Diagnosis of neuropathic pain: challenges and possibilities

Abstract

Neuropathic pain complicates many diseases of the peripheral and central nervous system and is frequently encountered in clinical practice. The mechanisms underlying its occurrence and chronification remain poorly understood. As a consequence, symptomatic treatment is frequently the only available therapeutic option. An appropriate diagnostic workup is an important prelude to treatment. Moreover, identification of the site of damage in the somatosensory pathway represents a mandatory step in the process of deciding on a disease-modifying therapy for any given patient. The recent revision of the definition of neuropathic pain has introduced the concept of a nosologic-based approach to the diagnosis, which is expected to be supported by the demonstration of a relationship between the clinical picture and a lesion or disease. This underscores the need for precise diagnostic assessment of the patient. In the last decade, a number of tools including validated scales, psychophysical tests and morphometric analysis of small nerve fibers carrying thermal and nociceptive sensation have been developed; these can provide important information about the quality and intensity of the multiple features that characterize neuropathic pain. More recently, advances on the recognition of a molecular substrate for neuropathic pain, both in terms of susceptibility and novel gene mutations, have provided the potential for new diagnostic perspectives and a path toward a better comprehension of the pathogenetic mechanisms. This editorial addresses briefly the impact of these developments on the diagnosis of neuropathic pain in clinical practice.

Keywords::

• diagnosis
• neuropathic pain
• neuropathy
• nociceptive
• quantitative sensory testing
• skin biopsy
• sodium channels
• somatosensory

1. Neuropathic pain in clinical practice

1.1 How definition impacts patient care

Neuropathic pain is a relatively common condition, occurring in about 5% of the population [1]. It can follow central and peripheral nervous system diseases, leading to distinct clinical pictures that reflect the anatomic–functional impairment through the somatosensory pathway. This editorial will focus on neuropathic pain associated with peripheral nervous system diseases.

Its definition has been long debated and was recently revised in response to the need for more narrow criteria that can be used in clinical practice for the differential diagnosis with other forms of pain [2]. Indeed, the definition formerly adopted by the International Association for the Study of Pain (IASP) of a ‘pain initiated or caused by a primary lesion or dysfunction in the nervous system’ [3] did not address the issue of the localization of the site of injury or the nature of pain triggered by motor dysfunction (e.g., the pain that can accompany spasticity due to pyramidal tract lesions). In 2007, a panel of experts proposed the revised definition of neuropathic pain as ‘pain arising as a direct consequence of a lesion or disease affecting the somatosensory system’ [2]. This new definition replaced an ill-defined term with a nosologic-based approach to diagnosis and introduced the boundaries that required identification of injury or dysfunction of a structure involved in the transduction and perception of sensation, from the most distal nerve fibers in the skin to the brain, as prerequisite for the diagnosis of neuropathic pain.

A grading system was introduced to provide a level of certainty from possible to probable and definite based on the plausibility of the distribution of symptoms and signs and the demonstration of their relationship with the lesion or disease, using validated confirmatory tests (e.g., neurophysiological or neuropathological exams). Overall, this approach emphasized the need for an adequate workup, aiming to provide an understanding of the etiology of the lesion, which in turn could facilitate the choice of an appropriate disease-modifying therapy. Examples of how this impacts patient care are provided by paraneoplastic or immune-mediated painful sensory neuronopathies [4] and vasculitic neuropathies [5], in which a delay in the diagnosis means a delay in initiating life-saving therapies. Conversely, patients not meeting the criteria for neuropathic pain should be referred to a different diagnostic workup, including rheumatologic and psychiatric evaluation, before starting a pain therapy.

1.2 Screening tools

Several questionnaires have been validated and used as discriminative or descriptive instruments. Among the first, one example is the Neuropathic Pain Questionnaire (NPQ), which was reported to differentiate neuropathic from non-neuropathic pain patients with a degree of sensitivity and specificity around 70%. Another is the 10-item questionnaire DN4 (Douleur Neuropathique 4 questions), which combines two questions on symptoms (seven items) and two questions on clinical examination (three items) limited to negative signs (touch and pinprick hypoesthesia) and mechanical dynamic allodynia (brushing). The DN4 has been proposed as a diagnostic tool for neuropathic pain based on the ability to correctly classify 86% of patients. [6]. Sensitivity and sensibility of DN4 were reported to be high (above or about 80%) either for the complete form or for the seven-item format limited to the interview. Based on the cut-off value of 4 for the 12-item format (3 for the 7-item one), and the high discriminative value, the authors indicated that the clinical examination might not be mandatory for the diagnosis of neuropathic pain per se, although necessary to diagnose a nervous lesion. However, the validation study, based on the comparison between patients with diseases known by definition to cause neuropathic pain (e.g., nerve trauma, postherpetic neuralgia, polyneuropathies, spinal cord injury) and non-neuropathic pain (e.g., osteoarthritis, inflammatory arthropathies, mechanical low back pain), did not include patients with mixed painful conditions (e.g., mono- or multiple radiculopathy) or conditions mimicking neuropathic pain such as fibromyalgia or somatoform disorders. This may explain the much higher values of sensitivity and specificity compared those found with the NPQ. The DN4 has been recently validated in diabetic painful neuropathy [7].

Among descriptive tools, the Neuropathic Pain Symptom Inventory (NPSI) has been validated and used in several countries. Although developed to evaluate the different features of neuropathic pain [8], the NPSI was tested to demonstrate its ability in correlating symptoms to etiology or localization of pain. Focused studies could not prove this hypothesis. A study including 482 patients with pain associated with peripheral or central lesions failed to demonstrate the association between symptoms and etiologies, types of lesions or pain localization with the exceptions of trigeminal neuralgia and postherpetic neuralgia [9]. A more recent German study on 68 patients with neuropathic pain caused by central or peripheral nervous system diseases and 169 patients with non-neuropathic pain (osteoarthritis and headache) showed that the NPSI could distinguish between neuropathic and other pain syndromes only on group basis, but not for individual patients [10]. Based on these findings, the proposed rationale for subgrouping etiologically different neuropathic pain patients into specific categories for therapeutic management [9] using the NPSI does not seem to be feasible in clinical practice.

In addition, all these questionnaires, like most of the outcome measures used in health care, are constructed as multi-item ordinal scales based on the classical test theory [11] assuming that i) the sum scores are at the linear level; ii) all the items have equal relevance and weight; and iii) given differences in sum score across the metric used have equal meaning. Concerns have been raised about inappropriate analysis of the generated summed scores, which are erroneously assumed to be at the interval level [11]. To overcome these limitations, modern scientific methods, such as the Rasch, have been adopted to transform ordinal scores into interval-level variables [12]. Future studies in pain should consider selecting or revising measures based on one of the modern clinimetric methods.

1.3 Analysis of sensory thresholds

Quantitative sensory testing (QST) has been used to determine the functional impairment of large or small caliber sensory fibers by the detection of vibratory or warm and cooling thresholds, respectively. Vibratory threshold can be reliably replaced by the 128 Hz tuning fork [13], making this evaluation much easier at the bedside. Several works demonstrated thermal threshold abnormalities in patients with damage of small nerve fibers and pain phenomena. QST was used to investigate whether the individual pattern of signs and symptoms reflecting specific alteration in nociceptive processing allowed to stratify patients based on their somatosensory profile rather than the underlying etiology, thus providing a mechanism-based classification of neuropathic pain. The analysis of 13 QST parameters including measurements of negative (e.g., hypoesthesia) and positive (e.g., hyperesthesia) sensory thresholds performed in a cohort of 1236 patients of neuropathic pain patients did not confirm this hypothesis [14]. Indeed, there was a remarkable phenotypic heterogeneity across the major neuropathic pain syndromes that included an overlap between central and peripheral nervous system diseases. For example, polyneuropathy and central pain were both reported to have low rates of positive and high rates of negative sensory signs, with similar frequencies of paradoxical heat sensation. On the other hand, this study demonstrated a quite high incidence of mechanical allodynia in postherpetic neuralgia, reflecting what is typically reported by the patients and observed at the neurological examination. The conclusion that postherpetic neuralgia cannot be considered a model for all other types of neuropathic pain should be regarded with particular attention in the design of future clinical trials, which typically include this condition and painful diabetic neuropathy to assess the efficacy of new treatments.

A further interesting observation was that in about one-fourth of patients with trigeminal neuralgia, the QST battery could detect subclinical sensory abnormalities. A similar finding was reported in a study on patients with trigeminal neuralgia with laser-evoked potentials, which, however, demonstrated abnormal latencies even on the nonpainful side compared with age-matched controls and significant correlation with the carbamazepine dose [15]. QST seems to have a relatively poor power in terms of localization of the injury in the somatosensory system and the interpretation of results comparing with normative data in individual patients is difficult [16]. An example is given by the results of a recent study showing no differences in terms of thermal thresholds between two groups of patients with nerve injury complaining or not of pain, and higher frequency of hyperphenomena (light touch allodynia and reduced mechanical pain thresholds) in the pain group only [17]. It appears that the analysis of sensory thresholds by QST may provide information on the rearrangement of the nociceptive pathway functions after a lesion or disease and provides a valid complement to the bedside examination. However, it should be used in relation to the clinical context and in conjunction with other tests, and not alone for the diagnosis of a neurological lesion [16]. Researchers should also strive to get international consensus on QST aspects such as type of instruments and assessments, standardized conditions during assessment and proper definition of normative reference values.

1.4 Analysis of skin nerve fibers

Skin biopsy with morphometric assessment of intraepidermal nerve fibers (IENF) was introduced into clinical practice about 15 years ago. This technique permits demonstration of the predominant and even exclusive degeneration of unmyelinated nerve fibers carrying thermal and nociceptive sensations. IENF have somatic functions and widely express receptors such as TRPV1, which enable them to function as distal nociceptors [18]. The density of IENF can be reliably assessed based on age- and gender-adjusted normative values [19] in order to confirm the diagnosis of possible or probable small fiber neuropathy [20,21]. The relationship between neuropathic pain and IENF has been investigated in several studies including HIV, diabetic and idiopathic painful neuropathies. What emerged is that the most severe loss of IENF suggests more severe neuropathy and higher risk of neuropathic pain, whereas regeneration of these distal axons is a sign of recovery from neuropathy and is associated with reduction in pain [22]. Examples are the recovery from small fiber neuropathy and pain observed in hypothyroidism after hormonal therapy [23], in steroid responsive neuropathy [24] and in impaired glucose tolerance after metabolic improvement [25]. However, there is no correlation between the density of IENF and either intensity or features of neuropathic pain [26]. This emphasizes that skin biopsy is a tool to diagnose small fiber damage rather than neuropathic pain per se. Indeed, a profound skin denervation can be found also in hereditary sensory and autonomic neuropathy type IV with insensitivity to pain [27] as well as in Ross syndrome characterized by altered sweating but no pain complaints [28].

The approach to the diagnosis of neuropathic pain using skin biopsy in patients with no clinical and neurophysiologic evidence of large fiber neuropathy demonstrated once more time that the localization of the site of injury must be regarded as the first step of the diagnostic workup. Indeed, the correct diagnosis of small fiber neuropathy provided a basis for identification of previously undiagnosed diabetes or impaired glucose tolerance at 2-year follow-up in about 20% of patients with symptoms of neuropathic pain [26]. This approach also provided a basis for identification of novel mutations of the SCN9A gene, encoding the Nav1.7 subunit of sodium channels, in about 30% of patients with idiopathic small fiber neuropathy [29]. These mutations induced a distinct profile of sensory neuron hyperexcitability, which underlies channelopathy-associated small fiber neuropathy. Identification of this new syndrome enlarges the spectrum of channelopathy-related pain disorders, along with primary erythromelalgia, paroxysmal extreme pain disorder and channelopathy-associated insensitivity to pain [30], could help in shedding light on some pathophysiological mechanisms of neuropathic pain and will hopefully create the background for targeted analgesic therapies.

2. Expert opinion

The term neuropathic pain represents the synthesis of a constellation of symptoms and clinical signs and should be considered the epiphenomenon of a neurological disease or lesion perturbing the transduction of sensory stimuli throughout the somatosensory system. Therefore, neuropathic pain should not be regarded as a disease per se, but as a condition indicating the existence of a possible damage in the nervous system. It follows that the primary need is to identify the location of the lesion and, therefore, its etiology. This approach should lead clinicians to seek the possibility of a causal therapy before or along with symptomatic treatment. The recent redefinition of neuropathic pain offers a guideline that builds on these theoretical bases and is expected to have a direct impact on the clinical practice among neurologists and pain specialists.

The diagnostic workup in patients with neuropathic pain can certainly benefit from validated questionnaires, which, however, should be regarded as screening instruments and do not substitute for either a thorough clinical evaluation or for the exams needed to identify the site of injury in the central or peripheral nervous system. Confirmatory tests should be performed with a dynamic perspective in order to examine the hypothesis of the plausible relationship between symptoms and putative lesion or disease that is raised by the clinical examination. Among them, the analysis of the biophysical properties of primary sensory neurons has provided important information about patients housing mutations in voltage-gated channels. Indeed, sensory neuron hyperexcitability and not IENF degeneration per se can explain the occurrence of neuropathic pain in small fiber neuropathy.

The current armamentarium, largely consisting of nonspecific analgesic treatments, reflects the lack of evidence that mechanisms underlying peculiar symptoms (e.g., paroxysmal pain) or categories of symptoms (e.g., spontaneous or evoked pain) can be attributed to specific diseases (e.g., diabetic neuropathy, spinal cord injuries, multiple sclerosis). A more mechanism-based approach to the diagnosis of neuropathic pain, along with recent discoveries on its pathophysiology, will hopefully provide base for more effective management of patients in clinical practice and for the development of more efficacious drugs in the future.

Declaration of interest

The authors state no conflict of interest and have received no payment in preparation of this manuscript.