Taxanes are among the most effective agents for use against breast cancer. Their inclusion into adjuvant regimens translate into significant benefits for disease-free and overall survival.1,2 For metastatic disease, first- or second-line taxanes also improve progression-free and overall survival.3 Although myelosuppression is the major toxicity for docetaxel,4 paclitaxel is more likely to be associated with dose-limiting peripheral neuropathy.3 Paclitaxel-induced peripheral neuropathy (PIPN) usually begins with paresthesia and numbness.5 The neurotoxicity is cumulative and may result in functional impairment. Mild or moderate sensory symptoms usually improve after paclitaxel discontinuation, but severe neuropathy may be irreversible in some patients.6,7 PIPN affects quality of life and is a potential source of functional impairment in long-term breast cancer survivors.8 PIPN is also an escalating issue because of the recent adoption of weekly paclitaxel (wP), which is associated with a higher incidence of significant peripheral neuropathy.9
It is well-known that the grade of neuropathy depends on cumulative dose, duration of infusion, age, and treatment schedule.6,7 The contribution of other factors to PIPN, such as diabetes mellitus (DM) or alcohol,10 is not yet completely understood.11,12 DM is a major health problem, particularly in elderly women,13 and several population-based studies have found that between 7% and 16% of patients with breast cancer also have DM.14,15 Because DM is itself a cause of peripheral neuropathy, PIPN might be more prevalent and severe in patients with breast cancer and DM. However, the available evidence is limited for considering DM a risk factor for PIPN. Several case reports and small uncontrolled series16 suggested a higher grade of paclitaxel neurotoxicity in women with DM.10,17,18 In contrast, 2 recent publications have not consistently found an increased risk or a worse outcome of PIPN in patients with breast cancer and DM,11,12 although, as far as we know, no studies specifically addressing this issue have been published. Because most studies only include patients treated with paclitaxel 3 times weekly, the lack of evidence is even larger for wP, the currently preferred schedule for breast cancer. Only a recent analysis of the wP arm of the E1199 trial showed an association between hyperglycemia and significant PIPN, but no information regarding DM diagnosis or neuropathy evolution was available.19
The goal of this study was to evaluate the impact of preexisting DM in the development and evolution of weekly PIPN, and to test the hypothesis that DM would associate with more severe and prolonged neuropathy.
Methods
We performed a retrospective, unmatched, case-control study including patients with and without DM who had breast cancer treated with wP at University Hospital Morales Meseguer.
Patient Eligibility and Case-Control Study
A search in the hospital pharmacy database was performed to find all patients with breast cancer treated with wP between March 2007 and December 2012. For inclusion in the study, the following criteria had to be met: age older than 18 years, pathologically confirmed breast cancer, treatment with wP (80–100 mg/m2/wk, 1-hour infusion) as the only chemotherapy, more than 6 months of follow-up after the last cycle of paclitaxel, and no previous taxanes or cisplatin. Concomitant treatment with antibodies (trastuzumab or bevacizumab) was allowed. Patients with previous neuropathy were excluded. Preexisting DM was diagnosed according to American Diabetes Association (ADA) criteria (Table 1).20 Controls were randomly selected at a rate of 1:2. Informed consent was obtained from patients, and the study was approved by the hospital ethics committee.
Data Collection and Neurotoxicity Assessment
The following information was obtained from the hospital database and patient medical records: age, performance status (ECOG scale), history of DM and previous neuropathy, tumor data, chemotherapy regimen, initial dose level of paclitaxel, total dose of paclitaxel, granulocyte colony-stimulating factor (G-CSF) administration, date of first documented PIPN, maximum grade of PIPN, delayed or reduced doses of chemotherapy, date of disappearance of PIPN, and grade of residual PIPN at the last follow-up. Grade of PIPN was classified according to NCI-CTC 2011
Summary of American Diabetes Association Criteria for the Diagnosis of Diabetes
Statistical Analysis
Chi-square and Fisher exact tests were used for comparison of categorical variables, and the t test was used for continuous variables. Multivariate logistic regression analysis was performed to determine the contribution of potential risk factors to the appearance of PIPN and significant PIPN. Follow-up time was measured from the date of the first cycle of wP. Analysis of time to onset of PIPN (time from the first paclitaxel dose to the first documentation of peripheral neuropathy) and of duration of PIPN (time from appearance of PIPN to its complete resolution) was performed using the Kaplan-Meier method; log-rank test was used for comparison between groups. Multivariate analysis of PIPN evolution was performed with Cox proportional hazard regression models. P level for significance was less than .05, and all tests were 2-sided. SPSS software (SPSS Inc., Version 17.0; Chicago, IL) was used for statistical analysis.
Results
Patient Characteristics and Treatment
Between March 2007 and December 2012, a total of 296 consecutive patients with breast cancer were treated with single-agent wP (80–100 mg/m2/week, 1-hour infusion) and met eligibility criteria for study inclusion. Among these, 43 patients (33.3% of the final sample) had preexisting DM according to ADA criteria (Table 1)20; another 86 patients (66.7%) without DM were randomly selected as controls. In the DM group, 9 patients (20.9%) were diet-controlled, 25 (58.2%) were on oral hypoglycemic agents, and 9 (20.9%) were undergoing on insulin treatment. Depending on the chemotherapy protocol, 2 dose levels of wP were used: 80 and 100 mg/m2/wk, although no significant differences were seen in total cumulative dose (P=.92) of paclitaxel between both groups.
Baseline characteristics of the study population are listed in Table 2. No significant differences between patients with and without DM were found for age or other clinical characteristics, except for a trend toward worse performance status in the DM group (P=.07). Dose level, median total dose, and dose intensity of wP were also well balanced between groups. Finally, neither administration of G-CSF (P=.49) nor the use of antibodies (P=.61) differed between patients with and without DM.
Incidence and Risk Factors for PIPN
PIPN developed in 59.7% of patients after a median number of 6 doses of chemotherapy (Table 3). More patients with DM developed PIPN of any grade
Patient Characteristics
We evaluated the risk factors for PIPN (any grade) and significant PIPN. For the whole group, only DM was associated with PIPN of any grade (P=.018). No association was found for age older than 65 years (P=.13), ECOG performance status (P=.17), initial dose level (80 vs 100 mg/m2; P=.42), total dose (mg/m2; P=.25), dose intensity (mg/m2/wk; P=.43), number of cycles (P=.76), G-CSF administration (P=.48), and concurrent antibodies (P=.43). Among patients with DM, no differences in the incidence of wP-related PIPN were found for the type of treatment (insulin vs oral agents vs diet; P=.45). In a logistic regression multivariate model including age (>65 years), DM, and G-CSF administration, DM was the only independent predictor for PIPN (odds ratio [OR], 2.54; 95% CI, 1.12–5.76; P=.025).
We next analyzed predictive factors for significant neuropathy. In a multivariate model for development of grade 2 to 3 PIPN, DM (OR, 3.34; 95%
PIPN in Patients With Breast Cancer With and Without DM
Time Course of PIPN
The median time to onset of PIPN in women with and without DM was similar (4 vs 5 weeks; log-rank test, P=.871; Figure 1A). However, after a median follow-up of 23.5 months, 81.8% of patients with DM showed persistent PIPN compared with 40.9% in the control group (P<.001). Kaplan-Meier analysis of persistence of neuropathy showed a significant delay in resolution of PIPN in patients with DM when compared with those without DM (log-rank test; P=.001; Figure 1B). Among patients with a follow-up of more than 12 months from the first cycle of paclitaxel, persistent PIPN was observed in 68.0% of women with DM and 33.3% of those without (P=.002). The difference was still evident after 24 months of follow-up, with 68.7% of women with DM (vs 29.2% without) showing PIPN (P=.01). This delay in recovery for women with DM was also observed for significant PIPN: after 1 year of follow-up, 29.4% of women with DM still experienced grade 2 to 3 PIPN, whereas patients without DM showed no grade 2 to 3 neuropathy (P=.002). At 2 years, a significant recovery had occurred, but grade 2 to 3 PIPN was still evident in 18.2% of patients with DM. When we analyzed other factors related to PIPN duration, no association was found with age (P=.83), performance status (P=.19), G-CSF use (P=.97), or initial dose level (P=.97). A longer recovery was only seen for patients with grade 2 to 3 PIPN (log-rank test; P=.03).
In a multivariate Cox regression model also including age (>65 years) and grade 2 to 3 neuropathy, DM was the only independent negative predictor for PIPN recovery (hazard ratio [HR], 0.17; 95% CI, 0.05–0.56; P=.004; Table 4). No direct association with time to recovery of PIPN was observed for age older than 65 years (P=.65), whereas grade 2 to 3 neuropathy showed a nonsignificant trend for longer duration of PIPN (HR, 0.52; 95% CI, 0.25–1.11; P=.09).
Discussion
This report presents a case-control retrospective study aimed to determine the relation between DM and PIPN. Our data demonstrate a substantial delay
Kaplan-Meier plots assessing the association between diabetes mellitus and the time course of paclitaxel-induced peripheral neuropathy (PIPN). (A) Time to PIPN (wk). (B) Time to complete recovery of PIPN (mo).
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 13, 4; 10.6004/jnccn.2015.0057
Kaplan-Meier plots assessing the association between diabetes mellitus and the time course of paclitaxel-induced peripheral neuropathy (PIPN). (A) Time to PIPN (wk). (B) Time to complete recovery of PIPN (mo).
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 13, 4; 10.6004/jnccn.2015.0057
Kaplan-Meier plots assessing the association between diabetes mellitus and the time course of paclitaxel-induced peripheral neuropathy (PIPN). (A) Time to PIPN (wk). (B) Time to complete recovery of PIPN (mo).
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 13, 4; 10.6004/jnccn.2015.0057
Toxic neuropathy was found in 59.7% of our patients, in accordance with previous reports.21,22 The frequency of PIPN of any grade was significantly higher in the DM group, and this difference was even larger when only functionally relevant (grade 2–3) neuropathy was considered. Our data showing the independent association of DM with PIPN incidence partially agree with previous studies: Nurgalieva et al12 did not find relevant differences in the HR for PIPN11 after stratification by DM (HR, 2.92 vs 3.33 for DM) in a large cohort of elderly patients treated with taxanes and/or platinum. Another retrospective study including 219 patients treated with adjuvant paclitaxel (mostly with the every-3-weeks schedule) found a 97% incidence of PIPN.11 In contrast with our results, these authors found neither a longer duration nor a higher grade of PIPN in patients with DM, although their series only included 19 women with DM. The same limitation (only 5 patients with DM) applies to a recently published study by Kanbayashi et al,23 in which no association was found between DM and PIPN.
However, in agreement with our results, the largest study addressing PIPN, which included a wP (80 mg/m2) arm from a large clinical trial, found that hyperglycemia was associated with significant PIPN.19 As the authors point out, treatment-related hyperglycemia is probably an imperfect surrogate for DM, which limits a direct comparison with our data. In that same study, and in contrast with previous reports, age was not associated with development of significant neuropathy, which is also in agreement with our results.19,24
Two other factors had an impact on the appearance of grade 2 to 3 neuropathy: a higher dose level (100 mg/m2/wk) and G-CSF concurrent administration. No differences in the total cumulative dose were found between 80- and 100-mg/m2/wk dose levels, but the higher dose-intensity (ie, the 100 mg/m2 dose) might explain the development of more severe neuropathy. Our data on the association of G-CSF with severe PIPN represent an unexpected finding, because only one report has been published documenting an increased neurotoxicity after administration of G-CSF with vincristine.25 A large interindividual variability has been described for paclitaxel pharmacokinetics and toxicity.26,27 Because a direct neurotoxic effect of G-CSF is improbable, we speculate
Multivariate Cox Regression Model for PIPN Duration (N=68)
Long-term PIPN is a potential concern for breast cancer survivors after adjuvant treatment. However, few studies have evaluated the duration of chemotherapy-induced peripheral neuropathy. One study reported persistent neurotoxicity 1 year after adjuvant taxanes, although less than 5% of patients had significant symptoms,8 which is in agreement with our data in the control group. Other studies have reported higher frequencies (64%–81%) of long-term neurotoxicity in patients with breast cancer receiving adjuvant paclitaxel, which were even longer with wP.28,11 The contribution of DM to the time course of PIPN is even less known, especially for wP. An analysis of neuropathic pain in breast cancer survivors suggested a role for DM in persistent neurotoxicity. Tanabe et al11 demonstrated a longer-than-expected duration of PIPN (median, 727 days), but did not find an association between PIPN duration and DM. However, their series included only one case of DM within the subgroup of patients (n=24) treated with wP. In our series, which was homogeneously treated and enriched in patients with DM, we found that wP was associated with prolonged peripheral neuropathy in patients with breast cancer and DM. DM was the main predictor of residual PIPN, even after adjustment for grade of PIPN and age, which were previously reported as predictive factors for PIPN duration. The data showing a significant fraction of patients with DM experiencing grade 2 to 3 PIPN after 1 and 2 years also suggest the functional relevance and the probable impact on quality of life of delayed recovery from PIPN.
The exact mechanism of taxane-induced peripheral neuropathy is still unclear.6 Animal models of PIPN have shown a variety of pathophysiologic mechanisms leading to axonal degeneration of the peripheral nerves and dorsal root ganglia changes, such as neurotrophic factors deficit, proinflammatory activation, and microvascular damage.29,30 Some of these mechanisms of paclitaxel-induced neural damage are probably shared with diabetic neuropathy.31 Thus, both the concurrence of DM-induced and paclitaxel-induced nerve injury and the DM-related impairment of nerve regeneration could explain a delayed recovery after paclitaxel-induced nerve damage.32,33 Glycemic control might also modulate the degree of nerve injury, as suggested by the data reported by Schneider et al.19 Alternative explanations for the increased duration of PIPN in patients with DM are the preexistence of subclinical diabetic neuropathy or a higher systemic exposure to paclitaxel.34 The understanding of the underlying mechanisms is a key issue for designing clinical strategies to limit the impact of PIPN on patients with DM. Depending on the mechanisms, potential approaches might include either a detailed screening for diabetic neuropathy with electrophysiologic testing before chemotherapy administration, or stricter glycemic control during chemotherapy administration. Other approaches to ameliorate the development of peripheral neuropathy have been proposed, such as changes in time of infusion or concurrent administration of neuroprotective agents, but these have had little success.7 The alternative formulation of nanoparticle albumin-bound paclitaxel is not approved for adjuvant treatment and has been associated with a higher incidence of grade 3 PIPN when administered in 3-times-weekly regimens. Although lower incidence and faster recovery time have been shown in a clinical trial using weekly schedules of paclitaxel, no data are available regarding its toxicity in women with DM.35 Lastly, according to our results, the choice of paclitaxel schedules with lower weekly doses (80 mg/m2 instead of 100 mg/m2) might also limit the severity of PIPN in patients with DM, although no changes in incidence or duration would be expected.
Our work is limited by its retrospective nature, the short follow-up, and the pragmatic toxicity assessment. The retrospective design may have biased the detection of preexisting diabetic neuropathy or the evaluation of neurotoxicity. However, because comorbidity is higher in patients with DM,36 and those patients might be underrepresented in clinical trials, our clinical practice–based approach is probably better for determining the prevalence of DM and the course of PIPN in unselected patients with breast cancer. In addition to the homogeneous treatment schedule and infusion time, the comparable total dose and dose-intensity in cases and controls and the use of wP allow for the exclusion of other confounding variables and increase the clinical applicability of our findings.
In conclusion, we demonstrate a prolonged duration of significant peripheral neuropathy after wP in patients with breast cancer and preexisting DM. The incidence of significant (grade 2–3) PIPN was also higher in women with DM and associated with treatment-related factors (G-CSF use and dose level of paclitaxel). Given the expanding burden of breast cancer in elderly patients and the increasing incidence of DM among this population,13 a widening number of patients might be at risk for developing PIPN that will impair their daily functions for a long period. Before administering wP in patients with DM, the benefits and risks should be balanced carefully and alternative chemotherapy schedules should be evaluated.
The authors would like to acknowledge José Antonio López Oliva for his assistance in data retrieval.
References
- 1.↑
Henderson IC, Berry DA, Demetri GD et al.. Improved outcomes from adding sequential paclitaxel but not from escalating doxorubicin dose in an adjuvant chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol 2003;21:976–983.
- 2.↑
Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Peto R, Davies C et al.. Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term outcome among 100,000 women in 123 randomised trials. Lancet 2012;379:432–444.
- 3.↑
Ghersi D, Wilcken N, Simes RJ. A systematic review of taxane-containing regimens for metastatic breast cancer. Br J Cancer 2005;93:293–301.
- 4.↑
Jones SE, Erban J, Overmoyer B et al.. Randomized phase III study of docetaxel compared with paclitaxel in metastatic breast cancer. J Clin Oncol 2005;23:5542–5551.
- 5.↑
Lipton RB, Apfel SC, Dutcher JP et al.. Taxol produces a predominantly sensory neuropathy. Neurology 1989;39:368–373.
- 6.↑
Lee JJ, Swain SM. Peripheral neuropathy induced by microtubule-stabilizing agents. J Clin Oncol 2006;24:1633–1642.
- 7.↑
Scripture CD, Figg WD, Sparreboom A. Peripheral neuropathy induced by paclitaxel: recent insights and future perspectives. Curr Neuropharmacol 2006;4:165–172.
- 8.↑
Shimozuma K, Ohashi Y, Takeuchi A et al.. Taxane-induced peripheral neuropathy and health-related quality of life in postoperative breast cancer patients undergoing adjuvant chemotherapy: N-SAS BC 02, a randomized clinical trial. Support Care Cancer 2012;20:3355–3364.
- 9.↑
Seidman AD, Berry D, Cirrincione C et al.. Randomized phase III trial of weekly compared with every-3-weeks paclitaxel for metastatic breast cancer, with trastuzumab for all HER-2 overexpressors and random assignment to trastuzumab or not in HER-2 nonoverexpressors: final results of Cancer and Leukemia Group B protocol 9840. J Clin Oncol 2008;26:1642–1649.
- 10.↑
Chaudhry V, Chaudhry M, Crawford TO et al.. Toxic neuropathy in patients with pre-existing neuropathy. Neurology 2003;60:337–340.
- 11.↑
Tanabe Y, Hashimoto K, Shimizu C et al.. Paclitaxel-induced peripheral neuropathy in patients receiving adjuvant chemotherapy for breast cancer. Int J Clin Oncol 2013;18:132–138.
- 12.↑
Nurgalieva Z, Xia R, Liu CC et al.. Risk of chemotherapy-induced peripheral neuropathy in large population-based cohorts of elderly patients with breast, ovarian, and lung cancer. Am J Ther 2010;17:148–158.
- 13.↑
Wild SH, Roglic G, Green A et al.. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047–1053.
- 14.↑
Louwman WJ, Janssen-Heijnen ML, Houterman S et al.. Less extensive treatment and inferior prognosis for breast cancer patient with comorbidity: a population-based study. Eur J Cancer 2005;41:779–785.
- 15.↑
Lipscombe LL, Goodwin PJ, Zinman B et al.. The impact of diabetes on survival following breast cancer. Breast Cancer Res Treat 2008;109:389–395.
- 16.↑
Gogas H, Shapiro F, Aghajanian C et al.. The impact of diabetes mellitus on the toxicity of therapy for advanced ovarian cancer. Gynecol Oncol 1996;61:22–26.
- 17.↑
Rowinsky EK, Chaudhry V, Forastiere AA et al.. Phase I and pharmacologic study of paclitaxel and cisplatin with granulocyte colony-stimulating factor: neuromuscular toxicity is dose-limiting. J Clin Oncol 1993;11:2010–2020.
- 18.↑
Jerian SM, Sarosy GA, Link CJ Jr et al.. Incapacitating autonomic neuropathy precipitated by taxol. Gynecol Oncol 1993;51:277–280.
- 19.↑
Schneider BP, Zhao F, Wang M et al.. Neuropathy is not associated with clinical outcomes in patients receiving adjuvant taxane-containing therapy for operable breast cancer. J Clin Oncol 2012;30:3051–3057.
- 20.↑
Kirkman MS, Briscoe VJ, Clark N et al.. Diabetes in older adults. Diabetes Care 2012;35:2650–2664.
- 22.↑
Freilich RJ, Balmaceda C, Seidman AD et al.. Motor neuropathy due to docetaxel and paclitaxel. Neurology 1996;47:115–118.
- 23.↑
Kanbayashi Y, Hosokawa T, Kitawaki J, Taguchi T. Statistical identification of predictors for paclitaxel-induced peripheral neuropathy in patients with breast or gynaecological cancer. Anticancer Res 2013;33:1153–1156.
- 24.↑
Akerley W, Herndon J, Egorin M et al.. Weekly, high dose paclitaxel in advanced lung carcinoma. Cancer 2003;97:2480–2486.
- 25.↑
Weintraub M, Adde MA, Venzon DJ et al.. Severe atypical neuropathy associated with administration of hematopoietic colony-stimulating factors and vincristine. J Clin Oncol 1996;14:935–940.
- 26.↑
de Graan AJ, Elens L, Smid M et al.. A pharmacogenetic predictive model for paclitaxel clearance based on the DMET platform. Clin Cancer Res 2013;19:5210–5217.
- 27.↑
Lichtman SM, Hollis D, Miller AA et al.. Prospective evaluation of the relationship of patient age and paclitaxel clinical pharmacology: Cancer and Leukemia Group B (CALGB 9762). J Clin Oncol 2006;24:1846–1851.
- 28.↑
Hershman DL, Weimer LH, Wang A et al.. Association between patient reported outcomes and quantitative sensory tests for measuring long-term neurotoxicity in breast cancer survivors treated with adjuvant paclitaxel chemotherapy. Breast Cancer Res Treat 2011;125:767–774.
- 29.↑
Ledeboer A, Jekich BM, Sloane EM et al.. Intrathecal interleukin-10 gene therapy attenuates paclitaxel-induced mechanical allodynia and proinflammatory cytokine expression in dorsal root ganglia in rats. Brain Behav Immun 2007;21:686–698.
- 30.↑
Kirchmair R, Tietz AB, Panagiotou E et al.. Therapeutic angiogenesis inhibits or rescues chemotherapy-induced peripheral neuropathy: taxol- and thalidomide-induced injury of vasa nervorum is ameliorated by VEGF. Mol Ther 2007;15:69–75.
- 31.↑
Zychowska M, Rojewska E, Przewlocka B, Mika J. Mechanisms and pharmacology of diabetic neuropathy: experimental and clinical studies. Pharmacol Rep 2013;65:1601–1610.
- 32.↑
Singh B, Singh V, Krishnan A et al.. Regeneration of diabetic axons is enhanced by selective knockdown of the PTEN gene. Brain 2014;137:1051–1067.
- 33.↑
Kennedy JM, Zochodne DW. Impaired peripheral nerve regeneration in diabetes mellitus. J Peripher Nerv Syst 2005;10:144–157.
- 34.↑
Mielke S, Sparreboom A, Steinberg SM et al.. Association of paclitaxel pharmacokinetics with the development of peripheral neuropathy in patients with advanced cancer. Clin Cancer Res 2005;11:4843–4850.
- 35.↑
Gradishar WJ, Krasnojon D, Cheporov S et al.. Significantly longer progression-free survival with nab-paclitaxel compared with docetaxel as first-line therapy for metastatic breast cancer. J Clin Oncol 2009;27:3611–3619.
- 36.↑
Kalyani RR, Saudek CD, Brancati FL, Selvin E. Association of diabetes, comorbidities, and A1C with functional disability in older adults: results from the National Health and Nutrition Examination Survey (NHANES), 1999-2006. Diabetes Care 2010;33:1055–1060.
