Seventy-five patients were seen at the UK LAM Centre for evaluation of TSC-LAM between 2011 and 2022. All were women and had a mean (SD) age of 39 (12) years at first assessment, 73 of these patients had LAM. A control cohort of patients with sporadic LAM seen at the same centre over the same period were used for comparison with the TSC-LAM cohort. The sporadic LAM cohort comprised 175 patients with definite LAM according to ATS/JRS criteria who were participating in the BioLAM observational research cohort. Sporadic LAM patients were all women, with a mean (SD) age of 47 (11) years at first assessment. The rate of referral to the service for patients with TSC-LAM was similar over the study period and did not change after the introduction of CT screening for LAM in 2013 (Fig. 1).

Referral rates of TSC and sporadic LAM cohorts. 2011 includes existing patients transferred to the LAM Centre

The mean age of diagnosis with TSC was 17 years (median 11 years, range 0 to 70). One third of these patients were not diagnosed with TSC until adulthood including 12% after 40 years of age. Of these patients, a history of epilepsy was present in 51%, learning difficulty in 24%, TAND in 15% and SEGA in 11%. 95% of the cohort had one or more renal angiomyolipoma and 19% had renal cysts. LAM was present in all but two subjects (97%). Multifocal micronodular pneumocyte hyperplasia (MMPH) was present in 8% and co-existed with LAM in all cases. 19 patients had undergone TSC genotyping with 89% of these patients having a mutation in TSC2 (Table 1).

We compared the TSC-LAM patients evaluated at the LAM Centre with 2093 contemporaneous TSC patients currently receiving care for TSC from 31 countries recorded in the Tuberous Sclerosis registry to increase disease awareness (TOSCA) [24]. Compared with patients registered in TOSCA, our cohort being evaluated for LAM tended to have been diagnosed with TSC at an older age, had a lower prevalence of neurological and cognitive manifestations, were more likely to have angiomyolipoma and more likely to have TSC2 mutations than TSC1 or no identifiable mutation (Fig. 2).

Comparison of clinical features of TSC-LAM cohorts with the TOSCA registry. 73 patients with TSC-LAM are compared with 2093 patients from the TOSCA registry. Median age at assessment and age at diagnosis of TSC in the left panel and percentage of cohort with each disease related feature in the right panel. TOSCA the Tuberous Sclerosis registry to increase disease awareness. TSC tuberous sclerosis complex. TSC1 TSC1 mutation detected. TSC2 TSC2 mutation. NMI no mutation identified. TAND Tuberous sclerosis associated neuropsychiatric disorders. MMNPH Multifocal micronodular pneumocyte hyperplasia

We examined the presenting problems leading to a diagnosis of LAM in TSC. Most common was the discovery of LAM due to workup or management of angiomyolipoma which occurred in over a quarter of patients. Pneumothorax and dyspnoea were also common modes of presentation. CT scanning performed to investigate other respiratory and unrelated symptoms also led to the diagnosis in a smaller number of subjects (Table 2). One fifth of patients had been diagnosed after CT screening for LAM. Of all 75 patients with TSC, CT screening for LAM was not performed in 48 and screening status was unknown in a further 10 patients. Seventeen patients were screened for LAM, of these, LAM was present in 13 including four patients diagnosed with TSC as adults, one patient had negative screening at age 33 and developed LAM symptoms at age 41, a further patient had negative screening as a teenager and went on to develop LAM symptoms age 25. Two patients screened did not have LAM.

As most evidence and clinical guidelines for the management of LAM are derived from patients with sporadic LAM [25,26,27], we compared the clinical phenotype of TSC-LAM with that of sporadic LAM to determine how similar the disease manifestations were. Compared with sporadic LAM, patients with TSC-LAM were younger at their first symptom, less likely to present with dyspnoea, more likely to have angiomyolipoma, and almost twice as likely to require embolization or surgery for angiomyolipoma and surgical treatment for pneumothorax. FEV1 and DLCO at first assessment were not significantly different between the two groups (Table 2).

Patients with TSC-LAM were followed up over a mean duration of 49.2 months and sporadic LAM 82.3 months. Over this period, six (8.2%) patients with TSC-LAM died, in only one case due to LAM and one (1.4%) received a lung transplant. The risk of lung transplantation or death due to LAM in the TSC-LAM patients was 0.67% / year. Patients with TSC had a higher overall mortality than patients with sporadic LAM (p < 0.0001). Whilst only one patient in each group died due to LAM, the combined risk of death or transplant per year of observation was statistically higher in patients with TSC-LAM (Table 3).

Fifty-three percent required treatment with an mTOR inhibitor, 36% with sirolimus for LAM and 17% with Everolimus for angiomyolipoma (Table 3). mTOR inhibitors were initiated for TSC-LAM at a mean (SD) age of 39.2 (11) years, a mean of five years earlier than those with sporadic LAM with a mean (SD) age of 44 (9) years (p = 0.009).

Serial lung function was measured as part of clinical care. Despite lung function being attempted by experienced lung function technicians, 11 of the 75 patients with TSC were unable to perform lung function reliably and are not reported. For those patients with serial lung function data, FEV1 fell by a mean of 81 ml/year and DLCO fell by 0.309 mmol/ml/kPa/year in patients not being treated with an mTOR inhibitor.

Rates of mTOR inhibitor treatment for LAM were similar between the two groups. As some patients with TSC also required mTOR inhibitor treatment for angiomyolipoma, patients with TSC-LAM were more likely to be treated with an mTOR inhibitor overall. The choice of mTOR inhibitor used was largely determined by treatment indication: patients with active lung disease were prescribed sirolimus and patients with renal angiomyolipoma requiring treatment were generally treated with Everolimus. Of the 46 patients with TSC prescribed an mTOR inhibitor, 16 (21%) were prescribed Everolimus, all but one of these for renal angiomyolipoma, with one patient taking the drug for combined renal and lung disease. Thirty (40%) were prescribed sirolimus, 25 of whom were treated for LAM, the remainder for angiomyolipoma and of these 2 had changed from Everolimus to sirolimus (one due to funding issues and one after an adverse reaction). Lung function was lower in patients treated with sirolimus (mean (SD) percent predicted FEV1 59.9 (23), DLCO 48.0 (18)) compared to those treated with Everolimus (FEV1 80.4 (20), DLCO 66.8 (27), both p = 0.01). For those treated with either sirolimus or Everolimus, mean FEV1 fell by 3 ml/year and DLCO increased by 0.032 mmol/ml/kPa/year. Lung function decline was similar between patients with TSC-LAM and sporadic LAM in both mTOR inhibitor treated and untreated patients (Table 3).