State of the Art Oncology in EuropeFont: aaa

Chemotherapy-Induced Peripheral Neurotoxicity: an Update

1. INTRODUCTION

The peripheral nervous system can be vulnerable to the toxic action of several drugs since it is not protected as effectively as the central nervous system from noxious exogenous agents. Drug-induced neurotoxicity can affect the nerve fibers or the neuronal bodies (generally the dorsal root ganglia (DRG) of the primary sensory neurons). The clinical features of such neurotoxicity are dependent on the type of agent involved and the site of action – ranging from motor, to sensory-motor or almost exclusively sensory neuropathies, with or without autonomic impairment. Among the neurotoxic drugs antineoplastic agents represent a major clinical problem, given their widespread use and the potential severity of their toxicity. In fact, the peripheral neurotoxicity (PN) of antineoplastic agents frequently represents one of their dose-limiting side effects. Moreover, even when antineoplastic agents’ peripheral neurotoxicity is not dose-limiting, its onset may severely affect the quality of life of cancer patients and cause chronic discomfort.

We will review the features of chemotherapy-induced PN resulting from the administration of the most widely used and better investigated compounds, such as platinum drugs, taxanes, vinca alkaloids and thalidomide, with a focus on new classes of promising antineoplastic agents, such as epothilones and proteasome inhibitors (Fig. 1).

Figure 1 – Chemical structure of the most widely used neurotoxic antineoplastic drugs*
* Source:  http://pubchem.ncbi.nlm.nih.gov/
Abbreviations:
a = cisplatin; b = carboplatin, c = oxaliplatin; d = paclitaxel; e = docetaxel; f = vincristine;
g = ixabepilone; h = sagopilone; i = bortezomib; j = thalidomide; k = lenalidomide;
l = pomalidomide

Chemotherapy-Induced Peripheral Neurotoxicity-fig1

2. PLATINUM DRUGS

Since platinum compounds were identified as antineoplastic agents (cisplatin has been used since 1970), their use has been increasingly adopted in routine oncological clinical practice. While the toxicity profile differs among the different drugs, platinum-induced PN is a common feature.

2.1 Pathogenesis

DRG represent the main target of platinum compound-induced damage (Thompson 1984; Krarup-Hansen 1999; Gregg 1992; Meijer 1999; Cavaletti 2001) and platinum levels in the DRG of treated patients correlate with the severity of their PN (Gregg 1992; Dzagnidze 2007). Although platinum compounds differ in several chemical properties, the primary mechanisms involved in platinum-induced PN are probably similar (Cavaletti 1998a; Cavaletti 2001).

Two main mechanisms have been proposed to explain the physiopathology of platinum-induced PN. Firstly, the platinum compounds form intrastrand adducts and interstrand crosslinks which alter the tertiary structure of the DNA (Ta 2006; McDonald 2005). This effect on DNA promotes alterations in cell-cycle kinetics resulting in the upregulation of cyclin D1 expression and hyperphosphorylation of the retinoblastoma gene product, with an attempt of differentiated postmitotic DRG neurons to re-enter into the cell cycle resulting in an induction of apoptosis (Gill 1998). The second mechanism proposes an involvement of oxidative stress and mitochondrial dysfunction as the trigger of neuronal apoptosis (Zhang 2007).  PN could be modulated by a reduction in the activity of enzymes involved in DNA base excision, repair of oxidative damage and in redox regulation (Jiang 2008). PN could also involve apoptosis, mediated by p53 increased activity and mitochondrial release of cytocrome-c pathway, independent of fas receptor activation (Mcdonald 2002). Another recently described modulator of proteins involved in platinum-induced apoptosis is the activation of p38 and ERK1/2 (Scuteri 2009).

Acute oxaliplatin neurotoxicity (see below) is thought to be caused by a dysfunction of nodal axonal voltage-gated Na+ channels (Krishnan 2005), probably the Ca2+-dependent channels. This effect is likely to be due to the oxalate chelating effect on both Ca2+ and Mg2+ which could interfere with channel kinetics (Adelsberger 2000) and reduce the overall Na+ current (Grolleau 2001).

2.2 Clinical and electrophysiological characteristics

The earliest signs of PN observed in platinum-treated patients are a decreased vibratory sensitivity in the toes and loss of ankle jerks, associated with numbness, tingling or paresthesias in finger and toes. Prolonged treatment may worsen symptoms and signs, with generalized loss of deep tendon reflexes (DTR) and more proximal vibratory sensitivity impairment. Pin and temperature sensation, joint position and light touch perception are less severely affected. In the worst cases the loss of propioception may result in an ataxic gait. Lhermitte’s phenomenon secondary to DRG cell degeneration and spinal cord dorsal columns damage can be occasionally observed (Eeles 1986; Jurado 2008).

Oxaliplatin-induced PN presents as two distinct clinical syndromes: an acute, cold-induced and transient syndrome characterized by paresthesias in the distal extremities and perioral region that usually appears during or within hours after infusion, and a chronic cumulative sensory neuropathy with the typical features of platinum drug-induced PN. A smaller number of patients present with slurred speech, jaw pain during chewing and paresthesias in the extremities or calf cramps with walking that tend to persist for days to weeks (Wilson 2002; de Gramont 2000; Park 2009).

Nerve conduction studies performed in patients treated with platinum drugs consistently demonstrated sensory axonal damage with reduced amplitude of the sensory nerve action potentials (SNAP) (Krarup-Hansen 2007; Thompson 1984; Daugaard 1987; Cavaletti 1991; Krishnan 2005; Argyriou 2007). Motor nerve conduction velocities (NCV), compound muscle action potentials (CMAP) and F-wave latencies remain unchanged during treatment. Conflicting findings have been observed in autonomic testing (Hansen 1990; Bogerd 1990; Earl 1998; Krarup-Hansen 2007).  Neurophysiological studies performed within 24-48h after oxaliplatin infusion showed neuromyotonic discharges and repetitive compound muscle action potentials. All these abnormalities resolved within 3 weeks after oxaliplatin administration (Lehky 2004).

2.3 Outcome

PN usually develops during platinum drug treatment, but symptoms and signs may progress for 2 to 6 months after cessation of chemotherapy (the so called “coasting” effect, occurring in up to 30% of cisplatin-treated patients) (Siegal 1990; Schlippe 2001) and frequently recovery is incomplete.  Recently, a cross-sectional study of testicular cancer patients re-evaluated between 23 and 33 years after finishing treatment, showed that PN remains detectable in up to 20% of patients, being symptomatic in 10% of them (Glendenning 2010). Similar results were found in another study that evaluated cisplatin-treated patients after a median follow-up of 15 years: 38% and 28% patients had asymptomatic and symptomatic neuropathy respectively, which was disabling in 6% (Strumberg 2002). The cumulative dose of cisplatin is the main risk factor associated with the persistence of neurotoxicity (Schlippe 2001).

Chronic oxaliplatin PN is partially reversible in about 80% of patients and completely resolved in about 40% of them at 6-8 months after treatment discontinuation (de Gramont 2000; Andre 2004; Wilson 2002). However, coasting may also occur with this drug. Two studies have reported persistence of neuropathy in almost 35% of patients 5-6 years after cessation of oxaliplatin treatment (Pietrangeli 2006; Brouwers 2009). It has been reported that PN may be exacerbated by surgery just after finishing treatment with oxaliplatin. In these patients surgery-induced haemolysis is thought to release oxaliplatin that has accumulated in erythrocytes (Gornet 2002).

2.4 Incidence, severity and risk factors

Platinum-induced PN is related to the total cumulative dose (Gregg 1992; van der Hoop 1990; Thompson 1984; Glendenning 2010) and to the dose-intensity of treatment (Cavaletti 1992; Mollman 1988). When using cisplatin, the onset of PN is expected to occur after the administration of 250-350 mg/m 2 (Thompson 1984; Glendenning 2010), while at a cumulative dose of 500-600 mg/m2, almost all patients have objective evidence of neuropathy (Roelofs 1984), being severely disabling in at least 10% of patients (Van der Hoop 1990)  (Table 1). Recently, reduced neurotoxicity with the use of a liposomal formulation of cisplatin (Lipoplatin) has been reported (Boulikas 2009).

Carboplatin induced-PN is less frequent and severe at conventional doses than is in cisplatin-treated patients (Van der Hoop 1990; Gurney 1990; Adams 1989).

In combination schedules, the use of cisplatin or carboplatin with taxanes has been reported to induce higher incidence of PN, even at lower cumulative dose of each antineoplastic agent (Hilkens 1997; Chaudhry 1994; Argyriou 2005 and 2007; Rowinsky 1993), but this result has not always been confirmed (Cavaletti 1997; Verstappen 2003). Retreatment with neurotoxic drugs is a feasible option in several patients (van den Bent 2002). Combinatory therapies with other neurotoxic agents and low magnesium serum levels have also been related to a higher incidence and severity of neuropathy (Glendenning 2010; Bokemeyer 1996). Other risk factors including alcohol consumption, smoking, familial diabetes, serum creatinine levels, age, tumour characteristics or performance status are not considered as being related to the onset or the severity of PN (Mollman 1988; Van der Hoop 1990). Polymorphisms in genes involved in the platinum biotransformation (McWhinney 2009) and in the mechanisms for repairing platinum-DNA adducts (Dzagnidze 2007) could modulate and partially explain the individual differences in the severity of the PN induced by these drugs (Kweekel 2005).

Oxaliplatin-induced acute symptoms are transiently present in the vast majority of patients. Sensory symptoms are usually reversible over hours to days. Conversely, chronic PN persists between cycles and severity increases with cumulative dose (Rothenberg 2003; de Gramont 2000; Argyriou 2007; Al-Batran 2008; Park 2009). Severe PN has been observed with cumulated doses ranging from 510-765 mg/m2 in up to 10% of patients, but it can affect approximately 50% of the patients receiving doses higher than 1000 mg/m2 (de Gramont 2000; Souglakos 2002).Table 2 summarizes phase III and neurotoxicity-focused oxaliplatin trials published in the literature. It has been suggested that sensory excitability techniques provide an early predictor of the chronic neurotoxicity of oxaliplatin (Park 2009). Furthermore, sustained thermal hyperalgesia after the third oxaliplatin cycle has also been identified as an early predictor of chronic PN (Attal 2009). Few studies have attempted to detect distinctive gene polymorphisms that identify patients at high-risk of developing PN (Kweekel 2005; McWhinney 2009). Glutathione S-transferase P1 polymorphism (Ile105Val) and SCNA 1A polymorphism (T1067A T/T) have been suggested as identifying subjects at increased risk of developing oxaliplatin-induced PN. Integrin beta-3 (IGTB3) polymorphism at residue 33 might also represent a specific biomarker that predicts the incidence and severity of chronic PN (Antonacopoulou, 2010).

Table 2. Incidence of oxaliplatin-induced neuropathy. Selection criteria: phase III studies and prospective studies with specific neurologic assessment tools.
FACT/GOG-Ntx: Functional Assessment of Cancer Therapy Scale/Gynecologic Oncology Group-Neurotoxicity scale. NCI-Sanofi: oxaliplatin Sanofi-developed specific questionnaire. NCI-CTC: National Cancer Institute-Common Toxicity Criteria. WHO: World Health Organization. QLQ-C30: European Organization for the Treatment of Cancer quality of life questionnaire-30 items PN: peripheral neuropathy. NSS: Neurological Symptom Score. NDS: Neurological Disability Score. FGS: Functional Grading Scale. NCS: nerve conduction studies. TNS: total neuropathy score. For a description of the scales see Cavaletti et al. (2010).

Author

(Year)

Patients

(n)

Chemotherapy schedule

Dose-Intensity planned (mg/m2/week)

Neurological

Assessment

Neuropathy

Incidence

PHASE III STUDIES

Land

(2007)

189

85 mg/m2 at week 1,3,5 of a 8 weeks cycle for 3 cycles

31.875

FACT/GOG-Ntx NCI-CTC

NCI-Sanofi grade

At week 12:

NCI-Sanofi. Grade 1-2: 68.3%

Grade 3-4: 3%

NCI-CTC Grade 1-2:61.3%

Grade 3-4:2.6%

Giacchetti (2000)

100

125 mg/m2 every 21 days, as a continuous 6-hour intravenous infusion

41.67

Specific

scale of peripheral neuropathy

Grade 1: 32%

Grade 2: 13%

Grades 3-4: 10%

Tournigand (2006)

309

303

85 mg/m2 every 14 days for 12 cycles

130 mg/m2 every 21 days x6 cycles, stop OXL 6 weeks and reintroduction other 6 OXL cycles

42.5

37.14

NCI-CTC

Grade 1: 34%

Grade 2: 37%

Grade 3: 18%

Grade 4: 0 %

Grade 1: 36%

Grade 2: 42%

Grade 3: 13%

Grade 4: 0 %

de Gramont (2000)

209

85 mg/m2 every 14 days,

until progression disease

42.5

NCI-CTC

Grade 1: 20.6 %

Grade 2: 29.2 %

Grade 3: 18.2 %

Grade 4: 0%

André

(2004)

1108

85 mg/m2 every 14 days for 12 cycles

42.5

NCI-CTC

Grade 1: 46%

Grade 2: 33.6%

Grade 3: 12.4%

Rothenberg

(2003)

153

150

85 mg/m2 every 14 days, median cycles 3 (1-18)

85 mg/m2 every 14 days, median cycles 3 (1-16)

42.5

specific scale

(similar to

NCI-CTC)

Acute: Grade 1-2: 51%

Grade 3-4: 7%

Chronic: Grade 1-2: 49%

Grade 3-4: 2%

Acute: Grade 1-2: 50%

Grade 3-4: 3%

Chronic: Grade 1-2: 48%

Grade 3-4: 3%

Allegra

(2009)

1321

1326

85 mg/m2 every 14 days for 12 cycles

85 mg/m2 every 14 days for 12 cycles, plus bevacizumab

42.5

NCI-CTC

Grade 2: 29%

Grade 3: 14%

Grade 2: 33%

Grade 3: 16%

Al-Batran (2008)

112

85 mg/m2 every 14 days,

until progression disease

42.5

NCI-CTC

OXL-specific scale

Grade 1-2: 49.2%

Grade 3-4: 14.3%

Wang

(2007)

44

85 mg/m2 every 14 days

42.5

WHO

NCS

Grade 1-2: 40.9%

Grade 3-4: 31.8%

Ishibashi

(2010)

16

85 mg/m2 every 14 days

42.5

NCI-CTC

Debiopharm neurotoxicity scale

NCI-CTC

Grade 1: 94%

Grade 2: 6%

Grade 3-4: 0%

Cassidy

(Gibson 2006)

324

85 mg/m 2 every 14 days

42.5

NCI-CTC

Grade 1: 38%

Grade 2: 18.8%

Grade 3-4: 16.7%

Cassidy (2008)

655

649

130 mg/m2 every 21 days, maximum 16 cycles

85 mg/m2 every 14 days, maximum 24 cycles

43.33

42.5

NCI-CTC

Equal in both groups

Grade 1: 11 %

Grade 2: 5 %

Grade 3: 4 %

Grade 4: 0 %

Cunningham (2008)

452

130 mg/m 2 every 21 days, maximum 8 cycles

43.33

NCI-CTC

QOL-QC30

Grade 1-2: 75.2%

Grade 3-4: 6.4%

Cathomas (2009)

20

130 mg/m2 every 21 days, maximum 6 cycles

heated at 37ºC

43.33

NCI-CTC

Neurological symptoms questionnaire (NCI)

NCI-CTC

Grade 1: 85%

Grade 2: 15%

Questionnaire :

Dysestesias/ Paresthesias/ Acute

Grade 1: 35% / 40%/ 35%

Grade 2: 10% / 20%/ 5%

Grade 3: 30% / 20%/ 30%

Grade 4: 5% / 10%/ 20%

Qvortrup (2010)

69

70

130 mg/m2 every 21 days for 8 cycles

130 mg/m2 every 21 days for 8 cycles, chronotherapy

43.33

NIC-CTC

Grade 2 : 42%

Grade 3: 16%

Grade 2: 27%

Grade 3: 19%

Poplin

(2009)

272

100 mg/m2 every 14 days, until progression disease

50

NCI-CTC

Grade 3: 9.5%

Louvet

(2005)

157

100 mg/m2 every 14 days, until progression disease

50

NCI-CTC

Grade 1-2: 76%

Grade 3: 19.1%

Cascinu

(2002)

26

100 mg/m2 every 14 days

50

NCI-CTC

After 8 cycles:

Grade 1: 21.1%

Grade 2: 31.6%

Grade 3-4: 26.3%

  Table 2. Studies with specific neurological assessment tools.
STUDIES WITH SPECIFIC NEUROLOGICAL ASSESSMENT TOOLS

Krishnan

(2005)

16

100 mg/m2 every 14 days

50

NCI-CTC

NSS

Oxaliplatin-specific neurotoxicity score

NCS

NCI.CTC:

Grade 1-2: 12.5%

Grade 3-4: 37.5%

NSS:

1: 25%

2: 12.5%

3: 12.5%

Argyrou (2006)

20

85 mg/m2 every 14 days for 12 cycles

42.5

NSS

NDS

NCS

PN: 75%

TNS: 11.2±9.05

NDS:20±23.1

NSS: 1.5±1.3

Argyrou (2007)

25

85 mg/m2 every 14 days for 12 cycles

42.5

NSS

NDS

FGS

NCS

TNS

WHO CIPN (1-3)

TNS:

1-11 (mild): 24%

12-23 (moderate): 32%
≥ 24 (severe): 8%

Mean NDS: 21.1±17.5

Mean NSS: 1.8±0.8

 

2.5 Options for neuroprotection

Several chemoprotective agents have been tested for their ability to limit or prevent the PN induced by cisplatin or carboplatin. A recent Cochrane Library review involving amifostine, diethyl-dithio-carbamate, glutathione, Org 2766 and vitamin E concluded that data were insufficient to support their effectiveness in cisplatin PN (Albers 2007). A more recent trial with vitamin E provided class II evidence about its benefit in preventing cisplatin-induced PN (Pace 2010), but the results need to be confirmed in a larger setting. Until then, dose-reduction, or a longer interval between cycles, represent the only way to ameliorate cisplatin-induced PN.

Similarly, schedule modification and eventually, treatment withdrawal and later reintroduction when the PN has been resolved, are the methods to minimize oxaliplatin-induced PN (Mattioli 2005; Maindrault-Goebel 2004). “Chronomodulation” (i.e. a dosing schedule based on circadian rhythm) has recently been evaluated in a meta-analysis, but results are conflicting (Liao 2010). Several agents have been tested as neuroprotectants against oxaliplatin-induced PN (Argyriou 2008), including glutathione (Milla 2009; Cascinu 2002), glutamine (Wang 2007), xalopriden (Gibson 2006) and oxcarbazepine (Argyriou 2006). Another, still controversial, pharmacological approach is the intravenous administration of calcium gluconate and magnesium sulfate before and after oxaliplatin infusion (Gamelin 2008; Ishibashi 2010). A large phase III trial (Combined Oxaliplatin Neuropthy Prevention Trial, CONCEPT) was terminated early because it has been suggested that Ca/Mg intravenous supplementation may interfere with the cytotoxic activity of oxaliplatin-based schedules (Hochster 2007). However, a later results review of this study performed by an independent board did not find differences in response rate between patients receiving placebo or calcium and magnesium infusion. In line with this, a multicentre study, reported as an abstract, confirmed the safety use of calcium and magnesium salts and their benefit in lowering as the frequency and severity of chronic oxaliplatin neuropathy (Gamelin 2008), although these positive results have not been confirmed in another recently published study (Ishibashi 2010).

3. VINCA ALKALOIDS

This group of chemotherapeutic agents includes both natural alkaloids, such as vincristine and vinblastine, and semi-synthetic compounds, such as vindesine, vinorelbine and vinflunine. These drugs have a broad spectrum of indications in the treatment of haematologic and lymphatic malignancies as well as of solid tumours such as breast, ovarian, testicular, brain and non-small cell lung tumours and sarcomas (Chu 2009). Vinca alkaloids have a different toxicity profile; vincristine being the most neurotoxic drug.

3.1 Pathogenesis

Vinca alkaloids exert their antineoplastic effect by inhibiting microtubule dynamics in mitotic spindles, resulting in an arrest of dividing cells at the metaphase stage and ultimately leading to cell death (Jordan 1992). Vinca alkaloids form a stable complex in the GTPase domain of b-tubulin, inhibiting the GTP hydrolysis, which prevents its polymerization from soluble dimers into microtubules. The affinity for tubulin differs among vinca alkaloid compounds (vincristine > vinblastine > vinorelbine > vinflunine) and this biochemical property could explain the distinct neurotoxic profile of these drugs (Lobert 1996). The effect on tubulin dimers produce loss of axonal microtubules and alteration in their length, arrangement and orientation leading to axonal swelling in both myelinated and unmyelinated fibers (Sahenk 1987; Tanner 1998; Topp 2000). These alterations of the neuronal cytoskeleton lead to abnormalities in axonal transport, build up of neurofilaments in the cell bodies and proximal axons and progressive accumulation of axoplasmic organelles and vesicles (Shelanski 1969; Schlaepfer 1971; Sahenk 1987; Topp 2000). Secondary to axonal damage, reduction of myelin thickness (Callizot 2008), shortening of inter-nodal length and segmental demyelination can occur (McLeod 1969; Gottschalk 1968).

3.2 Clinical and electrophysiological characteristics

The first manifestation of vincristine-induced PN is the decrease/loss of DTR (DTR) followed by paresthesias (Sandler 1969, Casey 1973). If treatment continues muscular weakness can occur. Objective touch and two-point discrimination sensory loss are usually infrequent or mild and limited to fingers and toes; however 70% of asymptomatic patients have increased touch detection thresholds (Postma 1993). Vibration perception is rarely more than mildly impaired (Sandler 1969; Casey 1973; DeAngelis 1991), but when measured with quantitatively methods it is reduced in 20% of patients (Verstappen 2005). Joint position sense usually remains intact (Sandler 1969; Verstappen 2005). Pain can be observed in some patients, usually restricted to the glabrous skin of the fingertips and toes, with elevated sharpness and warm detection thresholds in these areas (DeAngelis 1991; Dougherty 2007). When motor weakness appears, it is most evident in the dorsiflexors of ankles and toe/fingers extensors muscles (Sandler 1969; Casey 1973; DeAngelis 1991; Verstappen 2005). In high-intensity treatments muscle cramps are frequent (Haim 1994 and 1991). Symptoms attributable to severe autonomic dysfunction, e.g. colicky abdominal pain and constipation, can occur even within a few days of drug administration and precede paresthesias or DTR reduction (Haim 1994; Sandler 1969). Impotence has been reported to occur in between 15-24% of patients (Haim 1994; Kornblith 1992). Other dysautonomic manifestations such as urinary retention and orthostatic hypotension have been reported anecdotally (Haim 1994, Wheeler 1983).

At neurophysiological examination, both SNAP and CMAP have been reported to decrease with treatment duration. Spontaneous fibrillation associated with reduction in the interference pattern is found in all distal muscles (Casey 1973). Although sensory symptoms and signs improve when the treatment is withdrawn, SNAP remains altered in most patients (Casey 1973; Ramchandren 2009). Only slight reduction of NCV is demonstrated in either sensory or motor fibers even in the setting of severe PN (Casey 1973; DeAngelis 1991; McLeod 1969).

3.3 Incidence, severity and risk factors

Although vincristine is one of the oldest neurotoxic antineoplastic drugs, the exact incidence of PN is still unknown due to the heterogeneity in chemotherapeutic regimens that are employed for the treatment of different type of tumours. Table 3 reports the results obtained in the most relevant trials.

Table 3. Incidence of vinca alkaloids-induced neuropathy. Selection criteria: prospective studies including information about neuropathy and neurological assessment tools. 

NA: not available. NCI-CTC: National Cancer Institute-Common Toxicity Criteria. PN: peripheral neuropathy. WHO: World Health Organization. NCS: nerve conduction studies. SWOG: South West Oncology Group. QLQ-C30: European Organization for the Treatment of Cancer quality of life questionnaire-30 items. VPT : Vibration perception threshold. NSS: Neurological Symptom Score. For a description of the scales see Cavaletti et al. (2010)

Author

(year)

Patients (n)

Schedule

Dose-intensity Planned

(mg/m2/week)

Neurological

Assesment

Neuropathy

Incidence

VINCRISTINE

Holland

(1973)

393

Induction:

75 µg/kg/week for 4 months.

NA

Not validated neuropathy related symptoms scale, similar to NCI-CTC

Grade 1-2: 23%

Grade 3: 33%

Grade 4: 25%

Grade 5: 10%

Watkins

(1978)

10

13

23

9

5

Mean dose:

0.28 (0.06-0.74) mg/kg

0.23 (0.05-0.75) mg/kg

0.3 (0.05-0.57) mg/kg

0.42 (0.07-1.47) mg/kg

0.23 (0.06-0.51) mg/kg

NA

Neurologic clinical assessment

Patients with PN:

60%

61.5%

4.3%

22%

40%

Haim

(1994)

104

1.4 mg/m2 every cycle

NA

WHO

Sensory:

Grade 1: 52.8%

Grade 2-3: 11.5%

Motor:

Grade 2: 17.3%

Grade 3: 11.5%

DeAngelis

(1991)

27

0.5 mg/m2/day during 4 days every 7 days for 12 cycles

2

NCS

Neurologic clinical assessment

All patients had moderate to severe signs and symptoms of  sensory-motor neuropathy.

Rea

(2006)

31

2 mg every 7 days for 4 cycles

2

NCI-CTC

Grade 2-3: 45%

Reinders-Messelink (2000)

11

1.5 mg/m2/week for 8 cycles

1.5

WHO

VPT

NCS

Grade 1: 9%

Grade 2: 73%

Grade 3-4: 0%

33% sensory abnormalities

22% VPT abnormalities

Broun

(1993)

32

1.4 mg/ m2 was weekly as iv bolus for 4 weeks, then every other week

1.4

NCI-CTC

Grade 2-3: 34%

Verstappen

(2005)

47

67

2 mg every 21 days

4 mg every 21 days

0.67

1.33

Neuropathy scales

Symptoms scales

VPT

Paresthesias 34%

Numbness 43%

Pain 14%

Paresthesias  60% (10%severe)

Numbness 70%  (4% severe)

Pain  62 %  (16% severe)

Jackson

(1984)

25

0.5 mg plus 0.25 mg/m2/ day during 5 days every  21 days

0.58

Neurologic clinical assessment

Patients with PN: 48%

Depression reflexes: 36%

Motor weakness: 4%

Sensory disturbances: 24%

Taylor

(1997)

27

1.6 mg/m2 every 21 days x 6 cycles

0.53

SWOG criteria

Grade 1-2: 22% (mild paresthesias)

Sandler

(1969)

50

2 mg/m2 every 14 days

0.5

NCS

Neurologic clinical assessment

Depression reflexes: 100%

Motor weakness: 34%

Sensory disturbances: 46%

Powles

(1991)

105

1.4 mg/m2 every 21 days

0.47

WHO

Grade 1:24%

Grade 2:14%

Grade 3-4:5%

Walewski

(2010)

24

1.4 mg/m2 every 21 days for 6-8 cycles

0.47

NCI-CTC

No peripheral neuropathy

Thant

(1982)

11

1.4 mg/m2 twice every 21 days for 3 cycles

0.47

NCS

Neurologic clinical assessment

Grade 1-2: 63.6%

Grade 3-4: 36.4%

Klasa

(2002)

44

1.2 mg/m2 every 21 days

0.4

WHO

QLQ-C30

Grade 1: 54%

Grade 2: 11%

Grade 3-4: 2%

Katsumata

(2003)

23

1.2 mg/m 2 every 21 days

0.4

NCI-CTC

Grade 1:      0%

Grade 2:    26.1%

Grade 3-4:   0%

Leighl

(2006)

31

1.2 mg/m 2 every 21 days

0.4

NCI-CTC

Grade 3-4: 13%

van Kooten

(1992)

15

1.4 mg/m2 twice every 28 days for 6 cycles

0.35

VPT

Neurologic clinical assessment

Patients with PN: 80%

 Table 3. Liposomal Vincristine
LIPOSOMAL VINCRISTINE

Sarris

(2000)

35

2 mg/m2 every 14 days for 12 cycles

1

NCI-CTC

Grade 3-4: 31.4%

Rodriguez

(2009)

119

2 mg/m2 every 14 days for12 cycles

1

NCI-CTC

Grade 3-4: 32%

VINBLASTINE

Druker

(1989)

24

6 mg/m 2 on days 1 and 8 every 28 days

3

Neurologic clinical assessment

mild PN: 8%

VINORELBINE

Pace

(1996)

23

25 mg every 7 days for 24 cycles

NA

NSS

NCS

WHO

WHO>2 : 0%

Moderate (5-10): 66.6%

Severe (>10): 33%

The severity of vincristine-induced PN is dose-related. All patients receiving at least 4 mg/m2 have reduction or loss of ankle reflexes and most patients who have received total doses of 2-6 mg/m2 also report mild distal paresthesias. A relevant number of patients receiving a total dose of 8 mg/m2 develop motor weakness or gait impairment (Sandler 1969). Neuropathic pain is relatively frequent in high-dose vincristine-treated patients (Sandler 1969; Dougherty 2007).

From the literature, two conditions emerged as possible high-risk situations for a severe course and early onset of vincristine PN: unrecognized hereditary peripheral neuropathies (Graf 1996) and hepatic insufficiency (Sandler 1969). Age and poor nutritional status have not been well demonstrated as risk factors (Thant 1982).

A few cases of severe fulminating neuropathy during vincristine treatment, mimicking Guillain-Barré syndrome features, have also been described (Bakshi 1997).

3.4 Outcome

Vincristine neuropathy is usually reversible when therapy is discontinued. The median duration of paresthesias and motor weakness after treatment discontinuation is 3 months (Haim 1994). However, a coasting effect has been reported in the first month after finishing therapy in up to 30% of patients, being more prevalent in high-dose intensity regimens (Verstappen 2005). Although most DTR reappear, ankle reflexes recovery is uncommon (Sandler 1969; Casey 1973; Postma 1993).

3.5 Options for neuroprotection

To minimize the potential neurotoxic effects of vincristine the usual recommended vincristine dose is 1.4 mg/m2 per single dose with an upper limit of 2 mg on single doses, regardless of body surface area.

In addition, there are no pharmacological treatments to reduce or prevent vinca alkaloid-induced PN. Randomized, double-blind trials assessing the efficacy of Org2766 (Koeppen 2004) and gangliosides (DeAngelis 1991) have not shown benefit. Up to now, reduction in dose level and frequency of administration, or even treatment discontinuation, are the only proven methods to ameliorate and stop the progression of vinca alkaloid-induced PN.

3.6 Other vinca alkaloids

Vinblastine-induced PN is similar, but less severe, to that observed with vincristine; however haemathological toxicity associated with vinblastine administration usually precedes neurotoxicity, and represents the main dose-limiting adverse effect (Chu 2009).

Vinorelbine-treated patients usually develop a mild, distal axonal sensory-motor neuropathy, involving small and large fibers without evidence of C-fiber dysfunction, as evaluated by sympathetic skin response. The most common symptoms and signs include DTR impairment (94%), paresthesias (50%) and hypopallaesthesia (9%) with mild overall neurotoxicity (Pace 1996). Vinorelbine-induced PN is also dose-dependent and reversible after drug discontinuation (Pace 1996).

NCI-CTC scale grade 1-2 sensory-motor neuropathy has been reported in 10% of patients treated with vinflunine. Notably, autonomic neuropathy characterized by constipation and abdominal pain is the most frequently reported feature, occurring in 20% of patients, and being severe grade in 8% (Krzakowski 2010; Culine 2006).

Liposomal vincristine is a novel formulation developed with the aim of improving the efficacy and pharmacokinetic profile without increasing neurotoxicity. Conflicting results about its PN safety have been reported in phase I and II trials. Given at doses ranging between 2 and 2.8 mg/m2, sensory-motor neuropathy has been reported in 12 to 55% of treated patients, being severe in 7-34% (Sarris 2000; Rodriguez 2009; Gelmon 1999).

4. TAXANES

Taxanes (paclitaxel and docetaxel) belong to the family of chemotherapy agents classified as microtubule-stabilizing agents (MTSAs) and they are effective in treating various types of solid tumours. PN is considered as being the main non-haematological toxicity of taxanes, usually resulting in dose modification (Windebank 2008).

4.1 Pathogenesis

Through their action of disrupting microtubules of the mitotic spindle and the subsequent interference in axonal transport, taxanes are able to affect the soma of sensory neurons as well as axons. In addition, it has been demonstrated that they evoke a “dying back” process starting from distal nerve endings followed by disturbed cytoplasmatic flow in the affected neurons (Cavaletti 1995; Persohn 2005).
Injury of neuronal and non-neuronal cells within the peripheral nervous system, macrophage activation in both the DRG and peripheral nerve, and microglial activation within the spinal cord are all cellular processes that also seem to contribute greatly to the genesis of taxane-induced PN through signal transduction-mediated events (Peters 2007). Finally, results from a recently published experimental study showed that paclitaxel leads to massive polar reconfiguration of axonal microtubules, accompanied by impaired organelle transport in cultured Aplysia neurons (Shemesh 2010).

4.2 Clinical and electrophysiological characteristics

Taxane-based therapy usually induces paresthesias, numbness and/or pain in a stocking-and-glove distribution. Decreased vibration perception and sense of position, loss of pain and temperature sensation and DTR impairment are also evident at clinical examination (Argyriou 2008). Electrophysiological abnormalities, mainly involving the decrease of SNAP or abolishment of sensory responses, indicate axonal sensory PN. Reduction of CMAP occurs at the highest cumulative doses, while sensory and motor NCV are usually spared. Motor involvement with distal or proximal weakness and myopathy are less frequently seen (Windebank 2008).

4.3 Incidence, severity and risk factors

Besides the National Cancer institute – Common Toxicity Criteria (NCI-CTC), the 11-item neurotoxicity subscale Functional Assessment of Cancer Therapy/Gynaecological Group Neurotoxicity  (FACT/GOG-Ntx) and the FACT/Taxane scale have also been employed to evaluate the incidence and grade the severity of taxanes-induced PN. Recently, the Total Neuropathy Score (TNS) (Cavaletti 2003; Cavaletti 2006; Cavaletti 2007) as well as other modifications of the TNS have been used to assess the neurotoxicity secondary to paclitaxel-based treatment (Argyriou 2007).
The administration of paclitaxel-containing regimens is consistently associated with an increased incidence of PN compared with docetaxel (Chon 2009). The incidence and severity of taxanes-induced PN is mainly related to the cumulative dose. Generally, severe taxane-induced PN occurs in patients receiving cumulative doses around 1000 mg/m2 for paclitaxel and 400 mg/m2 for docetaxel (Lee 2006). Table 4 summarizes phase III and neurotoxicity-focused taxanes trials.

Table 4. Incidence of taxanes-induced neuropathy. Selection criteria: phase III studies with specific neurologic assessment tools.
NCI-CTC: National Cancer Institute-Common Toxicity Criteria. TNS: Total Neuropathy Score. NCS: nerve conduction studies. PN: peripheral neuropathy. For a description of the scales see Cavaletti et al. (2010).

Author

(Year)

n

Chemotherapy schedule

Neurological

Assessment

Neuropathy

Incidence

Vassey

(2004)

532

Paclitaxel 175 mg/m2 and carboplatin AUC 5 every 21 days for 6 cycles

NCI-CTC

Neurotoxicity Score

Sensory Grade 1-2: 70 pts

Grade 3-4: 8 pts

Motor Grade 1-2: 14 pts

Grade 3-4: 2 pts

Winer

(2004)

152

Paclitaxel 210 mg/m2 administered as a 3-hour infusion every 3 weeks

NCI-CTC

Grade 3-4 sensory: 19%

Grade 3-4 motor: 11%

Winer

(2004)

149

Paclitaxel 250 mg/m2 administered as a 3-hour infusion every 3 weeks

NCI-CTC

Grade 3-4 sensory: 33%

Grade 3-4 motor: 14%

Smith

(1999)

278

Paclitaxel 250 mg/m2 administered as a 3-hour infusion every 3 weeks

NCI-CTC

Grade 3-4 sensory: 13%

Grade 3-4 motor: 9%

Smith

(1999)

279

Paclitaxel 250 mg/m2 administered as a 24-hour infusion every 3 weeks

NCI-CTC

Grade 3-4 sensory: 7%

Grade 3-4 motor: 7%

Argyriou (2005)

21

Paclitaxel 175 mg/m2 and carboplatin AUC 5 every 21 days for 6 cycles

TNS

NCS

Overall neurotoxicity: 14/21 pts (66.6%).

TNS:

1-11 (mild): 28.6%

12-23 (moderate): 50%
≥ 24 (severe): 21.4%

Krzakowski

(2010)

275

Docetaxel 75 mg/m2 1-hour infusion every 21 days for 6 cycles

NCI-CTC

Overall neurotoxicity:

Grade 1-4: 15%.

Kruijtzer

(2000)

253

Docetaxel 100 mg/m2 1-hour infusion every 21 days for 6 cycles

NCI-CTC

Grade 3-4 sensory: 2%

Grade 3-4 motor: 1%

Vassey

(2004)

539

Docetaxel 75 mg/m2 1-hour infusion every 21 days for 6 cycles

NCI-CTC

Neurotoxicity Score

Sensory Grade 1-2: 44 pts

Grade 3-4: 2 pts

Motor Grade 1-2: 8 pts

Grade 3-4: 1 pts

Prior or concomitant administration of platinum compounds and pre-existing PN due to various medical conditions are also considered to increase the incidence and severity of taxane-induced PN. Moreover, the risk appears to be related to treatment schedules (weekly versus every three weeks treatment schedule) for paclitaxel (Sparano 2008). However, a recent metanalysis does not support this view as it has been demonstrated that the incidence of serious adverse events, including PN, was significantly lower in weekly taxane schedules (Mauri 2010). Recently published data show that elderly cancer patients are not at increased risk of developing taxanes-induced PN (Argyriou 2006).

4.4 Outcome

Generally, symptoms of taxane-induced PN improve or resolve within the first 3-6 months after the discontinuation of treatment (Argyriou 2005a). However, severe symptoms may persist for a much longer time.

4.5 Options for neuroprotection

Strategies to prevent taxane-induced PN evaluated the use of several neuroprotective drugs, including thiols, neurotrophic factors and antioxidants, but none of them proved effective in large, controlled, clinical trials (Park 2008). The clinical efficacy of amifostine has been tested with conflicting overall negative, results (Moore 2003; Argyriou 2008). It was demonstrated in a small phase II clinical trial that glutamine is potentially able to preventing high dose paclitaxel-induced PN (Stubblefield 2005). In a pilot phase II clinical trial, acetyl-l-carnitine significantly reduced the severity of paclitaxel-induced PN (Maestri 2005). Two phase II randomized controlled trials, have shown that vitamin E at a daily dose of 600mg bid may also exert neuroprotective effects in patients either treated with paclitaxel-based regimens alone or with paclitaxel/cisplatin regimens (Argyriou 2006; Argyriou 2005b). The results of a large international clinical phase III trial that sought to assess the efficacy of vitamin E against chemotherapy-induced PN, including taxane-induced PN, have recently been reported. In that setting, vitamin E did not appear to reduce the incidence of sensory neuropathy in patients receiving neurotoxic chemotherapy (Kottschade, in press). However, the negative results of this study conducted by researchers of the Mayo Clinic should be considered with caution, because of some significant methodological problems (Argyriou and Kalofonos, in press).

5. EPOTILONES

It was only in 1995 that a second class of MTSAs was discovered (Bollag 1995; Gerth 1966). These macrolides were secondary metabolites produced by myxobacteria and they were called “epothilones” to reflect their basic structural features, which include an epoxide moiety, a thiazole-containing side chain and a single ketone function. Two major fermentation products were reported in the myxobacterium Sorangium cellulosum Soce 90, epothilone A and epothilone B (also known as patupilone) (Altmann 2000; Hardt 2001).

In addition to these principal compounds, other metabolites were subsequently isolated from the fermentation of Soce 90: epothilones C-F. Of these, only the 12, 13-desoxyepothilones, called respectively epothilone C (desoxyepothilone A) and D (desoxyepothilone B, also known as KOS-862), were produced in significant amounts.

Although epothilones were originally described as natural fungicidal macrolides, interest significantly increased when their capacity to induce microtubule polymerization at submicromolar concentrations was identified in an in vitro screening program. These in vitro studies indicated that epothilones have a taxane-like effect on tubulin, inducing microtubule bundling, the formation of multipolar spindles, and mitotic arrest (Chou 1998; Kamat 2003).

5.1 Pathogenesis

Epothilones, like paclitaxel, are able to induce in vitro the polymerization of tubulin dimers in microtubules in the absence of either guanosine triphosphate (GTP) and/or microtubule-associated proteins, and to stabilize preformed microtubules against conditions favoring depolymerization, including dilution, cold temperatures or Ca2+ (Bollag 1995; Kowalski 1997; Altmann 2000).

Competition studies with paclitaxel have demonstrated that epothilones share either the same or an overlapping binding site on tubulin (Bollag 1995; Altmann 2000). Kinetic experiments have revealed that these compounds are competitive inhibitors of paclitaxel binding to tubulin polymer. These findings suggest that the microtubule binding sites of paclitaxel and epothilones are largely overlapping or even identical, thus supporting the view that the mechanism of their neurotoxicity might be similar.

5.2 Clinical and electrophysiological characteristics

Similar to the observations performed with taxanes since the first phase I studies, PN has emerged as a potentially epothilone-associated severe toxicity (Cavaletti 2004; Lee 2006). However, its real impact and clinical features, as well as the identification of possible risk factors for severe neurotoxicity and the assessment of the final outcome of symptoms and signs, are still unclear. Despite the fact that sensory peripheral neurotoxicity is definitely the main toxic effect on the nervous system, the occurrence of motor or autonomic neuropathies also has occasionally been reported in the clinical trials (mostly phase II, Table 5).

Table 5. Incidence of bortezomib-induced neuropathy. Selection criteria: phase III studies with specific neurologic assessment tools.
 FACT/GOG-Ntx: Functional Assessment of Cancer Therapy Scale/Gynecologic Oncology Group-Neurotoxicity. NCS: nerve conduction studies. TNS: Total Neuropathy Score; TNSc: TNS-clinical versión; TNSr: TNS reduced version PN: peripheral neuropathy. NCI-CTC: National Cancer Institute-Common Toxicity Criteria. For a description of the scales see Cavaletti et al. (2010).

Author

(Year)

Patients

(n)

Chemotherapy schedule

Neurological

Assessment

Neuropathy

Incidence

Richardson (2006)

256

1.0 or 1.3 mg/m2 iv bolus on days 1, 4, 8, and 11, every 21 days, for up to eight cycles

FACT/GOG-Ntx

Neurological examination

Grade 1-2: 22%

Grade 3-4: 13.4%

dose reduction in 12% and discontinuation in 5% of pts

Richardson (2005; 2009)

331

1.3 mg/m2 iv bolus on days 1, 4, 8, and 11 for eight three-week cycles, followed by treatment on days 1, 8, 15, and 22 for three five-week cycles

FACT/GOG-Ntx

Neurological examination

Any grade: 124 (37%)

Grade 3-4: 30 (9%)

discontinuation in 8% of pts

Richardson (2009b)

64

1.3 mg/m2 on days 1, 4, 8, and 11, for up to eight 21-day cycles

Neurologic and neurophysiological examination,

measurement of intraepidermal nerve fiber density

64% of patients (grade 3 in 3%)

Chaudhry

(2008)

27

1.3 mg/m2 on days 1, 4, 8, and 11, for up to eight 21-day cycles; median cumulative dose of 35.6 mg/m2

TNSr

NCS

26/27 pts; median TNSr 10, range 1–24; 11 were grade 1, 10 grade 2, 5 grade 3, and

none grade 4 (TNSr 2–8= grade 1; TNSr 9–16=grade 2; TNSr 17–24=grade 3; and TNSr 25–32=grade 4)

Velasco

(2010)

24

1.3 mg/m2 on days 1, 4, 8, and 11, for up to eight 21-day cycles; cumulative dose of 27.37 ± 11.42

TNSc

TNSr

NCS

NCI-CTC

Grade 1-2: 10(41,7%)

Grade 3-4: 4(16.6%)

dose reduction in 8(33%)

median TNSc 7 (range 5–12)

median TNSr 11 (range 5–16)

 

No neurophysiological investigation has so far been reported in epothilone-treated patients.

5.3 Incidence, severity and risk factors

In the case of epothilones, PN has also generally been assessed using common toxicity scales, although the real incidence is difficult to assess since most of those patients exposed to the drug had been previously treated with other potentially neurotoxic drugs.

Moreover, this family of compounds is new and therefore the data reported so far are relatively scarce. Most of them concern ixabepilone (BMS-247550), a semi-synthetic epothilone B used in refractory breast cancer. A partial description of the ixabepilone-induced PN has been reported in more than 1,000 patients (Eng 2004; Denduluri 2007a and 2007b; Perez 2007; Roche 2007; Thomas 2007a and 2007b; Vansteenkiste 2007; Dreicer 2007).

In these studies the incidence of severe (i.e. grade 3/4 according to the NCI-CTC scale) sensory neuropathy ranged between 6-21%, while overall sensory neuropathy was reported in up to 71% of exposed patients. The relationship existing between the schedule of ixabepilone administration and the incidence of sensory neuropathy is not clear: for  example, using the same schedule (40 mg/m2 delivered via a 3-hour infusion every 3 weeks) an incidence of grade 3/4 sensory neuropathy of 6% has been reported in one study (Dreicer 2007), while this value increased  to 12% and 20-21% in other studies (Eng 2004, Denduluri 2007a and 2007b; Perez 2007; Roche 2007; Thomas 2007a and 2007b). A possible reason for this discrepancy may reside in the schedule modifications that occurred during the course of the clinical trials, based on the emerging safety data.

The important issue of predicting the risk of severe ixabepilone-induced peripheral neurotoxicity was specifically addressed in breast cancer patients (Lee 2006), but several methodological limitations in this study prevent the results obtained being considered as generally applicable in the management of epothilone-treated patients.

5.4 Outcome

A very important issue of ixabepilone-induced sensory neuropathy relates to its long-term outcome, since evidence  from the published data show that symptoms and signs are reversible in most of the  patients displaying neuropathy within weeks of drug treatment withdrawal, i.e. much earlier that in taxane-treated patients. However, the numerical data supporting this assumption are still relatively limited and observations should be extended. Moreover, no details regarding the outcome of patients with the poorest neurological profile are available, so that identification of potential risk factors for an unfavorable long-term course is not possible. In one study, it was stated that ixabepilone PN is irreversible, but the duration of observation after drug withdrawal was not indicated (Eng 2004).

5.5 Other epothilones

The synthesis of epothilones has been the subject of extensive efforts by chemists and currently there have been more than 30 reports of synthesis of epothilones A and B in the literature.

Only one phase II clinical trial has been performed using KOS-862; in that study 71% of the treated patients developed sensory neuropathy (NCI-CTC grade 3 = 10%). Motor neuropathy was present in about 5% of the patients and ataxia (of unknown origin) in 8% of the patients (Beer 2007).

The first fully synthetic third-generation epothilone B derivative, ZK-EPO (sagopilone) is currently under clinical investigation in a trial designed to assess its activity and PN.

6. BORTEZOMIB

Bortezomib, a modified dipeptidyl boronic acid, is one of the first-line treatments in multiple myeloma (MM) patients. Bortezomib inhibits the 20S proteasome complex and acts by disruption of various cell signaling pathways. However, its use is frequently associated with the development of significant neurotoxic effects. Bortezomib-induced PN is, in most cases, painful and when it occurs it can potentially lead to dose modification and severe disability, thereby compromising the outcome of patients with MM (Argyriou 2008).

6.1 Pathogenesis

The pathogenesis of bortezomib-induced PN remains largely unknown and only recent results from experimental studies have contributed to a better understanding. Overall, it seems that DRG represent the anatomical structures which are particularly susceptible to significant cellular changes secondary to bortezomib administration. This is because of the structure of their capillaries that allow free passage of molecules between the circulation and the extracellular fluid (Argyriou 2008; Cavaletti 2007a).

This hypothesis has been supported by a study in animal models demonstrating that bortezomib exerts significant neuronal dysfunction characterized by interference with transcription, nuclear processing and transport, and cytoplasmic translation of mRNAs in DRG neurons (Casafont 2010). Histopathological and neurophysiological findings show that the neuronopathy secondary to bortezomib use mostly evokes a dose-dependent reduction of large and C-fibers with abnormal vesicular inclusion body in unmyelinated axons (Bruna 2010; Meregalli 2010).

Mitochondrial and endoplasmic reticulum damage and also dysregulation of neurotrophins, mainly by either bortezomib-induced activation of the mitochondrial-based apoptotic pathway or inhibition of the transcription of nerve growth factor-mediated neuron survival (through interference with the NF-kB pathway), have also been proposed as potentially playing a significant role in the genesis of bortezomib-induced PN (Landowski 2005; Montagut 2006).

6.2 Clinical and electrophysiological characteristics.

Distal painful sensory PN represents the clinical hallmark of bortezomib-induced PN. Affected patients experience neuropathic pain, distal sensory loss in a stocking-and-glove distribution, hyporeflexia or even areflexia and disturbed proprioception (Badros 2007). Most cases will self-report neuropathic pain as of moderate to severe intensity with a mean VAS score of 8 (Cata 2007).

Quantitative sensory testing shows abnormalities consistent with a painful neuropathy due to dysfunction in all three major fiber types (i.e. Aβ, Aδ, and C) of the sensory nerves (Cata 2007). Common findings from the conventional nerve conduction studies are SNAP changes (Argyriou 2008). Myelin changes with distal slowing of sensory NCV may also be present as part of degeneration of fast-conducting fibers secondary to an initial dysfunctional neuronopathy (Bruna 2010). Motor nerve conduction studies evidencing reduced CMAP can be occasionally observed.

6.3 Incidence, severity and risk factors

Most of the available studies assessing bortezomib-induced PN have employed the NCI-CTC criteria or the FACT/GOG-Ntx subscale (Calhoun 2003; Richardson 2006). Recent studies have also adopted different TNS versions to assess the severity of bortezomib-induced PN (Lanzani 2008; Velasco 2010).

The reported data (Tab. 6) indicate that bortezomib-induced PN (any grade) may occur in about half of the patients receiving treatment with this drug (Gilardini 2008). The cumulative dose of bortezomib has also been consistently associated with an increased incidence of bortezomib-induced PN. However, patients with recurrent disease may develop more severe bortezomib-induced PN than those with newly diagnosed disease. In fact, pooled safety data reveal that up to 75% of patients receiving bortezomib for recurrent disease are at risk of developing NCI-CTC grade 1-2 bortezomib-induced PN as opposed to 33% of patients with newly diagnosed MM. Severe neurotoxicity (grade 3-4) may occur in up to 30% of patients with recurrent MM as opposed to about 20% of patients with newly diagnosed disease (Mohty 2010).

Several risk factors have been proposed to be associated with increased incidence of bortezomib-induced PN, including MM-associated neuropathy, pre-existing neuropathy of other origin, age and comorbidities such as diabetes mellitus (Anderson 2006; Mateos 2006; Richardson 2006; Lanzani 2008). A recently published study enrolling 58 relapsed/refractory MM patients treated with bortezomib highlighted that that absence of neurological monitoring and prior treatment with vincristine were associated with greater risk of bortezomib-induced PN. From this study it also emerged that patients with a baseline total neurology score – clinical version (TNSc) score above 2 are at increased risk for developing more severe bortezomib-induced PN (Velasco 2010).

6.4 Outcome

Symptoms of bortezomib-induced PN usually improve or completely resolve after three to four months following discontinuation of treatment, as evidenced by the long-term data of the phase III APEX trial analyzed to assess the reversibility of bortezomib-associated PN in MM patients. This study demonstrated that 64% of patients with at least NCI-CTC grade 2 bortezomib-induced PN experienced improvement or resolution of symptoms compared to baseline at a median of 110 days (Richardson 2009).

6.5 Options for neuroprotection

Several pharmacological agents have been tested to treat bortezomib-induced PN, including the administration of various opioids, tricyclic antidepressants, anticonvulsants, serotonin-norepinephrine reuptake inhibitors, non-steroidal anti-inflammatory agents, vitamins and nutritional supplements, given either as monotherapy or in combination regimens (Argyriou 2008), but no evidence-based conclusions can be provided regarding their effectiveness.

Since there is no evidence for any effective prophylactic treatment, strict adherence to the dose modification guidelines, as presented in Table 7, is recommended to reduce the risk of severe bortezomib-induced PN (San Miguel 2006).

7. THALIDOMIDE

Thalidomide was first introduced into European markets in the 1950s as a sleep aid and antiemetic drug for pregnant women. It was withdrawn from the market soon after when its teratogenic effects were discovered. It has re-emerged recently as an effective treatment for several dermatological, gastrointestinal, and oncological conditions. In May 2006 the US Food and Drug Administration (FDA) granted approval to thalidomide for use in combination with dexamethasone in newly diagnosed MM patients. PN is now recognized as one of the most significant side effects of this medication.

7.1 Pathogenesis

The exact mechanism of the anti-malignancy action of thalidomide is not known, but it is possible that it acts through angiogenesis inhibition, immunomodulation and cytokine modulation – individually or in combination. Similarly, the pathogenesis of thalidomide-induced PN damage is also unknown. The structure of thalidomide is characterized by a 3-substituted glutarimide ring and a phthalimide ring. Both rings are prone to enzymatic or non-enzymatic hydrolysis (Williams 1964; Lepper 2006). Several in vitro and in vivo studies have been performed to elucidate the structures of the metabolites formed, identify the enzymes responsible for their production and assess interspecies differences in metabolism, which may account for differences in the activity and toxicity of thalidomide. However, no conclusive results have been achieved regarding PN.

7.2 Clinical and electrophysiological characteristics

The most common presentation of thalidomide-induced PN is distal paresthesias and/or dysesthesias with or without sensory loss (Cundari and Cavaletti 2009). Physical examination may be normal or show mild reduction in sensation in distal limbs. Strength is usually preserved, although mild weakness may be present. DTR, particularly ankle jerks, may be depressed or absent. Symptoms can be disabling and often necessitate discontinuation of the drug despite disease control. Onset is usually delayed after initiating thalidomide, mostly depending on the schedule of treatment. On a neurophysiological basis, reduction in SNAP with relative preservation of CMAP and NCV are typical findings (Isoardo 2004), suggesting a sensory axonal neuropathy as the predominant pathological event.

Alternatively, some evidence suggests that thalidomide may also cause a DRG neuronopathy. In these patients the clinical presentation is rather different, with an early involvement of all four limbs. Using MRI, T2 hyperintense, non–mass-like, non-enhancing lesions may be seen in the posterior columns of the spinal cord of these patients, indicating the involvement of the centripetal branch of the primary sensory neuron axons (Isoardo 2004; Giannini 2003).

7.3 Incidence, severity and risk factors

Somewhat conflicting results regarding the relationship of thalidomide dosage and incidence of PN have been reported. Although some studies found a relationship between cumulative dose and occurrence of PN (Steurer 2004; Cavaletti 2004), others failed to do so (Briani 2004; Tosi 2005; Bastuiji-Garin 2002). Alternatively, increased risk of thalidomide-induced PN has been related to daily dose of thalidomide (Bastuji-Garin 2002). Risk factors for PN in thalidomide-treated patients have not been elucidated and most studies failed to find a correlation between age, thalidomide administration and occurrence of CIPN (Briani 2004; Tosi 2005; Bastu Ji-Garin 2002). Preliminary clinical data suggest that the thalidomide analogues lenalidomide and pomalidomide are more potent and have a better toxicity profile, including reduced risk of inducing CIPN.

7.4 Outcome

The issue of the long-term course and final outcome of thalidomide-induced PN has not yet been extensively investigated. However, when specifically studied it appeared that the neuropathic symptoms may frequently improve on discontinuation of thalidomide (Isoardo 2004; Steurer 2004).

 7.5 Options for neuroprotection

No effective pharmacological intervention is available to prevent thalidomide-induced PN and dose reduction or withdrawal is warranted if it occurs.

7.6 Immunomodulatory Drugs (IMiDs)

Newer classes of a-phthalimidoglutarimides are designated as IMiDs and they include lenalidomide and pomalidomide.

In 2005, lenalidomide was FDA-approved for the treatment of myelodysplastic syndrome in patients with deletion 5q cytogenetic abnormality and in 2006 for use in combination with dexamethasone in patients with relapsed MM. Pomalidomide is currently in clinical development for the treatment of haematological malignancies.

An open-label randomized phase II study in 102 heavily pre-treated, relapsed or refractory MM patients evaluated the efficacy and safety of two lenalidomide dosing regimens, 15 mg twice daily or 30 mg once daily on days 1-21 of a 28 day cycle (Richardson 2006).  The incidence of treatment-emergent PN was 23% in the twice-daily arm versus 10% for the once-daily arm, with 3% NCI-CTC grade 3 neuropathy.

The use of lenalidomide plus dexamethasone was also investigated in two phase III studies.  In the lenalidomide/dexamethasone groups the incidence of grade 3-4 PN was 1.7% vs. 1.1% in the placebo/dexamethasone arm in one study and 6.9% in both arms in the second ones (Dimopoulos 2007; Weber 2007).

In a single-centre study, 78 patients with recurrent MM who had received bortezomib were retrospectively reviewed for the incidence, severity and risk factors for PN (Badros 2007).  In 6 out of 9 patients who experienced PN with bortezomib, two weeks after switching treatment to lenalidomide, a significant unexpected improvement in PN symptoms was observed, including 3 patients who were able to stop analgesics, although a clear causal relationship between lenalidomide treatment and PN course could not be established.

8. CONCLUSION

From this review it appears that CIPN still represents an unmet clinical need in the modern approach to cancer patients. Its importance has become even clearer in view of the better results obtained with more effective schedules of treatment allowing longer survival and cancer cure, thus making the occurrence of a disabling and long-lasting peripheral neurotoxicity even less tolerable.

A deeper knowledge should, therefore, be achieved and this goal may be accomplished only through an extensive collaboration between all the healthcare professionals involved in cancer patient management, focused on the common aim to optimize cancer treatment while minimizing potentially disabling side effects such as CIPN.

Table 6. Incidence of ixabepilone-induced neuropathy. Selection criteria: prospective studies with neurologic assessment. NCI-CTC: National Cancer Institute-Common Toxicity Criteria. For a description of the scale see Cavaletti et al. (2010)

Author

(year)

Patients

(n)

Schedule

Neurological

Assessment

Neuropathy

Incidence

Eng

(2004)

25

40-50 mg/m 2, 1-hour infusion/3-hour infusion, every 3 weeks

NCI-CTC

Sensory Grade 1: 28%

Grade 2: 4%

Grade 3: 20%

Denduluri (2007a)

12

8-10 mg/m2/day 1-hour infusion

for 3 days, every 3 weeks

NCI-CTC

Sensory Grade 1: 16%

Grade 2: 25%

Denduluri (2007b)

23

6 mg/m2/day 1-hour infusion for 5 days, every 3 weeks

NCI-CTC

Sensory Grade 1: 39%

Grade 2: 13%

Motor   Grade 2:  4%

Grade 3:  4%

Autonomic    Grade 2: 4%

Dreicer (2007)

45

40 mg/m2, 3-hour infusion,

every 3 weeks

NCI-CTC

Sensory Grade 3-4: 6%

(no data about other grades)

Roche (2007)

65

40-50 mg/m2, 1-hour infusion/3hour infusion , every 3 weeks

NCI-CTC

Sensory  Grade 1-2: 51%

Grade 3: 20%

Motor Grade 2: 1%

Grade 3 : 5%

Perez

(2007)

126

40 mg/m2, 3-hour infusion,

every 3 weeks

NCI-CTC

Sensory Grade 1: 29%

Grade 2: 30%

Grade 3: 13%

Motor Grade 1: 1%

Grade 2: 7%

Grade 3: 1%

Vansteenkiste (2007)

146

50-40-32 mg/m2, 3-hour infusion,

every 3 weeks

or

6 mg/m2/day 1-hour infusion for 5 days, every 3 weeks

NCI-CTC

Sensory Grade 3: 5%

Grade 4: 1%

Any grade sensory neuropathy: 40%

Thomas (2007a)

49

40-50 mg/m2, 1-hour infusion/3-hour infusion, every 3 weeks

NCI-CTC

Sensory Grade 1: 18%

Grade 2: 33%

Grade 3: 12%

Motor Grade 2: 2%

Autonomic Grade 2: 2%

Thomas (2007b)

369

40 mg/m2, 3-hour infusion,

every 3 weeks

NCI-CTC

Sensory Grade 1: 17%

Grade 2: 27%

Grade 3: 20%

Grade 4: 1%

Motor Grade 1: 4%

Grade 2: 7%

Grade 3: 5%

Burtness

(2008)

85

6 mg/m2/day 1-hour infusion for 5 days, every 3 weeks

or

20 mg/m2 /day 1-hour infusion on days 1, 8 and 15, every 4 weeks

NCI-CTC

Sensory Grade 3: 6%

Sensory Grade 3: 20%

Motor Grade 3: 11%

Bunnell

(2008)

62

40 mg/m2, 3-hour infusion,

every 3 weeks

NCI-CTC

Sensory Grade 1: 24%

Grade 2: 31%

Grade 3: 19%

Motor Grade 3: 2%

O’Connor

(2008)

28

25 mg/m2, 1-hour infusion weekly for 3 weeks,

every 4 weeks

NCI-CTC

Sensory Grade 1: 21%

Grade 2:   0%

Grade 3:   7%

Baselga

(2009)

161

40 mg/m2, 3-hour infusion,

every 3 weeks

NCI-CTC

Overall neuropathy

Grade 1:26%

Grade 2:14%

Sensory Grade 3: 1%

Ott

(2010)

24

20 mg/m2/day 3-hour infusion on days 1, 8 and 15, every 4 weeks

NCI-CTC

Overall neuropathy

Grade 1:29%

Grade 2:8%

Grade 3: 4%

Huang

(2010)

87

6 mg/m2/day 1-hour infusion for 5 days, every 3 weeks

NCI-CTC

Sensory Grade 1: 32%

Grade 2:   40%

Grade 3:   16%

Table 7: Recommended* schedule modifications for bortezomib-related neuropathy.
*Based on schedule modifications in Phase II and III multiple myeloma studies and post-marketing experience.

Severity of neuropathy

Schedule modification

Grade 1 (paresthesias, weakness and/or loss of reflexes) with no pain or loss of function

No action

Grade 1 with pain or Grade 2 (interfering with function but not with activities of daily living)

Reduce to 1.0 mg/m2

 

Grade 2 with pain or Grade 3 (interfering with activities of daily living)

Withhold bortezomib treatment until symptoms of toxicity have resolved. When toxicity resolves re-initiate bortezomib treatment and reduce dose to 0.7 mg/m2 and change treatment schedule to once per week.

Grade 4 (sensory neuropathy which is disabling or motor neuropathy that is life threatening or leads to paralysis) and/or severe autonomic neuropathy

Discontinue bortezomib

from: http://www.ema.europa.eu/humandocs/PDFs/EPAR/velcade/emea-combined-h539en.pdf

INDEX