The scratch collapse test in the diagnosis of compression of the median nerve in the proximal forearm

Introduction

The clinical manifestations of proximal entrapment of the median nerve (PEMN) in the forearm include aching pain in the anterior forearm, loss of key and tip pinch strength, including the typical inability to make an ‘OK’ sign (Chi et al., 2010), and loss of fine motor skills and a sense of clumsiness (including dropping objects). The severity and completeness of these motor symptoms vary substantially (Pham et al., 2014). Less frequently, paraesthesias in the median nerve innervated region of the hand similar to a carpal tunnel syndrome (CTS) may be also present (Hagert, 2013; Hagert and Hagert, 2008; Lee et al., 2012).

Up to four different situations have been identified: the anterior interosseous nerve syndrome (Kiloh and Nevin, 1952); the pronator syndrome (Seyffarth, 1951); and more recently, the superficialis and the lacertus tunnel syndromes (Hagert, 2013). The existence of these as individual syndromes has been questioned (Lafon et al., 2013; Miller-Breslow, 1990).

The diagnosis of PEMN is not easy and requires a high index of suspicion. Nerve conduction studies (NCS) are usually normal and thus unhelpful in PEMN (Pham et al., 2014). Ultrasound has been reported to identify the site of compression by virtue of the change in the diameter of the nerve (Kodama et al., 2015). However, in the absence of a mass or haematoma causing compression, it may not be easy to make a diagnosis with ultrasonography, since the fibrous bands or scar tissue may be too small to visualize (Choi et al., 2015). Since there are no objective tests to verify the diagnosis, we must rely on clinical examination.

Our purpose was to assess the usefulness of the scratch collapse test (SCT) before and after treatment of PEMN.

Methods

After institutional review board approval, all patients diagnosed with PEMN and surgically treated from June 2013 to December 2015 with a minimum follow-up of 6 months were included. We reviewed the electronic records of all cases and we contacted all patients for a new assessment in the office.

The diagnosis was made based on the patient’s symptoms and signs (weakness of flexor pollicis longus and flexor digitorum profundus including inability to make the ‘OK’ sign, aching pain in the anterior forearm) and a positive SCT in the affected forearm performed on three consecutive visits to the office. The contralateral non-affected forearm was used as control for the SCT except in one patient who had bilateral PEMN. Associated CTS (two of 18 patients) was diagnosed from the patients’ symptoms and signs and confirmed by NCS. Magnetic resonance imaging (MRI) was done in all patients to exclude other space-occupying lesions that could cause the symptoms.

All preoperative and follow-up examinations and data collection were carried out by the two authors. The gender, age, time from onset of symptoms to diagnosis and clinical features reported by the patient and assessed by the clinician at the office were recorded. The postoperative SCT at first and subsequent consultations, complications occurring during the follow-up and subjective assessments were noted. All patients completed the DASH questionnaire (from 0 to 100 points) (Rosales et al., 2002) and the Visual Analog Scale for pain (VAS; from 0 to 10 points) before surgery and at final follow-up.

Surgical treatment was carried out by the senior author (PJD) (level of expertise 4, specialist – highly experienced; Tang and Giddins, 2016) and operative findings were recorded.

The SCT was done as described by Cheng et al. (2008) for the ulnar nerve at the elbow and median nerve at the wrist but scratching the skin over the median nerve at the proximal forearm level instead. The test was carried out with the patient facing the examiner, with arms adducted, elbows flexed, and both hands outstretched with wrists at neutral position. The patient was asked to carry out simultaneous external rotation of both shoulders, keeping the arms adducted and to sustain steady resistance while the examiner pushed against both of the patient’s forearms. The patient was then asked to maintain the same position of the arms. The examiner then supported the arm to be tested and scratched the skin overlying the median nerve along its course in the proximal forearm with the fingertips. The first step was then immediately repeated. The SCT was considered to be positive if the patient demonstrated a momentary loss of external resistance tone on the affected side after scratching over the nerve. This loss of muscle resistance was brief, with the patient regaining strength almost immediately on repeat resistance testing.

Results

Eighteen consecutive cases in 17 patients met the inclusion criteria. There were ten affected arms in nine men and eight arms in eight women. The mean age at surgery was 29 years (range 10–56). There were 12 right affected forearms and the dominant arm was affected in 12 of 18 arms. Symptoms most frequently reported by patients were aching pain in the anterior forearm (13 of 18 arms) and weakness affecting both the flexor pollicis longus and flexor digitorum profundus to the index finger (12 of 18 arms). The time from onset of symptoms to diagnosis was 18 months (range 2–96). NCS were reported as abnormal only in the two patients who had both PEMN and CTS. MRI showed denervation signs in the pronator quadratus in two cases. Follow-up was 17 months (range 6–36).

The SCT was positive in the affected forearm in three preoperative consecutive consultations in all cases and it was negative at the first consultation 5–7 days after surgery and subsequent postoperative consultations in all patients. It was negative in the contralateral forearm in all clinical assessments before and after surgery, except for the patient who had bilateral involvement.

In all cases, an anatomical reason for nerve compression was identified at operation. Isolated entrapment of the median nerve at the level of the lacertus fibrosus was identified in 2/18 patients, at the vascular leashes across the nerve in 1/18 patients and at the fibrous arch of the flexor digitorum in 3/18 patients. More than one area of entrapment was identified in 12/18 of cases.

The mean preoperative DASH questionnaire score was 39 points (95% confidence interval [CI] 24–53) and the mean postoperative score was 5 points (95% CI 0–15). The p-value with Student’s t-test for paired samples was p < 0.001.

The mean preoperative VAS score for pain was 7.1 points (95% CI 3.8–10) and the postoperative score was 0.9 (95% CI 0–3.3). The p-value with Student’s t-test for paired samples was p < 0.001.

Of 12 cases presenting with weakness at the time of diagnosis, seven achieved full recovery of strength at final follow-up scoring 5 on the Medical Research Council (MRC) 5-point scale (Medical Research Council, 1976) while making the ‘OK’ sign. Five showed improved strength, scoring 4 on the MRC scale compared with the contralateral side.

All patients returned to their normal activities without sequelae, but three developed temporary dysaesthesias in the territory of the lateral cutaneous nerve of the forearm which spontaneously resolved; a keloid scar developed in two cases.

The result was rated as excellent or very good in 16/18 arms and as good in the other two.

Discussion

In our series, the SCT was positive in the affected forearm in three consecutive preoperative consultations in all cases and it was negative at the first and subsequent postoperative consultations in all patients. We believe that the SCT is useful as a diagnostic tool in PEMN.

Peripheral nerve injury may result in neuropathic pain. Clinical and animal studies have shown that this occurs with skin stimulation over the territory of the injured nerve. Cutaneous stimulation has been noted to cause a period of inhibition in tonic voluntary muscle activity in humans, with a period of electrical silence at the NCS that has been named the cutaneous silent period (Cheng et al., 2008; Uncini et al., 1991). The SCT, the mechanism of which is still poorly understood, probably depends on this cutaneous silent period. It has previously been shown to be an effective and reproducible test in diagnosing carpal tunnel, ulnar nerve entrapment at the elbow, the level of ulnar nerve compression at the elbow, median nerve entrapment at the lacertus tunnel and also in peroneal nerve entrapment diagnosis, with an even higher sensitivity than traditional clinical tests (Brown et al., 2010; Cheng et al., 2008; Gillenwater et al., 2011; Hagert, 2013). Its usefulness in diagnosing patients with potential secondary gain has been reported (Cheng et al., 2008). In addition, it could be also used in children as a game in the office. However, the SCT should not be used indiscriminately without having any idea about the underlying condition. Anatomical knowledge is mandatory to perform it correctly and an associated learning curve has been also reported (Cheng et al., 2008).

The most useful application of the SCT is in the diagnosis of neuropathies in which other clinical manoeuvres are unspecific, the NCS are usually negative or there is no reference standard method for its diagnosis, as in PEMN (Pham et al., 2014). To the best of our knowledge, there is no reference standard for the diagnosis of the PEMN with which the accuracy of the SCT can be compared. Nevertheless, in those patients with positive symptoms but negative electrodiagnostic studies, we believe the SCT has a role in confirming the diagnosis.

Our work has limitations: it is a non-randomized retrospective study and the SCT has not been validated for the diagnosis of PEMN, although it has been used in a previous study (Hagert, 2013) and it has been validated in other peripheral nerve entrapments (Cheng et al., 2008). All clinical examinations were carried out by the authors and not by independent researchers and this could produce an evaluation bias. We assessed patient satisfaction using a closed satisfaction survey, but this is not a validated system and it could also produce an evaluation bias.