Early diagnosis of peripheral nervous system involvement in Fabry disease and treatment of neuropathic pain: the report of an expert panel

Burning, nagging pain with a symmetric distribution pattern in the palms of the hands and soles of the feet of children or adolescents and peripheral dysesthesias are the first symptoms to raise alarm. Recurrent attacks of pain ("pain crises"), described as "excruciating", "agonizing", "lightning" or "stabbing" pain, often begin in the distal parts of the extremities and may radiate proximally [12, 32]. They can be so intense that the patient is confined to bed [12], and episodes can occur daily [4]. They may be triggered by a rapidly changing core body temperature (e.g., fever, stress, physical activity) presumably due to a decreased ability to sweat [12, 32], although they have been noted even in patients with hyperhidrosis [33]. Additional triggering factors include sudden exposure to cold, rapid changes in humidity, and fatigue [34]. Deep ache, significant joint pain, bouts of unexplained fever, and elevated erythrocyte sedimentation rate can accompany a pain crisis [12, 35, 36].

Neuropathic pain was reported by ~60 and 80% of the affected boys enrolled in the two Fabry disease registries, usually beginning in childhood, and by ~40 to 60% of affected girls, often a few years later than in boys [1, 2]. Most adult males and 60 to 90% of females experience some form of Fabry pain [13, 37‐39]. Over time, neuropathic pain appears to diminish, perhaps due to progressive loss of nerve fibers, but can also become more severe [40, 41].

It has recently been proposed to abandon the use of "acroparesthesia" in the description of Fabry neuropathic pain because, by definition, paresthesias are painless tingling sensations [42]. In fact, paraesthesias appear to be rather uncommon in Fabry disease as they were reported by only ~10% of affected adults [42].

In children with Fabry small fiber neuropathy, reductions in feelings of well-being and in school attendance and a reluctance to participate in sports, gymnastics, and leisure activities are of particular concern. As the patient ages, the symptoms lead to increasing neurological disability and impairment of quality of life in both males and females [7, 10, 43‐45].

As in other conditions with chronic neuropathic pain, pain medication dependence (e.g., opioids) and substance abuse can occur [46]. Patients may develop social-adaptive or psychological functioning deficits [47, 48] and many become depressed, with the presence of neuropathic pain being the strongest predictor of depression [49]. Suicidal ideation has been reported in patients with Fabry disease [49‐51].

The pattern of small fiber dysfunction in Fabry disease is seen in various other diseases characterized by peripheral neuropathy, including diabetes mellitus [52, 53]. However, if provoked crises of excruciating pain occur, the diagnosis of Fabry disease should be highly suspected. The neuropathy in Fabry disease is a length-dependent neuropathy [12, 54] typically associated with a marked reduction of thinly myelinated Aδ fibers (mediating sharp pain, cold perception), unmyelinated C fibers (pain, warmth perception), and loss of unmyelinated autonomic fibers [17, 19, 21, 23, 24, 40, 55‐59]. Female patients, even when asymptomatic, may also be affected by loss of small fiber function [39, 54, 60].

The thermal sensation deficits are initially more pronounced in the feet than in hands, and gradually progress to more proximal parts (Figure 4) [12, 39, 59, 61]. In early stages, impairment of thermal stimuli primarily involves cold perception (Aδ fibers), rather than warmth perception (C fibers) [62], suggesting that the thinly myelinated Aδ fibers are more vulnerable to GL-3-induced damage [36, 40, 63, 64].

Disappearance of pain may not necessarily reflect nerve recovery but rather degeneration of nerve fibers and the sensory neuropathy may progress to numbness and hypoalgesia. Recently, an association between the severity of small fiber neuropathy, determined by cold perception in the feet, and progressive sensorineural hearing loss has been reported, both in male and female patients [65].

Autonomic small fiber damage in Fabry disease is likely to be related to the patients' gastrointestinal dysmotility (e.g., abdominal cramps, bloating, diarrhea, nausea), hypohidrosis, impaired pupillary constriction, decreased tear and saliva formation, Raynaud phenomena, sensory losses, reduced heart rate acceleration upon exercise and, in advanced stages, orthostatic hypotension [17, 37, 66]. The cutaneous response to scratch can be diminished [17, 67], and baroreflex-mediated vasoconstriction can be deficient due to sympathetic vasomotor nerve fiber dysfunction [68].

Pancreatic parasympathetic dysfunction may cause post-prandial diarrhea after consumption of a high-fat meal [69]. Both autonomic sudomotor nerve fibers and sweat gland function are impaired in untreated patients [37, 70]. Hyperhidrosis can also be recognized in patients with Fabry disease, mainly in females. If present, it often manifests in childhood or adolescence [33].

Of note, the generally accepted assumption that autonomic neuropathy plays an important role in the pathophysiology of Fabry disease has recently been questioned by one group of investigators [71]. The observation of nearly normal male sexual function and autonomic control of the cardiovascular system in patients with Fabry disease led the investigators to suggest that end-organ damage, rather than autonomic dysfunction, might play a prominent role in Fabry disease.

As large Aα and Aβ fiber modalities (e.g., motor function, position, and vibration sensation) are generally spared in early stages of Fabry disease, it is not surprising that the results of motor and sensory nerve conduction tests have mostly been normal [19, 21, 34, 36, 40, 57, 61, 66, 72, 73]. With progression of the disease, however, large fiber involvement with associated nerve conduction abnormalities may develop. Reductions of compound muscle action potential amplitudes are seen more often than the slowed conduction velocities [40]. Impaired vibration perception may be a consequence of disturbed transmission of the synchronized impulse volleys in large myelinated afferent Aβ fibers, or by GL-3-induced dysfunction of intradermal corpuscles (mainly Pacinian corpuscles and Meissner bodies [74, 75]) mediating the vibratory stimuli [40, 73]. With age, more severe abnormalities involving large fibers are expected to occur as demonstrated by a significant influence of age on vibratory perception [40].

The mechanisms contributing to development of large nerve fiber abnormalities in a subset of patients are as yet unclear. It has been suggested that Schwann cell pathology due to GL-3 accumulation might account for slowing of conduction velocities [40]. Uremia due to Fabry end-stage renal disease and overuse of alcohol may add to the large fiber pathology [40, 55, 66, 76].

Median nerve conduction abnormalities may indicate median nerve entrapment caused by GL-3 accumulation in carpal tunnel structures. A prevalence of carpal tunnel syndrome of ~25% has been reported for both male and female patients with Fabry disease [36, 55].

The early neurologic manifestations of Fabry disease are initially often subtle and affected children are frequently misdiagnosed as having rheumatism, viral infection, growing pains, "bone problems", psychogenic pain, cryptogenic pain, food intoxication, or non-specific gastrointestinal pain [77]. Although first symptoms appear in childhood, correct diagnosis may be delayed until well into adulthood. Data from the Fabry Registry shows that the mean diagnostic delay was approximately 11 years for both genders [10]. While still undiagnosed and untreated, the disease increasingly impairs the young patient's well-being.

Children and young adults with unexplained peripheral neuropathic pains may be referred to the pediatrician, or the pediatric/general rheumatologist or neurologist. Therefore, they play a critical role in recognizing the array of distinct early features of Fabry disease (Table 1), of which several are related to the small fiber neuropathy. The importance of their role is underpinned by a recent Fabry screening study among individuals with small fiber neuropathy of unknown aetiology; several previously unrecognized patients were diagnosed with Fabry disease [53]. The differential diagnosis of peripheral neuropathy can be significantly narrowed by taking a detailed medical history (Table 2). This should include information on the onset, quality, intensity and distribution of pain symptoms, abnormal responses to normally non-painful stimuli, triggering and exacerbating factors, psychological factors, changes of symptoms over time, and coping skills. Previous and current use of pain medications (prescription, non-prescription and recreational drugs) and smoking habits should be documented. The patient's general status, energy level, co-morbidities, school, work and sport performance, and lifestyle should be explored. Gastrointestinal problems (abdominal pain, post-prandial diarrhea, number of bowel movements per day, bloating, nausea) and hearing problems need to be documented, as well as the family history (e.g., renal/cardiac failure, transient ischemic attack (TIA) or stroke, early death).

A neuropathy pain assessment scale should be used at the initial assessment and follow up examinations. The Neuropathy Symptoms and Change Questionnaire evaluates the number, severity and change of symptoms, as well as motor, autonomic, large fiber and small fiber sensory nerve function [78]. Other available neuropathic pain evaluation tools include the Leeds Assessment of Neuropathic Symptoms and Signs, Neuropathic Pain Questionnaire, Neuropathic Pain Symptom Inventory, Douleur Neuropathique en 4 questions, pain DETECT, and ID-pain [79]. The Total Symptoms Score has been used for grading of neuropathic pain in clinical studies in diabetes and Fabry disease patients with peripheral neuropathy [73, 80].

The patients can also be asked to keep a (weekly) pain diary. A more general pain assessment can be performed by using, for example, the Pediatric Pain Questionnaire or the McGill Pain Questionnaire. The quick version, the Short-form McGill Pain Assessment Questionnaire [81], is particularly easy to administer. Quality of life can be assessed by using the Short Form-36 (SF-36^®) Health Survey (http://​www.​sf-36.​org) or Pediatric Quality of Life Inventory (PedsQL™) [82].

A complete physical examination should be performed as it may provide clues to the diagnosis of Fabry disease (Table 2). It should include measurement of height and weight, heart rate (irregular pulse) and supine and standing blood pressure to determine orthostatic hypotension (defined as a drop in systolic blood pressure by at least 20 mmHg, or in diastolic blood pressure by at least 10 mmHg, within the first three minutes upon active standing up from the sitting or supine position, or upon passive head-up tilt) [83]. The skin should be inspected for presence of vascular skin lesions (angiokeratoma, Figure 1), particularly in the bathing trunk area. Cornea verticillata (Figure 2) is readily detectable by slit lamp examination and present in almost all males and 70% of females with Fabry disease [32, 84]. Routine blood work-up is usually normal. Albuminuria may be detected in young patients and is an antecedent to progressive Fabry nephropathy [85].

The neurologic examination in children, adolescents and young adults should focus on sensations mediated by peripheral small fibers, as overt central nervous system involvement is usually not yet present (Table 2). Relevant, easy to perform and non-invasive sensory perception tests include cold, heat, light touch, pinprick, pressure and vibration perception testing. Assessment of allodynia and hyperalgesia should preferably be carried out in the area of maximal pain. All tests should be performed under comfortable ambient conditions and temperature at the distal lower limbs.

Early in the course of small fiber neuropathy, impaired appreciation of thermal stimuli primarily involves cold perception rather than warm perception [39, 62]. A rather crude test of thermal perception consists of placing tubes filled with cold and warm water, a cold and warm tuning fork, the handle of a reflex hammer, or thermal discs with a polyvinyl surface on one side and a metal surface on the other side (similar to the "Minnesota Thermal Discs" [86]) on the patient's foot. Cold sensation can also be assessed by using acetone spray, or by immersing one leg up to the mid thigh (or one forearm) for 30 seconds into ice-cold water [34]. Normal subjects report little discomfort to a brief immersion of a limb in ice water, but the Fabry patient experiences intense burning pain within 10-30 seconds.

Pain perception and hyperalgesia can be tested by applying just enough pressure with a sharp pin to indent the skin. Loss of discrimination between pinprick sensation and application of a blunt pressure sensation points towards small fiber neuropathy.

Perception of light touch and mechanical allodynia can be tested by lightly rubbing the skin with a cotton swab or by lightly squeezing the patient's leg. Vibration can be assessed by placing a scaled (8/8), vibrating 128 Hz tuning fork to the first metatarsal bone [87].

Symmetrical sensorineural hearing loss can be explored by performing the Rinne test. The stem of a vibrating tuning fork (512 or 256 Hz) is placed on the mastoid process until sound becomes inaudible; the fork is then immediately placed outside the ear. In case of sensorineural hearing loss, bone conduction and air conduction are both equally reduced. However, to reliably assess or rule out hearing loss, three-tone audiometric testing needs to be performed [88].

Abnormal results of the tests described above may lead to more comprehensive examinations trying to document the sensory abnormalities qualitatively and quantitatively. A standardized quantitative sensory testing (QST) protocol to assess the somatosensory profile within 30 minutes and diagnose a small fiber neuropathy is available [53, 56, 89]. The QST battery consists of 7 tests measuring thermal detection, thermal pain and mechanical detection thresholds and mechanical pain sensitivity. An abbreviated version of this protocol assessing cold and warm detection thresholds can be applied within approximately 10 minutes and may facilitate the development of new screening strategies for Fabry disease. Methods to test nerve conduction velocities, sympathetic skin response, quantitative sudomotor axon reflex sweating, saliva and tear formation, and other autonomic dysfunction are not further discussed in this manuscript that focuses on simple bedside tests. Given the availability of an accurate diagnostic laboratory test, nerve or skin biopsies are not required for diagnosing Fabry disease, although skin biopsy can detect small fiber disease in yet asymptomatic patients and may be used to quantify loss of skin innervation [39].

Measuring α-galactosidase A enzyme activity in leukocytes is the gold standard assay of diagnosis of Fabry disease in males [4]. Many females with Fabry disease, however, may have enzyme activity levels within the normal range. Diagnosis in females, therefore, requires genetic studies to detect or confirm the Fabry mutation. Extended familial assessment identified, on average, five family members with Fabry disease for every proband [90]. Therefore, the importance of taking a family history and genetic testing in all members of the family of a newly identified patient cannot be underestimated.

Evaluation and monitoring of the small fiber neuropathy are important as the elicited signs and symptoms may provide insight into possible pathophysiological mechanisms and, in turn, may provide guidance on optimal pain management. Given the progressive nature of Fabry disease, regular objective and comparable monitoring of the neuropathy and its progression using reproducible measures of staging is required for therapy adjustment. Particularly useful may be the neuropathy staging scheme developed by Dyck [91] as it scores sensory, autonomic, and motor components separately, and provides an overall stage. Other available methods to score severity include the Total Symptom Score for grading of pain [73, 80], and the Neuropathy Impairment Score [92].

Current management of pain in pediatric patients with Fabry disease appears to be inadequate. This is as illustrated by the fact that the majority of children in the Fabry Registry who had neuropathic pain reported not to have received any pain medication [1].

Given the episodic and provoked nature of many of the painful crises, it should be emphasized that identification of individual provocateurs is important for conservative management. Patients can change their lifestyle and take preventive measures to avoid exposure to, for example, rapid changes in temperature. The goal of pharmacological management of neuropathic pain should be to decrease pain to a tolerable and functional level and, thereby, to improve the patient's quality of life, activities of daily life, and psycho-social function.

We are not aware of any randomized controlled trial of an analgesic for the treatment of painful peripheral neuropathy in Fabry disease; only some empiric use of drugs has been reported [93‐95]. However, a substantial body of evidence exists for the treatment of other types of painful neuropathy and this may serve as a guide for the physician treating the Fabry disease patient [96‐99]. Pain treatment should generally consider the most likely mechanism in each case. However, in the attempt to achieve optimal pain management, multiple medications targeting different aspects of the complex pathways might be considered in patients with advanced disease. It is recommendable to start pain medication(s) at low dose, and to evaluate the tolerability and effectiveness of a change in medication(s) after 2-3 weeks.

Anticonvulsants are commonly used to reduce pain in neuropathic pain disorders, including diabetic neuropathic pain [96‐99]. Carbamazepine alone [95], or in combination with pregabalin (rather than gabapentin) [99], is recommended as first-line treatment in Fabry neuropathic pain. Symptoms suggestive of autonomic dysfunction should be monitored as dose-dependent exacerbation of autonomic dysfunction has been reported [95]. Antidepressants, particularly dual reuptake inhibitors of both serotonin and norepinephrine (SNRIs; venlafaxine, duloxetine; black-box warning for use in pediatric age group) are also viable options. Because of their anti-cholinergic effect, tricyclic anti-depressants have potential concomitant and difficult side-effects in Fabry patients (e.g. accentuation of autonomic instability). Prescription of opioids carries the risk of drug dependence or abuse. Although they should only be given if other therapeutic options are ineffective, opioids may be helpful in the acute management of intolerable pain crises [96‐99]. Topical application of anesthetics may provide relief of burning pain or hyperalgesia.

With increasing disability and unpredictability and difficulty of controlling the pain, patients are at risk of chronic depression [49]. Patients should be evaluated for psychosocial and behavioral deficits and appropriately treated.

Common non-neurologic symptoms of Fabry disease and risk factors should be carefully monitored and treated as outlined in the general guidelines for the management of Fabry disease patients [100]. Co-morbidities must be addressed and smoking discouraged as it may aggravate airflow obstruction commonly found in patients with Fabry disease [101], and further increase the risk of cardiovascular and cerebrovascular events. Anti-platelet drugs (aspirin, clopidogrel) are recommended for patients at risk of thrombotic events. Dyslipidemia should be treated with statins [100], and recently the use of statins to prevent stroke in Fabry disease has been discussed [102].

Enzyme replacement therapies (ERTs) are available for treatment of patients with Fabry disease and several clinical studies have reported pain outcomes [45, 73, 103‐112]. For example, the Total Symptom Score improved with biweekly ERT at the dose of 1 mg/kg for 18 or 23 months [73], and some adults were able to discontinue their pain medication(s), or doses of pain medications could be tapered [105, 112]. Improvements in the Bodily Pain scale of the SF-36 during 2 to 3 years of treatment with ERT (1 mg/kg), particularly in younger patients, have recently been reported [45]. Signs of recovery of small nerve fibers have been noted. Significant improvements in vibration, heat-pain, and cold perception thresholds, indicating recovery of Aβ, Aδ- and C-fiber function, have been reported in adults treated biweekly with 1 mg/kg agalsidase beta for 18 or 23 months [73].

As for the renal [107, 113] and cardiac complications [114] of Fabry disease, early initiation of treatment may provide a better outcome, i.e., improvement or more optimal preservation of peripheral nerve function. Additional studies are needed in pediatric patients to assess if early-stage treatment can prevent the often devastating peripheral neurologic burden of Fabry disease.