Background

Breast cancer is the most commonly diagnosed malignancy globally, with 2.3 million new cases in 2020 [1]. Outcomes have generally improved particularly in higher income nations, including Canada [2]. This is largely attributed to better screening, improved local therapies, and advances in systemic treatment. In addition, a further appreciation of the breast cancer subtypes and associated disparate biology has facilitated several new approaches to multidisciplinary care that have changed the paradigm of breast cancer management [3]. In particular, practice-changing data available in the last several years has resulted in an increased momentum for the pre-operative, or neoadjuvant chemotherapy (NAC) approach to breast cancer treatment [4,5,6,7,8].

A Canadian national consortium for the neoadjuvant treatment of breast cancer (NABC) has existed since 2010. The most recent meeting of this group of national multidisciplinary experts was in May 2019 (Ontario, Canada). This group assembles national multidisciplinary experts in breast cancer to discuss and disseminate emerging evidence-based guidance across the country, and in particular focus on areas that have incomplete evidence and require expert opinion to help direct practice. Various members of this group have previously published meeting reports and one expert consensus guideline, with a significant focus on the utility of NAC for locally advanced breast cancer (LABC) [4,5,6]. However, over the last few years, research has increasingly demonstrated the important prognostic and predictive implications of treating certain subtypes of early breast cancer (HER2 positive and triple negative) with NAC [3], irrespective of upfront clinical stage or operability. The routine use of NAC for early stage breast cancer that is operable on presentation is a paradigm shift of great importance, with significant therapeutic and resource implications [7, 8]. Rapidly evolving evidence, the paucity of long-term data in some studies, the use of variable patient endpoints, and drug funding disparities within the country, can create some uncertainty in therapeutic approaches, but also opportunities for ongoing clinical trials [9]. Considering all this, there is an ongoing need for expert opinion to help consolidate the approach to NABC patient management. This is paramount to achieving the best possible uniform outcomes for Canadian breast cancer patients, particularly considering the publicly funded healthcare landscape. International breast oncology guidelines often embed the use of NAC within larger documents pertaining to breast cancer management [10]. In addition, some of the NAC recommendations are resource and practice-setting specific; there also remains some debate around the impact of certain research findings on clinical care (example using pathologic complete response rate as a practice-changing endpoint). Finally, most guidelines, including the recently published American Society of Clinical Oncology (ASCO) document, focus on systemic therapy alone [11], and do not include the subsequent implications of systemic therapy on surgical and radiation therapy decision-making. We, therefore, developed a contemporary, evidence-based Canadian National Consensus on the Neoadjuvant Treatment of Breast Cancer, using validated consensus methodology. This is meant to capture the most up-to-date evidence on optimal patient management throughout the entire treatment journey, while aligning multidisciplinary expert opinion with practice-based consensus from clinicians across the country.

Guideline type: Evidence Based Consensus.

Intended users: Practitioners who treat invasive breast cancer (pathology, radiology, surgery, medical oncology, radiation oncology, and other involved health professions.)

Applicable resource setting: Upper middle to high income nations with access to advanced screening, diagnostic, pathologic, surgical, radiation, and systemic therapy options.

Methods

Expert guideline panel

An expert guideline steering committee was established at the most recent Canadian National NABC Consortium meeting (May 2019, Ontario, Canada). The committee was comprised of academic and clinical experts in breast cancer management in the following specialties: medical oncology, surgical oncology, radiation oncology, breast radiology and anatomic pathology. All committee experts practice in academic cancer centres for more than 5 years, treat more than 100 unique breast cancer patients per year, and have demonstrated research and academic impact in NABC (peer-reviewed publications, research grants/projects, clinical trials involvement, and/or academic meeting presentations.) Representation from multiple Canadian provinces was sought.

Systematic evidence review

A systematic review of the literature was performed. As management for the neoadjuvant treatment of locally advanced breast cancer and general treatment principles of early breast cancer are well-established [5, 10, 11], the focus of the review was to update established principles of NABC care and highlight areas of new or evolving evidence that in particular would benefit from consensus to help improve practice. The overall focus was defined as the comprehensive management of breast cancer with a neoadjuvant therapy approach, including specific attention to the domains of multi-disciplinary assessment, diagnosis, monitoring, systemic therapy and local treatment. To maintain scope and feasibility, a single database search (PUBMED) was performed with the following parameters: invasive breast cancer, neoadjuvant, limited to phase 3 or 4 studies, meta-analysis, systematic review, or guidelines published in the English language. To focus mainly on new developments in this area, the search was limited to the past 5 years (initially October 2015 to October 2020 inclusive); the search was then repeated for November 2020 to May 2021 prior to manuscript preparation, to ensure no new relevant evidence had been published (Fig. 1). In December 2020, July 2021, October 2021, and December 2021, a targeted online gray literature search was completed to review any updated evidence as presented at four high impact oncology meetings (San Antonio Breast Cancer Symposium 2020 and 2021, American Society of Clinical Oncology 2021, European Society of Medical Oncology 2021), and any new published guidelines. The guideline panel decided it was important to capture any relevant new evidence with a select few additional recommendations, to ensure the guideline was most up-to-date. It was decided a priori that if the new evidence did not change the relevance or accuracy of existing recommendations, or greatly change the guideline’s scope or impact, these few select recommendations would not be sent for consensus to prevent delays in final guideline submission.

Fig. 1
figure 1

Literature search consort diagram. There were 78 studies included for analysis after applying the inclusion and exclusion criteria

Full size image

Consensus statements and consensus process

The steering committee developed recommendations for consensus based on the evidence review and discussion of important principles of NABC care, as established during and since the last Canadian NABC consensus statement in 2015 [5]. Discussions were held virtually (telephone), and via email correspondence. The statements were further reviewed by five additional clinical experts in breast surgical, radiation, and medical oncology; these expert reviewers were identified from past Canadian NABC Consortium involvement. Representation from multiple provinces was again ensured.

The Modified-Delphi approach is well recognized as a robust consensus methodology, particularly for consensus development in healthcare [12, 13]. This anonymous, survey-based consensus guideline model has several advantages compared to more traditional expert-based or nominal group methods; the latter rely solely on the opinions of a select group of individuals, and can be more subject to bias or the influence of the most vocal members of a guideline committee. Using the modified-Delphi approach, final statements were emailed to potential physician respondents using a secure online survey platform (Survey Monkey Inc., San Mateo, California, USA). Potential respondents were identified as being probable breast cancer clinicians by national or provincial medical society membership, provincial cancer centre affiliation, description of medical practice as available in public domain (example: institutional websites), recommendation by invited colleagues, and/or previous attendance at Canadian National NABC Consortium meetings. Invited participants were instructed only to respond if they had enough clinical expertise and experience in the neoadjuvant treatment of breast cancer to have an opinion. To achieve a broad practice-based consensus on expert recommendations, the guideline committee preferred not to restrict responses by years of practice or number of patients, and assumed respondents would only engage in the survey if they felt comfortable with the subject matter. Invitation to participate in the consensus process was carried out using email addresses for these individuals, as available in the public domain or through personal solicitation from the steering committee, or as shared by invited colleagues. Given the focus of this guideline was an practice-based physician consensus, other health care disciplines, and patient representatives were not included in the statement development or consensus process itself. Widespread physician representation was targeted, including multidisciplinary providers in both academic and community centers and in all Canadian provinces with comprehensive cancer programs.

Responses were anonymous; only respondent demographics including discipline, geographic area of practice, and years in practice were collected. Respondents were asked to indicate agreement, disagreement, or neutrality (i.e., "no opinion") to each statement. Given that multiple oncology specialties were involved, respondents were asked to indicate "no opinion" only if the statement was outside of their area of direct practice and not because they had no opinion about a statement related to their specialty. Reminders were sent twice over an 8-week period. Respondents were required to provide detailed qualitative feedback regarding statements they disagreed with. Specifically, respondents were instructed that if they did not agree with a statement in its entirety, to indicate disagreement, and provide detailed feedback regarding the elements they did not agree with. As per Modified-Delphi process, statements that did not achieve consensus were reviewed by the steering committee and modified based on the qualitative feedback as collected by the survey. These statements were emailed for a second round of survey; this was emailed to the same participants. Participants were instructed to respond only if they had responded to the first survey; one reminder was sent over 6 weeks. A third round was planned if required (Fig. 2).

Fig. 2
figure 2

Modified-Delphi Process. Consensus statements were developed by the steering committee based on evidence and relevant discussion. Statements were reviewed by 5 additional experts, finalized, and emailed as a secure online survey. Two rounds were required for consensus of > 79% for all statements

Full size image

Consensus analysis

Agreement statistics were calculated for each statement based on the total number of responses. The denominator for each statement (N, Agree + Disagree) was calculated as the sum of respondents who agreed and disagreed. Blank responses and those who indicated "no opinion" were excluded from the total number of responses for each question. The numerator (n, Agree) corresponded to the number of respondents who indicated "agree" for each statement. The proportion (n/N) was converted to a percent value (%) to determine the consensus value. A threshold value was determined a priori; consensus was defined as statements with 80% or more of respondents in agreement; statements with consensus > 79.5% were rounded up to 80%. Statements with less than an 80% (i.e., ≤ 79.5%) agreement level were marked for modification in the next round of survey, as per Modified Delphi methodology. Qualitative feedback was collected from respondents who indicated disagreement with particular statements; this feedback was utilized to modify statements that did not achieve consensus with the initial round.

Grading of recommendations

The final statements were ranked using the GRADE recommendations for guidelines (Strong or Conditional) [14], with consideration of the four domains within the framework for a recommendation’s direction and strength, which include: estimates of effect for desirable and undesirable outcomes of interest, confidence in the estimates of effect, estimates of values and preferences, and resource use [15]. In considering this framework, recommendations were generally considered strong if they were based on positive data and had level 1 or 2 evidence as per the GRADE framework for ranking evidence [14], and met the threshold for consensus. If a recommendation was lacking updated level 1–2 evidence in the last 5 years (acknowledging the review was limited to this timeframe), was deemed imperative to patient care and received a very high level of consensus (> 89%), it was also rated as strong. Recommendations deemed less impactful to patient care, with level 3 or 4 evidence, with preliminary (short term) level 1 or 2 evidence, high resource implications/lack of public funding, or with no published evidence and consensus < 90%, were marked as conditional. The term conditional was preferred over “weak” to indicate these statements may have evolving data thatmay strengthen the recommendation over time) and/or the statement may still be impactful for patient care, particularly in certain contexts. Available evidence was linked to recommendations in the “Grading” column (Tables 1 and 2). As the consensus statements may have been based on several sources of evidence with varying strengths, and to illustrate the grading of recommendations was not based on the strength of evidence alone (as discussed above), formal grading of the evidence was not integrated into the guideline table. Finally, additional statements added by the panel (Table 3) were all graded as “conditional” to reflect that they were not sent for consensus.

Table 1 Round one consensus
Full size table
Table 2 Round 2 consensus
Full size table
Table 3 Additional statements—not sent for consensus
Full size table

Engagement of further stakeholders

The final guideline was reviewed by a pharmacist, breast cancer patient representative, and a neoadjuvant nurse navigator, all affiliated with the Sunnybrook Health Sciences Centre. Their feedback on the premise of the guideline, agreement with recommendations, and on implementation was sought.

Results

A total of 47 recommendations were initially created by the steering committee and integrated into a consensus survey. Email invitations to complete the survey were sent to 391 clinicians in October 2020. There were 109 participants who completed the survey, for a response rate of 28%. Surgical oncology represented the largest respondent group (41/109; 37.6%), followed by medical oncology (29/109; 26.61%) and radiation oncology (21/109; 19.27%) (Fig. 3). Respondents were predominantly within their mid-career level of practice. Geographical representation was achieved from across several Canadian provinces, although the majority of respondents were located in central Canada (66.0% Ontario and Quebec), and at academic health institutions (77.0%). A summary of all respondent profiles is presented in Fig. 3.

Fig. 3
figure 3

Summary of Expert Respondents' Information. A Participants Clinical Specialty. B Years of Experience. C Practice Setting. D Geographic Region

Full size image

Consensus agreement: round 1

During the first survey round, 89.4% of questions (42/47) achieved 80% or greater consensus (agreement). A summary of all statements with the levels of consensus is presented in Table 1. Consensus was not reached for five statements under the following domains: patient selection, neoadjuvant systemic therapy, and local–regional management after neoadjuvant systemic therapy. The five statements that did not receive 80% agreement were modified based on qualitative feedback from the survey. These five statements were re-structured into 12 statements and sent for a second round of survey in December 2020. One additional new statement was integrated into the second survey to encompass new data regarding adjuvant therapy that became available as part of the targeted gray literature search in at that time.

Consensus agreement: round 2

In the second round, there were 81 respondents (81/109; 74.3% of respondents from round 1). A summary of the modified statements in round 2 is outlined in Table 2. All of the modified statements reached ≥ 80% consensus. The new statement on adjuvant therapy achieved consensus. A third round of survey was therefore not required, given the second round achieved complete consensus on this new statement, and all revised statements.

Systematic review

There were 389 citations found on systematic review; 311 were excluded based on abstract review; criteria are shown in Fig. 1. There were many early phase studies, and those focused on biomarker assessment, a rapidly evolving area of research in the neoadjuvant realm. As much of this data is exploratory or early, these studies were excluded. There were also many studies evaluating imaging response modalities for neoadjuvant therapy; some of these were included and matched to the statement regarding their investigational use. In general, studies that had negative results or did not meet primary efficacy endpoints, or with early phase data only, or therapies that had subsequent or conflicting data demonstrating a lack of meaningful impact on patient care, were excluded. The 78 included citations were fully reviewed and matched with guideline statements. For ease of readability and clarity, detailed descriptions of the evidence were not included in the recommendation table itself. Some recommendations did not have associated citations, as they were based on data published before the systematic review timeframe.

Further evidence review and additional statements

Targeted gray literature review did not demonstrate any impact on the accuracy or relevance of existing consensus statements. However, five additional recommendations were created by the expert guideline panel to reflect important areas of practice deemed not to be captured in the initial or revised statements This included statements on sentinel lymph node biopsy for N0 disease [16], nab-paclitaxel for neoadjuvant therapy [17, 18], and two additional adjuvant therapies, neratinib [19] and olaparib [20]. Finally, the rapidly evolving impact of molecular gene profiling on NABC was decided to be more clearly addressed after final external review. The five additional statements are presented in Table 3. Footnotes were included for a few consensus recommendations to clarify concepts as suggested by external review.

Additional stakeholder feedback

The nursing, patient, and pharmacist feedback sought demonstrated agreement with the recommendations overall, and in particular with the multi-disciplinary approach to NAC care. Suggestions to disseminate the guideline in patient, nursing, and pharmacy forums were made.

Discussion

There was a high level of agreement on 59 final statements encompassing the complex, multidisciplinary care pathway of neoadjuvant breast cancer patients. Five additional statements were not sent for consensus but were integrated to reflect the most up-to-date evidence pertaining to NABC at the time of manuscript preparation. Important highlights of this guideline include the recommendation to use neoadjuvant systemic therapy for early (operable) stage HER-2 positive and triple negative breast cancer, and the subsequent use of additional adjuvant therapies for those patients with residual disease after definitive surgery. In addition, this guideline demonstrates the importance of multi-disciplinary collaboration throughout the patient care journey. Finally, this consensus guideline demonstrates a balance between improving patient outcomes in an evidence-based manner while seeking to minimize toxicities, with a focus on individualized decision making, including clinical trial enrollment, particularly where evidence is less robust or still accumulating. This is particularly relevant for local–regional treatment approaches where evidence continues to accumulate from ongoing studies.

Assessment of resource implications

Given the broad scope of this guideline, including many treatment modalities, formal assessment of cost-effectiveness for individual therapies was outside the scope of this guideline. The committee acknowledged that most of the recommendations were applicable to High-middle and High income countries. Due to the robust health technology assessment for cancer drug funding recommendations nationally (pan-Canadian Oncology Drug Review through CADTH) [21], provincially funded systemic therapies for cancer generally have a cost-effectiveness backing within the Canadian healthcare landscape. Therefore, Health Canada approved agents that do not have wide-spread public funding or remain under evaluation (such as pertuzumab) were acknowledged within the recommendations as potentially having resource constraints at this current time. Corresponding statements suggested that accessibility and resources should be considered in particular for these drugs.

Limitations

Limitations of this guideline include only a 29% response rate to the consensus survey, and potential sampling and non-response bias. Primarily academic physicians responded to the survey, as such, the opinion of breast cancer clinicians in community practice settings may be under-represented. In addition, medical oncologists and surgeons comprised the largest group of respondents, and the opinion of other specialties may not completely be captured. There were also approximately 25% of initial participants who did not respond to the second round of survey, potentially impacting the results. However, this is unlikely, given the high levels of initial consensus on these statements with the first round (range 66–79%). Patient and other health care professionals were not engaged in the initial development of recommendations; their feedback was only sought on the final guideline and implementation plan. Five additional statements were created but not sent for consensus to prevent delay on the timely dissemination of this guideline; however, they are unlikely to impact on the scope and relevance of the guideline in general.

Summary and knowledge dissemination plan

This work represents an updated Canadian National Consensus on the Neoadjuvant treatment of breast cancer, across all parts of the therapeutic patient journey. A systematic review of recent literature and formal grading of recommendations was also achieved. The evidence was reviewed several times during guideline preparation, ensuring the most updated data was incorporated in a meaningful manner. The neoadjuvant treatment of breast cancer is a rapidly evolving area of clinical and academic interest; data can change quickly and uptake in clinical settings can falter based on sub-optimal knowledge dissemination or hesitancy to change practice. We believe our approach demonstrates Canadian consensus on key areas of neoadjuvant care, integrating available evidence, expert opinion, and practice-based consensus. We believe this guideline can help optimize patient outcomes across the country, by synthesizing the evidence into comprehensive recommendations for clinical care. Furthermore, the presence of national practice guidelines may help to foster clinical and policy change within healthcare organizations and health networks, with the hope of achieving uniformity of practice and thus patient outcomes. Given the importance of ensuring patient management is aligned with best practice, and to help optimize the use of resources and expertise in this area, we hope to achieve broad dissemination of this consensus guideline. A particular strength of this work is the inclusion of all elements of the patient treatment journey, formal grading of recommendations, and also achieving high levels of consensus, particularly in areas where evidence is lacking or evolving. This may help implementation and uptake of practice elements that can standardize Canadian breast cancer care as the neoadjuvant landscape continues to rapidly evolve. This guideline will be disseminated at the next Canadian National NABC Consensus meeting (planned for mid-2022), and ideally at national and international academic forums. There is also much interest in this document from national, provincial and hospital-based cancer programs in Canada to help guide local practice and resource allocation. We hope this guideline will be a strong addition to the published literature in this important area.