• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jrsocmedLink to Publisher's site
J R Soc Med. Aug 2003; 96(8): 379–383.
PMCID: PMC539565

Identifying neuropathic pain in patients with head and neck cancer: use of the Leeds Assessment of Neuropathic Symptoms and Signs Scale

Jean Potter, BSc MRCP, Irene J Higginson, PhD FFPHM, John W Scadding, MD FRCP,1 and Columba Quigley, MD MRCP2


The Leeds Assessment of Neuropathic Symptoms and Signs Scale (LANSS) is a simple bedside test in two parts—a patient-completed questionnaire and a brief clinical assessment. Its diagnostic capabilities have never been tested in patients with cancer pain. To determine these we conducted a prospective study in outpatients with head and neck cancer. All patients with pain completed the LANSS and underwent a medical assessment with a palliative care physician, whose findings were then reviewed by a pain specialist blinded to the LANSS scores. We assessed acceptability and understanding of the LANSS by patients and calculated the sensitivity and specificity of total LANSS scores and subscores derived from the patient-completed section.

Of 130 patients approached, 125 took part. 25 (20%) of these had cancer related pain, mean score on an 11 point numerical rating scale 6.3 (median 6.0, range 3-10). Average age was 60 years (median 60, range 27-84); 56% were male. LANSS completion time was about five minutes, and the procedure was acceptable to all patients. The pain specialist diagnosed neuropathic pain in 14/25 patients, in 13 of whom the neuropathic pain was part of a mixed pain picture. The LANSS correctly identified 11 of these cases (sensitivity 79%; specificity 100%). The patient-completed section alone had a sensitivity of 86% and a specificity of 91%.

The LANSS is a simple and suitable screening test for neuropathic pain in patients with head and neck cancer related pain, although some modifications might improve it.


Neuropathic pain is pain initiated or caused by a primary lesion or dysfunction of the nervous system.1 The distinction of neuropathic pain from other pathophysiological types of pain is fundamental in the management of patients with cancer because of the implications for investigation and treatment.2-4 Conventionally, the diagnosis of neuropathic pain relies on a detailed pain assessment conducted by a pain specialist. The ‘gold standard’ diagnostic test is therefore individual expert opinion. This method is labour-intensive and, since case definition may vary between experts, hinders comparisons between studies.

The Leeds Assessment of Neuropathic Symptoms and Signs Scale (LANSS) is a standardized bedside test to identify patients in whom neuropathic pain mechanisms predominate.5 This tool has two parts—a patient-completed section and a brief physical assessment. Five questions in the patient-completed section (maximum score 16) identify those who are experiencing phenomena associated with neuropathic pain: ‘pins and needles’ (paraesthesia); ‘red skin’ (autonomic changes); ‘sensitive skin’ (evoked dysaesthesia); ‘electric shock pain’; and ‘burning pain’ (spontaneous dysaesthesia). The physical assessment (maximum score 8) is designed to identify allodynia by stroking cotton wool over the painful and the anatomically equivalent non-painful area, and altered pinprick threshold (PPT) by use of a 23 gauge needle to assess perception of pinprick in the same areas. The LANSS identifies patients with neuropathic pain by combining the scores of a patient's verbal description of pain and the results of neurological examination. A cut-off score of 12 points or more (out of a total of 24), when compared with expert opinion, had a sensitivity of 83% and a specificity of 87%.5

Pain affects 40-90% of patients with malignant disease,6 and neuropathic pain occurs in about one-third of patients with cancer-related pain.7 It is usually a mixed nociceptive-neuropathic pain phenomenon.6,8 The LANSS tool was developed and validated by assessing patients with pure neuropathic pain of largely non-malignant origin. A simple and quick test to identify patients with neuropathic pain in the general oncology clinic would be an important first step in improved pain control. However, the diagnostic capabilities of the LANSS in cancer-related mixed-pathophysiology pain is unknown.

Although only around 5% of all cancers arise in the head and neck,9 their highly erosive nature, the rich innervation of the head and neck and the intensive treatment regimens instigated mean that they are a common cause of pain,7,10-11 and neuropathic pain in particular.7,12-13 Between 25% and 60% of patients with pain from head and neck cancer experience neuropathic pain.7,12-14

We therefore aimed to assess the acceptability and understanding of the LANSS by patients with head and neck cancer, and to calculate the sensitivity and specificity of the LANSS by comparison with medical expert opinion in these patients. In busy clinics a screening tool may be most useful when unaided patients can complete all parts; therefore the patient-completed subscores were also analysed separately to calculate a cut-off score with the highest sensitivity and specificity.


Design and patients

Patients attending the head and neck cancer outpatient clinic at a London hospital were approached by a palliative care doctor (JP) in the clinic waiting area or by telephone survey. Potentially eligible patients were screened with a standardized format that explained the study and asked a screening question based on the first question of the Brief Pain Inventory.15 Those who experienced pain ‘more severe or troublesome than everyday aches and pains’ completed the LANSS and underwent a clinical assessment with the investigator (JP).

Patients were included in the study if they were over the age of 18, were able to understand and speak English, and had head and neck cancer. Those patients with head and neck cancer with small-volume disease (staged at T1N0M0) and those who received only minor treatment (local excision and/or brachytherapy) were excluded.

The study was approved by the local hospital research ethics committee. Written informed consent was obtained from all patients.

Measures and procedures

A single investigator, JP, collected all data. Demographic and medical data (regarding disease severity, sequelae and treatments) were collected on all surveyed patients from their medical notes, by use of standardized proformas.

Initially all patients who agreed to take part completed the LANSS questions unaided. The investigator then reviewed the questions with each patient to verify patient understanding. If this review revealed questions that had initially been misunderstood by the patient, then the investigator clarified the answer with the patient and scores were amended accordingly. Where the question could not be answered by the patient either alone or in conjunction with the doctor, then the default answer was no (no score). Again this led in some cases to amended LANSS scores.

The investigator, using a standardized format, took a medical history and conducted a neurological examination including the sensory assessment component of the LANSS. Sensation (fine touch, pinprick and allodynia), power and appropriate tendon reflexes were assessed in all patients.

A pain specialist (JWS) who was blinded to the LANSS scores reviewed medical details of each case (demographic and medical data plus clinical assessment findings). The expert opinion of this pain specialist was used as the gold standard for identifying neuropathic pain in these patients.


Acceptability to patients was expressed as a percentage:

equation M1

The sensitivity and specificity of the LANSS in these patients, compared with expert opinion, were calculated by use of the cut-off score of 12 suggested in the original validation study.5 The individual components of the LANSS were assessed by χ2 tests to identify which components were chosen more often by patients with neuropathic pain than those with non-neuropathic pain. Fisher's exact test was used to assess any association between two categorical variables when expected counts calculated for χ2 were below 5. The LANSS patient-completed section alone was analysed to calculate a cut-off score with the highest sensitivity and specificity.

Analysis was conducted with the Statistical Package for the Social Sciences version 10.1.


Of 130 patients who were approached, 125 took part. 26 (21%) of these had had pain in the previous 24 hours that was ‘more severe than everyday aches and pains’. One patient was found to have pain that was unrelated to the cancer and was excluded from the analysis. The mean score on an 11-point numerical rating scale for cancer-related pain was 6.3 (median 6.0; range 3-10). Table 1 shows the patients' demographic and medical characteristics.

Table 1
Demographic and medical characteristics of patients

Acceptability and understanding by patients

The LANSS was completed by all patients who were invited to do so. Completion time was about five minutes, including the sensory assessment. The full assessment could be conducted with the patient fully clothed in most cases. 6 patients (24%) required help with the patient-completed section. Nine answers were amended on review of these 6 patients. In total, 25 patients answered five questions each. Amendment was necessary in 11 patients (9%) (Table 2).

Table 2
LANSS patient-completed section: difficulties with understanding

Diagnosis of neuropathic pain

Expert opinion diagnosed neuropathic pain in 14/25 patients (56%). In 13/14 cases neuropathic pain was part of a mixed pain picture. Each pain quality assessed by the LANSS was experienced more often by patients with neuropathic pain than by those with non-neuropathic pain, but this reached statistical significance only for items related to ‘sensitive skin’, ‘electric shock pain’, allodynia and altered PPT (Table 3).

Table 3
Presence of pain qualities according to underlying pathophysiology

With the suggested cut-off score of 12, the LANSS correctly identified 11/14 cases of neuropathic pain (sensitivity 79%; specificity 100%). Amendment of responses changed the total LANSS scores of 6 patients, but this had no effect on the sensitivity or specificity.

A cut-off score of 6 or more, with data from the unaided patient-completed section, gave sensitivity and specificity scores of 86% and 82%, respectively. When the amended patient-completed scores were used the sensitivity was unchanged but the specificity rose to 91%.


This study indicates that the LANSS is a valid test for identifying neuropathic pain in patients with pain from head and neck cancer. A total score of 12 or more had a sensitivity of 79% and a specificity of 100%.

Some of the LANSS questions were difficult for patients to complete accurately. For example, skin redness was associated with either pain or treatment. Patients with damaged mucous membranes experienced sensitive skin and burning sensations but these were not, on medical assessment, related to neuropathic pain mechanisms. Despite these drawbacks, a cut-off score of 6 (maximum 16) on the patient-completed subscore alone had a sensitivity and specificity of 86% and 82%, respectively. Amendment of patients' answers by giving no score to questions regarding sensitive skin and burning sensations in patients with damaged mucous membranes, and by asking whether skin colour change occurred only when pain was particularly severe, had no effect on the sensitivity but improved the specificity.

In this study, 1 in 5 patients attending a head and neck outpatient clinic had pain, and over half of these had neuropathic pain. Similar prevalence values have been found by others.2,7,12-14,16-19 The total LANSS score correctly identified 86% of these patients, and the patient-completed data offered the same sensitivity with only slight loss of specificity. Identification of patients with neuropathic pain is an important first step towards control of this difficult symptom. A shortened version of the LANSS, comprising a modified patient-completed section alone, could be used as a screening tool for neuropathic pain in the waiting area of busy oncology clinics.


The study was supported by the Mrs Coco Markus Clinical Research Fellowship.


1. Merskey H, Bugduk N. Classification of Chronic Pain. Descriptions of Chronic Pain Syndromes and Definitions of Pain Terms. Seattle: IASP Press, 1994 [PubMed]
2. Robertson MS, Hornibrook J. The presenting symptoms of head and neck cancer. N Z Med J 1982;95: 337-41 [PubMed]
3. Twycross R. Symptom Management in Advanced Cancer. Oxford: Radcliffe Medical Press, 1997
4. Twycross R. Cancer pain classification. Acta Anaesthesiol Scand 1997;41: 141-5 [PubMed]
5. Bennett M. The LANSS Pain Scale: the Leeds assessment of neuropathic symptoms and signs. Pain 2001;92: 147-57 [PubMed]
6. Portenoy RK. Cancer pain. Epidemiology and syndromes. Neurologic Clinics 1993;63: 2298-307 [PubMed]
7. Grond S, Zech D, Diefenbach C, Radbruch L, Lehmann KA. Assessment of cancer pain: a prospective evaluation in 2266 cancer patients referred to a pain service. Pain 1996;64: 107-14 [PubMed]
8. Caraceni A. Clinicopathologic correlates of common cancer pain syndromes. Haematol/Oncol Clin N Am 1996;10: 57-78 [PubMed]
9. Sanderson RJ, Ironside JAD. Squamous cell carcinomas of the head and neck. BMJ 2002;325: 822-7 [PMC free article] [PubMed]
10. Caraceni A. An international survey of cancer pain characteristics and syndromes. Pain 1996;82: 263-74 [PubMed]
11. Brescia FJ, Adler D, Gray G, Ryam MA, Cimino J, Marntani R. Hospitalised advanced cancer patients: a profile. J Pain Symptom Management 1990;5: 221-7 [PubMed]
12. Chua KSG, Reddy SK, Lee M, Patt RP. Pain and loss of function in head and neck cancer survivors. J Pain Symptom Management 1999;18: 193-202 [PubMed]
13. Vecht CJ, Hoff AM, Kansen PJ, de Boer MF, Bosch DA. Types and causes of pain in cancer of the head and neck. Cancer 1992;70: 178-84 [PubMed]
14. Grond S, Zech D, Lynch J, Diefenbach C, Schug SA, Lehmann KA. Validation of the World Health Organisation guidelines for pain relief in head and neck cancer. Ann Otolaryngol Rhinol 1993;102: 342-8 [PubMed]
15. Daut RL, Cleeland CS, Flanery R. Development of the Wisconsin Brief Pain Questionnaire to assess pain in cancer and other diseases. Pain 1983;17: 197-210 [PubMed]
16. Greenslade R, Portenoy RK. Pain syndromes in head and neck cancer. J Pain Symptom Management 1988;3(suppl): S21
17. Keefe FJMGBA, Crisson J. Pain in the head and neck cancer patient: changes over treatment. Head Neck Surg 1986;8: 169-76 [PubMed]
18. Chaplin JM, Morton RP. A prospective, longitudinal study of pain in head and neck cancer patients. Head Neck 1999;21: 531-7 [PubMed]
19. Aird DW, Bihari J, Smith C. Clinical problems in the continuing care of head and neck cancer patients. Ear Nose Thorat J 1983;62: 30 [PubMed]

Articles from Journal of the Royal Society of Medicine are provided here courtesy of Royal Society of Medicine Press
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...