ORIGINAL ARTICLE

Can pregabalin prevent paclitaxel-associatedneuropathy?—An ACCRU pilot trial

Shivani S. Shinde1 & Drew Seisler2 & Gamini Soori3 & Pamela J. Atherton2&

Deirdre R. Pachman1& Jacqueline Lafky1 & Kathryn J. Ruddy1 & Charles L. Loprinzi1

Received: 5 March 2015 /Accepted: 8 June 2015 /Published online: 9 July 2015# Springer-Verlag Berlin Heidelberg 2015

AbstractPurpose Paclitaxel can cause an acute pain syndrome (P-APS), considered to be an acute form of neuropathy andchronic chemotherapy-induced peripheral neuropathy(CIPN). Anecdotal reports suggested that gabapentin may behelpful in the prevention of these toxicities. The purpose ofthis pilot study was to obtain data to support or refute theutility of pregabalin for the prevention of P-APS and CIPN.Methods Patients scheduled to receive weekly paclitaxel(80 mg/m2/dose) were randomized 1:1 to receive pregabalin75 mg or a placebo, twice daily, during the 12 weeks of che-motherapy. Patients completed the European Organization ofResearch and Treatment of Cancer Quality of Life (EORTCQLQ) CIPN20 questionnaire at baseline, prior to each dose ofpaclitaxel and monthly for 6 months post-treatment. Patientscompleted a post-paclitaxel questionnaire for 6 days after eachdose of paclitaxel and an acute pain syndrome symptom ques-tionnaire on day 8. The primary end point was to determinethe effect of pregabalin on the maximum of the worst acutepain scores for the week following paclitaxel administrationfor cycle 1.Results Forty-six patients were randomly assigned to thetreatment or placebo arm. There was no suggestion of a

difference between the two study arms with regard to P-APSmeasures. While there was a suggestion that pregabalin de-creased numbness, there was no suggestion that it decreasedtingling, pain, or the EORTC QLQ-CIPN20 subscale scores.There were no evident toxicity differences between the twostudy arms.Conclusions The results of this pilot trial do not support thatpregabalin is helpful for preventing P-APS or paclitaxel-associated CIPN.

Keywords Paclitaxel-associated acute pain syndrome .

Chemotherapy-induced peripheral neuropathy . Pregabalin .

Prevention

Introduction

Paclitaxel, a commonly used chemotherapeutic drug, cancause peripheral numbness, tingling, and pain, resulting in asyndrome known as chemotherapy-induced peripheral neu-ropathy (CIPN). It can also cause an acute pain syndrome,described in up to 58 % of patients [1]. Historically, this syn-drome, characterized by subacute onset of aches and paindeveloping within 1 to 3 days of paclitaxel administration,has been commonly referred to as paclitaxel-inducedarthralgias/myalgias [1]. Although historically described asmyalgia/arthralgia, the paclitaxel-induced acute pain syn-drome (P-APS) has been hypothesized to be a type of neuro-toxicity, possibly due to preferential sensitization of mechan-ical nociceptive neurons [2]. Consistent with this hypothesis,patients with severe P-APS appear to be at an increased risk ofdeveloping CIPN [3].

Gabapentin and pregabalin are effective in treating manyforms of neuropathic pain [4–6]. Their role for the treatment ofCIPN, however, is not clearly established. A randomized

Electronic supplementary material The online version of this article(doi:10.1007/s00520-015-2807-5) contains supplementary material,which is available to authorized users.

* Charles L. Loprinzicloprinzi@mayo.edu

1 Department of Oncology, College of Medicine, Mayo Clinic, 200First Street SW, Rochester, MN 55905, USA

2 Department of Biostatistics, Mayo Clinic, Rochester, MN, USA3 Medical Oncology, Nebraska Cancer Specialists, Omaha, NE, USA

Support Care Cancer (2016) 24:547–553DOI 10.1007/s00520-015-2807-5

controlled trial failed to show significant benefit of usinggabapentin to treat established symptoms caused by CIPN[7]. While these medications have not been tested in a ran-domized clinical trial to treat P-APS, two case series,consisting of two [8] and ten [9] patients each, suggested thatgabapentin could diminish the severity of P-APS. Additional-ly, there are animal models to support that gabapentin candecrease paclitaxel-induced nerve injury [10]. A review ofseveral clinical trials showed that perioperative administrationof gabapentin and pregabalin decreased the incidence of post-surgical chronic pain, i.e., pain persistent for 2 or moremonthsafter surgery [11]. These data led to the hypothesis that agabapentinoid agent would decrease the incidence and/or se-verity of the P-APS and paclitaxel-associated CIPN.Pregabalin was chosen over gabapentin for the current studyas it has some improved analgesic properties and is moreconvenient in terms of dosing [12, 13]; additionally, Pfizerwas willing to provide this agent and a placebo for this clinicaltrial.

The current clinical trial was developed to provide pilotdata regarding the potential role of pregabalin for the preven-tion and management of P-APS, to hopefully provide enoughpositive data to support the conduct of a subsequent largedefinitive randomized placebo-controlled trial.

Methods

Study design

This was a multi-centric, randomized, double-blinded, pilottrial. Institutional Review Board approval was obtained ateach of the 11 registering institutions.

Eligibility characteristics

Inclusion criteria included age≥18 years, an Eastern Cooper-ative Oncology Group (ECOG) performance status of lessthan or equal to 1, and life expectancy greater than 6 months.All of the participants were women scheduled to receiveweekly paclitaxel at a dose of 80 mg/m2, in the adjuvant(post-operative) or neo-adjuvant setting for breast cancer, fora planned course of 12 weeks, without any other concurrentcytotoxic chemotherapy (trastuzumab and/or other antibodyand/or small molecule treatments being allowed, except forpoly-ADP ribose polymerase [PARP] inhibitors). Participantsalso needed to have the ability to complete questionnaires bythemselves or with assistance and have the ability to provideinformed written consent.

Exclusion criteria included pregnant or nursing women,previous diagnosis of diabetic or other peripheral neuropathy,fibromyalgia, seizure disorder, significant renal insufficiency,prior exposure to neurotoxic chemotherapy, and a history of

allergic or other adverse reactions to gabapentin or pregabalin.Patients receiving gabapentin, pregabalin, or a PARP inhibitorwithin the preceding 6 months were also excluded.

Protocol treatment

Patients were randomized 1:1, using the Pocock-Simon dy-namic allocation procedure, to either the pregabalin or placeboarm. After the treatment assignment had been ascertained, thepatient’s study medication code number was listed on the con-firmation of registration screen. The blinded data manager/nurse/pharmacist at the patient’s institution contacted theACCRU Registration Office for a code number when studyproduct was needed for the patient. The treating physicianswere also blinded. Patients received pregabalin 75 mg or pla-cebo twice daily, starting on the first night of chemotherapyand continuing through the planned 12 weeks of chemothera-py. During the 13th week, the dose was decreased to once aday at bedtime, after which patients went off-study. Patientswere instructed to use acetaminophen 500mg every 6 h and/oroxycodone 5 mg every 1–2 h as needed for breakthrough painassociated with the P-APS.

Treatment evaluation

At baseline, patients had a history and physical exami-nation and completed a pre-paclitaxel questionnairewhich addressed the presence of symptoms related tobaseline pain and symptoms that might be related topotential pregabalin toxicities (to obtain baseline dataprior to starting pregabalin). A day after receiving pac-litaxel, patients were asked to complete a daily post-paclitaxel questionnaire for 6 days. This questionnaireincluded questions to assess worst, least, and averagepain scores and the need for analgesics over the last24 h. On day 8 after paclitaxel treatment, just prior toreceiving the next dose of paclitaxel, patients were re-quired to complete an acute pain syndrome symptomsummary questionnaire, which summarized symptomsand analgesic use over the preceding week and evaluat-ed patients for potential pregabalin toxicities. CIPN wasmeasured using the European Organization for Researchand Treatment of Cancer Quality of Life (EORTC-QLQ)CIPN20 questionnaire at baseline, prior to each paclitax-el cycle, and then every 30 days following completionof paclitaxel treatment, for 6 months. The EORTCQLQ-CIPN20 has been shown to be reliable, valid,and responsive to change [14, 15]. The tools used toassess the P-APS were the ones used to define the syn-drome [1–3]. Adverse events were monitored per thepatient questionnaires noted above, along with NationalCancer Institute (NCI) Common Terminology Criteriafor Adverse Events (CTCAE) version 4.0.

548 Support Care Cancer (2016) 24:547–553

Statistical methods

The primary goal was to assess the effectiveness of pregabalinon the P-APS. The primary end point was the maximum of theworst pain scores for the week following the first cycle ofpaclitaxel administration, as measured by a question on thedaily post-paclitaxel questionnaire. A non-parametricKruskal-Wallis test was used to compare primary end pointsbetween the two study arms. Secondary end points includedthe maximum of average pain and area under the curve (AUC)of the worst, least, and average pain following the first cycle ofpaclitaxel; opioid, and non-prescription drug use; maximumEORTC CIPN20 subscales; and item responses over thecourse of treatment and toxicities associated with pregabalin.The EORTC QLQ-CIPN20 was scored by a standard scoringalgorithm and converted to a 0–100 scale, where higher scoresrepresent better quality of life. Linear growth curve models,accounting for repeated measures from patients over thecourse of treatment, and AUC analysis were used to assessthe difference in CIPN20 subscales and items relating to tin-gling, numbness, and pain in the extremities by treatment arm.Longitudinal plots were also employed to depict differences inpatient-reported data over time.

Results

Baseline characteristics

This study accrued 46 patients between March 2012 and No-vember 2013, from 11 individual sites. A patient consort dia-gram is provided (Fig. 1). Five patients (four pregabalin; oneplacebo) either cancelled or withdrew or did not provide base-line pre-paclitaxel data andwere not evaluable for primary endpoint analyses. Baseline demographics were balanced amongthe remaining 41 patients. All the patients were females and23 (56 %; 53 and 59 % of pregabalin and placebo patients,

respectively) were over age 50 years (mean age 53.7, standarddeviation 13.7).

Neuropathy data

Paclitaxel acute pain syndrome

The primary end point of worst pain scores over the first cycleof treatment was not significantly different between arms (p=0.56). Patient-reported acute pain syndrome data, for 6 daysafter each chemotherapy dose, are illustrated in Table 1 andFig. 2. The maximum of average pain, as well as the AUCs ofthe worst, average, and least pain over the first cycle of treat-ment, were also not significantly different between arms (p=0.48, 0.62, 0.22, and 0.07, respectively). There, likewise, wasno significant difference between the study arms in the pro-portion of patients who used opioid or non-prescription med-ications for control of the paclitaxel acute pain syndrome. Thepercentages of patients who took non-prescription analgesicsin the 7 days after the first dose of paclitaxel, for paclitaxel-associated pain, were 53 and 50 % for the pregabalin andplacebo arms, respectively, while 16 and 18 % used opioids.

Chronic peripheral neurotoxicity

Growth curve models, as well as AUC analysis, show nosignificant differences in the EORTC CIPN20 sensory sub-scale (p=0.88 and p=0.46, respectively) between arms(Fig. 3). Similarly, there were no differences in the motorneuropathy or autonomic neuropathy subscales. Data fromsix individual questions regarding numbness, tingling, andshooting/burning pains, in fingers/hands and in toes/feet, areillustrated in Fig. 4, with a suggestion of a potential smalldifference for numbness symptom but not for tingling orshooting/burning pain. None of the other 14 individualEORTC CIPN20 instrument questions provided any sugges-tion of a difference between the study arms.

Evaluation of pregabalin toxicity

There were no statistically significant or clinically apparentdifferences between the two study arms with regard to multi-ple toxicities (including increased swelling, sleepiness,

Fig. 1 Consort diagram

Table 1 Effect of pregabalin on P-APS pain scores

P-APS (pain score over the first 6 days following initiation of paclitaxel;higher scores are worse)

Placebo(N=22)

Pregabalin(N=19)

p value

Worse pain: mean (SD) 3.2 (3.0) 2.6 (2.5) 0.56

Average pain: mean (SD) 2.2 (2.6) 2.6 (2.2) 0.48

Support Care Cancer (2016) 24:547–553 549

nausea, fatigue, trouble sleeping, nervousness, negative moodchanges, and dizziness or difficulty walking) which were eval-uated using weekly symptom questionnaires. Likewise, therewere no apparent differences in adverse events per NCICTCAE criteria.

Discussion

Findings from the current pilot trial were unable to support thepre-study hypothesis that this study would provide data tosupport the conduct of a placebo-controlled phase III trial, totest the efficacy of pregabalin for the prevention of paclitaxel-

associated neuropathy. The results are in concert with reportedfindings of another randomized, double-blind, placebo-controlled trial that assessed the efficacy of pregabalin forthe prevention/treatment of CIPN [16]. This clinical trial, con-ducted in patients with advanced colorectal cancer who werereceiving oxaliplatin-based chemotherapy, enrolled 64 pa-tients. The primary end point of this trial was the duration-adjusted average paresthesia change measured by a numericalrating scale. Patients received flexible pregabalin dosing, from150 to 600 mg/day versus placebos. The trial was terminatedearly, as an interim analysis found that there were no suffi-ciently positive data to continue the trial, based on a Bconditionalpower to detect a difference in treatment groups.^ To

Fig. 2 Worse pain (a) andaverage pain (b) scores over6 days following paclitaxel dosesfor each cycle. Higher scoresrepresent more pain

550 Support Care Cancer (2016) 24:547–553

our knowledge, there are no other reported trials of pregabalinor gabapentin, to prevent chemotherapy-induced neuropathy.

Similarly, there are no positive randomized clinical trials ofa gabapentinoid drug for the treatment of established CIPN.The one published randomized placebo-controlled double-blinded trial that studied this topic did not provide any sug-gestion of benefit for gabapentin for the treatment ofestablished CIPN [17].

Recently published American Society of Clinical Oncolo-gy guidelines acknowledge that gabapentinoids are notestablished as being beneficial. Nonetheless, these guidelinessuggested that it is reasonable to try gabapentinoids as a treat-ment option for selected patients with CIPN, given (1) theestablished efficacy of gabapentin and pregabalin for otherforms of neuropathic pain [18], (2) that only a single negativerandomized, placebo-controlled double-blinded trial for thisagent was completed [7], and (3) the limited treatment optionsfor CIPN.

While the current trial was established to study the role ofpregabalin in prevention of paclitaxel-induced neuropathy, itcan be argued that, indirectly, the results point against its util-ity for treating established CIPN. If the drug was effective fortreatment of established neuropathy, then it should have led toan improvement in CIPN scores in patients who developedCIPN. Although this trial had a low statistical power to dem-onstrate such a difference, the lack of even a trend in favor ofpregabalin suggests that it is not effective for treating CIPN.

It should be noted that pregabalin patients reported fewernumbness symptoms in both lower and upper extremities,throughout treatment and follow-up, though not significantlyso. It is not known whether this is a true benefit that can beattributed to pregabalin or if this occurred by chance, given thelarge number of items examined in this trial. However, the fact

that these results are consistent in both lower and upper ex-tremities supports that there may be a real benefit frompregabalin in preventing numbness. Of note, numbness isone of the more common problematic symptoms associatedwith paclitaxel neurotoxicity [19]. Despite the suggestion thatpregabalin might influence numbness, the current data are notstrong enough to recommend that pregabalin be used to pre-vent paclitaxel-associated numbness in clinical practice. Like-wise, these data, in and of themselves, are not strong enoughto support another clinical trial to evaluate the efficacy ofpregabalin for this situation.

A potential weakness of the current trial is that pilot trialreports of gabapentin suggested a benefit for this agent, whilethe current trial evaluated pregabalin, not gabapentin. Al-though there are no direct comparisons betweengabapentin and pregabalin for any situation, to the bestof our knowledge, both these drugs are commonly usedfor treatment of neuropathic pain. In fact, pregabalin hasbeen found to have distinct pharmacokinetic advantagesover gabapentin. It has higher bioavailability (90 versus33–66 % as compared to gabapentin) and is rapidlyabsorbed, reaching peak concentrations within 1 h.Gabapentin, unlike pregabalin, exhibits a non-linear re-lationship between drug dose and plasma concentration,making its pharmacokinetics less predictable. Also, basedon the population pharmacokinetic-pharmacodynamic model,pregabalin achieves a greater treatment effect in post-herpeticneuralgia and epilepsy [20]. For these reasons, we felt thatpregabalin would have had as reasonable a chance atpreventing P-APS in this patient population as gabapentinwould have.

A second potential weakness is related to the dose chosento be studied in this trial. In a recent Cochrane review fortreatment of neuropathic pain, in situations other than CIPN,pregabalin was found to be more beneficial at doses rangingbetween 300 and 600 mg daily, while 150 mg daily was lesseffective [21]. Three points can be made in defense of thecurrent trial’s dose: First, this Cochrane review reported som-nolence and dizziness as common side effects at higher doses,requiring treatment discontinuation in 18–28 % of patients.Second, the dose chosen for the current trial was one thatwas hoped that would lead to efficacy, without providing sub-stantial additional toxicity over and above that caused by thepaclitaxel. Third, the previous gabapentin CIPN treatment trialdid use large drug doses (2700 mg/day) without reported ben-efit [7].

A third potential weakness might be that it was chosen tostudy weekly paclitaxel doses at 80 mg/m2/dose, as opposedto every 3-week dose of 175 mg/m2/dose. While each of theseschedules result in similar CIPN toxicities, the lower individ-ual doses cause substantially less P-APS symptoms [1, 2].While it could be argued that studying the higher individualdoses would have allowed a bigger opportunity for observing

Fig. 3 CIPN scores during 12 weeks of treatment and over 6-monthfollow-up for the EORTC CIPN20 sensory neuropathy scores. Higherscores represent fewer symptoms

Support Care Cancer (2016) 24:547–553 551

Fig. 4 EORTC CIPN20 selected individual item scores during treatmentand over 6-month follow-up for tingling fingers/hands (a), tingling toes/feet (b), numbness of fingers/hands (c), numbness of toes/feet (d)

shooting/burning pain of fingers/hands (e), and shooting/burning pain oftoes/feet (f). Higher scores represent fewer symptoms

552 Support Care Cancer (2016) 24:547–553

a beneficial result for the P-APS, the weekly dose was chosenfor the following reasons: (1) The weekly paclitaxel dose isfrequently used for adjuvant breast cancer therapy, given that arandomized trial supported that this schedule was more effec-tive in this setting than an every 3-week dose schedule [22],and (2) it could be argued that the effect of an effective agent,such as was hoped with pregabalin, might be even more pro-nounced with a paclitaxel dose that caused less P-APSseverity.

In total, the data from this randomized placebo-controlledpilot study are unable to provide enough support to warrantthe conduct of a more formal phase III clinical trial to test theefficacy of pregabalin for reducing or preventing symptoms ofP-APS or paclitaxel-induced CIPN. In addition, the data fromthis trial provide further information that fails to support thevalue of gabapentinoids as an effective treatment ofestablished CIPN. Such data may lead to subsequent guidelinerecommendations that do not support the use ofgabapentinoids for treating CIPN.

Acknowledgments This project was funded by Pfizer and conductedby the Academic and Community Research United (ACCRU) group.

Conflict of interest Pfizer funded this study but had no input into thewriting of this manuscript. We have control of the primary data, which areavailable to the journal to review, if indicated.

References

1. Loprinzi CL, Reeves BN, Dakhil SR et al (2011) Natural history ofpaclitaxel-associated acute pain syndrome: prospective cohortstudy NCCTG N08C1. J Clin Oncol 29(11):1472–1478

2. Loprinzi CL, Maddocks-Christianson K, Wolf SL et al (2007) Thepaclitaxel acute pain syndrome: sensitization of nociceptors as theputative mechanism. Cancer J 13(6):399–403

3. Reeves BN, Dakhil SR, Sloan JA et al (2012) Further datasupporting that paclitaxel-associated acute pain syndrome is asso-ciated with development of peripheral neuropathy: North CentralCancer Treatment Group trial N08C1. Cancer 118(20):5171–5178

4. Caraceni A, Zecca E, Bonezzi C et al (2004) Gabapentin for neu-ropathic cancer pain: a randomized controlled trial from theGabapentin Cancer Pain Study Group. J Clin Oncol 22(14):2909–2917

5. Ross JR, Goller K, Hardy J et al (2005) Gabapentin is effective inthe treatment of cancer-related neuropathic pain: a prospective,open-label study. J Palliat Med 8(6):1118–1126

6. Guay DR (2005) Pregabalin in neuropathic pain: a moreBpharmaceutically elegant^ gabapentin? Am J GeriatrPharmacother 3(4):274–287

7. Rao RD,Michalak JC, Sloan JA et al (2007) Efficacy of gabapentinin the management of chemotherapy-induced peripheral neuropa-thy: a phase 3 randomized, double-blind, placebo-controlled, cross-over trial (N00C3). Cancer 110(9):2110–2118

8. van Deventer H, Bernard S (1999) Use of gabapentin to treattaxane-induced myalgias. J Clin Oncol 17(1):434–435

9. Nguyen VH, Lawrence HJ (2004) Use of gabapentin in the preven-tion of taxane-induced arthralgias andmyalgias. J Clin Oncol 22(9):1767–1769

10. Matsumoto M, Inoue M, Hald A, Xie W, Ueda H (2006) Inhibitionof paclitaxel-induced A-fiber hypersensitization by gabapentin. JPharmacol Exp Ther 318(2):735–740

11. Clarke H, Bonin RP, Orser BA, Englesakis M, Wijeysundera DN,Katz J (2012) The prevention of chronic postsurgical pain usinggabapentin and pregabalin: a combined systematic review and me-ta-analysis. Anesth Analg 115(2):428–442

12. Tarride JE, Gordon A, Vera-Llonch M, Dukes E, Rousseau C(2006) Cost-effectiveness of pregabalin for themanagement of neu-ropathic pain associated with diabetic peripheral neuropathy andpostherpetic neuralgia: a Canadian perspective. Clin Ther 28(11):1922–1934

13. Athanasakis K, Petrakis I, Karampli E, Vitsou E, Lyras L,Kyriopoulos J (2013) Pregabalin versus gabapentin in the manage-ment of peripheral neuropathic pain associated with post-herpeticneuralgia and diabetic neuropathy: a cost effectiveness analysis forthe Greek healthcare setting. BMC Neurol 13:56

14. Lavoie Smith EM, Barton DL, Qin R, Steen PD, Aaronson NK,Loprinzi CL (2013) Assessing patient-reported peripheral neuropa-thy: the reliability and validity of the European Organization forResearch and Treatment of Cancer QLQ-CIPN20 questionnaire.Qual Life Res 22(10):2787–2799

15. Postma TJ, Aaronson NK, Heimans JJ et al (2005) The develop-ment of an EORTC quality of life questionnaire to assesschemotherapy-induced peripheral neuropathy: the QLQ-CIPN20.Eur J Cancer 41(8):1135–1139

16. ClinicalTrials.gov. prevention and treatment of chemotherapy-induced peripheral neuropathy in subjects with advanced colorectalcancer. https://Clinicaltrials.gov/ct2/show/NCT00380874.Accessed 4014 Dec 25

17. Hershman DL, Lacchetti C, Dworkin RH et al (2014) Preventionand management of chemotherapy-induced peripheral neuropathyin survivors of adult cancers: American Society of ClinicalOncology clinical practice guideline. J Clin Oncol 32(18):1941–1967

18. Moulin D, Boulanger A, Clark AJ et al (2014) Pharmacologicalmanagement of chronic neuropathic pain: revised consensus state-ment from the Canadian Pain Society. Pain Res Manag 19(6):328–335

19. Hershman DL, Weimer LH, Wang A et al (2011) Association be-tween patient reported outcomes and quantitative sensory tests formeasuring long-term neurotoxicity in breast cancer survivors treat-ed with adjuvant paclitaxel chemotherapy. Breast Cancer Res Treat125(3):767–774

20. Bockbrader HN,Wesche D,Miller R, Chapel S, Janiczek N, BurgerP (2010) A comparison of the pharmacokinetics and pharmacody-namics of pregabalin and gabapentin. Clin Pharmacokinet 49(10):661–669

21. Moore RA, Straube S, Wiffen PJ, Derry S, McQuay HJ (2009)Pregabalin for acute and chronic pain in adults. CochraneDatabase Syst Rev (3):CD007076. doi:10.1002/14651858.CD007076.pub2

22. Sparano JA, Wang M, Martino S et al (2008) Weekly paclitaxel inthe adjuvant treatment of breast cancer. N Engl J Med 358(16):1663–1671

Support Care Cancer (2016) 24:547–553 553