Lymphangioleiomyomatosis and tuberous sclerosis: where is the border?

Pulmonary lymphangioleiomyomatosis (PLAM) is a rare disease, characterized by an abnormal proliferation of smooth muscle throughout the lung; it occurs exclusively in women, generally of reproductive age [1]. The muscle cells proliferating in PLAM show little or no atypia at histological level. Therefore, PLAM has been considered to be a hamartomatous lesion rather than a true neoplastic process. However, in spite of the bland cytological features, the disease causes a progressive structural remodelling of the lung, leading to serious impairment of pulmonary function. Whilst survival curves now seem better than previously believed and hormonal treatment has been introduced with encouraging results [2], some patients progress to a condition necessitating lung transplantation. A disease with overlapping morphological features is occasionally encountered within the neurological syndrome of tuberous sclerosis (TS), also called Bourneville's disease. Indeed, the first description of PLAM, in 1918, was by LAUTENBACHER [3], in a female patient with the stigmata of TS. Subsequently, various authors have reported examples of PLAM, in patients without clinical evidence of TS. The question of whether PLAM is a "forme fruste" of tuberous sclerosis or a distinct disease has since been debated. In 1975, CORRIN et al. [1] reviewed 34 cases in the literature at that time and added 23 previously unpublished cases. In this work, the authors discussed the possibility of a relationship between tuberous sclerosis and PLAM, concluding that the question must be considered to be unresolved. When we consider the clinical settings of PLAM and TS, important differences emerge. Notably, PLAM develops exclusively in females, whilst TS affects both sexes (but lung lesions in TS develop almost exclusively in females), and familial history is common in TS patients, whilst it is absent in PLAM. These findings led STOVIN et al. [4] to conclude that "lymphangiomyomatosis and tuberous sclerosis are probably not different aspects of the same disease process but seem to be two different disorders". Other authors have expressed an opposite view, as in a recent paper by CASTRO et al. [5] reporting the clinical features of nine TS patients with pulmonary involvement. The authors concluded that "... TS is similar to PLAM in terms of its natural history, clinical features and response to treatment". The key to addressing this unresolved question seems to be the definition of the diagnostic criteria of TS. The classical Vogt triad (seizures, mental retardation and facial angiofibroma) was the first important attempt to clinically define the syndrome. This diagnostic triad was universally accepted and was, for a long time, considered the hallmark of the disease. Whilst useful for clinicians, this approach led to a delay in the recognition of other diagnostic signs of TS, particularly the presence and classification of visceral lesions, such as angiomyolipoma, so characteristic in this hereditary disease of autosomal dominant transmission. With the accumulation of knowledge on TS and the introduction of new diagnostic techniques, the Vogt triad has lost much of its importance. The Vogt's diagnostic triad is now considered obsolete and too strict to encompass all cases of TS. GOMEZ [6] revised the criteria for diagnosis of TS, showing that TS is much more common than previously estimated, and that only 29% of cases had the complete Vogt's triad. A detailed analysis of the hierarchy of clinical and imaging features necessary for a definitive, presumptive or suspect diagnosis of TS can be found in the work of GOMEZ [6]. In this revised classification of diagnostic criteria, PLAM is considered sufficient for a presumptive diagnosis of TS, when alone; and sufficient for a definitive diagnosis when in combination with at least one other feature of the same hierarchical diagnostic level (such as renal angiomyolipoma). TS is, by definition, a disease with multiorgan involvement. Most, if not all organs, can present abnormal growths of hamartomatous/neoplastic nature, thus leading to a great variety of signs and symptoms. The protean manifestations of TS account for the difficulties in establishing diagnostic criteria. It has always been difficult to place PLAM in the context of TS. In fact, PLAM has most often been observed as an isolated disease in patients without other symptoms or signs characteristic of TS. In the work of CORRIN et al. [1], only 2 of the 28 cases evaluated had evidence of renal angiomyolipoma. At that time, however, most cases could not be thoroughly studied with modern imaging techniques. In the interesting study by MAZIAK et al. [7] published in this issue of the Journal, the authors show a much stronger association between PLAM and renal angiomyolipoma than previously believed. Eight of 14 patients (57%) had renal tumours consistent with angiomyolipomas, five EDITORIAL