Conventional Therapy in Adults With XLH Improves Dental Manifestations, But Not Enthesopathy

X-Linked hypophosphatemia (XLH; OMIM no. 307800) is an X-linked dominant disease due to inactivating mutations in the PHEX gene that increase fibroblast growth factor 23 (FGF23) expression in bone, resulting in renal phosphate wasting and inappropriately normal calcitriol concentrations (1). Although there are currently no approved therapies for XLH, high-dose calcitriol and phosphate salts have been the mainstay of therapy for the past three decades (1, 2). There is general agreement that most affected children should be treated (if the administration and safety of therapy can be adequately monitored), but there is a lack of consensus regarding when to treat adults (1, 2). Features of XLH that frequently affect adults include short stature, lower extremity deformity (from rickets during childhood), bone pain, pseudofractures, osteoarthritis, stiffness from enthesopathy, and tooth abscesses. Obviously, treating adult patients will not correct lower extremity deformities or short stature, and except for patients with pseudofractures, XLH patients do not have increased fracture risk (3). However, some investigators have reported improvements in bone pain and osteomalacia in treated adults (4). In this issue of the JCEM, Connor et al (5) report retrospective observations made in 52 XLH patients regarding two important disease manifestations, enthesopathy and dental disease. Unfortunately, there are no large prospective trials on the effects of medical therapy on enthesopathy or dental disease. As for many other rare diseases, large, prospective, randomized, controlled trials are not feasible. The present study is the largest to date on the effect of therapy on these two disease manifestations, and it provides important information for the care of persons having XLH.

Predictors of Disease Severity

Nonrandomized studies are subject to unrecognized confounding variables and false assumptions of causality. Nonetheless, the baseline data from the present study give useful information that may challenge some assumptions about factors influencing XLH disease severity. Some of these factors are genotype, sex, biochemical variables, and the possibility that patients who received more treatment during adulthood may have been either more severely affected at the outset or more adherent to therapy (compliant patients might have better oral hygiene). In this regard, the patients who had longer treatment periods as adults had a greater number of osteotomies, favoring the view that they may have had more severe disease, which would tend to obscure true differences between treatment groups (making therapy look less effective). Although a discussion about genotype/phenotype correlations is beyond the scope of this editorial, the limited data available do not lend strong support to there being a strong genotype/phenotype correlation.

Although biochemical variables before treatment are not available, PTH area under the curve was associated with the number of sites of enthesopathy with borderline significance. FGF23 concentrations at study end did not predict either enthesopathy or dental disease. However, in accord with other studies (6–8), FGF23 concentrations were much higher in the patient groups that received therapy the longest. Interestingly, Whyte et al (9) found that height Z-score was inversely correlated to tubular maximum for phosphate corrected for glomerular filtration rate, so perhaps some biochemical variables might predict clinical severity.

Although Connor et al found in their multivariate-adjusted model that enthesopathy was more common in men than in women (Tables 3 and 4 in Ref. 5), Hardy et al (10) found that it was slightly more common in women (74 vs 60%), but men had more than twice the number of affected sites. Polisson et al (11) did not observe a difference in enthesopathy between the sexes. It is often assumed that males have more severe disease than females; however, biochemical profiles are similar between affected men and women (9) and Hyp mice (12, 13), and height Z-scores were similar between affected boys and girls (9). Studies comparing sex differences regarding dental disease are few, and most studies had inadequate power to detect gender differences (14–16) or failed to analyze data by sex (3, 17). In the Connor et al study, these differences reached statistical significance only after substantial adjustment (Table 6 in Ref. 5), despite a large population. Thus, gender differences are not as profound as one might expect. Although there may be slight differences between males and females for some features of the disease, the variance in phenotypic severity between patients of the same sex (and frequently from the same family) is often much greater than the variance between the genders. Therefore, it may be wise to decrease emphasis on male/female differences, particularly in the setting of genetic counseling.

Dental Disease

Of note, in the study of Connor et al (5), most of the sample (61.5%) had severe dental disease (defined as more than five tooth abscesses) vs nonsevere disease (five or fewer abscesses). Impressively, 75% of patients who were not treated in adulthood had severe dental disease, whereas “only” 47% of those who were treated their entire adult life had severe dental disease. These findings emphasize how common dental abscesses are in XLH. Of note, age is a confounding factor in this analysis, with older patients being more likely to have more than five abscesses; perhaps this results simply from longer disease duration. Nonetheless, those with a greater proportion of time treated during adulthood (or during total lifetime) had lower odds ratios for having more than five dental abscesses.

Tooth abscesses likely result from profound abnormalities of the dentin (reviewed in Ref. 18). In brief, dentin is severely undermineralized in XLH patients' teeth. Chaussain-Miller et al (16) observed large interglobular spaces between unmerged calcospherites in untreated children with XLH. Minor enamel defects over poorly mineralized dentin allow bacteria to invade the pulp, leading to a tooth abscess in the absence of trauma or tooth decay (16). Chaussain-Miller et al (16) also showed that defects in dentin were improved in three children treated with active vitamin D analogs and phosphate. These findings, taken together with those of Connor et al (5) in adults, support the concept that medical therapy decreases dental pathology and the resulting abscesses in XLH patients. However, despite state-of-the art XLH therapy (these patients were medically managed in a tertiary care center by expert physicians), tooth abscesses were still very common. Although appropriate dental care will not eradicate tooth abscesses in XLH patients, poor dental care and hygiene leading to caries are likely to worsen an already dismal situation. Therefore, all XLH patients should be encouraged to maintain optimal dental care and hygiene.

Enthesopathy

Enthesopathy (calcification of joint capsule, tendon insertions, and ligaments) is a very common feature of XLH, particularly as patients age, causing pain, stiffness, and limited mobility (5, 10, 11). It can be one of the most disabling features of the disease, particularly in older patients. The cause of enthesopathy in XLH is not entirely clear, but it occurs in untreated patients (10, 11) and in the mouse model of XLH, the Hyp mouse (19). Critically, enthesopathy is also found in the FGF23-overexpressing transgenic mouse (20), indicating that FGF23 and/or the metabolic effects of FGF23 excess lead to enthesopathy. Because enthesopathy can be one of the most disabling features of adult XLH, it is disappointing that our current therapy does not appear to alter the course of this feature substantially. The ability to prevent or mitigate enthesopathy would be a major breakthrough for patients having XLH. In the future, therapy with FGF23 neutralizing antibody (in clinical trials) might theoretically prevent enthesopathy (whether the pathophysiology is due to FGF23 itself or its metabolic consequences). However, such studies in patients would take years to complete, and extended randomized studies are likely too costly and impractical to be performed. The presence of enthesopathy in the Hyp mouse provides an ideal opportunity for proof-of-concept studies testing whether treatment to block FGF23 activity could reduce or eliminate enthesopathy.

For now, therapy with high-dose calcitriol and phosphate does not appear to prevent, alleviate, or exacerbate enthesopathy. Enthesopathy is associated with increased body mass index (BMI), suggesting that maintaining optimal weight might decrease the burden of enthesopathy. However, evidence for weight loss/maintenance as a management strategy is lacking, and it is possible that enthesopathy contributes to higher BMI through impaired mobility and decreased activity. However, it seems prudent to advise patients to maintain a normal BMI.

In summary, Connor et al (5) report clinical data on the largest XLH cohort examined to date. Their data suggest that current standard XLH treatment is associated with reduced risk for severe dental disease but does not influence the occurrence of enthesopathy.