Incidence of paraneoplastic polyneuropathies associated with non-Hodgkin lymphoma in the province of Asti, Italy

Abstract

Sirs: Non-Hogdkin lymphoma (NHL) may be associated with different types of peripheral nervous system diseases [2, 4, 10–12]. Some patients may be affected with neuropathy unrelated to direct infiltration of nerves or spinal roots, vasculitis, toxic effect of chemotherapic drugs and infections [2, 11]. These polyneuropathies may be defined as possible paraneoplastic diseases according to current criteria [5]. In this study, we prospectively evaluated the incidence of paraneoplastic polyneuropathies in patients with NHL in the province of Asti, Italy. All patients diagnosed as having NHL between March 31, 2004 and October 21, 2005 at the Hematology Department of the Hospital of Asti were included. All patients underwent standardised clinical and neurophysiological assessment [6] before chemotherapy. NHL was classified according to the Revised European-American Lymphoid Neoplasms (REAL) classification [1]. Grading was defined according to the Ann Arbor classification [1]. All patients underwent cerebrospinal fluid (CSF) examination. We excluded patients with previous history of acquired or inherited neuromuscular diseases, previous or current exposure to toxins or drugs known to cause polyneuropathies, cachexia, diabetes or glucose intolerance, Vitamin E or B12 deficiency, hypothyroidism, paraproteinemia, detectable onconeural antibodies [5], other autoimmune diseases, vasculitis, primary cerebral lymphoma. No patients had evidence of lymphomatous cells, increase in the protein content or cell count in the CSF. In the province of Asti, all patients with haematological diseases have been usually referred to the Haematology Department of the Hospital of Asti. It is feasible to consider the number of patients with NHL newly diagnosed in this department, as a realistic estimate of the incidence of NHL in that province. The Asti’s province population, according to Piedmont BDDE’s database, was 220,333 in the year 2004. The results were adjusted for the number of the observation/month. In order to determine the relative risk of polyneuropathy associated with NHL, an ageand sex-matched control group of subjects without NHL or diseases associated with polyneuropathy underwent clinical and neurophysiological assessment. Eighteen NHL patients were enrolled. Demographic, clinical and haematological data are summarised in the Table 1. In five patients (27.7%, 3 males, 2 females, mean age 75.4 years), nerve conduction studies showed the presence of sensorimotor polyneuropathy of the axonal type. Four of them had also clinical evidence of polyneuropathy: one complained of mild paresthesias at the feet, whereas signs of mild sensory involvement of the lower limbs were present in all, as described in the Table 1. No patients had detectable weakness, muscle wasting, ataxia or significant pain. Rankin score was 0–1 in patients with polyneuropathy. No significant worsening of the polyneuropathy was evident after chemotherapy. The estimated incidence rate of paraneoplastic polyneuropathies associated with NHL was 0.99 per 100,000 population. Two subjects in the control group (11.1%, 1 male and 1 female, ages: 81 and 79 years) had clinical and neurphysiological signs of a mild axonal mainly sensory polyneuropathy. The relative risk of having an axonal polyneuropathy in patients with NHL was 2.5 (95% CI: 0.56–11.25). Some observations are needed based on our results. None of our patients had chronic inflammatory demeylinating polyneuropathy, probably because of exclusion of patients with paraproteinemia. Even if the absence of sural nerve biopsy specimens did not allow to rule out the presence of vasculitis or infiltration by lymphomatous cells, clinical and electrophysiological G. Isoardo (&) SOS Neurofisiologia Clinica ASO CTO-Maria Adelaide Via Zuretti 29 10126 Torino, Italy Tel.: +390116933881 E-Mail: gianlucaisoardo@yahoo.it