Tailored axillary surgery with or without axillary lymph node dissection followed by radiotherapy in patients with clinically node-positive breast cancer (TAXIS): study protocol for a multicenter, randomized phase-III trial

1Worldwide more than 2 million patients are diagnosed with breast cancer every year. It accounts for one third of all cancer diagnoses among women, and causes more than 600,000 deaths per year. Still about 1200 patients per year in Switzerland need axillary lymph node dissection (ALND) as part of their surgical treatment. ALND is indicated primarily for node-positive breast cancer. Patients are identified either by pre-operative evaluation of the axilla or by intraoperative sentinel lymph node biopsy (SLNB).

2ALND has traditionally been considered standard care for all patients with breast cancer. Due to the morbidity of the procedure (ALND vs. SLNB Risk Ratio = 3.07 in a meta-analysis of 98 studies) and decreasing axillary node involvement over time, a series of randomized controlled trials have been initiated in the 1990s and early 2000s to question this paradigm in patients with clinically negative axillary lymph nodes. These trials can be divided into four categories:Omission of any surgical axillary staging in selected patients,Omission of axillary dissection in all patients with negative sentinel lymph node (SLN) proceduresOmission of axillary dissection in selected patients with limited nodal disease in the SLNAxillary radiation vs. axillary dissection or observation

3Omission of any surgical axillary staging in selected patients,

4Omission of axillary dissection in all patients with negative sentinel lymph node (SLN) procedures

5Omission of axillary dissection in selected patients with limited nodal disease in the SLN

6Axillary radiation vs. axillary dissection or observation

7The trial findings supported the current trend in clinical practice toward decreased rates of axillary dissection in patients with non-palpable axillary lymph node metastases and showed that axillary radiation is a valid alternative to dissection in selected patients.

8The important landmark trial ACOSOG Z0011 has been criticized for several reasons. One limitation of ACOSOG Z0011 was the lack of standardization and detailed documentation of adjuvant radiation fields and the angles of the tangents, which makes it impossible to know how much of the axilla was irradiated. Since the no-axillary dissection arm in Z0011 was categorized as “no further axillary treatment” and defined by “no axillary dissection and no third-field nodal irradiation,” this question became relevant. However, an attempt to reconstruct the radiation fields of Z0011 resulted in the receipt of only 30% of detailed radiotherapy (RT) records for centralized review, and produced evidence of improperly applied axilla irradiation in at least 20% of cases.

9In parallel with the trend toward less axillary surgery, radiation oncologists have been broadening the indication for extended regional lymph node irradiation based on evidence from two large phase-III trials showing improved disease-free survival (DFS) and a large population-based cohort study showing improved overall survival in node-positive patients after ALND. In addition, recent data from the latest Early Breast Cancer Trialists’ Collaborative Group meta-analysis confirmed that post-mastectomy radiotherapy (PMRT) for patients with one to three positive nodes reduced recurrence and breast cancer mortality. If given, PMRT includes the chest wall and regional nodes in most patients. Consequently, the optimal contemporary loco-regional management of node-positive patients has become controversial, while ALND remains standard of care for patients with high-volume (i.e., large tumor load in the lymph nodes) or treatment-resistant nodal disease (i.e., residual disease in the lymph nodes after neoadjuvant systemic therapy).

10Two surgical trials have been initiated to provide evidence for the safety of omitting any axillary surgery in selected clinically node-negative patients (SOUND and INSEMA trials, please see Table 1 below).

11Several other ongoing randomized controlled trials primarily aim at validating the Z0011 protocol in different countries with several minor protocol modifications (Z0011-China, SERC/IPC 2012–001 in France, SENOMAC in Sweden, POSNOC in the United Kingdom, BOOG 2013–07 in the Netherlands).

12POSNOC allows axillary RT as an alternative to ALND in the control arm. Similarly, BOOG 2013–07 allows axillary RT or ALND to complete axillary treatment in the control arm, but includes patients with one to three positive SLN undergoing mastectomy. Since the Z0011 study population is not included in our proposed trial, there is no or only minimal overlap between these trials and the present protocol.

13The most progressive ongoing clinical trial on axillary management, partially overlapping with the proposed protocol, is Alliance A011202. It compares ALND with axillary radiation in patients with residual disease after chemotherapy, which was an exclusion criterion in both the AMAROS and the Z0011 trials. The trial tests the hypothesis that the AMAROS protocol in combination with extended regional nodal irradiation works in these patients without ALND to the extent that chemotherapy-resistant lymph node metastases are as radiosensitive as chemotherapy-naive disease.

14The optimal treatment of the regional lymph nodes in clinically node-positive patients is currently unclear, with less surgery, more RT or a combination thereof being under discussion. In the era of increasing post-mastectomy and extended regional node irradiation as well as effective systemic therapy, the indication for ALND in patients with clinically positive nodes and confirmed disease at the time of surgery has to be investigated in a clinical trial. Most of the above-mentioned trials excluded patients with residual metastases after neoadjuvant chemotherapy, and the omission of axillary dissection in patients with imaging-detected lymph node metastases is controversial since they may predict a larger volume of axillary disease.

15TAXIS is designed to determine whether ALND may be no longer necessary for confirmed nodal disease at first diagnosis of breast cancer given the prevalence of extended regional nodal irradiation in clinically node-positive patients in the adjuvant setting or incomplete response of nodal disease in the neoadjuvant setting. In many patients undergoing ALND, the number of negative lymph nodes removed exceeds the number of tumor-affected nodes. Removing multiple unaffected lymph nodes increases morbidity with no therapeutic benefit. It is a consequence of the commitment of radical surgery to the principle of complete tissue removal within the anatomical borders of the axilla. Benefits and harms of this approach must be re-assessed. TAXIS investigates tailored axillary surgery (TAS), a tailored approach focusing on the removal of tumor-affected lymph nodes while limiting the extent of surgery to minimize the number of negative nodes removed. TAS may offer non-inferior oncological outcomes with less morbidity than ALND in patients with clinically positive nodes at first presentation and confirmed nodal disease at surgery with or without neoadjuvant therapy.

16TAXIS is an international, multicenter, phase-III, non-inferiority randomized controlled trial with 34 study sites from four different countries currently planning to participate (Switzerland (Schweizerische Arbeitsgemeinschaft für Klinische Krebsforschung (SAKK)): 17; Austria (ABCSG): 10; Italy (IBCSG): 6; Hungary (IBCSG): (1). A list of sites and investigators participating in the trial can be downloaded from the public section of the SAKK website: http://sakk.ch/en/sakk-provides/our-trials/breast-cancer/sakk-2316-taxis/). Enrollment of patients started in July 2018 and will stop after the randomization of the targeted total of 1500 patients, which is expected in Q4 2023. End-of-trial treatment is expected for Q3 2024. All patients will be followed up for 20 years after randomization of the last patient. The trial will be terminated after the last visit of the last patient, which is expected to be in 2043. See Fig. 1 for the trial flow chart.

17The main objective of the trial is to test the hypothesis that treatment with TAS and axillary radiotherapy (RT) is non-inferior to axillary lymph node dissection (ALND) in terms of DFS of breast cancer patients with positive nodes at first presentation in the era of effective systemic therapy and extended regional nodal irradiation.

18Prior to enrollment, eligible patients are informed about the aims, procedures and possible risks of the trial, as well as its confidentiality policy regarding patient data, and are given sufficient opportunity and time to consider whether or not to participate. All patients are made aware that participation is voluntary and that they are allowed to refuse further participation in the trial whenever they want. Written informed consent is obtained from each patient before enrollment and prior to any trial-specific procedures.

19Eligible for inclusion in this trial are female and male patients aged ≥ 18 years if they meet the following criteria:

20Node-positive breast cancer (histologically or cytologically proven both in primary tumor and in lymph node) American Joint Committee on Cancer/Union for International Cancer Control (AJCC/UICC) stage II–III (all molecular subtypes allowed), with or without neoadjuvant treatment plannedNode-positivity detected by imaging (iN+) and confirmed by pathology or detected by palpation (cN1–2) and confirmed by pathologyNode-positivity initially detected by imaging (negative on palpation) and residual disease intraoperatively confirmed by pathology (in SLN or non-SLN during surgery) in case of neoadjuvant treatmentNode-positivity initially detected by palpation and residual disease intraoperatively confirmed by pathology in case of neoadjuvant treatmentEligible for primary ALND or SLN procedure with frozen section and either (1) newly diagnosed or (2) isolated in-breast recurrence or second ipsilateral breast cancer (at least 5 years disease free and no prior axillary surgery or loco-regional RT)Most suspicious axillary lymph node clippedAdequate condition for general anesthesia and breast cancer surgeryAbility to understand and complete the quality of life (QoL) questionnaireWHO performance status 0–2

21Node-positivity detected by imaging (iN+) and confirmed by pathology or detected by palpation (cN1–2) and confirmed by pathology

22Node-positivity initially detected by imaging (negative on palpation) and residual disease intraoperatively confirmed by pathology (in SLN or non-SLN during surgery) in case of neoadjuvant treatment

23Node-positivity initially detected by palpation and residual disease intraoperatively confirmed by pathology in case of neoadjuvant treatment

24Eligible for primary ALND or SLN procedure with frozen section and either (1) newly diagnosed or (2) isolated in-breast recurrence or second ipsilateral breast cancer (at least 5 years disease free and no prior axillary surgery or loco-regional RT)

25Most suspicious axillary lymph node clipped

26Adequate condition for general anesthesia and breast cancer surgery

27Ability to understand and complete the quality of life (QoL) questionnaire

28WHO performance status 0–2

29Patients are excluded on any of the following grounds:Stage IV breast cancerClinical N3 breast cancerClinical N2 breast cancer (if limited to the internal mammary nodes only)Contralateral breast cancerPrior axillary surgery (except prior sentinel node procedure in case of in-breast recurrence)Prior regional RTHistory of hematological or primary solid tumor malignancy, unless in remission for at least 5 years from pre-registration with the exception of adequately treated cervical carcinoma in situ or localized non-melanoma skin cancerConcurrent treatment with any experimental drug within 30 days of pre-registrationConcomitant use of other anticancer drugs or RT

30Stage IV breast cancer

31Clinical N3 breast cancer

32Clinical N2 breast cancer (if limited to the internal mammary nodes only)

33Contralateral breast cancer

34Prior axillary surgery (except prior sentinel node procedure in case of in-breast recurrence)

35Prior regional RT

36History of hematological or primary solid tumor malignancy, unless in remission for at least 5 years from pre-registration with the exception of adequately treated cervical carcinoma in situ or localized non-melanoma skin cancer

37Concurrent treatment with any experimental drug within 30 days of pre-registration

38Concomitant use of other anticancer drugs or RT

39Patients are, furthermore, excluded if they have any serious underlying medical, psychiatric, psychological, familial or geographical condition, which, in the judgment of the investigator, may interfere with the planned staging, treatment, and follow-up, affect patient compliance or place the patient at high risk from treatment-related complications.

40A total of 1500 patients will be randomized at a ratio of 1:1 (750 patients per each treatment arm) to either receive TAS followed by ALND and regional nodal irradiation excluding the dissected axilla as a target volume (arm A), or receive TAS followed by regional nodal irradiation including the full axilla (arm B). Randomization is performed using the minimization method with 80% allocation probability according to the following stratification factors:Responsible surgeonType of positive node detection:○ Node-positivity detected by imaging (iN+) and confirmed by pathology○ Node-positivity detected by palpation (cN1–2) and confirmed by pathology○ Node-positivity initially detected by imaging (negative on palpation) and residual disease confirmed by pathology (in SLN or non-SLN during surgery) after neoadjuvant treatment○ Node-positivity initially detected by palpation and residual disease confirmed by pathology after neoadjuvant treatmentNewly diagnosed vs. isolated in-breast recurrence or second ipsilateral breast cancer (at least 5 years disease free and no prior axillary surgery or loco-regional RT)Normofractionated vs. hypofractionated RTMale / female patient

41Responsible surgeon

42Type of positive node detection:○ Node-positivity detected by imaging (iN+) and confirmed by pathology○ Node-positivity detected by palpation (cN1–2) and confirmed by pathology○ Node-positivity initially detected by imaging (negative on palpation) and residual disease confirmed by pathology (in SLN or non-SLN during surgery) after neoadjuvant treatment○ Node-positivity initially detected by palpation and residual disease confirmed by pathology after neoadjuvant treatment

43○ Node-positivity detected by imaging (iN+) and confirmed by pathology

44○ Node-positivity detected by palpation (cN1–2) and confirmed by pathology

45○ Node-positivity initially detected by imaging (negative on palpation) and residual disease confirmed by pathology (in SLN or non-SLN during surgery) after neoadjuvant treatment

46○ Node-positivity initially detected by palpation and residual disease confirmed by pathology after neoadjuvant treatment

47Newly diagnosed vs. isolated in-breast recurrence or second ipsilateral breast cancer (at least 5 years disease free and no prior axillary surgery or loco-regional RT)

48Normofractionated vs. hypofractionated RT

49Male / female patient

50Randomization is usually performed during the breast surgery as soon as the operating surgeon has certified that all inclusion criteria are met, including in particular the pathological confirmation of node-positivity and residual disease for patients who had received neoadjuvant treatment. If node-positivity cannot be confirmed on frozen section during primary surgery, the randomization takes place as soon as possible after the positive pathology results are available. Randomization is exclusively done online by using the centralized electronic data capture system secuTrial®. Patients who are not randomized will not enter into the trial and no further data will be collected except retrospective data for the two subprojects described in section “Performance characteristics of TAS” and “Patterns of use of neoadjuvant systemic treatment” of Appendix 3. Those patients will be treated according to the current best standard of care according to investigator’s decision, which commonly involves ALND.

51The investigational trial treatment consists of tailored axillary surgery (TAS), which is defined by the SLN procedure in combination with the selective removal of all palpable disease and documentation of the removal of the initially biopsy-proven and clipped lymph node metastasis by specimen radiography. If the clip is not documented in the specimen radiography, the patient is excluded and undergoes ALND.

52All patients undergo adjuvant whole-breast irradiation after breast-conserving surgery and chest-wall irradiation after mastectomy. While patients allocated to arm A (control arm) receive regional nodal irradiation excluding the dissected axilla as a target volume (levels (II)/III; medial supraclavicular; internal mammary lymph nodes), patients allocated to arm B (investigational arm) receive regional nodal irradiation including the full axilla (levels I–III; medial supraclavicular; internal mammary lymph nodes).

53If indicated, patients may undergo adjuvant systemic treatment.

54A video of the intervention has been uploaded as Additional file 1.

55All patients undergo breast-conserving surgery or mastectomy and the SLN procedure. The SLN procedure should preferably be performed by dual mapping, which includes all nodes that are either blue (blue dye), hot (technetium Tc-99 m), fluorescent (indocyanine green) or magnetic (superparamagnetic iron oxide particles), according to local SLN procedure practices. In addition to the removal of the sentinel nodes, all palpably suspicious nodes, defined as either hardened or irregular or very large or a combination thereof, are removed.

56Surgical evaluation of the levels I–III by palpation is mandatory to ensure that there is no palpably suspicious disease left behind in the axilla after TAS. The length of the incision and the opening of the clavipectoral fascia has to be chosen accordingly. Therefore, a minimum incision length of 2–3 cm is recommended. In case of palpably suspicious nodes left behind in the axilla after TAS, the patient must be excluded und typically undergoes ALND according to the decision of the treating surgeon.

57Surgeons who perform the SLN procedure as part of their clinical routine are allowed to perform TAS. TAS is feasible for all surgeons who master the SLN procedure since the only difference is the selective removal of palpable disease, which we expect to be a frequent step in the clinically node-positive patient population of this trial. However, thorough palpation of the axilla is also mandatory during the regular SLN procedure in current clinical practice to remove suspicious findings and minimize the false-negative rate. The difference is that during the regular SLN procedure that is performed as a staging procedure, this step is rarely necessary since it is only performed in clinically node-negative patients.

58The selective removal of the clipped lymph node by image-guided localization, a procedure increasingly referred to as targeted axillary dissection (TAD), is encouraged to increase the chances of successful clip removal. However, TAD is not a mandatory part of TAS, since it is technically challenging and, therefore, quite controversial, and the clipped lymph node corresponds to one of the SLNs in the majority of patients. TAD has recently emerged as an effective strategy to reduce the false-negative rate of the SLN procedure in patients with initially confirmed node-positive breast cancer that showed a complete clinical response in the nodes after neoadjuvant treatment. The metastatic node is marked with a clip during biopsy, or shortly after the lymph node metastasis has been confirmed by pathology or cytology, and is then selectively localized and removed during the procedure. Any method is allowed for localization of the clipped node, such as the use of wire, iodine-125 or magnetic seeds, radioguided occult-lesion localization, ultrasound or a tattoo. If the clip is not documented in the specimen radiography, the patient is excluded and undergoes ALND, since confirmation of the removal of the lymph node with the initially biopsy proven is not possible.

59Importantly, residual suspicious lymph nodes detected by imaging before the end of adjuvant treatment does neither demand nor prohibit take-back surgery for completion ALND or selective removal of these nodes or an additional RT boost. The study group is fully aware that some patients will have non-palpable residual disease in the axilla after TAS. The hypothesis of TAXIS is that this residual disease does not progress to recurrence. The clinical scenario of residual disease after axillary surgery detected by imaging, i.e., performed for RT treatment planning or staging, to rule out distant disease, has become more frequent after implementation of the ACOSOG Z0011 and the EORTC-AMAROS protocols. Histological evaluation and adjustment of axillary treatment (such as completion ALND, selective removal of suspicious nodes or additional RT boost) are allowed and interdisciplinary consensus-based decisions are encouraged.

60After randomization, the patients in arm A will be treated according to the current standard of care by ALND. ALND is defined by the intention of the surgeon to radically remove the entire soft tissue within the anatomical borders of the axilla. To ensure applicability of the findings of the present study to clinical practice, we purposefully refrained from further defining the exact technique or number of removed lymph nodes. One common way to perform ALND is described here. Standard ALND clears levels I and II. A clip may be applied by the surgeon to mark the medial border of dissection, commonly between levels II and III, and documented in the surgical report. A full level-III dissection above the pectoralis minor muscle is carried out when there is gross nodal disease. The latissimus dorsi muscle is identified and followed until it is crossed by the axillary vein. The surface of the vein is then cleared of fat. Dissection inferior to the vein is carried out, dividing the fat and controlling the branches of the axillary vein entering the specimen. The thoracodorsal bundle is the deep lateral branch. Once identified, the pectoralis minor is retracted and the level-III nodes are dissected – if indicated – from the space below the axillary vein. The long thoracic nerve is identified against the chest wall and dissected free from the specimen. The fat between the thoracodorsal and long thoracic nerves is encircled with a clip or bipolar vessel-sealing system, divided, and bluntly swept inferiorly. Branches of the thoracodorsal vessels entering the specimen are dissected, and the specimen is freed from its remaining attachments to the inferior chest wall. A closed suction drain is standard care for most surgeons. The dermis is re-approximated and the skin is closed.

61All patients undergo adjuvant whole-breast irradiation after breast-conserving surgery or chest-wall irradiation after mastectomy, as well as regional nodal irradiation, excluding the dissected axilla as a target volume in arm A (levels (II)/III; medial supraclavicular; internal mammary lymph nodes) and including the full axilla (levels I–III; medial supraclavicular; internal mammary lymph nodes) as a target volume in arm B. In case of extensive nodular involvement in arm A, the target volume includes the area at risk of the dissected levels. Inclusion of internal mammary nodes is recommended in this high-risk group of patients irrespective of treatment arm. RT should start preferably within 8 weeks from the last breast surgical procedure and not later than 12 weeks. In case chemotherapy was applied, RT should start within 6 weeks after the end of the last cycle of chemotherapy and not later than 8 weeks. Dose to the breast/thoracic wall as well as the regional nodal pathways: 50 Gy in 25 fractions of 2 Gy or 50.4 Gy in 28 fractions of 1.8 Gy; daily, 5 days a week. Hypofractionated schedule allowed: 40 Gy in 15 fractions of 2.67 Gy to the same volume.

62For a detailed description of RT, see Appendix 6.

63Where adjuvant systemic anticancer treatment is indicated, all drugs to be used are locally chosen according to international guidelines, including the sequence of systemic therapy in relation to surgery (neoadjuvant vs. adjuvant setting). Handling of chemotherapy, anti-HER2 treatment and endocrine therapy is subject to the approved product information in each respective country. The decision of the indication of adjuvant chemotherapy could be different in the two treatment arms, because of a higher number of lymph nodes removed in arm A (ALND) compared to the group with TAS only. However, this belongs to a strategic trial like this and cannot be avoided with pre-defined rules in the protocol.

64For a description of all evaluations performed before, during, and after the trial treatment, see Appendix 1.

65The primary endpoint of the trial is disease-free survival (DFS), defined as time from randomization until one of the following events, whichever comes first:Local recurrence, regional recurrence, distant recurrenceSecond breast cancerDeath from any cause

66Local recurrence, regional recurrence, distant recurrence

67Second breast cancer

68Death from any cause

69Patients not experiencing an event will be censored at the date of the last available assessment.

70The secondary endpoints of the trial are:QoL

71QoL

72The primary objective is to compare patient-reported arm problems in the short (after 9 months), intermediate (after 24 months), and long term (after 60 and 120 months) in breast cancer patients with clinically positive nodal disease in the era of extended regional nodal irradiation who are randomized to TAS followed by axillary radiation vs. TAS followed by ALND. The primary QoL endpoint is the change in the ARM subscale of the FACT-B + 4 from baseline to 24 months after randomization.

73Secondary objectives include the comparison of (1) short-, intermediate-, and long-term QoL and of (2) short-, intermediate-, and long-term effect of arm problems on daily and social activities between randomized groups.

74For a detailed description of the QoL assessment procedure, see Appendix 2.Overall survivalBreast cancer-specific survivalTime to local recurrenceTime to regional recurrenceTime to distant recurrencePhysician-reported morbidity outcomes○ Lymphedema: the change in ipsilateral upper-extremity circumference, corrected for any change in the contralateral upper extremity, will be calculated using the following formula:

75Overall survival

76Breast cancer-specific survival

77Time to local recurrence

78Time to regional recurrence

79Time to distant recurrence

80Physician-reported morbidity outcomes○ Lymphedema: the change in ipsilateral upper-extremity circumference, corrected for any change in the contralateral upper extremity, will be calculated using the following formula:

81○ Lymphedema: the change in ipsilateral upper-extremity circumference, corrected for any change in the contralateral upper extremity, will be calculated using the following formula:

82where I indicates ipsilateral upper-extremity circumference, C indicates contralateral upper-extremity circumference, a indicates assessments during trial treatment and follow-up, and b indicates baseline assessment. L will be calculated for both upper arm and forearm, and lymphedema is defined as present if L > 2 cm for either location. The proportion of patients with lymphedema will be calculated at each assessment○ Decreased range of shoulder motion: the change in shoulder motion (flexion, abduction, internal rotation, and external rotation) assessed by goniometric measurement of arm movement (Appendix 6) on both sides will be calculated for each assessment

83where I indicates ipsilateral upper-extremity circumference, C indicates contralateral upper-extremity circumference, a indicates assessments during trial treatment and follow-up, and b indicates baseline assessment. L will be calculated for both upper arm and forearm, and lymphedema is defined as present if L > 2 cm for either location. The proportion of patients with lymphedema will be calculated at each assessment

84○ Decreased range of shoulder motion: the change in shoulder motion (flexion, abduction, internal rotation, and external rotation) assessed by goniometric measurement of arm movement (Appendix 6) on both sides will be calculated for each assessment

85Adverse events (AEs) according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) v4.03

86Clipping-related AEs and specific AEs related to the surgical procedure and RT will be assessed according to NCI CTCAE v4.03Late RT-related AEs

87Late RT-related AEs

88Late AEs related to the RT will be assessed according to the Late Effects in Normal Tissues—Subjective, Objective, Management, and Analytic (LENT-SOMA) scaleSurgical site infections (SSI)

89Surgical site infections (SSI)

90SSI will be assessed according to the Centers for Disease Control and Prevention Surgical Site Infection Classification System.

91The following seven subprojects have been pre-specified:Immune profile of axillary lymph nodesEvaluation of the prognostic value of multigene testsPerformance characteristics of TASPatterns of use of neoadjuvant systemic treatmentImpact of TAS on adjuvant systemic treatment decisionsQuality assessment of delineation and dose planning and impact on treatment outcomes in adjuvant RTImpact of a bolus on outcome and RT-related AEs in patients with mastectomy

92Immune profile of axillary lymph nodes

93Evaluation of the prognostic value of multigene tests

94Performance characteristics of TAS

95Patterns of use of neoadjuvant systemic treatment

96Impact of TAS on adjuvant systemic treatment decisions

97Quality assessment of delineation and dose planning and impact on treatment outcomes in adjuvant RT

98Impact of a bolus on outcome and RT-related AEs in patients with mastectomy

99For a detailed description of the subprojects, see Appendix 3.

100The sample size is based on the primary endpoint DFS. With a type I error of 5% and a power of 80%, 385 events will be needed to show non-inferiority of TAS and axillary RT in comparison to ALND with a non-inferiority hazard ratio of 1.289 (corresponding to a DFS at 5 years of 80% in the ALND arm and 75% in the TAS and axillary RT arm). The sample size needed is 1500 patients (750 per arm). One interim analysis for efficacy/futility is foreseen after 20% of the required events have occurred. A report including the interim efficacy results as well as patient characteristics, treatment administration and safety results (AEs, SSI, and serious AEs) will be presented to an independent Data Monitoring Committee appointed by the SAKK Board. The SAKK Board will decide on the continuation/modification/early stopping of the trial based on the recommendations of the committee. All efficacy endpoints will be analyzed based on the per-protocol set. For the primary endpoint, DFS, the hazard ratio and one-sided 95% confidence interval will be calculated using a Cox regression model with the treatment arm as independent variable and the stratification factors as strata. The median DFS and the corresponding 95% confidence intervals using the Kaplan-Meier method will be presented for each treatment arm.

101Planned subgroup analyses for the primary endpoint:Type of positive node detection at first diagnosis: cN1 vs. iN1Neoadjuvant vs. adjuvant systemic treatment vs. bothycN1 vs. ycN0BMI < 25 kg/m2 vs ≥ 25 kg/m2Normofractionation vs. hypofractionationBreast-conserving surgery vs. mastectomyTriple-negative vs. HER2+ vs. ER- and/or PR-positive, and HER2-

102Type of positive node detection at first diagnosis: cN1 vs. iN1

103Neoadjuvant vs. adjuvant systemic treatment vs. both

104ycN1 vs. ycN0

105BMI < 25 kg/m2 vs ≥ 25 kg/m2

106Normofractionation vs. hypofractionation

107Breast-conserving surgery vs. mastectomy

108Triple-negative vs. HER2+ vs. ER- and/or PR-positive, and HER2-

109No imputation of missing data will be performed. A row denoted “Missing” will be included in count tabulations if necessary to account for drop-outs and missing values. For continuous variables a column with the number of available observations will be added. Patients lost to follow-up before reaching the primary endpoint will not be replaced.

110Patients will be instructed by the investigator to report the occurrence of all AEs. The following events are reported by the investigator within the trial:Any AE related to the clipping of the most suspicious axillary lymph node andSpecific AEs related to the trial surgical procedure or the RT (possible, probable, definite) as pre-defined in the original trial protocol

111Any AE related to the clipping of the most suspicious axillary lymph node and

112Specific AEs related to the trial surgical procedure or the RT (possible, probable, definite) as pre-defined in the original trial protocol

113The investigator assesses and records the AEs observed during the AE-reporting period: from the date of patient consent signature up to 20 years after randomization of the last patient. However, no AEs related to neoadjuvant and/or adjuvant systemic therapy will be reported.

114For a description of the study management and administration, see Appendix 4.

115For a description of the study documentation, see Appendix 5.

116Designed to test the hypothesis that TAS in combination with RT is non-inferior to ALND in terms of curing patients and preventing recurrences, yet is significantly superior in terms of patient morbidity, this trial has the potential to establish a new worldwide treatment standard in breast cancer surgery. If found to be non-inferior to standard treatment, TAS may contribute significantly to reduce morbidity in breast cancer patients by avoiding surgical overtreatment. For the SPIRIT checklist applicable to this study protocol, see Additional file 2.

117Protocol no: SAKK 23/16 / IBCSG 57–18 / ABCSG-53

118Protocol version: final version 1.0, 6 April 2018

119EudraCT no: 2018–000372-14

120Trial type: clinical trial with other health intervention

121Categorization: risk category B according to the Swiss Human Research Act and its ordinance KlinV/Oclin

122Trial sponsor: SAKK (Swiss Group for Clinical Cancer Research), Effingerstrasse 33, 3008 Bern, Switzerland

123TAXIS was registered at ClinicalTrials.gov, on 1 May 2018, under NCT03513614 and on the Swiss National Clinical Trials Portal (SNCTP) at www.kofam.ch/de/studienportal/suche/38529/studie/42266, on 17 June 2018, under NCT03513614.

124The trial was activated on 31 July 2018. The first patient was randomized on 7 August 2018.