Pathogenesis, detection and treatment of Achilles tendon xanthomas

Authors


  • Department of Internal Medicine, University of Ioannina, Greece (S. G. Tsouli, D. N. Kiortsis, M. S. Elisaf); Department of Radiology, University of Ioannina, Greece (M. I. Argyropoulou); Department of Molecular Pathology and Clinical Biochemistry, Royal Free Hospital, London, UK (D. P. Mikhailidis).

Moses Elisaf, MD, FRSH, FASA, FISA, Department of Internal Medicine, University of Ioannina, 451 10 Ioannina, Greece. Tel.: +30 2651 0 97509; fax: +30 2651 0 97016; email: egepi@cc.uoi.gr

Abstract

Tendon xanthomatosis often accompanies familial hypercholesterolaemia, but it can also occur in other pathologic states. Achilles tendons are the most common sites of tendon xanthomas. Low-density lipoprotein (LDL) derived from the circulation accumulates into tendons. The next steps leading to the formation of Achilles tendon xanthomas (ATX) are the transformation of LDL into oxidized LDL (oxLDL) and the active uptake of oxLDL by macrophages within the tendons. Although physical examination may reveal Achilles tendon xanthomas (ATX), there are several imaging methods for their detection. It is worth mentioning that ultrasonography is the method of choice in everyday clinical practice. Although several treatments for Achilles tendon xanthomas (ATX) have been proposed (LDL apheresis, statins, etc.), they target mostly in the treatment of the basic metabolic disorder of lipid metabolism, which is the main cause of these lesions. In this review we describe the formation, detection, differential diagnosis and treatment of ATX as well as the relationship between tendon xanthomas and atheroma.

Introduction

Tendinous xanthomas are accumulations of collagen and macrophages which contain cholesterol esters (foam cells) [1]. Tendinous xanthomatosis is usually associated with hyperlipidaemia. Except for Achilles tendons, the other tendons where xanthomas usually appear are the tendons of the hand (extensor tendons) and the elbow. It should be mentioned at this point that some of the most common xanthomatous lesions, besides tendinous xanthomas, are tuberous xanthomas, eruptive xanthomas, xanthelasma palpebranum and xanthoma planum. Ectopic xanthomas have also been described in patients with familial hypercholesterolaemia (FH) [2]. With the exception of tendinous xanthomas, the other xanthomas are histologically characterized by cholesterol ester accumulation in dermal foam cells [3]. Progress in the understanding of xanthomas was associated with the progress of technology, which allowed the determination of the patients’ lipid profile. The lipid composition and the lipid physical state of the xanthomatous lesions differ significantly [4]. The main characteristic of tendinous xanthoma is the exceptionally high composition in free cholesterol and total cholesterol [3]. Achilles tendon xanthomas (ATX) are usually accompanied by an increase in tendon size, caused not only by the intratendinous lipids but also by the oedema and inflammation of the area. This pathologic state may interfere with the tendons’ function and may cause achillodynia, cosmetic effects and rarely spontaneous tendon rupture [5]. In this paper we review the pathologic states related with tendinous xanthomas, metabolic pathways used for their formation, proposed methods for detecting these lesions and finally the best treatment for their regression.

Differential diagnosis of tendon xanthomas

In most cases tendinous xanthomas are characteristically associated with hyperlipidaemic states. The size of the xanthoma correlates with the degree of hypercholesterolaemia. The most common state associated with tendinous xanthomatosis is FH [6]. It is already described that tendon xanthomas appear in homozygous FH patients from their childhood, while heterozygous FH patients develop tendon xanthomas (TX) by the age of 20 years [7]. The proportion of patients with TX increases with age. It can be observed in 75% of patients with FH as they grow older. Most of the relatively severe hereditary hyperlipidaemias associated with ATX are caused by a defect of the low-density lipoprotein (LDL) receptor [8,9]. On the other hand, ATX can also be combined with most of types of hyperlipidaemia or some unusual chromosome mutations. For example, drug-induced hyperlipidaemia (antiretroviral therapy) or familial recessive hypercholesterolaemia are some pathologic states which have been reported to be accompanied by tendon xanthomas [10–12]. Several studies have clearly demonstrated that nonfamilial hypercholesterolaemia or FH–like syndromes can be associated with tendon xanthomas. For example, diabetes mellitus has been reported to induce overt hyperlipidaemia and accumulation of lipids in tissues in subjects with ApoE7 [13]. Rarely, tendon xanthomas are observed in patients with increased β very low-density lipoproteins (β VLDL) in familial dysbetalipoproteinaemia [14,15].

We should also mention another entity that is not associated with elevated LDL levels: plexiform xanthomatous tumour [16]. This tumour can be located on the knee, foot, hand or Achilles tendon. Plexiform xanthomatous tumour may represent a morphologic variant of tuberous or tendinous xanthoma, yet its exclusive occurrence in men, the absence of personal or familial history of hyperlipidaemia in most of the patients and the relative paucity of inflammation and cholesterol clefts may make this tumour a distinctive entity [16].

Although in all the above cases xanthomas were associated with hyperlipidaemia, they can also be occasionally developed in normolipidaemic patients. In fact, other forms of xanthomas (palmar, planar, etc.) may sometimes be associated with normolipidaemic states, notably monoclonal gammopathy [17]. However, ATX are very rarely present in normolipidaemic patients. Nevertheless, we should always take into consideration the presence of compositional lipoprotein abnormalities in cases of normocholesterolamic tendon xanthomas. Apolipoprotein E3 deficiency has been reported to be associated with tendon xanthomatosis [18,19]. Overproduction of apolipoprotein B, possibly combined with a defect in high-density lipoprotein (HDL) metabolism, is also associated with tendon xanthomas in some cases [18,20]. A cysteine-containing truncated apolipoprotein A-I variant associated with HDL deficiency has also been reported to be accompanied with ATX [21].

One category of normolipidaemic patients with xanthomas are those with cerebrotendinous xanthomatosis (CTX); an autosomal recessive lipid storage disease. Patients with CTX are typically normolipidaemic but they have elevated blood cholestanol and prominent xanthomas in tendons and the brain [22]. The basic metabolic defect behind this disease is the lack of the cytochrome P450 enzyme, sterol 27 hydroxylase (CYP271). We will refer to this process in more detail later [23].

Hyper-b-sitosterolaemia is another unusual normolipidaemic disease with elevated plasma cholestanol and lipid storage in tendons [24,25].

Finally, it is important to interpret Achilles tendon (AT) thickening as cholesterol accumulation only after exclusion of other disorders. Tendonitis, peritendinitis or bursitis, trauma, nodules from rheumatic arthritis or gout tophi are rather frequent conditions which may lead to Achilles tendon (AT) thickening, and in such cases differential diagnosis with ATX may sometimes be difficult [26,27]. Detailed medical history, ultrasonography of the affected tendon and other biochemical examinations may help significantly in the differential diagnosis.

Composition of tendon xanthomas

The composition of ATX could provide useful information about the process of lipid storage. Furthermore, it could help us understand the similarities of this mechanism with the formation of atherosclerotic plaques; another form of lipid storage responsible for acute occlusive events and a great number of deaths nowadays. The results from this comparison could be useful for better understanding of the mechanisms of lipid accumulation in the human body.

The major constituents of xanthomas are lipids (33% of dry weight) and collagen (24% of dry weight) [28]. Detailed lipid analysis showed that the lipid composition of tendon xanthomas is 55% free cholesterol, 28% cholesterolesters and 13% phospholipids. This lipid composition did not resemble any other xanthomatous lesion except the atherosclerotic plaques and the adult fatty streaks [3]. Staining tissue sections with special dyes revealed that unesterified cholesterol accumulated predominantly extracellularly, separately from the esterified cholesterol that accumulated both extra- and intracellularly [23,29].

Many investigators have been trying to determine with accuracy the mechanism of lipid deposition in tendon xanthomas. Several experiments administrating isotopic cholesterol (99 m technetium-labelled LDL) suggested total and rapid exchangeability of cholesterol between plasma and xanthomas [30]. Thus, lipids in tendon xanthomas seem to be derived from the circulation rather than from local synthesis, secretion or cell death. These findings also seem to support previous biopsy data indicating active uptake of LDL by macrophages within xanthomas [31]. Immunohistochemical studies provided further evidence of lipid accumulation into macrophages. Using specific antibodies, the investigators found that oxidatively modified low-density lipoproteins (oxLDLs) had a similar distribution in xanthomata to that of macrophages, which appear to be the main storage sites for oxLDL [32]. Furthermore, these studies suggested that oxLDL was associated with macrophages and occurred intracellularly. In contrast, LDL had a different distribution from oxLDL and was detected extracellularly [32]. Therefore, we can assume that LDL derived from plasma is trapped in collagen and glycosaminoglycans of the tendon matrix and it can be oxidized at these sites. The reaction of LDL with macrophages or other cells of the tendon can explain the modification of LDL into oxLDL. After this modification, oxLDL is taken up mainly by macrophages, thereby promoting the formation of foam cells [1,33,34]. Interestingly, it has been shown that the titres of autoantibodies against oxLDL are significantly increased in patients with ATX compared with controls [35].

In addition, using antibody labelling, the immunohistochemical studies mentioned above detected also apo (a), which had the same distribution with LDL in the dermis and subcutaneous tissues in the xanthomas. However, this observation does not prove that lipoprotein (a) [Lp(a)] may also play a role in xanthoma formation and further studies are needed [32].

We should also, however, briefly refer to the mechanisms of xanthoma formation in normolipidaemic states. There are certain differences in xanthoma formation in each of these states. For example, in untreated CTX patients there is an overproduction of cholestanol [22]. Part of the explanation of accumulation of cholestanol in tendon and brain xanthomas of CTX patients may be the lack of the CYP27A1-dependent mechanism for the efflux of the steroids [23].

On the other hand, the study of apolipoprotein E (ApoE) function has shown that ApoE accelerates the efflux of cholesterol from macrophages and prevents them from transforming into foam cells. A defect in this mechanism seems to be the main reason for atherosclerosis and the formation of tendon xanthomas in patients with ApoE deficiency [19].

Concerning patients with overproduction of ApoB, it has been demonstrated that they have transient periods of lipidaemia and abnormal lipoproteins (small, apo B-rich VLDL) that promote tissue deposition. Furthermore, the apolipoprotein B ‘saturates’ normal clearance of LDL and promotes the formation of foam cells [20].

A defect in HDL function could also explain tendon xanthomatosis [21]. Several studies have demonstrated that a high level of HDL3 cholesterol was an independent factor in the smaller size of ATX [36].

Profile of patients with Achilles tendon xanthomas

Many studies have tried to determine the clinical characteristics of patients with ATX and FH, which is the most common situation associated with ATX. Most of the recent studies have demonstrated that there is a positive correlation between Achilles tendon thickness (ATT) and serum cholesterol concentration [37–40]. The same positive correlation has been described between ATT and age [37–40]. On that basis, some investigators proposed the term cholesterol year score, which is calculated by the age and the mean serum cholesterol concentration of the patients. Achilles tendon width was found to be best correlated with this score in patients with homozygous FH (r = 0·86, P < 0·001) [38]. On the other hand, some of these studies demonstrated that ATT was not significantly correlated with the serum HDL-cholesterol concentration [38,39], while others proposed that ATT was inversely correlated with serum levels of HDL lipid parameters, serum phospholipids, serum cholesterol ester transfer protein (CETP) or other serum parameters [36,41–43]. Further multivariate analysis indicated that only HDL3 cholesterol level was an independent negative predictor of ATT (P = 0·004) [36]. Another interesting finding from the previous study was that ApoA1 was also significantly inversely correlated with ATT (P = 0·028). More studies are needed in order to confirm these relationships.

Moreover, interesting data demonstrated that another serum parameter that correlated significantly with ATX is the titre of autoantibodies against oxidized low density lipoproteins (P < 0·01) [35]. Further confirmation of these observations is also needed. In addition, the results of many studies regarding the correlation of ATX with sex are controversial [37,40,44].

Increased serum levels of cholestanol is the main pathological finding in patients with CTX and hyper-b-sitosterolaemia [25]. This measurement could help the differential diagnosis of ATX in some difficult cases. On the other hand compositional lipoprotein abnormalities, which could be associated with normocholesterolaemic tendon xanthomas, can be easily detected by the measurement of serum apolipoproteins (apo A-I, apoB, apoE) [18,18–20].

Detection of Achilles tendon xanthomas (Table 1)

Table 1.  Detection of Achilles tendon xanthomas
MethodAdvantagesDisadvantages
  1. FH, familial hypercholesterolaemia.

Physical examinationno exposure to radiationsubjective
low sensitivity to detect lesions at early stages
Radiographylow costexposure to radiation
detection of calcificationslack of information regarding tendon's internal structure
possibility for measurements of the Achilles tendon 
non time-consuming 
Ultrasonographyno exposure to radiationlack of a neighbouring tissue reference to qualify
low costdiffuse abnormality of the tendon's echogenicity
detection of focal lesions 
possibility for measurements of the Achilles tendon 
non time-consuming 
Computed tomographypossibility for measurements of the Achilles tendonexposure to radiation
high cost 
lack of differences in tendon's density between 
normo-cholesterolaemic and FH subjects 
Magnetic resonance imagingno exposure to radiationhigh cost
possibility for measurements of the Achilles tendoninability to detect focal lesions
characteristic diffuse stippled pattern in FH patientstime-consuming

Achilles tendon xanthomas may be detected by clinical examination and imaging studies. Clinical evaluation, which consists of inspection and palpation of the Achilles tendon, has disadvantages, as it is more or less subjective, it cannot detect ATX at the early stages of their formation and it can not determine the exact size of ATX [27,45–48].

Plain radiographs are the first imaging modality that has been used for the evaluation of the ATXs [49,50]. Lateral views, using appropriate parameters to visualize the soft tissues of the ankle joint, can depict calcifications of the xathomatous lesions and offer the possibility to measure the anteroposterior diameter of the Achilles tendon [49–51]. However, conventional radiography is a projectional technique, which does not detect structural abnormalities before tendon enlargement occurs [51]. Finally another disadvantage of this method is the exposure of the patient to radiation.

Computed tomography (CT) scan, which is a cross-sectional method, depicts the Achilles tendon as a soft tissue structure surrounded by the very low-density subcutaneous fat [52–54]. The good contrast between the Achilles tendon and the surrounding fat tissue offers the possibility to perform accurate size measurements [52–54]. Although size measurements are useful for the initial evaluation of ATX and for controlling the response to treatment, density measurements of the Achilles tendon fail to demonstrate differences between normocholesterolaemic and hypercholesterolaemic subjects [52,54]. An additional parameter that has to be considered for CT scan is the exposure of the patient to radiation.

Ultrasonography is the most widely used, cross-sectional, imaging modality for the evaluation of ATX [9,27,45–48,51,52,55–58]. The sonographic appearance of the normal Achilles tendon is characterized by the presence of multiple parallel linear echoes that give a fibrillar appearance (Fig. 1a). Based on histologic studies, these parallel linear echoes are thought to represent anaclastic interfaces between the endotendineum septa, made of loose connective tissue and the bundles of collagen fibres [47]. Although, histologic proof is missing, because tendon biopsy is precluded on clinical and ethical grounds, focal hypoechoic areas (Fig. 1b) or a diffuse heterogeneous echo pattern (Fig. 1c) has been considered to represent ATX, provided that the patient has no history suggesting trauma or tendonitis [47,55,57,59]. Sonograpy is a reliable method for the detection of focal hypoechoic areas inside the hyperechogenic Achilles tendon. In contrast, evaluation for diffuse changes in tendon echogenicity is very subjective, as no neighbouring tissue can serve reliably as a reference to qualify the echogenicity of the tendon [9]. Thickening of the Achilles tendon has been considered as another finding of ATX. According to different studies, in patients with (familial hypercholesterolaemia) FH, an anteroposterior diameter of the tendon > 5·8–7·1 was considered abnormal [9,47,55,58]. However, in only one study was the diagnosis of FH genetically confirmed and in this study a cut-off point of 5·8 mm was found to be a useful threshold [9]. Ultrasonography has not only been used for the initial diagnosis of ATX but also for evaluating the response to treatment [58]. Indeed, the anteroposterior diameter of the Achilles tendon of patients with FH has been found to decrease significantly after treatment [58].

Figure 1.

Sonographic appearance of the Achilles tendon. (a) A 22-year-old normal male subject – sagittal sonogram of the Achilles tendon (→) shows a fibrillar pattern characterized by the presence of multiple parallel linear echoes. (b) A 21-year-old patient with familial hypercholesterolaemia (FH) – axial sonogram of the Achilles tendon demonstrates a focal hypoechoic lesion (white markers) suggestive of Achilles tendon xanthomas (ATX). (c) A 25-year-old male patient with FH – sagittal sonogram of the Achilles tendon (black arrowheads) demonstrates a diffuse heterogeneous echo pattern.

Magnetic resonance (MR) imaging has also been used for the evaluation of ATX [51,55,60,61]. On MR the normal Achilles tendon exhibits a homogeneous low signal intensity in all pulse sequences and its anterior margin appears straight [51] (Fig. 2a). Abnormal MR findings of the Achilles tendon of patients with FH are increased anteroposterior diameter, a convex anterior margin and a diffuse stippled appearance (Fig. 2b) [51]. This stippled pattern consists of many low-signal structures of equal size surrounded by high-signal material. Although histologic confirmation is missing, these low signal intensity structures are thought to represent collagen fibres while the high signal material may represent infiltrating foam cells and the associated inflammatory response [51]. Besides classic MRI sequences, spin-wrap MRI techniques have been used to evaluate Achilles tendon signal intensity on fat and water images [60]. In all images, normal tendons showed a very low signal intensity, approaching that of the back-round noise. In patients with FH, the signal intensity of tendons was fourfold and tenfold higher than the background noise, in fat and water images, respectively [60]. However, no matter the type of MR sequence used, MRI does not depict discrete nodules, compatible with xanthomas [55,60,61].

Figure 2.

Magnetic resonance (MR) appearance of the Achilles tendon. (a) A 22-year-old normal male subject (same subject as in Figure 1a) – gradient echo T2-weighted (TR 428 ms, TE 18 ms, flip angle 25°) sagittal scan; the Achilles tendon has a straight anterior border (arrowheads) and exhibits a homogenous low signal. (b) A 21-year-old patient with FH (same patient as in Figure 1b) – gradient echo T2-weighted (TR 428 ms, TE 18 ms, flip angle 25°) sagittal scan; the Achilles tendon is enlarged, has a convex anterior border (arrowheads) and shows a diffuse stippled pattern. No focal lesion is detected with this imaging modality.

Considering all the above imaging techniques ultrasonography may be proposed as the modality of choice because it is a nonradiating cross-sectional imaging method, is not expensive and time-consuming and depicts focal lesions that others fail to detect.

Treatment of Achilles tendon xanthomas

The first treatment of Achilles tendon xanthomas was the surgical resection of the affected tendon. The methods used were total excision and reconstruction of the defect with fascial grafts or subtotal resection. The second method was found to have less complications [62].

However, it is obvious that the best approach towards ATX treatment is treating the metabolic disorder of lipid metabolism, which is the main cause of these lesions. Many recent trials have been trying to define the adequate pharmaceutical treatment that would produce the best results in terms of regression of ATX [63–71].

Probucol, not available in most European countries owing to severe adverse reactions, was one of the first drugs used for this purpose. It has been demonstrated that probucol decreases the levels of serum total cholesterol and HDL cholesterol [63,64]. A marked regression of xanthomas in patients treated with probucol was observed [64–67].

Another proposed lipid-lowering treatment for tendon xanthomas is fibrates. Significant regression of ATX was noted in many studies using several fibrates (e.g. bezafibrate, clofibrate) [68,69]. These drugs have a positive effect on the patient's (FH or nonFH) lipidaemic profile, on ATX regression and are well tolerated and safe [68,69].

It is essential to refer to the results obtained with the most used hypolipidaemic drug category: hydroxyl-methyl-glutaryl-acetyl-CoA (HMG-CoA) reductase inhibitors. The impressive effects of these drugs on atherosclerosis have already been defined. Several new studies have demonstrated that they can also lead to a significant decrease in the size of ATX.

Pravastatin and lovastatin have already been successfully tried for that purpose [36,70]. Moreover, treatment with simvastatin, atorvastatin and rosuvastatin decreased significantly the size of tendon xanthomas [71]. Statins act by changing the lipidaemic profile and are therefore associated with mobilization of cholesterol from tissue stores. In addition to lowering LDL and total cholesterol levels, statins can also increase HDL levels. It is also worth mentioning that higher levels of serum HDL3-triglyceride appear to be a common predictor of the regression of ATX after treatment with pravastatin [36].

Another drug that can affect ATX size is oestrogens. The impact of these hormones on xanthomas has not since been clearly defined. Results from published reports are controversial [69,72].

Finally we should mention the effect of modalities used in homozygous FH patients with ATX, such as plasmapheresis and in particular LDL apheresis. Cutaneous and tendinous xanthomas have been demonstrated to regress after plasmapheresis and LDL apheresis [58,71,73]. Liver transplantation is another rarely used method in homozygous FH patients. After the surgery xanthomas disappeared rapidly [74]. The common mechanism of action of these treatments is the rapid improvement of the patient's lipid profile.

Relationship between tendon xanthomas and atheroma

The main cause of death in dyslipidaemic patients with ATX is IHD-related events [37,75–77]. Several studies have found a strong correlation between the presence of ATX and IHD in FH patients [37,75–77]. Thus, we can assume that these two mechanisms of lipid accumulation have a lot in common.

The first clue indicating a possible correlation between atherosclerosis and formation of ATX is that patients with both lesions have a similar clinical and biochemical profile [43]. In addition, ATX width and area may be good predictors for coronary and aortic root calcification in FH homozygotes [38].

Histochemical analysis studies reported that tendinous xanthomas have an exceptional lipid composition. The only lipid accumulation that resembles the above pattern is the core of advanced atherosclerotic plaques [3]. It is noteworthy that noninvasive imaging experiments with 99 m Technetium-labelled LDL have demonstrated that in both cases the accumulated lipids are derived from the circulation [30,31,78]. However, xanthomas have more rapid exchangeability of cholesterol with plasma [30]. Additionally, lipid staining of tissue sections from both lesions revealed some very interesting data concerning the similarity of the lipid composition in both cases. In both lesions unesterified cholesterol accumulated predominantly in the extracellular space, while esterified cholesterol accumulated both extra- and intracellurarly [29,79].

The main cell that accumulates cholesterol in both lesions is the macrophage. Immunohistochemical studies have demonstrated that oxLDL is likely to play a pathogenic role in lipid uptake in both cases. oxLDL was detected intracellurarly and LDL extracellurarly in both lesions [32]. Furthermore, the titres of the antibodies against oxLDL were significantly higher in patients with ATX and also in patients with coronary atherosclerosis than in controls [35,80].

Another very interesting conclusion from the detailed histochemical analysis of tissues from both lesions is that atheromas and xanthomas seem to use similar defense mechanisms against a high cholesterol load (the enzyme sterol 27-hydroxylase which is contained in the macrophages) [23,81]. Another histochemical observation is that calcification can occur in both atheroma and tendon xanthomas [29,79].

It is also essential to compare the regression of these two lesions. We have already mentioned that the similar lipoprotein composition of ATX and atheromas make them equally difficult to regress. However, it has been shown that certain hypolipidaemic treatments, such as HMG CoA-reductase inhibitors (e.g. lovastatin, pravastatin) or fibrates (e.g. bezafibrate), can decrease the size of ATX, while their positive effect on the regression of atheroma is already proven [36,68–71].

Conclusions

Tendon xanthomas are white nodules that appear in tendons and cause local thickening. They are formed by collagen and foam cells and are associated mainly with FH. Achilles tendon xanthomas can be detected by physical examination and several imaging methods. However, the method of choice is ultrasonography. A range of lipid-lowering modalities has been tried effectively to reduce the size of ATX.

Ancillary