Lower Limbs Axonal Neuropathy in Polyarteritis Nodosa with Onset of Neuropathic Symptoms

Objective The retrospective study aimed to investigate the most common electrophysiological changes in patients with polyarteritis nodosa (PAN) of peripheral neuropathy onset and supply some clinical data to neurologists to pay attention to PAN. Methods ten and and investigated in our


Introduction
Polyarteritis nodosa (PAN) is a necrotizing vasculitis involving medium and small arteries 1, 2 . PAN is a systemic disease and affects patients of any sex or ethnic background. Before vaccination against hepatitis B virus (HBV) was available, more than one third of adults with PAN were infected by HBV. The incidence rate of PAN decreased signi cantly 3,4 , as there is widespread vaccination all over the world and increased awareness and improved diagnostic techniques to help doctors to recognize other systemic necrotizing vasculitis (i.e., ANCA-associated vasculitis, cryoglobulinemic vasculitis) as distinct entities.
Although PAN is not fully understood, available evidence suggests that the pathogenesis is different than 30 years ago. Necrotizing in ammation of the blood vessels probably involve both the innate and adaptive immune systems 5 , induced by chronic viral infections 6 , paraneoplastic syndrome 7 , immune de ciencies 8, 9 , gene mutant (DADA2 10,11 , SAVI 12 , FMF 13 ) etc. The most widely implicated mechanism is the development of lesions induced by immune complexes. PAN is becoming a rare disease, together with the highly variable clinical presentation of PAN, prompt identi cation of PAN is important because they are associated with an increased risk of mortality.
Peripheral neuropathy is often the most frequent and earliest symptom of PAN 14 . Vasculitis neuropathy as a usual manifestation of PAN occurs in 50-75% of PAN patients 15,16 . Nerve conduction studies (NCS) are an essential tool in the evaluation of the peripheral nervous system. Although it is well known that axonal neuropathy affects one or more nerves in system vasculitis neuropathy, more detailed peripheral nerve injury needs to be determined to identify PAN more quickly. Therefore, we conducted the retrospective study to investigate the most common electrophysiological changes in patients with PAN of peripheral neuropathy onset and supply some clinical data to neurologists to pay attention to PAN.

Study population
We Totally, 10 PAN patients were included in our study. Clinical characteristics and electrophysiological ndings were extracted from the medical records. And 13 matched healthy controls were recruited from volunteers and all wrote informed consents.

Electrophysiological Assessment
Electrophysiological assessment was performed with a Dantec Keypoint EMG machine and done following standard procedures 17 . Nerve conduction studies (NCS) were performed on the more severe side, if abnormal involved both sides in PAN patients. In addition, the right side was a routine option for patients with normal extremities and healthy controls. Sensory nerve action potential (SNAP) by anterograde recording method in all subjects were performed on median, ulnar and sural nerves. Motor NCS were conducted on median, ulnar and peroneal nerves. While recording the compound muscle action potential (CMAP), the median nerve was stimulated at the wrist and elbow; the ulnar nerve was stimulated at the wrist and proximal elbow; and the peroneal nerve was stimulated at the ankle and above bula head. Measurements included distal latency, nerve conduction velocity (NCV) and amplitude (value from baseline to negative peak). The F-wave latency (FWL) and frequency (FWF) were calculated after 20 stimulations on median, ulnar and peroneal nerves. Room temperature was maintained to ensure that the skin temperature remained at > 31℃ during all recording time.

Statistical Analysis
Patient characteristics are reported as the number and percentage for categorical variables. Continuous variables were expressed as mean ± standard deviation (SD), whereas categorical data were presented as proportions. Differences in data between the 2 groups were analyzed by Mann-Whitney U-test. Twotailed P value < 0.05 was statistically signi cant. All analyses were conducted with IBM SPSS Version 21(Ehningen, Germany).

Clinical characteristics
The 10 PAN patients included 6 females, and clinical characteristics were shown at Table 1. The median age on visit was 55.8 ± 15.47 years. Time from disease onset to electrophysiology examination was 3.77 ± 2.34 months. The frequent onset pattern was subacute (80%) followed by acute type (20%). In 9 cases (90%), onset symptoms of neuropathy were con rmed at low limbs by history or examination, and in only 1 (10%) case, symptoms were start from both upper and lower limbs. All patients suffered asymmetric peripheral neuropathic symptoms, and 7 patients (70%) had more severe symptoms in the distal limb than the proximal. There were 13 matched-healthy persons in the control group, including 5 males and 8 females, with a mean age of 55.17 ± 9.67 years.   Table 2. Compared with controls, CMAP amplitudes were signi cantly decreased (P < 0.01) at wrist and elbow points of median nerve and at ankle and above Fibula head points of peroneal nerve in PAN patients (Fig. 1A). No difference of CMAP amplitudes was found at both distal and proximal points of ulnar nerve (Fig. 1A). In addition, sensory nerves were more damaged than motor nerves, SNAP amplitudes of median, ulnar and sural nerves were all much signi cantly decreased in PAN patients compared to controls (p < 0.001, Fig. 1B). Only 2 cases (10%) had abnormal amplitude of CMAP of median nerve in upper limbs, yet the abnormal rate of peroneal nerve (6 cases, 60%) and sural nerve (5 cases,50%) in lower limb was much higher than that in upper limb (Table 2), suggesting more severe changes in the lower limb. Besides, there were no signi cant changes in the values of NCV and distal latency of all recorded motor and sensory nerves, and the similar results were observed in FWL and FWF of motor nerves (Supplement Table 1-3).

Discussion
PAN can affect almost any visceral organ, but it targets the peripheral nerves more often than other organ systems 3, 18 . So far, few studies focused on peripheral neuropathy due to PAN. Some of them concentrated on peripheral neuropathy in systemic vasculitis. According to our study, all PAN patients suffered asymmetric peripheral neuropathy at onset stage, which were in accord with mononeuropathy or multiplex mononeuropathy 3,19 . As the disease progress, multiple nerves were rapidly involved resulting in generalized symmetric polyneuropathy, just like previous studies on primary systemic vasculitis [20][21][22] . Fever is not part of the ACR diagnostic criteria nor the three indices proposed by the French Vasculitis Study Group (FVSG) 23 . Most patients with PAN develop organ involvement within weeks of developing fever 24 . In our study, 80% of patients had fever before the onset symptoms or as an accompanying symptom. Thus, PAN need to be kept in mind when patient with mononeuropathy or multiplex mononeuropathy presented weeks of fever.
In our study, the characteristics of peripheral neuropathy in PAN patients are consistent with asymmetric axonal neuropathy that has a predilection for the lower extremities, affects distal limbs more severely than proximal, and involves both motor and sensory nerves. Previous studies have shown that sensorimotor abnormalities on NCS and the presence of a pure axonal neuropathy were most consistent with pathologically con rmed vasculitis 25,26 , which is consistent with our ndings. The lower limb is the most susceptible to nerve injuries in PAN patients, since the high frequent of disease onset rate appeared at lower limb in the clinical data and the abnormal amplitudes of peroneal and sural nerve were detected in NCS. Despite the less rates of site of disease onset and abnormality of CMAP amplitude of median nerve, the mean values of CMAP amplitude of median nerve and SNAP amplitude of median and ulnar nerves were signi cantly declining, which are considered to exist subclinical changes in upper limbs of patients with PAN. Ulnar motor nerve was selectively spared in our result, which need to be further veri ed through enlarging the sample size of patients.
PAN can make arterial aneurysm and thrombosis in many organs, including brain 27 , heart 28 , kidney 29 , extremities 30 etc., due to in ammatory lesions of the blood vessels and necrosis of the vessel wall. It's associated with an increased risk of mortality, understanding the outstanding features of nerve conduction might enable us to quickly recognize PAN to start speci c treatments as early as possible.
Our ndings suggest that the asymmetric motor and sensory axonal neuropathy is the most common style, lower limbs are the most frequent site of involvement, and the severity of peripheral nerve involvement is more signi cant in the distal limb than in the proximal in patients with PAN of peripheral neuropathy onset. It is advisable for neurologists to consider PAN in the face of patients demonstrating above clinical and nerve conductive features as well as weeks of fever.

Declarations
Ethics approval and consent to participate The study was approved by the central institutional review board at The Second Hospital of Hebei Medical University. All participants or their legal guardians gave written informed consent to participate in the study.

Figure 1
The compound muscle action potential (CMAP) and sensory nerve action potential (SNAP) amplitudes of motor nerves and sensory nerves in patients with PAN and controls. (A) CMAP amplitudes of median and peroneal nerves in patients with are signi cantly lower than those of controls. (B) SNAP amplitudes of median, ulnar and sural nerves in patients with PAN are also signi cantly lower than those of controls.
Star scatters on the left of boxes are the values of each person, the bars show the ±SD, the boxes show the values ranged from 25% to 75%, the dots inside boxes showed the mean value of each group, and the lines inside boxes showed the median value of each group.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. supplement.doc