed with suspended inspiration from the lung apices to the adrenal glands with a pitch of 1.0. Axial 0.625 mm high resolution CT scans were performed at six equally spaced intervals in the chest during inspiration. Scans were viewed on soft tissue window settings (window level: 40; window width: 400) and lung window settings (window level: −600; window width: 1500). The original available scan was retrospectively reviewed for each patient. All chest CT scans were reviewed independently by two board certified radiologists (AS, 11 years experience thoracic imaging; VM, 16 years experience thoracic imaging). Both were blinded to patients’ mutational status. Where interpretation differed, consensus was reached by repeat review of the case.

The cystic involvement of the lung was graded using a previously published scale.9 14^–16 Three equally sized divisions of the lung (upper, middle, and lower) were graded independently based upon the degree of cystic involvement in that zone: grade 0 (no cystic involvement), grade 1 (less than one third involvement), grade 2 (between one third and two thirds), or grade 3 (greater than two thirds). The predominant cyst size was recorded as either <0.5 cm, between 0.5 cm and 1.0 cm, or >1.0cm, in keeping with previous studies.14 17

Following initial review, it was determined that the above scale for cystic involvement (grade 0, 1, 2, or 3) was not sufficiently sensitive to capture the differences in cystic involvement present in this population. Thus, a post hoc evaluation was performed by a single radiologist (AS, still blinded to mutational status) counting the total number of pulmonary cysts as <10, between 11 and 100, or >100.

Statistical analysis

Either χ² tests or Fisher’s exact tests, depending on applicability, were used to explore the relationship between cystic changes on CT consistent with LAM and (1) TSC1 vs TSC2, (2) TSC1 vs NMI, and (3) TSC2 vs NMI. A Mann–Whitney test was used to explore the relationship between (1) predominant cyst size, (2) degree of cystic involvement, and (3) cyst number, as defined above, and (1) TSC1 vs TSC2, (2) TSC1 vs NMI, and (3) TSC2 vs NMI. A t test was used to explore the relationship between patient age and (1) TSC1 vs TSC2, (2) TSC1 vs NMI, and (3) TSC2 vs NMI. Due to their substantially lower rate of LAM, the cyst positive men were excluded from each of these analyses. Alpha was set at 0.05.

RESULTS

Patients and imaging

Patient characteristics are presented in table 1.

On chest CT, a total of 22 women (49%) and two men (10%) had cystic findings consistent with LAM. Predominant cyst size and cyst counts for all LAM women are provided in table 2 along with statistical relationships to mutation type.

The majority of patients’ cysts, in all mutation groups, were classified as grade 1 (less than one third involvement of lung), necessitating the post hoc analysis of cyst number, which was more sensitive to the observed differences. Both cyst positive men had fewer than 10 cysts of predominantly <5 mm in size, which occupied less than one third of the lung. One of these men (TSC2 mutation) had at least one cyst in all three lung zones, whereas the other (NMI) had cysts only in the upper lung zones.

Mutational analysis

Twenty-two (34%) patients had a TSC1 mutation, 34 (52%) a TSC2 mutation, and nine (14%) NMI. Disease-causing mutations in the cyst positive patients are listed in table 3.

Several of these mutations were also observed in patients without LAM. One of these was a mother–daughter pair in which the mother, in her 60s, had cysts while the daughter, in her 40s, did not. This study included three additional related pairs of patients with identical mutations (one father–daughter pair with TSC1, one sister–brother pair with TSC1, and one sister–sister pair with TSC2), but none of these six patients had LAM. Statistical analyses revealed that TSC1 women with LAM tended to have (1) a lower number of cysts than TSC2 women with LAM (p = 0.010), and (2) predominantly smaller cysts than NMI women with LAM (p = 0.041). Additionally, TSC2 women in this study were notably younger than either TSC1 (p = 0.065) or NMI (p = 0.011) women.

DISCUSSION

LAM is characterised by a proliferation of abnormal smooth muscle cells throughout the lungs, with the cystic CT appearance described above.6 7 In this population of TSC patients with both chest CT scan and mutational testing available for review, we report a higher than previously recognised rate of LAM in women (49% compared to 34–39%), in part due to the fact that previous prevalence studies sampled only asymptomatic women.8 9 Additionally, the two cyst positive men (10% of our male population) identified in our study represent, by our literature review, the first TSC males identified with asymptomatic LAM.10^–12 These findings indicate the necessity of obtaining chest CTs to document cystic disease in TSC patients of both genders. This should allow for prospective documentation of the course of this disorder, which could in turn provide a comprehensive understanding of TSC related LAM.

Among TSC patients, the TSC2 mutation has been associated with more severe epilepsy, mental retardation, autism, and hamartomatous involvement of the brain, skin, kidneys, and eyes.18^–23 While previous work has indicated LAM to be primarily a disease of women with TSC2, we found a comparable rate of LAM in our TSC1 women as well.13 24 25 However, our post hoc analysis revealed the number of cysts present in TSC2 women to be significantly greater than those in TSC1 women. Assuming a relationship between degree of disease on CT and degree of pulmonary compromise,14 15 17 one might then expect to find a greater rate of pulmonary dysfunction in the TSC2 population. If so, this could contribute to a greater rate of imaging, greater rate of detection, and subsequently greater presumed rate of LAM in TSC2, as has hitherto been the case. Whether or not the cystic changes observed in our TSC1 population can or will progress to the more pronounced presentation that was observed in TSC2 warrants further study. That our TSC1 women were actually older than our TSC2 women, however, suggests that the two mutations may have fundamentally different presentations of LAM.