Serum VEGF-D a concentration as a biomarker of lymphangioleiomyomatosis severity and treatment response: a prospective analysis of the Multicenter International Lymphangioleiomyomatosis Efficacy of Sirolimus (MILES) trial

Abstract

Background: VEGF-D is a lymphangiogenic growth factor that has a key role in tumour metastasis. Serum VEGF-D concentrations are increased in most patients with lymphangioleiomyomatosis, a rare neoplasm associated with mTOR-activating tuberous sclerosis gene mutations, lymphadenopathy, metastatic spread, and pulmonary cyst formation. We used data from the Multicenter International Lymphangioleiomyomatosis Efficacy of Sirolimus (MILES) trial to assess the usefulness of serum VEGF-D concentration as a marker of severity and therapeutic response to sirolimus in patients with lymphangioleiomyomatosis.

Methods: In the MILES trial, patients with lymphangioleiomyomatosis who had forced expiratory volume in 1 second (FEV1) of 70% or less of predicted were randomly assigned (1:1) to 12 months masked treatment with sirolimus or placebo. Serum VEGF-D concentrations were measured at baseline, 6 months, and 12 months. We used a linear regression model to assess associations of baseline VEGF-D concentrations with markers of disease severity, and a linear mixed effects model to assess the associations of VEGF-D concentrations with between-group differences in clinical, physiological, and patient-reported outcomes.

Findings: We included 42 patients from the placebo group and 45 from the sirolimus group in our analysis. Baseline VEGF-D concentrations in individual patients varied from 0·34 ng/mL to 16·7 ng/mL. Baseline VEGF-D concentrations were higher in patients who needed supplemental oxygen than in those who did not need supplemental oxygen (1·7 ng/mL [IQR 0·99–3·36] vs 0·84 ng/mL [0·52–1·39]; p<0·0001) and in those who had a bronchodilator response than in those who did not (2·01 ng/mL [0·99–2·86] vs 1·00 ng/mL [0·61–2·15]; 0·0273). Median serum VEGF-D concentrations were similar at baseline in the sirolimus and placebo groups, and fell from baseline at 6 and 12 months in the sirolimus group but remained roughly stable in the placebo group. Each one-unit increase in baseline log(VEGF-D) was associated with a between-group difference in baseline-to-12-month FEV1 change of 134 mL (p=0·0007). In the sirolimus group, improvement in baseline-to-12-month FEV1 occurred in 15 of 23 (65%) VEGF-D responders (ie, those in whom baseline-to-12-month VEGF-D concentrations decreased by more than they did in any patients in the placebo group) and four of 15 (27%) VEGF-D non-responders (p=0·0448).

Interpretation: Serum VEGF-D is a biologically plausible and useful biomarker in lymphangioleiomyomatosis that correlates with disease severity and treatment response. Measurement of serum VEGF-D concentrations could inform the risk–benefit analysis of sirolimus therapy in patients with lymphangioleiomyomatosis and reduce the numbers of patients needed for clinical trials.

Funding: National Institutes of Health, US Department of Defense.

Figure 1. Study population

MILES=Multicenter International Lymphangioleiomyomatosis Efficacy of Sirolimus.

Figure 2. Relation between serum VEGF-D concentrations and disease features at enrolment

Note that the y-axis is broken and that the scale changes. Serum VEGF-D concentrations at baseline were significantly higher in participants who used supplemental oxygen than in those who did not use supplemental oxygen (p<0·0001) and in those who had a positive bronchodilator response than in those who did not (0·0273). Patients with a history of pneumothorax had significantly lower baseline concentrations than did those without such a history (0·0015). Baseline concentrations were not significantly associated with menopausal status or historical presence or absence of angiomyolipoma.

Figure 3. Relation between baseline log(VEGF-D) and 12 month change in FEV₁ for individual participants in the placebo (A) and sirolimus (B) groups, and expected change from baseline in FEV₁ and FVC (in mL) on the basis of each one-log-unit change in baseline log(VEGF-D) (C)

FEV₁=forced expiratory volume in 1 s. FVC=forced vital capacity.

Figure 4. Distribution of baseline-to-12-month change in FEV₁ in participants in the placebo group (A), and relation between VEGF-D response and baseline-to-12-month change in FEV₁ in the sirolimus group (B)

Data are shown for all patients who had both VEGF-D concentrations and FEV₁ data at the 12 month study visit (one patient in the placebo group did not have a FEV₁ measurement at 12 months). Participants in the sirolimus group were stratified on the basis of VEGF-D response, which was defined as a decrease of greater than 42% in VEGF-D concentrations from baseline to 12 months—ie, the maximum change noted in the placebo group. FEV₁=forced expiratory volume in 1 s. *Denotes a participant who did not have a change in FEV₁ at 12 months. †Denotes a VEGF-D nonresponder whose FEV₁ decreased by 0·55% at 12 months.