Bb

APD

a

ARRAA

KBPUR

1

r(ehrstpcmzPdln

0d

Leukemia Research 34 (2010) 471–474

Contents lists available at ScienceDirect

Leukemia Research

journa l homepage: www.e lsev ier .com/ locate / leukres

ortezomib-induced peripheral neuropathy in multiple myeloma: A comparisonetween previously treated and untreated patients

lessandro Corso ∗, Silvia Mangiacavalli, Marzia Varettoni, Cristana Pascutto,atrizia Zappasodi, Mario Lazzarinoivision of Hematology, Fondazione IRCCS Policlinico San Matteo, University of Pavia, Viale Golgi 19, 27100 Pavia, Italy

r t i c l e i n f o

rticle history:eceived 2 May 2009eceived in revised form 12 July 2009ccepted 13 July 2009vailable online 11 August 2009

eywords:ortezomiberipheral neuropathyntreated patients

a b s t r a c t

Peripheral neuropathy (PN), with neuropathic pain as main symptom, represents the dose-limiting toxi-city of the proteasome inhibitor bortezomib. Aim of this study was to compare the incidence, risk factors,severity and outcome of PN and neuropathic pain in patient treated with bortezomib up-front or at relapse.We studied 55 patients with multiple myeloma (MM) who received bortezomib as first line therapy and70 pre-treated patients who received bortezomib in relapse or progression. Regarding PN, no differenceswere found among untreated and pre-treated patients in the incidence (55% vs 52%, p = 0.43), severity(NCI grade 3–4 9% vs 14%, p = 0.27), and outcome (improved/resolved 90% vs 91%, p = 0.58). Concerningneuropathic pain, the incidence was lower (50% vs 81%, p = 0.008) and solved earlier (35 days vs 91 days,p = 0.02) in untreated compared with pre-treated patients. Untreated patients needed dose modification

elapsed refractory myeloma less frequently (36% vs 73%, p = 0.012). No correlation was found between development of PN and priorexposure to potentially neurotoxic drugs such as thalidomide, vincristine, and cysplatin. Age representedthe main risk factor for PN (p = 0.036) with an increase in risk of PN amounting to 6% per year of age.In conclusion, incidence, severity and outcome of bortezomib-related PN are similar in untreated andpre-treated MM patients except for neuropathic pain which has lower incidence and shorter durationin untreated patients with less frequent need for bortezomib discontinuation. Age emerges as the most

riphe

relevant risk factor for pe

. Introduction

The proteasome inhibitor bortezomib is an effective drug forefractory/relapsed [1–5] and newly diagnosed multiple myelomaMM) patients [6–12]. The most commonly reported adverseffects are gastrointestinal symptoms, thrombocytopenia, posturalypotension and peripheral neuropathy with or without neu-opathic pain. Bortezomib-induced PN, predominantly sensory,hares some symptoms and signs (i.e. distal paresthesias, dises-hesia, tingling, suppression of deep tendon reflexes) with otherotentially neurotoxic drugs used in MM such as thalidomide, vin-ristine and platinum. Among symptoms of PN, neuropathic pain,ainly located at the feet or/and lower limbs, is peculiar to borte-

omib. Motor symptoms are rarely observed. Bortezomib-induced

N is dose-related and reversible after dose reduction or treatmentiscontinuation. Some studies have addressed the possible corre-

ations between risk of bortezomib-induced PN and pre-existingeurologic damages or comorbidities [13–18]. There is no agree-

∗ Corresponding author. Tel.: +39 0382 503595; fax: +39 0382 502250.E-mail address: a.corso@smatteo.pv.it (A. Corso).

145-2126/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.oi:10.1016/j.leukres.2009.07.022

ral neuropathy, with a risk increase for PN of 6% per year of age.© 2009 Elsevier Ltd. All rights reserved.

ment on the role of diabetes or prior exposure to potentiallyneurotoxic drugs such as thalidomide, cisplatin, and vincristine[13,14,16,17].

Aim of this study was to evaluate the incidence, risk factors,severity, and outcome of PN and neuropathic pain in patients withmultiple myeloma (MM) who received bortezomib as first line ther-apy, in comparison with pre-treated patients.

2. Patients and methods

Two groups of patients were compared: 55 untreated MM patients enrolled in anopen label prospective phase II protocol who received bortezomib as first line treat-ment; and 70 MM patients who received bortezomib at relapse. Untreated patientsreceived: 4 cycles of bortezomib plus dexamethasone, 2 cycles (with peripheral stemcells collection after the first) of DCEP chemotherapy (Dexamethasone, Cyclophos-phamide, Etoposide, Cys-Platinum), single autologous transplant with high-dosemelphalan 200 mg/m2 as conditioning regimen. Bortezomib was administered i.v.at 1.3 mg/m2 on days 1, 4, 8, 11 in combination with oral dexamethasone 40 mg ondays 1–4, 8–11 at 21 days interval. Pre-treated patients received a median of four

2

courses (range 2–8) of bortezomib i.v. at 1.3 mg/m on days 1, 4, 8, 11 with oraldexamethasone 20 mg on days 1–2, 4–5, 8–7; 11–12 at 21 days interval. Patientsin both groups provided written informed consent. The study was approved by thelocal Ethics Committee of Pavia and the procedures were in accordance with theHelsinki Declaration of 1975, as revised in 2000. Tables 1 and 2 summarize baselinecharacteristics and prior treatments in the two groups.

472 A. Corso et al. / Leukemia Research 34 (2010) 471–474

Table 1Patients characteristics and treatment history.

Untreated (55patients)

Pre-treated (70patients)

p

GenderMale 34 (62%) 39 (56%) 0.3Female 21 (38%) 31 (44%)

Age (range) 58 (37–65) 59 (38–69) 0.07

Type of myelomaIgG 31 (56%) 36 (51%) 0.2IgA 12 (22%) 18 (26%)Light chain 12 (22%) 15 (21%)

Median N of prior treatments (range) 1 (1–6)1 – 41 (58%)2–3 – 25 (36%)≥4 – 4 (6%)

Prior exposure toVincristine – 70 (100%)Cisplatin containing regimen – 50 (71%)

BD

2

nasu(iPCptPm

mqbbotaT

3

Ip4t5

TD

MO

T

M

Table 3Recommended dose modification for peripheral neuropathy.

Severity of peripheral neuropathy Modification of doseand schedule

Grade 1 (paresthesias or loss ofreflexes) without pain or loss offunction

No action

Grade 1 with pain or grade 2(interferes with function but notwith ADL)

Reduce to 1.0 mg/m2

Grade 2 with pain or grade 3(interferes with ADL)

Withhold treatmentuntil toxicity resolves,then reinitiate at a doseof 0.7 mg/m2 onceweekly

Thalidomide – 49 (70%)

aseline PN (NCI grade) – 25 (36%)iabetes mellitus – 6 (9%)

.1. Neurologic assessment

Neurologic assessment was performed on all patients at baseline, at the begin-ing of each bortezomib cycle, and subsequently at each follow-up, or in any momentppeared new neurological symptoms. Neurologic assessment consisted of: (a)pecific questions regarding neurological symptoms (sensory, motor) with partic-lar attention on neuropathic pain occurrence, and thermoregulatory disturbanceburning sensation or coldness), functional impairment interfering with daily liv-ng activities; (b) specific neurologic examination. Baseline PN, bortezomib-inducedN, and neuropathic pain were graded according to National Cancer Institute (NCI)ommon Toxicity Criteria (CTC) for Adverse Events (version 3.0). In pre-treatedatients were registered number and type of prior treatments, exposure to poten-ially neurotoxic drugs (vincristine, cysplatin, thalidomide), presence of baselineN, concomitant diabetes. The dose of bortezomib was adjusted according to doseodification guidelines as reported in the APEX trial (Table 3) [13].

Regarding statistical methods, numeric variables are summarized by theiredian and range. Categorical variables are described by counts and relative fre-

uencies (%). The onset of PN was the primary outcome, and was analysed as ainary categorical variable. Difference in the distribution of categorical variablesetween groups was investigated by means of the Fisher exact test for 2 × 2 tables,r the chi-squared test for larger tables. Multivariate logistic regression was appliedo analyse the risk factors for PN. Statistical significance for all analyses was defineds a p-value ≤ 0.05. All analyses were performed using Statistica 7.1 (Statsoft Inc.,ulsa, OK) and Microsoft Excel.

. Results

Features of bortezomib-induced PN are summarized in Table 4.n untreated patients the incidence of PN was 55% (30 out of 55

atients) with NCI grade 1–2–3–4 occurring in 8 (15%), 17 (31%),(7%), 1 (2%) patients, respectively. Neuropathic pain was regis-

ered in 15 of 30 patients with PN (50%) (10 with NCI grade 2 andwith grade 3–4 PN). Motor symptoms occurred in one patient

able 2etails of bortezomib therapy in untreated and pre-treated patients.

Untreated Pre-treated

edian N of bortezomib cycles 4 (2–4) 4 (2–8)RR (≥PR) 86% (47 of 55) 59% (41 of 70)

ype of responseCR/nCR 28 (52%) 5 (7%)VGPR 10 (18%) 25 (36%)PR 9 (16%) 11 (16%)SD 3 (5%) 19 (27%)Progression 4 (7%) 10 (14%)

edian follow-up (months) 12.7 (0.4–23) 12.9 (1.8–44.7)

Grade 4 (permanent sensory lossthat interferes with function)

Discontinue treatment

Abbreviation: ADL, activities of daily living.

(2%). Median time to PN occurrence was 70 days (range 21–133).Bortezomib therapy was discontinued for PN in 8 patients (27%)and reduced in 3 (10%). PN resolved or significantly improved in 27patients (90%), and remained unmodified in 3 (10%). Neuropathicpain resolved in all patients with a median time of 35 days (range10–111). Median time to complete recovery of PN was 169 days(29–408).

In pre-treated patients the incidence of PN was 52% (36 of 70patients) with NCI grade 1–2–3–4 occurring in 13 (19%), 13 (19%), 7(10%), and 3 (4%) patients, respectively. Neuropathic pain was regis-tered in 29 of 36 patients with PN (81%) (19 with NCI grade 2 and 10with grade 3–4 PN). Motor symptoms occurred in four patients (6%).Median time to PN occurrence was 68 days (range 29–184). Borte-zomib therapy was discontinued because of PN in 15 patients (39%)and reduced in 13 (34%). PN resolved or significantly improved in33/36 patients (91%) and remained unmodified in 3 (9%). Neuro-pathic pain disappeared in all patients in a median time of 91 days(range 17–276). Median time to PN complete recovery was 100 days(17–283).

Untreated and pre-treated patients were similar for gender andage distribution (M/F 62%/38% vs 56%/44%, p = 0.3; median age 58(range 37–65) years vs 59 (38–69) years, p = 0.07), median follow-up (12.7 months vs 12.9 months, p = 0.93), and median number ofbortezomib cycles (four cycle in both groups, p = 0.59).

Comparing untreated and pre-treated patients no differenceswere found in terms of incidence (55% vs 52%, p = 0.43), severity(NCI grade 3–4 9% vs 14%, p = 0.27), and outcome of bortezomib-related PN (improved/resolved 90% vs 91%, p = 0.58). Median timeto PN occurrence was similar in the two groups (70 and 68 days,respectively; p = 0.74). Among patients with PN, the incidence ofneuropathic pain was significantly lower in untreated than in pre-treated patients (50% vs 81%, p = 0.008). Median time to neuropathicpain recovery was significantly shorter in untreated patients (35days vs 91 days, p = 0.02), despite a longer time to PN resolution(169 days vs 100 days, p = 0.007). In patients with PN, bortezomibdose reduction or discontinuation was needed less frequently inuntreated than in pre-treated patients: 11/30 vs 28/36 (36% vs 73%,p = 0.012).

3.1. Risk factors for bortezomib-induced PN

In univariate analysis, PN incidence significantly correlated withage both in untreated (p = 0.035) and pre-treated patients (p = 0.04).

In a multivariate logistic regression with group (untreated/pre-treated), response to bortezomib and age as variables, only agesignificantly correlated with PN incidence (p = 0.036). In particular,the increase in risk of PN amounted to 6% per year of age. Table 5summarizes incidence and distribution of grades 3–4 PN by age

A. Corso et al. / Leukemia Research 34 (2010) 471–474 473

Table 4Features of bortezomib-induced peripheral neuropathy (PN) in untreated vs previously treated patients.

Untreated (55 patients) Pre-treated (70 patients) p

PN incidenceN patients (%) 30 (55%) 36 (52%) 0.43

PN NCI grade1 8 (15%) 13 (19%)

0.482 17 (31%) 13 (19%)3 4 (7%) 7 (9%)4 1 (2%) 3 (4%)

Neuropathic pain (% of patients with PN)Present 15 (50%) 29 (81%)

0.008Absent 15 (50%) 7 (19%)

Motor symptoms 1 (2%) 4 (6%) 0.35

Bortezomib dose modifications for PNDiscontinued 8 (27%) 15 (39%)

0.012Definitively 6 (20%) 13 (34%)Temporarily 2 (7%) 2 (5%)

Reduced 3 (9%) 13 (34%)Unmodified 19 (63%) 8 (26%)

PN outcomeResolved/significantly improved 27 (90%) 33 (91%)

0.58Persistent 3 (10%) 3 (9%)

Median time to PN occurrence (days) 70 (21–133) 68 (29–184) 0.74MM )

P

giotprooio

4

eueiolemcnApp

TFa

A

≤5≥

edian time to neuropathic pain recovery (days) 35 (10–111)edian time to PN recovery (days) 169 (29–408

N states for peripheral neuropathy.

roup (≤50 years, 50–60 years, ≥60 years). There was a significantncrease of incidence (p = 0.004) and severity (p = 0.02) in patientslder than 60 years. There was also a trend between age (as con-inuous variable) and occurrence of neuropathic pain (p = 0.06). Inre-treated patients, we did not find (both in univariate and logisticegression analysis adjusted by age) correlation between incidencer severity of PN and prior exposure to potentially neurotoxic drugsr the presence of baseline neuropathy. The only statistically signif-cant correlation was found between incidence of PN and numberf bortezomib cycles (p = 0.0045).

. Discussion

Bortezomib as a single agent or combined with either dexam-thasone or chemotherapy is effective in recurrent or previouslyntreated MM patients [1–12]. Among bortezomib-related sideffects, peripheral neuropathy (PN) with peculiar neuropathic pain,s the main dose-limiting toxicity. Knowledge of the pathogenesisf neural damage underlying bortezomib-related neuropathy is stillimited. Cavaletti et al. [19], in animal models found that the toxicffect is mainly directed towards satellite cells (Schawann cells,yelin), although axonal degeneration can also be present. Spe-

ific quantitative sensory studies showed that bortezomib-induced

europathic pain is associated with deficit of the afferent fibers�, A�, C [20]. An immuno-mediated pathogenetic mechanism oferipheral neural damage has also been hypothesized, especially inatients with motor dysfunction [21].

able 5requency of peripheral neuropathy (all grade and detailed by grade) according toge group.

ge group N patients PN p NCI PN grade p

Absent Present 1–2 3–4

50 years 22 77% 23% 0.004 18% 5% 0.020–60 48 46% 54% 44% 10%60 years 55 36% 64% 47% 17%

91 (17–276) 0.02100 (17–283) 0.007

The incidence of bortezomib-induced PN is variable among stud-ies. The lowest incidence has been reported by the PETHEMA study[12] that alternates bortezomib with dexamethasone in untreatedpatients (25% incidence of PN with no grade 3–4 events). The IFM2005/01 study [8] which uses bortezomib plus dexamethasone asfirst line for younger MM patients, reports a 35.3% PN incidencewith 17% grade 3–4. In the VISTA trial (bortezomib plus melphalanand prednisone in untreated elderly patients) [11] the incidence ofPN was 45% with 14% grades 3–4. In the studies by Richardson onpre-treated patients (phase II SUMMIT and CREST) the PN cumula-tive incidence was 35% (13.4% grade 3–4) [2–4], comparable rateswere found in phase III APEX study (PN overall incidence 37% with10% grade 3–4) [5]. The highest incidences of PN were reported byMateos et al. [10], in elderly untreated patients (55% with 17% grades3–4) and by Badros et al. [15] in relapsed patients (52%, with 22%NCI grades 3–4 events).

In the present study, incidence and severity of PN were not dif-ferent in untreated and pre-treated patients (overall incidence 55%vs 52%, p = 0.43; grades 3–4 events 9% vs 14%, p = 0.27). Also themedian time to the occurrence of PN was similar (70 and 68 days,respectively) confirming, together with the significant correlationbetween the number of bortezomib cycles and the incidence ofPN (p = 0.0045), that PN is dose-dependent [13,14]. The majority ofpatients in both groups had significant improvement or completeresolution of neurologic symptoms after completion of bortezomibtherapy (90 and 91% in untreated and pre-treated patients, respec-tively, p = 0.58). Among patients who developed symptoms of PN,neuropathic pain was less frequent (50% vs 81%, p = 0.008) andimproved more rapidly (35 days vs 91 days, p = 0.02) in untreatedpatients, with lower need for dose reduction or discontinuation(37% vs 73%, p = 0.012). The occurrence of neuropathic pain wasthe main reason for bortezomib dose modification or discontin-uation. The higher prevalence and the longer time for neuropathic

pain recovery observed in relapsed patients may be explained bypre-existing neural damage due to the longer disease history [22].

Differently from neuropathic pain, in this study, minor sensorysymptoms such as distal paresthesias, disesthesia, tingling, per-sisted for longer time in untreated than in pre-treated patients (169

4 a Rese

dtcu

npmut6rd1

optsbrtpa

C

R

[

[

[21] Ravaglia S, Corso A, Piccolo G, et al. Immune-mediated neuropathies inmultiple myeloma patients treated with bortezomib. Clin Neurophysiol

74 A. Corso et al. / Leukemi

ays vs 100 days, respectively, p = 0.007). This may be related tohe sequential exposure to potentially neurotoxic drugs, such asysplatin and high-dose melphalan, after bortezomib induction inntreated patients.

In pre-treated patients, the analysis of risk factors for PN didot show correlation with either baseline peripheral neuropathy orrevious exposure to potentially neurotoxic agents such as thalido-ide or cysplatin. PN was significantly correlated with age both in

ntreated and pre-treated patients, as in prior studies [10,15]. Inhis series PN rates range from 23% in patients aged less than 50 to4% in patients older than 60 years (p = 0.004) with an increase inisk for PN amounted to 6% per year. The severity of PN was also age-ependent, with 4% grades 3–4 PN in patients less than 50 years vs6% in patients older than 60 years (p = 0.02).

In conclusion, this study shows that the incidence, severity andutcome of bortezomib-related PN are similar in untreated andre-treated MM patients. Neuropathic pain, the peculiar symp-om of bortezomib-related neuropathy, has lower incidence andhorter duration in untreated patients with less frequent need forortezomib discontinuation. Neuropathic pain represents the maineason for dose modification or discontinuation, independently ofhe severity of PN. Age emerges as the most relevant risk factor foreripheral neuropathy, with a risk increase for PN of 6% per year ofge.

onflict of interest

None declared.

eferences

[1] Jagannath S, Barlogie B, Berenson J, et al. A phase 2 study of two doses ofbortezomib in relapsed or refractory myeloma. Br J Haematol 2004;127:165–72.

[2] Jagannath S, Barlogie B, Berenson JR, et al. Updated survival analyses after pro-longed follow-up of the phase 2, multicenter CREST study of bortezomib inrelapsed or refractory multiple myeloma. Br J Haematol 2008;143:537–40.

[3] Richardson PG, Barlogie B, Berenson J, et al. A phase 2 study of bortezomib inrelapsed, refractory myeloma. N Engl J Med 2003;348:2609–17.

[4] Richardson PG, Sonneveld P, Schuster MW, et al. Bortezomib or high-dose dex-amethasone for relapsed multiple myeloma. N Engl J Med 2005;352:2487–98.

[5] Richardson PG, Sonneveld P, Schuster M, et al. Extended follow-up of a phase3 trial in relapsed multiple myeloma: final time-to-event results of the APEXtrial. Blood 2007;110:3557–60.

[6] Jagannath S, Durie BG, Wolf J, et al. Bortezomib therapy alone and in combi-nation with dexamethasone for previously untreated symptomatic multiplemyeloma. Br J Haematol 2005;129:776–83.

[

arch 34 (2010) 471–474

[7] Harousseau JL, Attal M, Leleu X, et al. Bortezomib plus dexamethasoneas induc-tion treatment prior to autologous stem cell transplantation in patients withnewly diagnosed multiple myeloma: results of an IFM phase II study. Haema-tologica 2006;91:1498–505.

[8] Harousseau JL, Mathiot C, Attal M, et al. Bortezomib/dexamethasone versusVAD as induction prior to autologous stem cell transplantation (ASCT) in pre-viously untreated multiple myeloma (MM): updated from IFM 2005/01 trial. JClin Oncol 2008;26 [8505 abstr].

[9] Popat R, Oakervee HE, Hallam S, et al. Bortezomib, doxorubicin and dexametha-sone (PAD) front-line treatment of multiple myeloma: updated results afterlong-term follow-up. Br J Haematol 2008;141:512–6.

[10] Mateos MV, Hernandez JM, Hernandez MT, et al. Bortezomib plus melphalanand prednisone in elderly untreated patients with multiple myeloma: resultsof a multicenter phase II study. Blood 2006;108:2165–72.

[11] San Miguel J, Schlag R, Khuageva NK, et al. Bortezomib plus melphalanand prednisone for initial treatment of multiple myeloma. N Engl J Med2008;359:906–17.

12] Rosinol L, Oriol A, Mateos MV, et al. Phase II PETHEMA trial of alternat-ing bortezomib and dexamethasone as induction regimen before autologousstem-cell transplantation in younger patients with multiple myeloma. Effi-cacy and clinical implication of tumor response kinetics. J Clin Oncol 2007;25:4452–8.

[13] Richardson PG, Briemberg H, Jagannath S, et al. Frequency, character-istics, and reversibility of peripheral neuropathy during treatment ofadvanced multiple myeloma with bortezomib. J Clin Oncol 2006;24:3113–20.

[14] Richardson PG, Sonneveld P, Schuster MW, et al. Reversibility of symptomaticperipheral neuropathy with bortezomib in the phase III APEX trial in relapsedmultiple myeloma: impact of dose modification guideline. Br J Hematol2009;144:895–903.

[15] Badros A, Goloubeva O, Dalal JS, et al. Neurotoxicity of bortezomib therapyin multiple myeloma: a single centre experience and review of the literature.Cancer 2007;110:1042–9.

[16] Caravita T, Petrucci MT, Spagnoli A, et al. Neuropathy in multiple myelomapatients treated with bortezomib: a multicenter experience. Blood 2007;110[Abstract 4823].

[17] El-Cheikh J, Stoppa AM, Bouabdallah R, et al. Features and risk factors ofperipheral neuropathy during treatment with bortezomib for advance multiplemyeloma. Clin Lymph Myel 2008;8:146–52.

[18] Argyriou A, Iconomou G, Kalofonos H, et al. Bortezomib-induced peripheralneuropathy in multiple myeloma: a comprehensive review of the literature.Blood 2008;112:1593–9.

[19] Cavaletti G, Gilardini A, Canta A, et al. Bortezomib-induced peripheral neu-rotoxicity: a neurophysiological study in the rat. Exp Neurol 2007;204:317–25.

20] Cata JP, Weng HR, Burton AW, et al. Quantitative sensory findings in patientswith bortezomib induced pain. J Pain 2007;8:296–306.

2008;119:2507–12.22] Borrello I, Ferguson A, Huff CA, et al. Bortezomib and thalidomide treatment of

newly diagnosed patients with multiple myeloma: efficacy and neurotoxicity.Blood 2006;108 [Abstr 3528].