Emergent T-helper 2 profile with high interleukin-6 levels correlates with the appearance of bortezomib-induced neuropathic pain


Bortezomib, a nuclear factor kappa B (NFκB) inhibitor, has proven to be very effective for the treatment of multiple myeloma (MM). Peripheral neuropathy (PN) is one of the most common bortezomib-related side effects. The incidence of PN in MM ranges between 30% and 64% and is characterized by burning paresthesias and neuropathic pain (Argyriou et al, 2008). Despite its clinical relevance, information on the pathogenesis of bortezomib-related neural damage is scarce. Recent neurophysiological and histopathological studies identified a significant dysfunction of the afferent small fibres (Aβ, Aδ, C), particularly unmyelinated C fibres, responsible for pain conduction with signs of axonal degeneration being predominant (Cata et al, 2007;Cavaletti et al, 2007;Meregalli et al, 2009). All these characteristics are shared by diabetic neuropathy, so a common pathogenic mechanism is conceivable.

In diabetic neuropathy, there is some evidence that inflammation plays a role in causing axonal degeneration (Cameron & Cotter, 2008; Herder et al, 2009). A specific imbalance in CD4+ T helper (Th) cell sub-populations, with an increase of T helper type 2 (Th2) cells (responsible for allergen specific IgE immuno-mediated damage) with respect to T helper type 1 (Th1) cells (responsible for immunological damage through γ-interferon (IFN-γ)-mediated macrophage activation) was found during the course of various inflammatory neuropathies, particularly in chronic inflammatory demyelinating polineuropathy (CIPD) (Horiuchi et al, 2001; Ravaglia et al, 2008). Finally, in the course of diabetes, as in painful neuropathies of different aetiology, interleukin (IL)-6, one of the pro-inflammatory cytokines secreted by Th2 cells, was identified as the immune mediator most commonly associated with neuropathic pain development (Cameron & Cotter, 2008; Herder et al, 2009).

To investigate a possible role of inflammation in the development of bortezomib-induced PN, assessments of neurological and inflammatory status were prospectively scheduled in ten patients enrolled in a multicentre phase II study. Newly diagnosed MM patients entering the study received up to four cycles of bortezomib (Velcade®, Johnson & Johnson, New Jersey, USA) plus dexamethasone (Vel-Dex, Johnson & Johnson, New Jersey, USA) before stem cell mobilization and autotransplantation. The study was conducted according to the declaration of Helsinki. Institutional review board approval was obtained; each patient provided written informed consent. The neurological assessment consisted of: (i) recording sensory and motor neurological symptoms at baseline and before every bortezomib administration, and (ii) neurophysiological testing (including motor and sensory nerve conduction studies plus quantitative sensory testing) at baseline and after completion of the four cycles or when clinically indicated. Inflammatory status was assessed by: (i) peripheral blood flow cytometric evaluation of lymphocyte populations (CD4 and CD8 absolute counts and ratio), and (ii) evaluation of CD4+ sub-populations (Th1 and Th2) by means of two distinct flow-cytometric panels of cytokines for defining Th1 (IL-2, tumour necrosis factor-α [TNF-α], IFN-γ) and Th2 (IL-4, IL-5, IL-6, IL-10) profiles, respectively, employing the quantitative sandwich enzyme immunoassay technique. Using this method for dosing cytokines, Th1 and Th2 profiles were considered positive when at least two out three (IL-2, TNF-α, IFN-γ) and three out of four (IL-4, IL-5, IL-6, IL-10) cytokines, respectively, were above normal values. The immunological assessment was performed at baseline and after treatment with bortezomib. Table I summarizes the patients’ characteristics at baseline, their response to Vel-Dex, the incidence and severity of bortezomib-related PN, and electrophysiological and immunological findings.

Table I.   Characteristics of the patients at baseline, incidence and severity (according to NCI) of bortezomib-related peripheral neuropathy with significant electrophysiological findings, presence/absence of Th1/Th2 profile, Th1/Th2 ratio, and IL-6 concentrations at baseline and after bortezomib therapy.
PatientAge (years)SexMM typeResponsePNPN severityElectrophysiological findingsTh1Th2Th1/Th2IL6
  1. M, male; F, female; MM, multiple myeloma; PR, partial response; CR, complete response; VGPR, very good partial response; PN, peripheral neuropathy; α-SAP, amplitude of sensory action potential; CMAP, compound muscle action potential; Pre, baseline; Post, after bortezomib therapy. Th1+: at least two out of IL-2, TNF-α, IFN-γ above the normal value; Th2+ at least three out of IL-4, IL-5, IL-6, IL-10 above the normal value.

 160FLight chain λPRAbsentNormal++0·461·4319·15·8
 265MIgAκCRPresent2 with painReduction of α-SAP; slowing of sensory and motor velocities++1·220·474·821·7
 357FLight chain κPRAbsentNormal+1·051·142·54·4
 562MIgGκVGPRPresent3 with painReduction of α-SAP; slowing of sensory and motor velocities+0·860·198·887·2
 962FIgGκ (pleural)CRPresent4 with motor symptomsReduction of α-SAP and CMAP; slowing of sensory and motor velocities++0·850·321·225·8

As reported in Table I, all patients responded to Vel-Dex therapy: three achieved complete response, five had a very good partial response, two a partial response. Three out of the ten patients developed PN during Vel-Dex. In detail, one had National Cancer Institute (NCI) grade 2 PN with pain, one had grade 3 PN with pain, and one had grade 4 painful PN with motor involvement. All neurological symptoms resolved completely after discontinuation of bortezomib. Neurophysiological assessment showed alterations consistent with distal sensory axonal neuropathy in the three patients with PN, while slowing of motor conduction velocity was detected only in the patient with motor symptoms. All seven patients who did not develop neuropathy had normal neurophysiological findings either at baseline or at treatment completion. Regarding immunological findings, we found no imbalance of CD4 and CD8 absolute counts or of the CD4/CD8 ratio either at baseline or at the end-of-treatment evaluation. Comparing baseline to end-of-treatment Th1 and Th2 profiles, five patients had significant modifications. In particular, an imbalance in Th1/Th2 ratio with a shift towards Th2 and a concomitant significant increase of IL-6 levels were observed in the three patients with painful PN. Among the seven patients with no PN, five had normal IL-6 levels both at baseline and at the end of treatment, and two had high levels of IL-6 at baseline that had normalized by the end of treatment.

These neurophysiological findings confirm that bortezomib mainly damages the axon, reflecting the pattern seen in diabetic neuropathy (Cata et al, 2007; Cavaletti et al, 2007;Meregalli et al, 2009). The development of motor symptoms, reported in only one patient in this study, is consistent with a possible extension of the neural damage to the myelin sheath, with the appearance of slowed motor conduction velocity, typical of active demyelinating processes (Ravaglia et al, 2008). The Th2 shift, previously described in CIPD, together with the high levels of IL-6 detected in the three patients with overt PN, support the hypothesis that the neural damage could be related to inflammation (Horiuchi et al, 2001; Herder et al, 2009).

In conclusion, these results suggest that the immuno-modulating cascade triggered by bortezomib via NFκB inhibition, may play a role in causing PN. Sequential evaluation of cytokines characteristic of Th2 profile, particularly IL-6, could help in identifying patients at higher risk of developing painful neuropathy.