Clinical pattern and associations of oxaliplatin acute neurotoxicity

A prospective study in 170 patients with colorectal cancer

Authors


Abstract

BACKGROUND:

The objective of the current prospective, multicenter, international study was to trace the incidence and severity of acute oxaliplatin-induced peripheral neuropathy (OXLIPN) and to determine its clinical pattern. The authors also specifically tested whether patients who had more symptoms of acute OXLIPN eventually would develop a more severe chronic, cumulative form of OXLIPN.

METHODS:

One hundred seventy patients (mean ± standard deviation age, 63.7 ± 8.7 years) who were scheduled to receive either combined leucovorin, 5-fluoruracil, and oxaliplatin (FOLFOX) or combined capecitabine and oxaliplatin (XELOX) for metastatic colorectal cancer were monitored prospectively at baseline and were followed in 4 European sites. The incidence of hyperexcitability symptoms secondary to acute OXLIPN was assessed by using a descriptive questionnaire (yes/no question) at each clinical evaluation. Motor and neurosensory criteria according to version 3 of the National Cancer Institute's Common Toxicity Criteria were applied to clinically grade the severity of OXLIPN.

RESULTS:

Acute OXLIPN was present in 146 of 170 patients (85.9%). The vast majority of these patients manifested cold-induced perioral (95.2%) or pharyngolaryngeal (91.8%) dysesthesias. Severe acute OXLIPN that required prolongation of oxaliplatin infusion from 2 hours to 4 to 6 hours occurred in 32 of 146 patients (21.9%). The increased number of acute OXLIPN symptoms was correlated significantly (Spearman rho correlation coefficient [r]) with both the development (r = 0.602; P < .001) and the degree of the chronic, cumulative form (r = 0.702; P < .001).

CONCLUSIONS:

The current results indicated that the vast majority of patients with colorectal cancer who receive oxaliplatin-based chemotherapy will manifest symptoms of a transient acute syndrome soon after oxaliplatin administration. Patients who have a more complex combination of acute phenomena related to axonal hyperexcitability are those who eventually develop more severe OXLIPN. Therefore, it may be advisable to test agents against acute OXLIPN to verify their effects on the chronic form. Cancer 2013. © 2012 American Cancer Society.

INTRODUCTION

Colorectal cancer (CRC) represents the third most common type of cancer and the second leading cause of cancer-related deaths in the Western world.1 Therapeutic management of CRC is challenging and depends on the disease stage at diagnosis. Current knowledge indicates that oxaliplatin (OXL)-based combination regimens, in the form of either combined leucovorin, 5-fluoruracil, and OXL (FOLFOX) or combined capecitabine and OXL (XELOX), have demonstrated substantial cytotoxic synergy in the adjuvant setting and in the advanced or metastatic setting and, thus, are associated with prolonged disease progression-free and overall survival.2

However, peripheral neuropathy is currently recognized as 1 of the major nonhematologic toxicities of OXL with a possible negative impact on therapeutic outcomes, thus compromising survival and also negatively influencing patients' quality of life.3 OXL induces 2 clinically distinct forms of peripheral neuropathy; the neuromyotonia-like, acute, transient syndrome characterized by cold-induced distal or perioral paresthesias and pharyngolaryngeal dysesthesias and the chronic form that, in most cases, is a pure sensory, axonal neuropathy with a stocking-and-glove distribution.4

In clinical practice, to date, less attention is usually drawn to the acute form of OXL-induced peripheral neuropathy (OXLIPN) than to the chronic form. This is because the acute hyperexcitability syndrome is transient: It usually is reversible within hours or days and, in the vast majority of patients, does not require discontinuation of treatment or dose modification.5 By contrast, the chronic, cumulative form of OXLIPN may be detrimental to patients and may also persist for a long time, even after patients discontinue OXL.6

In light of recent evidence suggesting that the degree of acute nerve dysfunction may relate to the development of chronic neurotoxicity,7 the objective of our current study was to prospectively trace the incidence and severity of acute OXLIPN and to analyze its clinical pattern. Moreover, we specifically aimed at testing whether the degree of acute OXLIPN is related in some way to the development and severity of cumulative OXLIPN at the end of chemotherapy. The detailed monitoring of chronic OXLIPN was not included among the objectives of the current study.

MATERIALS AND METHODS

Study Design

This was a prospective, multicenter, international, collaborative study among experienced neurologists and oncologists at 4 sites in 3 European countries. The study protocol was approved by the corresponding institutional review boards of each site, and written informed consent was obtained from all patients before study entry.

Patient Selection

The study sample was recruited during an 18-month period. Patients who were included had a histologically confirmed diagnosis of metastatic CRC and were scheduled to receive OXL-based chemotherapy with either FOLFOX or XELOX. To be eligible for enrollment, patients had to be aged >18 years and had to have satisfactory liver and renal function, to have a life expectancy ≥9 months, to have a Karnofsky performance score ≥70, and to be able to fully understand the study information provided by the investigators. Patients also were required to be neurotoxic chemotherapy-naive and to be scheduled to receive adjuvant or first-line treatment. Patients with history of peripheral neuropathy and those with concomitant diseases (eg, diabetes, renal insufficiency, alcohol abuse [>5 IU daily]) that would interfere or complicate the clinical assessments were excluded.

Two hundred twenty-four patients initially were screened, and 170 eventually were included in the study. During screening, 54 patients were excluded for various reasons, including evidence of pre-existing neuropathy (n = 23), a change in treatment plan requiring the receipt of a chemotherapy regimen other than FOLFOX or XELOX (n = 12), the presence of conditions that would have complicated accurate assessment (n = 11), refusal (n = 6), and other reasons (n = 2). All included patients were evaluated clinically at baseline (Visit 1), which occurred at the time of the screening visit or up to 2 days after administration of the first chemotherapy course.

Outcome Measures

The incidence and severity of cumulative OXLIPN was graded by using version 3 of the National Cancer Institute's Common Toxicity Criteria (NCI-CTC v3) for sensory and motor neuropathy. In fact, this scale is considered the standard method for assessing chemotherapy-induced peripheral neuropathy (CIPN), and it grades the severity of CIPN from 0 (no toxicity) to 4 (most severe grade of CIPN).8

The frequency of the 11 most common hyperexcitability symptoms associated with acute OXLIPN,3 including cold-induced perioral paresthesias, cold-induced pharyngolaryngeal dysesthesia, shortness of breath, difficulty swallowing, laryngospasm, muscle cramps, jaw stiffness, fasciculations, voice changes, ptosis, and ocular changes, were assessed by a descriptive questionnaire (yes/no response format). The severity of acute OXLIPN was scored based on the number of symptoms reported by patients at each clinical assessment. An increase in the number of acute symptoms was considered suggestive of an increase in the severity of acute OXLIPN. In addition, severe acute neurotoxicity that required dose reduction and potentially subsequent treatment discontinuation was considered only in patients who had acute neurotoxicity lasting >2 weeks, even without motor symptoms or neurologic signs.5 Patients were asked about acute and chronic symptoms in the same encounter and by the same researcher.

Finally, conventional motor and sensory nerve-conduction studies at the ulnar, radial, peroneal, and sural nerves were carried out unilaterally on the nondominant side using widely accepted criteria for the identification of abnormalities.9 All of the aforementioned clinical and neurophysiologic evaluations were performed at baseline and were repeated after 6 courses (OXL planned dose, 510 mg/m2) and 12 courses (OXL planned dose, 1020 mg/m2) of the FOLFOX regimens and after 4 courses (OXL dose, 520 mg/m2) and 8 courses (OXL dose, 1040 mg/m2) of the XELOX regimen.

Chemotherapy Regimen and Dose Modification

The formal FOLFOX-4 regimen10 was administered to the majority of our patients (n = 88; 51.8%), consisting of leucovorin (LV) 200 mg/m2 as a 2-hour infusion on days 1 and 2 followed by 5-fluorouracil 400 mg/m2 on days 1 and 2 as an intravenously bolus. Then, 5-fluorouracil 600 mg/m2 was administered as a 22-hour infusion on days 1 and 2. OXL 85 mg/m2 was given as a 2-hour intravenously infusion concurrent with leucovorin on day 1. Therapy was repeated every 2 weeks. Routine antiemetic prophylaxis also was prescribed. Nine patients (5.3%) received FOLFOX-6, and 73 patients (42.9%) received XELOX, consisting of intravenous OXL 130 mg/m2 on day 1 followed by oral capecitabine 1000 mg/m2 twice daily (from the evening of day 1 to the morning of day 15) every 3 weeks.11

Dose adjustments of the study drugs or treatment delays were calculated according to toxicity grade. Dose modifications were determined according to the greatest degree of toxicity. Toxicities were graded using NCI-CTC v3. The OXL dose was reduced by 30% for persistent or temporary (at least 14 days) painful paresthesia, dysesthesia, or functional impairment. If, despite the 30% dose reduction of OXL, the grade 3 neurotoxicity persisted, then chemotherapy was omitted in subsequent cycles. No prophylactic or symptomatic treatment was given for neurotoxicity during the administration of chemotherapy. Patients with any grade 4 toxicity, except gastrointestinal or hematologic toxicity, had to be withdrawn from the study.

Statistical Analysis

Descriptive statistics were generated for all variables. The correlation between the severity of acute OXLIPN and the incidence and degree of chronic OXLIPN was examined with the Spearman rho correlation coefficient (r). All tests were 2-sided, and significance was set at P < .05. Statistical analyses were performed using the SPSS for Windows (release 17.0; SPSS Inc., Chicago, Ill).

RESULTS

The demographics and clinical characteristics of the patients are presented in Table 1.

Table 1. Baseline and Clinical Patient Characteristics (n = 170)
VariableNo. of Patients (%)
  1. Abbreviations: FOLFOX, leucovorin, 5-fluoruracil, and oxaliplatin; OXL, oxaliplatin; SD, standard deviation; XELOX, capecitabine and oxaliplatin.

Sex 
 Women67 (60.6)
 Men103 (39.4)
Age 
 Mean±SD [range], y63.7 ± 8.7 [38-80]
Height 
 Mean±SD [range], cm164.6 ± 8.4 [140-183]
Type of chemotherapy 
 FOLFOX97 (57.1)
 XELOX73 (42.9)
No. of single OXL doses per course: Median [range], mg 
 FOLFOX148 [80-190]
 XELOX220 [138-260]
No. of cumulative OXL doses: Median [range] 
 FOLFOX1653 [900-2280]
 XELOX1634 [848-2080]

Acute Oxaliplatin-Induced Peripheral Neuropathy Incidence

Acute OXLIPN was present in 146 of 170 patients (85.9%). The vast majority of patients manifested cold-induced perioral (95.2%) or pharyngolaryngeal (91.8%) dysesthesias. Table 2 summarizes the incidence of acute neurotoxic effects secondary to the receipt of OXL.

Table 2. The Incidence of Acute Neurotoxic Symptoms Secondary to Receiving Oxaliplatin in Patients Who Manifested Any Degree of Acute Oxaliplatin-Induced Peripheral Neuropathy (n = 146 of 170 Patients; 85.9%)
SymptomsNo. of Patients (%)
Cold-induced perioral paresthesias139 (95.2)
Cold-induced pharyngolaryngeal dysesthesia134 (91.8)
Shortness of breath58 (39.7)
Difficulty swallowing44 (30.1)
Laryngospasm0 (0)
Muscle cramps50 (34.2)
Jaw stiffness50 (34.2)
Visible fasciculations43 (29.5)
Voice changes9 (6.2)
Ptosis0 (0)
Ocular changes1 (0.7)

Acute Oxaliplatin-Induced Peripheral Neuropathy Severity

The median number of symptoms that patients with acute neurotoxicity reported was 4 symptoms (range, 1-7 symptoms). According to the total number of OXLIPN symptoms, the severity of OXLIPN at the final follow-up was grade 1 in 49 of 146 patients (33.6%), grade 2 in 41 of 146 patients (28.1%), and grade 3 in 56 of 146 patients (38.4%). The majority of patients (90 of 146; 61.7%) started developing acute OXLIPN between Courses 0 and 6 of FOLFOX (median cumulative OXL dose at Course 6, 888 mg) and between Courses 0 and 4 of XELOX (median cumulative dose at Course 4, 880 mg), whereas fewer patients (n = 56; 38.4%) developed acute neurotoxic effects between Courses 6 and 12 of FOLFOX (median cumulative final OXL dose at Course 12, 1653 mg) and between Courses 4 and 8 of XELOX (median cumulative final OXL dose at Course 8, 1634 mg).

In total, 146 of 170 patients (85.9%) developed acute OXLIPN, and 123 of 170 patients (72.4%) also eventually manifested various degrees of chronic, cumulative OXLIPN, as measured according to NCI-CTC v3 neurosensory criteria. Twenty-three patients with acute OXLIPN did not developed cumulative neurotoxicity as assessed at the end of treatment. The clinical and demographic data in these patients did not differ from those in the patients who developed both acute and chronic OXLIPN. This subgroup of patients included 11 men and 12 women with mean age (±standard deviation) of 64.8 ± 7.1 years (range, 51-76 years). Ten of these patients received FOLFOX, and 13 received XELOX. OXL was given as a 2-hours infusion at a median (±standard deviation) dose per single course of 171 ± 41 mg (range, 80-250 mg) and at a median cumulative dose of 1491 ± 330 mg (range, 940-2000 mg). The only characteristic of this subgroup was that 20 of 23 of patients (87%) developed mild, transient, grade 1 acute neuropathy, whereas the other 3 patients manifested grade 2 acute neuropathy. The median number of acute symptoms they reported was 2 (range, 1-3 symptoms) compared with a median of 4 symptoms (range, 1-7 symptoms) reported for all patients who manifested any grade of acute and chronic neurotoxicity. Twenty-four patients did not develop either acute or chronic OXLIPN. The characteristics of this subgroup were similar compared with those of the patients who manifested either the acute form alone or both acute and chronic OXLIPN.

Severe acute OXLIPN that required prolongation of OXL infusion from 2 hours to 4 to 6 hours was evident in 32 of 146 patients (21.9%). Prolongation of OXL infusion occurred after 2 courses in 7 of 32 patients (4.8%), after 3 courses in 2 of 32 patients (1.4%), after 4 courses in 8 of 32 patients (5.5%), after 5 courses in 6 of 32 patients (4.8%), and after 8 courses in 1 of 32 patient (0.7%); whereas another patient who received FOLFOX required prolongation of the OXL infusion after 8 courses of chemotherapy. However, none of the enrolled patients required dose reductions or discontinued treatment because of persistent grade 3 acute neurotoxicity.

Associations

On the basis of data at last follow-up, the worst severity of acute neurotoxicity was related to the cumulative OXL dose (Spearman rho [r] = 0.171; P = .026). Associations between the severity of acute OXLIPN and demographic data, such as age and sex, failed to reach significance. The increased number of acute OXLIPN symptoms, suggesting an increased severity of acute OXLIPN, was correlated significantly with both the development (r = 0.602; P < .001) and degree (r = 0.702; P < .001) of the chronic form (Fig. 1a,b).

Figure 1.

The y-axis in these box plots indicates to the number of symptoms reported by patients. (a) The x-axis in refers to the incidence of chronic oxaliplatin-induced peripheral neuropathy (OXLIPN). (b) The x-axis in b refers to the severity of OXLIPN, as graded according to version 3 of the National Cancer Institute's Common Toxicity Criteria (NCI-CTC v3). The box plot in a indicates that the increased number of acute OXLIPN symptoms is correlated significantly with the development of the chronic form, and the box plot in b indicates that the increased number of acute OXLIPN symptoms also is correlated significantly with the severity of the chronic form. Nbacute indicates the number of acute symptoms reported by patients; Overchrtox, the overall incidence of chronic OXLIPN; Oversevncictc, the severity of OXLIPN graded according to NCI-CTC v3 criteria.

However, based on data from the intermediate follow-up, the cumulative OXL dose was associated only with the development of chronic, cumulative neurotoxicity (r = 0.216; P = .005), but not with the acute syndrome (r = 0.097; P = .210). At the same time point, significant correlations emerged between the overall severity of acute OXLIPN with both the development (r = 0.513; P < .001) and the degree (r = 0.482; P < .001) of chronic, cumulative neurotoxicity at the end of chemotherapy.

Neurophysiologic monitoring revealed a significant longitudinal deterioration (from baseline to subsequent scores; ie, intermediate and final follow-up) in the amplitude of sensory action potentials of all 3 sensory nerves tested. The same comparison between motor-conduction variables failed to reach significance in both treatment groups. These abnormalities were in keeping with an axonal, sensory polyneuropathy and were linked to the chronic OXLIPN, because they were considered unrelated to the hyperexcitability acute syndrome secondary to OXL administration.

DISCUSSION

OXL, a third-generation organoplatinum compound, is capable of inducing 2 clinically distinct forms of peripheral neuropathy—acute and chronic—through mechanisms that have yet to be clearly defined.3 It is believed that the acute form is caused by a dysfunction of nodal axonal voltage-gated sodium channels, probably the calcium-dependent channels. In particular, the rapid chelation of calcium by OXL-induced oxalate and the subsequent alteration in voltage-gated sodium channel kinetics (reduction of the overall sodium current) may be the main response that generates the acute syndrome.12, 13

Conversely, the accumulation of OXL in dorsal root ganglia (DRG) cells may induce chronic neurotoxicity by damaging the DRG sensory neurons with axoplasmatic transport changes secondary to cell body damage.14 Mitochondrial dysfunction and oxidative stress also may contribute significantly to OXL-induced neuronal apoptosis.13

To date, research has been focused mainly on chronic OXLIPN, because it is often dose-limiting and long lasting.15 Therefore, the chronic form is considered to have more clinical relevance than the acute syndrome, which usually is reversible and, in most cases, apart from prolonging OXL infusion, does not require dose modification or treatment discontinuation.16 However, because it has been suggested that acute neurotoxicity is related to the onset and severity of chronic OXLPN,7 our objective was to analyze its frequency and features in a large, homogeneous cohort of patients.

In the current setting, the reported rate of patients who manifested acute (85.9%) and chronic (72.4%) OXLIPN after either FOLFOX or XELOX was quite similar to that previously reported.17, 18 The symptoms of acute toxicity are different from those of chronic toxicity. In addition, neurophysiology can distinguish the nature of symptoms (acute or chronic syndrome); thus, based on clinical examinations and neurophysiology, we were able to clearly differentiate symptoms of persistent acute neuropathy and symptoms that suggested the onset of chronic neuropathy.

In line with previous publications,17, 19 the majority of our patients experienced acute hyperexcitability symptoms that resembled disorders of voltage-gated ion channels.20 Most of our patients manifested cold-induced perioral (95.2%) or pharyngolaryngeal dysesthesias (91.8%), as outlined in Table 2. Results from an experimental study have suggested that cooling in the presence of OXL strongly mediates bursts of action potentials in myelinated A fibers of human and mouse peripheral axons and large-diameter DRG neurons, thus resulting in cold-induced hyperexcitability phenomena.21 In addition, there is evidence that cold exposure is able to further affect sodium channel kinetics, thus predisposing to ectopic activity.20

Fewer patients presented with other uncommon symptoms,7, 10 such as jaw spasm, difficulty swallowing, slurred speech, ocular changes, and calf cramps with walking that tended to persist for days. The occurrence of such symptoms may result from an acute, abnormal, OXL-induced hyperexcitability state of peripheral sensory and motor nerves fibers, including those of the cranial nerves.13, 22

Neurophysiologic monitoring revealed progressive deterioration in the amplitude of sensory action potentials, and this finding was linked to chronic neuropathy. Generally, it is believed that sensory nerve amplitude reduction is unlikely to be observed in acute OXL neuropathy. In support of this view are results from a previous neurophysiologic study assessing nerve-conduction studies and findings on needle electromyography in patients with metastatic CRC before and during treatment with OXL. That study revealed that repetitive compound muscle action potentials and neuromyotonic discharges were observed in the first 24 to 48 hours after OXL infusion but resolved by 3 weeks; whereas a decline in the amplitude of sensory action potentials was observed only after 8 to 9 treatment cycles.19

Most patients manifested acute neurotoxic effects without requiring prolongation of OXL infusion, whereas dose reduction or treatment discontinuation was not recorded. However, in the current study, a significant proportion of our patients (22%) required the infusion of OXL over a period >2 hours because of the severity of their acute OXLIPN. This rate is similar to the rate observed in some previous studies,17 although it is lower than that observed in other studies.23

The main finding of our study is that patients who had more severe acute OXLIPN, as determined by the number of symptoms reported by the patients, eventually also develop a more severe, chronic OXLIPN, suggesting that acute OXLIPN may predispose patients to the chronic, cumulative form. This result also is supported by the finding that 20 of 23 patients with acute OXLIPN who did not eventually developed chronic neurotoxicity experienced mild, transient, grade 1 hyperexcitability phenomena. These findings seem to be at odds with the different pathogenic mechanism of acute and chronic OXL neurotoxicity, but they may simply express a different individual, possibly genetic susceptibility to both acute and/or chronic OXL neurotoxicity. In addition, it has been demonstrated that patients who had alterations of axonal excitability in early OXL treatment were more prone to developing dose-limiting neurotoxicity, thus demonstrating that the degree of acute nerve dysfunction may relate to the development of chronic neurotoxicity.7

In support of our data, we have demonstrated that the cumulative dose of OXL also was related to the severity of acute hyperexcitability syndrome. A possible pathogenic hypothesis is that, in addition to the decreased cellular metabolism and axoplasmatic transport in DRG cells, the prolonged activation of voltage-gated sodium channels may induce cellular stress, further affecting the sensory nerve cells and contributing to the development of chronic OXLIPN.3

However, although this association may reflect the pathogenic relevance of acute OXLIPN, it may simply reflect the expression of a generic susceptibility to peripheral nervous system disturbance in a given patient. Therefore, further studies are needed to test whether neurophysiologic axonal changes (signs of chronic peripheral nerve damage) occur earlier in patients with acute OXLIPN versus those without, thus supporting the possibility that acute OXLIPN may be a predictor of chronic OXLIPN.

On clinical grounds, our results suggest that oncologists mostly should be alert and aware and should use prevention against the acute form of OXLIPN, because the acute hyperexcitability syndrome may be a strong mediator or predictor of OXL-induced chronic peripheral nerve damage. In the current setting, we did not adhere to NCI-CTC v3 criteria for adverse events to grade the severity of acute OXLIPN, because the severity of acute hyperexcitability symptoms is difficult to measure, and cold-related symptoms are not addressed in common neurotoxicity scales. Moreover, there is a lack of validated, specifically focused scales on this issue. Therefore, we applied scores from the scale, based on the sum of symptoms, reported by patients at each clinical assessment. We do acknowledge that the use of a nonvalidated measure to grade the severity of acute OXLIPN represents a limitation of our study design. Nevertheless, we still believe it was the most appropriate approach to assessing acute OXLIPN, because, to our knowledge, there is no validated instrument that could be substituted.

This limitation aside, our study was a multicenter, international study that reflected everyday clinical practice, and the results easily may be generalized and extended to daily clinical practice. Also in support of this view is our large sample size, probably the largest in the literature to be exclusively focused on acute OXLIPN in patients with CRC, which was prospectively studied at 4 European sites by experienced neurologists and oncologists on the topic of chemotherapy-induced neuropathies and supportive care in cancer.

To date, the ideal neuroprotective agent against OXLIPN is lacking. Oxalate chelators, such as calcium-magnesium infusion, may be effective in reducing the incidence of acute OXLIPN, but it is still not clearly proven that they do not interfere with OXL antitumor efficacy. Alternatively, drugs capable of blocking the voltage-gated ion channels deserve to be tested for their ability to significantly reduce the occurrence of acute and (eventually) chronic OXLIPN. In particular, certain anticonvulsants, such as oxcarbazepine and carbamazepine, because of their mode of action, may be able to target the central and peripheral sensitization mechanisms involved in neuropathic pain syndromes, thereby reducing the hyperexcitability of damaged peripheral nerves and, as such, interfering with the underlying mechanism involved in the genesis of OXLIPN.24-26

To summarize, the current study demonstrates that the vast majority of patients with CRC who receive OXL-based chemotherapy will manifest a transient, acute hyperexcitability syndrome that may contribute to the development of chronic peripheral neuropathy. Appropriate pharmacologic approaches to prevent the OXL-induced acute hyperexcitability syndrome, thus, also may have a positive effect on the chronic form of OXL-induced peripheral nerve damage.

FUNDING SOURCES

Spanish participation in this study was supported by grant PI070493 from Instituto de Salud Carlos III (ISCIII).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary