G. Brice and A. Child
Department of Cardiological Science
St. Georges Hospital Medical School, London

Introduction

Primary Lymphoedema is a chronic tissue swelling most commonly manifest in a limb, resulting from impaired lymph drainage and arising from a presumed intrinsic abnormality of lymph drainage (Mortimer 1995). If the lymph system is damaged, blocked or not formed correctly, protein rich fluid continued to leave the blood capillaries in the normal way but accumulates in the interstitial spaces, unable to return to the vascular compartment via the lymph system.

Insufficiency of the lymphatic system has two main consequences in addition to swelling; a predisposition to infection and, rarely, malignancy (Mortimer 1995). Most forms of primary lymphedema are assumed to result from a congenital abnormality of the lymphatics that may result in hyper or hypoplasia of the vessels. In the majority of the cases however, the clinical presentation is not until adolescence or later, and may be related to pregnancy or leg injury. Lymphedema may also be the predominant sign in congenital vascular disorders with unknown genetic basis.

Treatment

Lymphedema is largely incurable. Conservative management can either be surgical, with de-bulking operations, or medical, primarily using elastic compression hosiery in combination with gentle massage, exercise and pneumatic compression. Prevention of infection in the affected parts and good skin hygiene, particularly following an injury, are vital. If recurrent attacks of cellulites occur, then lifelong antibiotics may be necessary (Mortimer 1995).

Primary lymphedema can be classified in to two subsets: isolated, where the condition is not inherited, and familial, where there is a definite inherited trait. Further subdivisions can then be made according to the age of onset, which is most often pubertal or congenital. Rarely, the condition presents after the age of 35 when it is referred to as lymphedema tarda.

For many years the familial nature of lymphedema has been known, with both Milroy (1892) and Miege (1898) describing families affected by the condition. Milroy also gave the first description of incomplete penetrance in Lymphedema, with affected children born to clinically unaffected parents. More recently, in a study by Dale (1987), 35% of patients with primary Lymphedema attending a large teaching hospital had positive family history. An overall penetrance rate of 50% was estimated with the rate of expression much higher in females than males, 66% and 30% respectively. Though many cases of lymphedema appear to be inherited from an affected ancestor there are also many isolated cases. It is unclear if these cases are due to incomplete penetrance in previous generations, new mutations, recessive inheritance from two carrier parents, or if they have no genetic basis. It is envisaged that molecular genetics will help to define these phenotypes.

The population incidence of lymphedema is unknown with certainty; however, it has been estimated at 1:6000 (Dale 1985). Many people with the condition appear not to seek medical advice (particularly men) and the wide variety of diagnoses given to patients with swollen legs adds to the difficulty in calculating a realistic population risk figure.

Congenital hereditary lympedema

This condition was first described in 1892 (Milroy 1892) and given the title, Milroy's disease, by Sir William Osler in his work on the practice of medicine. Milroy's disease is frequently used as a blanket term to describe all primary lympedemas, however the term should be restricted to those rare cases fulfilling the criteria described by Milroy; congential onset of non-progressive swelling, familial occurrence, and lack of other associated features.

Recently, two independent groups have mapped a genetic locus for this type of lymphoedema to the telomere of chromosome 5q (Ferrell et al. 1998, Evans et al. 1999). Clinically, the families in the study by Ferrell et al. are described as having "early onset lymphoedema" and no other details are given. It is not clear if these families adhere strictly to the original criteria laid down by Milroy. All affected individuals in the family published by Evans et al. exhibited the same phenotype as Milroy's original family. In this study, linkage was also established to chromosome 5q35.3 and a penetrance of 84% was calculated, with the discovery of four male gene carriers in the family under study. Interestingly, one gene carrier, although having no obvious lymphedema, had suffered from a hydrocele, in common with seven other affected male family sufferers.

The vascular endothelial growth factor receptor gene VEGFR-3 (Flt-4) has been mapped to the same region on chromosome 5 (Galland et al. 1992) and proposed disease-causing mutations have been located in a number of families (Ferrell et al. 1998, Irrthum et al. 2000, Karkkainen et al. 2000). The expression of this gene becomes restricted to the lymphatic endothelium during development (Kaipainen et al. 1995) and is clearly a good candidate gene in all families mapping to this region. In a nuclear family, uninformative from the point of view of linkage analysis, Ferrell et al (1998) found a G-to-A transition at nucleotide 3360 of the Flt4 cDNA, predicted to cause a non-conservative amino acid substitution (pro11261leu) of the mature receptor. Subsequently, four other missense mutations in the VEGFR-3 gene have been found in primary lymphoedema families linked to chromosome 5q (Karkkinen et al. 2000), Irrthum et al. 2000). Our group, has also located a single mutation in this gene and one polymorphism. One large family, of Indian origin, with congenital edema of legs and hands, does not link to the chromosome 5q locus indicating that at least one other gene exists for congenital lymphoedema (Evans et al., unpublished data).

Whilst Milroy's family showed no signs of lymphedema except in the lower limbs, Esterley (1965) and Child (1999 personal communication) both describe families in which all four limbs were affected, at least in infancy. It remains to be seen whether this type of lymphedema is caused by mutations in the same gene as described for congenital lymphedema affecting only the lower limbs. The family examined by Esterley has recently been located and will be tested for linkage to the 5q locus.

Lymphedema distichiasis

The features of this condition are lymphedema of pubertal onset, preceded by the development of extra lashes arising from the meibomian gland openings. These can vary from a full extra set of lashes to the odd extra lash. They are normally soft and pale but often cause corneal irritation. Cases of apparently severe distichiasis in which patients have been unaware of the extra lashes have been reported (Kremer et al. 1986, Mangion et al. 1999) underlining the need for ocular examination in all patients in which there is a family history of distichiasis or lymphedema. Equally, patients referred with distichiasis alone should be examined for limb swelling and a family history sought. In our series of 10 families ascertained via an ophthalmic department there are no families with distichiasis alone. In most cases the extra lashes are clearly visible and slit lamp examination is not required to confirm the diagnosis. Treatment is by epilation, cryotherapy or lid splitting surgery (Shammas et al. 1979, Temple & Collin 1994).

Lymphedema distichiasis is typically associated with the uncommon lymphangiographic finding of hyperplasia, in which there are numerous dilated lymphatic vessels in the legs (Dale 1987). In addition it is often associated with other congenital abnormalities such as heart defects, cleft palate, ptosis and vertebral anomalies (Temple & Collin 1994).

Mangion et al. (1999) published linkage to chromosome 16q24.3 in three families with lymphedema distichiasis. The addition of further individuals from the largest of three families subsequently reduced the critical interval from 16cM to 2Mb. Eight further lymphoedema-distichiasis families were also consistent with linkage to the same 2Mb region (Bell et al. 2000). Recently, the FOXC2 (MFH-1) transcription factor gene was found to harbour mutations in families affected by lymphedema distichiasis. One nonsense mutation and one four base paid duplication were found in unrelated patients (Fang et al. 2000). Subsequently, our group has located 14 FOXC2 mutations in distichiasis families, 13 of which were insertions or deletions. Mutation sites are scattered throughout the gene but no apparent genotype-phenotype correlation was evident (Bell et al., submitted for publication). Further work will be carried out to determine the effect of the mutations on protein function.

Lymphedema hereditary type 2 (Meige lymphedema, lymphedema praecox)

In this type of lymphedema, swelling develops around the time of puberty and is particularly severe below the waist. First described in the 19th century, (Meige, 1898) this is the most common form of hereditary lymphedema. In our experience there is wide intra-familial variation in the degree of swelling but, as with Milroy's disease, no other associated features. Genetic analysis of families with this form of lymphedema has proved complex because of difficulties in defining the phonotype due these variations in expression within and between families. The phenotypic variety seen in this condition is likely to be influenced by gender, hormones, and other environmental factors. In order to clearly define the status of at risk family members, lymphoscintigraphy can be used to assess lymphatic function.

A panel of seventeen 2 & 3 generation UK families with pubertal onset lymphedema have been tested for linkage to both the published lymphedema loci without success (Sarfarazi, M. 1999, personal communication)>

Conclusion

In the period between the first descriptions of familial lymphedema in the late 19th century and today, little progress has been made in the treatment of the condition due primarily to the lack of understanding of the pathological processes involved. Why should functioning lymphatics fail in puberty? What is the connection between the development of the meibomian glands and the lymphatic system? It is hoped that the characterisation of the newly located genes will enable better, rational therapies to be developed for this often disabling and disfiguring condition. In the short term, the discovery of a mutation in a family will allow pre-symptomatic diagnosis and early management.

Correspondence:

Dr. Ann Child
Department of Cardiological Sciences
St Georges Hospital Medical School
Cranmer Terrace
London SW17 0RE
UK
E-Mail: g.brice@sghms.ac.uk

Published in the "Scope on Phlebology and Lymphology", No. 1, March 2001