Sequencing of anthracyclines and taxanes in neoadjuvant and adjuvant therapy for early breast cancer
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess whether the sequence in which anthracyclines and taxanes are administered, in an adjuvant or neoadjuvant setting, affects outcomes in women with early breast cancer.
Background
Description of the condition
Breast cancer is the most common malignancy and the second leading cause of cancer‐related mortality among women worldwide, and thus represents a significant healthcare burden (Ferlay 2015). Over the past few decades there have been substantial improvements in survival for women with early breast cancer following the introduction of adjuvant (after surgery) and neoadjuvant (before surgery) chemotherapy, endocrine therapy, and human epidermal growth factor receptor 2 (HER‐2)‐directed therapy (Cossetti 2015).
Description of the intervention
Anthracyclines and taxanes are active classes of chemotherapeutic agents used in the adjuvant and neoadjuvant treatment of women with early breast cancer.
Anthracyclines (e.g. doxorubicin, epirubicin, liposomal doxorubicin) exert their effect by complexing with DNA and topoisomerase II to induce apoptosis (i.e. cell death) and inhibit DNA and RNA synthesis. Potential toxicities of anthracyclines include cardiotoxicity, myelosuppression, and secondary malignancies (predominantly types of haematological cancer).
Taxanes (e.g. docetaxel, paclitaxel, nab‐paclitaxel) exert their effect by stabilising microtubules (fibrous shafts that assist chromosomes to divide), thereby inhibiting cell division and cell function. Potential toxicities of taxanes include neuropathy (i.e. tingling of the hands and feet), myelosuppression, and myalgia (muscle pain).
At present, standard clinical practice in women with early breast cancer is to administer a regimen of anthracycline‐based chemotherapy followed by a taxane. The reason for this established sequence appears to be historical rather than linked to outcomes. Anthracyclines were developed first and the benefit of anthracycline chemotherapy in early breast cancer was established prior to that of taxanes (Jones 2006; Levine 1998). However, one reason to assess the optimal sequence of anthracyclines and taxanes is the finding, in a large retrospective analysis involving approximately 1600 women with breast cancer who received paclitaxel or an anthracycline as adjuvant therapy, that outcomes were better when taxanes were given first (Alvarez 2010).
How the intervention might work
It is unknown whether the order in which taxanes and anthracyclines are administered results in significantly different outcomes in women with early breast cancer. It remains to be determined if the administration of taxanes first leads to better, worse, or no difference in treatment outcomes. The effect may also differ depending on the receptor status of the tumour.
Why it is important to do this review
The aim of this review is to assess whether the sequence in which anthracyclines and taxanes are administered, as adjuvant or neoadjuvant chemotherapy, affects outcomes in women with early breast cancer. The results of this review could potentially guide the management of chemotherapy sequencing in women with early breast cancer requiring adjuvant or neoadjuvant chemotherapy. A previous systematic review (Bines 2014) has explored this important topic, but since it was published further trials have been conducted. This Cochrane Review will complement the review by Bines 2014 by adding recent trial results and critically appraising the included studies.
Objectives
To assess whether the sequence in which anthracyclines and taxanes are administered, in an adjuvant or neoadjuvant setting, affects outcomes in women with early breast cancer.
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled trials examining the sequence of administration of anthracyclines and taxanes in women with early breast cancer receiving adjuvant or neoadjuvant chemotherapy.
Types of participants
Aged 18 years or older, with early breast cancer suitable for adjuvant or neoadjuvant chemotherapy.
Types of interventions
Intervention
Taxane (docetaxel, paclitaxel or nab‐paclitaxel) chemotherapy administered before an anthracycline‐based chemotherapy. We will include studies in which concurrent interventions with any other non‐anthracycline based chemotherapy, granulocyte colony stimulating factor, or trastuzumab were administered. We will not include studies in which concurrent interventions with radiotherapy or endocrine therapy were administered.
-
Docetaxel delivered intravenously at any dose weekly or every 21 days for 3 or 4 cycles
-
Paclitaxel delivered intravenously at any dose weekly for 12 weeks, every 14 days or 21 days for 3 or 4 cycles
-
Nab‐paclitaxel delivered intravenously at any dose weekly or every 21 days for 3 or 4 cycles
Comparator
Anthracycline (doxorubicin, epirubicin or liposomal doxorubicin)‐based chemotherapy administered before taxane chemotherapy. We will include studies in which concurrent interventions with any non‐taxane chemotherapy or granulocyte colony stimulating factor were administered. We will not include studies in which concurrent interventions with trastuzumab, radiotherapy, or endocrine therapy were administered.
-
Doxorubicin delivered intravenously at any dose every 14 days or every 21 days for 3 or 4 cycles
-
Epirubicin delivered intravenously at any dose every 14 days or every 21 days for 3 or 4 cycles
-
Liposomal doxorubicin delivered at any dose or frequency for 3 or 4 cycles
Types of outcome measures
Primary outcomes
Neoadjuvant and adjuvant setting
-
Overall survival, defined as the time from randomisation/study entry until death from any cause
Secondary outcomes
Neoadjuvant setting
-
Overall survival
-
Disease‐free survival, defined as time from surgery to first occurrence of recurrence of breast cancer at any site, development of new ipsilateral (same breast as previous breast cancer) or contralateral (different breast to previous breast cancer) breast cancer or second non‐breast malignant disease with the exception of basal cell or squamous cell carcinoma of the skin or carcinoma in situ of the cervix
-
Pathological complete response, defined as no invasive or in situ carcinoma in the breast or axillary lymph nodes (ypT0ypN0 (TNM staging; AJCC 2010)) after neoadjuvant therapy
-
Standardised Residual Cancer Burden score (RCB; MD Anderson Cancer Center)
-
Degree of response after neoadjuvant therapy
-
-
No invasive carcinoma in the breast or axillary lymph nodes (ypT0/isypN0)
-
No invasive carcinoma in breast (ypT0/isypN0/+)
-
No invasive carcinoma in axillary lymph nodes (ypN0)
-
Adjuvant setting
-
Overall survival
-
Disease‐free survival, defined as time from randomisation to first occurrence of recurrence of breast cancer at any site, development of new ipsilateral or contralateral breast cancer or second non‐breast malignant disease with the exception of basal cell or squamous cell carcinoma of the skin or carcinoma in situ of the cervix
Neoadjuvant and adjuvant setting
-
Adverse events classified according to the World Heath Organisation (WHO) or National Cancer Institute Common Terminology Criteria for Adverse Events (NCI‐CTCAE)
-
-
Febrile neutropenia
-
Cardiac toxicity
-
Pulmonary toxicity
-
Neurotoxicity
-
Haematological malignancy
-
Treatment‐related death
-
-
Treatment adherence, defined as delay in treatment or dose reductions, or both, or early cessation of treatment
-
Quality of life, measured using a validated instrument
Main outcomes of 'Summary of findings' table for assessing the quality of the evidence
The following outcomes will be included in a 'Summary of findings' table using the GRADE approach (Schünemann 2011).
-
Overall survival (mortality).
-
Disease‐free survival.
-
Pathological complete response.
-
Adverse events.
-
Quality of life.
Search methods for identification of studies
Electronic searches
We will search the following databases.
(a) The Cochrane Breast Cancer Group (CBCG)'s Specialised Register. Details of the search strategies used by the Group for the identification of studies and the procedure used to code references are outlined in the Group's module (mrw.interscience.wiley.com/cochrane/clabout/articles/BREASTCA/frame.html). We will extract and consider for inclusion in the review, trials with the key words "breast neoplasm; breast cancer; breast carcinoma; breast adenocarcinoma; breast tumour/tumor; adjuvant; neoadjuvant; anthracycline; taxane; chemotherapy; docetaxel; paclitaxel; nab‐paclitaxel; carbazitaxel; doxorubicin; epirubicin; daunorubicin; idarubicin and valrubicin".
(b) Cochrane Central Register of Controlled Trials (CENTRAL; latest issue) in the Cochrane Library. See Appendix 1.
(c) MEDLINE OvidSP (top up search to complement CBCG's Specialised Register, dates searched to be adjusted as necessary). See Appendix 2.
(d) Embase OvidSP (from 1974 onwards). See Appendix 3.
(e) The WHO International Clinical Trials Registry Platform (ICTRP) search portal (apps.who.int/trialsearch/Default.aspx) for all prospectively registered and ongoing trials. See Appendix 4.
(f) ClinicalTrials.gov (clinicaltrials.gov/). See Appendix 5.
We will re‐run searches within one month before publication of the review.
Searching other resources
(a) Bibliographic searching
We will try to identify further studies from the reference lists of identified relevant trials or reviews. We will obtain a copy of the full article for each reference reporting a potentially eligible trial. Where this is not possible, we will make attempts to contact authors to provide additional information.
(b) Searching conference proceedings
We will search the following conference proceedings in Embase (via OvidSP) from 2006 to the present to identify relevant abstracts.
-
American Society of Clinical Oncology Annual Scientific Meeting.
-
European Society for Medical Oncology Annual Scientific Meeting.
-
San Antonio Breast Cancer Symposium.
-
American Society of Clinical Oncology Breast Cancer Symposium.
-
European Breast Cancer Conference.
Data collection and analysis
Selection of studies
We will merge the search results using referencing software (e.g. Endnote) and remove duplicate results. Two review authors (MZ and AG) will independently screen titles and abstracts, and assess full‐text articles of potentially relevant studies for inclusion. We will resolve any disagreement about the eligibility of a study by discussion and, if required, by consulting a third review author (NW). We will record our reasons for the exclusion of any potentially relevant studies in the 'Characteristics of excluded studies' table. We will impose no language restrictions and will obtain translations of relevant studies, if required.
Data extraction and management
Two review authors (MZ and MW) will independently extract data using standard extraction forms tested and refined for this review. We will collect the following information: study design, participants, setting, interventions, follow up, sources of funding, notable conflicts of interest of trial authors, and outcomes.
We will extract at least the following items.
-
General information: title, authors, contact details, location, publication status, language, year of publication, source of funding.
-
Trial characteristics: study design, length of follow‐up.
-
Participants: inclusion and exclusion criteria, sample size, baseline characteristics and similarity at baseline, neoadjuvant/adjuvant setting, hormone receptor status, HER‐2 in‐situ hybridisation status, withdrawals, and losses to follow‐up.
-
Intervention and comparator: drug, dose, timing and number of cycles, dose reductions, dose omissions.
-
Adverse events and toxicities.
-
Outcomes: hazard ratios and 95% confidence intervals, log rank Chi2 statistic, log rank P values, number of events.
We will resolve any disagreement regarding the extraction of quantitative data for a study by discussion and, if required, by consulting a third author (NW, AG or DO'C). For studies with more than one publication, we will collate data from each publication into a single data collection form and will consider the final or updated version of each study the primary reference. We will record the selection process in a PRISMA flow diagram. We will include any excluded studies in the 'Characteristics of excluded studies' table.
Assessment of risk of bias in included studies
Two review authors (MZ and MW) will independently assess the risk of bias in each study using Cochrane's 'Risk of bias' assessment tool, as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Chapter 8.5; Higgins 2011). We will resolve any disagreements by discussion and, if needed, by consulting a third author (AG, NW or DO'C). We will assess the following sources of bias.
-
Sequence generation.
-
Allocation concealment.
-
Blinding of participants, personnel.
-
Blinding of outcome assessment for outcomes other than overall survival.
-
Incomplete outcome data.
-
Selective outcome reporting.
-
Other sources of bias.
We will outline the 'Risk of bias' assessments in a 'Risk of bias' table.
Measures of treatment effect
For dichotomous outcomes (variable with only two outcomes such as yes or no) ‐ treatment adherence, pathological complete response and adverse events ‐ we will report the treatment effect as a risk ratio with 95% confidence intervals. We will express pathological complete response as number needed to treat for an additional beneficial outcome and will estimate baseline risk based on the comparator arm.
For continuous outcomes (where measurement is continuous on a numerical scale) ‐ quality of life ‐ we will report the treatment effect as a standardised mean difference with 95% confidence intervals as quality of life is expected to be measured using different scales. If all data use the same scale, we will report the mean difference.
For time‐to‐event outcomes ‐ disease‐free survival and overall survival ‐ we will report the treatment effect as a hazard ratio with 95% confidence intervals. Where possible, we will extract the hazard ratio and associated variances directly from the trial publications. If this is not possible, we will obtain the data indirectly, using methods described by Parmar 1998 or Tierney 2007. We will record the use of indirect methods in the Notes section of the 'Characteristics of included studies' tables. We will report the ratios of treatment effects for response so that hazard ratios less than 1.0 will favour the administration of taxanes first and hazard ratios greater than 1.0 will favour the administration of anthracyclines first.
Unit of analysis issues
We anticipate no unit of analysis issues.
Dealing with missing data
We will contact authors of included studies in writing in order to obtain missing data (e.g. dosing or toxicity), which we will include where possible.
Assessment of heterogeneity
We will assess the degree of heterogeneity by visual inspection of forest plots, the I² statistic (Higgins 2003), and the Chi² test for heterogeneity (Cochran 1954). We will consider there to be substantial heterogeneity if the I² statistic is greater than 50% and the P‐value is less than 0.10 for the Chi² test for heterogeneity. If there is substantial heterogeneity we will use the random‐effects model (see Data synthesis) for pooling estimates across trials.
Assessment of reporting biases
We will assess publication or other bias by visual examination of funnel plot symmetry provided there are at least 10 studies in the meta‐analysis (Higgins 2011). Where possible we will review the protocols of included studies to assess outcome reporting bias.
Data synthesis
We will pool data using the fixed‐effect model if sufficiently similar (in terms of population and intervention) studies are available to provide meaningful results. We will performed all analyses using RevMan software (RevMan).
For dichotomous outcomes, we will use the fixed‐effect (Mantel‐Haenszel; Mantel 1959) method; if there is evidence of substantial heterogeneity, we will use the random‐effects (DerSimonian and Laird; DerSimonian 1986) method.
For continuous outcomes, we will use the fixed‐effect with inverse variance method (Deeks 2011); if there is evidence of substantial heterogeneity, we will use the random‐effects (DerSimonian and Laird with inverse‐variance) method.
For time‐to‐event outcomes, we will use the fixed‐effect with inverse variance method; if there is evidence of substantial heterogeneity, we will use the random‐effects (DerSimonian and Laird with inverse‐variance) method.
If we are concerned about the effect of small studies on the meta‐analysis and there is evidence of heterogeneity, we will compare the fixed‐effect and random‐effects estimates. If results from the random‐effects analysis suggest a more beneficial effect of treatment, we will perform sensitivity analyses to consider restricting the meta‐analysis to include the larger studies only.
'Summary of findings' table
We will use the GRADE approach to assess the quality of evidence for the following outcomes: mortality (overall survival), disease‐free survival, pathological complete response, adverse events, and quality of life. We will use GRADEproGDT software to develop the 'Summary of findings' table and follow GRADE guidance (Schünemann 2011). Two authors (MZ and MW) will grade the quality of the evidence.
To calculate the absolute risk for the control group for time‐to‐event outcomes, we will estimate the event rate at a specific time point (five years for overall survival and disease‐free survival) from Kaplan‐Meier curves or reported event rates. We will use a range for the baseline event rates (e.g. for low‐risk and high‐risk participants). We will enter these estimated values in GRADEproGDT and the corresponding absolute risks for the intervention group with low‐ and high‐risk subgroups at five years will be automatically populated by the GRADEproGDT software.
Subgroup analysis and investigation of heterogeneity
If data are available, we will perform subgroup analyses for the following subgroups.
-
Those receiving adjuvant versus neoadjuvant chemotherapy.
-
Those with positive versus negative HER‐2 status.
-
Those with positive, negative or triple negative hormone receptor status.
We will conduct tests for interaction to determine whether the sequence in which anthracyclines and taxanes are administered has a significantly different effect in subgroups.
Sensitivity analysis
We will perform the following sensitivity analyses.
-
Risk of bias: low versus high/unclear risk of bias. We will assign an overall unclear/high risk of bias to studies in which we have judged at least five of the eight domains to have unclear/high risk of bias.
