Abstract OT3-07-01: Update of the randomized, non-inferiority LORD trial testing safety of active surveillance for women with screen-detected low risk ductal carcinoma in situ (EORTC-1401-BCG/BOOG 2014-04, DCIS)

The introduction of population-based breast cancer screening and implementation of digital mammography have led to an increased incidence of ductal carcinoma in situ (DCIS) without a decrease in the incidence of advanced breast cancer. This suggests DCIS overdiagnosis exists.

We hypothesize that asymptomatic, low-grade DCIS can safely be managed by active surveillance. If progression to invasive breast cancer would still occur, this will be low-grade and hormone receptor positive with excellent survival rates. Also, breast-conserving treatment will still be an option, if no prior radiotherapy has been applied. Management by active surveillance also may save many low-grade DCIS patients intensive treatment.

Therefore, we will compare active surveillance with conventional treatment, being either mastectomy, wide local excision (WLE) only, or WLE plus radiotherapy, possibly followed by hormonal therapy for primary low-grade DCIS. For this, we conduct a phase III, open-label, non-inferiority, multi-center, randomized clinical trial sponsored by the European Organization for Research and Treatment of Cancer (EORTC-1401-BCG). The Dutch Centers are coordinated by the Dutch Breast Cancer Research Group (BOOG) (BOOG 2014-04). This trial is developed and implemented in close collaboration with patient advocates.

Randomization will be in a 1:1 ratio among one of the following arms: (1) active surveillance or (2) standard treatment per local policy. In total, 1,240 women (≥ 45 years) will be included without prior breast cancer, but with asymptomatic, pure, low-grade DCIS, based on a minimum of tissue harvested by biopsy from calcifications detected by population-based or opportunistic screening. Assuming 25% of randomized women qualified to enroll in the study will drop out or will be excluded from per protocol evaluation, at least 1,240 women need to be randomized to obtain the 930 patients required for the evaluation of the primary endpoint. The same follow-up scheme will be applied in both study arms, i.e. annual mammography for a period of 10 years. The primary end-point is ipsilateral invasive breast tumor-free rate at 10 years. Secondary end-points are among others: overall survival, breast cancer-specific survival, mastectomy rate, patient reported outcomes and cost-effectiveness. Accrual has started in the Netherlands in February 2017 and will start internationally in over 30 centers shortly.

Acknowledgements: This trial is funded by Pink Ribbon Netherlands, the Dutch Cancer Society and Dutch Cancer Society/Alpe d'HuZes, and Cancer Research UK.

Citation Format: Wesseling J, Elshof LE, Tryfonidis K, Poncet C, Aalders K, van Leeuwen-Stok E, Skinner V, Loo C, Winter-Warnars G, Bleiker E, Retèl V, Pijnappel R, Bijker N, Rutgers E, van Duijnhoven F. Update of the randomized, non-inferiority LORD trial testing safety of active surveillance for women with screen-detected low risk ductal carcinoma in situ (EORTC-1401-BCG/BOOG 2014-04, DCIS) [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr OT3-07-01.

Abstract P3-08-09: Impact of gene-expression profiling in patients with early breast cancer when applied outside the guideline directed indication area

Purpose

In Dutch guidelines gene expression profiles (GEP) are indicated in estrogen receptor positive early breast cancer patients in whom benefit of chemotherapy (CT) is controversial based on traditional prognostic factors alone. Aim of the current study is to assess the use and impact of GEP on administration of adjuvant CT in breast cancer patients who have according to national guidelines a clear indication to either use or withhold adjuvant chemotherapy (clinical high or low risk).

Methods

Clinical low- and high risk patients, according to Dutch breast cancer guidelines, diagnosed between 2011-2014 were selected from the Netherlands Cancer Registry (NCR). Influence of GEP use and GEP test result on CT administration was assessed with logistic regression.

Results

Overall, 26,425 patients were identified; 4.8% of patients with clinical low- risk (444/ 9,354), 7.5% of the patients with a clinical high-risk (1,281/ 17,071) received a GEP. GEP use was associated with a significantly increased odds of CT administration in clinical low-risk patients (OR=2.12 95%CI: 1.44-3.11). In clinical high-risk patients GEP use was associated with a decreased frequency of CT administration (OR=0.55, 95%CI: 0.48-0.63). Adherence to the GEP result was higher in clinical high-risk patients with a discordant GEP result as compared to clinical low-risk patients with a discordant GEP result: 71.7% vs. 52.2%, respectively.

Conclusion

GEP is frequently used outside the indicated area and significantly influenced the administration of adjuvant CT, although adherence to the test-result was limited.

Citation Format: Schreuder K, Kuijer A, Rutgers EJTh, Smorenburg CH, Van Dalen T, Siesling S. Impact of gene-expression profiling in patients with early breast cancer when applied outside the guideline directed indication area [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P3-08-09.

Abstract P1-07-02: Withdrawn

This abstract was withdrawn by the authors.

Abstract P3-12-08: Evaluation of treatment compliance during extended endocrine therapy; secondary analysis of the IDEAL trial

In the first clinical trial reports about extended endocrine therapy in early breast cancer, treatment compliance appeared as a major concern. Earlier, it was shown in the IDEAL trial that approximately 35% of all patients stopped therapy before the allocated time. This additional study was conducted to evaluate the factors contributing to early treatment discontinuation.

Methods: In the IDEAL trial, a total of 1824 patients were randomized between either 2.5 or 5 years of extended letrozole, after 5 years of any adjuvant endocrine therapy. Only eligible patients who started therapy were included in the analysis. Adverse events were collected until 30 days after last treatment dose Reasons for ending therapy were collected prospectively at the time of treatment discontinuation.

Results: The majority of early treatment discontinuation was caused by adverse events (AEs) (n=372, 20.4% of all patients, 58% of all early treatment discontinuations). The most frequently reported AEs associated to treatment discontinuation were arthralgia (n=71, 9.9% of AEs associated treatment discontinuation), fatigue (n=48, 6.7%), depression (n=47, 6.5%), hot flashes (n=47, 6.5%) and alopecia (n=39, 5.4%). Of all AEs associated to early discontinuation, 86% was grade 1 or 2 (table 1). All grade 5 events were not associated to therapy.

Table 1 - Overview of adverse events most frequently associated to early treatment discontinuation Grade 1Grade 2Grade 3Grade 4Grade 5TotalArthralgia2236121071Fatigue192610148Depression202160047Hot flashes162092047Alopecia28731039Total (all AEs)30231680136720

Furthermore, the influence of previous type of adjuvant endocrine therapy was evaluated. Of all patients initiallytreated with 5 years of tamoxifen, 29% stopped due to an AE. In contrast, patients who were treated with aromatase inhibitors during the first 5 years, either with monotherapy or after 2-3 years of tamoxifen, stopped due to AEs in 22% and 18% respectively (Pearson Chi-square p-value 0.001). The average number of AEs per patient per previous treatment group was 2.27 for tamoxifen monotherapy, 2.03 for AI monotherapy and 1.73 in the sequential group. Corrected for the number of AEs in each group, patients pre-treated with 5 years of tamoxifen had a chance of treatment discontinuation of 12.7% per AE, compared to 10.8% and 10.4% for AI monotherapy and sequential therapy respectively. Additionally, of patients that completed regular adjuvant therapy between 1 and 2 years before randomization, 34% stopped due to adverse events. In contrast, of patients that completed therapy within 6 months before randomization stopped in 19% of all cases (Pearson Chi-square p-value <0.001).

Conclusion: We have shown that adverse events are an important factor in early treatment discontinuation. Furthermore, the relation between adverse events and early discontinuation is influenced by the type of earlier therapy, with the highest rate of discontinuation for AI-naïve patients. This suggests that after 5 years of tamoxifen, patients are more inclined to stop therapy when encountering new AI-related adverse events compared to patients who were pre-treated with an AI.

Citation Format: Blok EJ, Kroep JR, Meershoek-Klein Kranenbarg E, Duijm-de Carpentier M, Nortier JWR, Rutgers EJTh, van de Velde CJH. Evaluation of treatment compliance during extended endocrine therapy; secondary analysis of the IDEAL trial [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P3-12-08.

Abstract P1-07-08: Young age and the risk of disease recurrence as assessed by the 70-gene signature – an analysis from the EORTC 10041/BIG 03-04 MINDACT trial

Purpose: Increased insight in tumor biology has revealed that not all young women are at high risk of disease recurrence. Therefore, in some patients extent of treatment could probably be safely scaled down. We aimed to evaluate the risk of breast cancer (BC) relapse according to the 70-gene signature (70-GS) result in relation to young age, in early-stage BC patients enrolled in the MINDACT trial.

Patients and Methods: The analyzed population consisted of enrolled BC patients in the MINDACT trial with available clinical (C), as per a modified version of Adjuvant!Online, and genomic (G), according to the 70-GS, risk assessments and known age (n=6693). Patients were categorized in three age groups; <45 (young), 45-55 (peri-menopausal) and >55 years (post-menopausal). Clinicopathological and treatment characteristics as well as gene expression were compared for the different age groups further split by corrected risk groups (C-low/G-low, C-low/G-high, C-high/G-low, C-high/G-high). Subsequently, the 5-year distant metastasis-free survival according to risk category was calculated.

Results: The study included 1100 patients <45 (16%), 2272 aged 45-55 (34%) and 3321 patients >55 years of age (50%). Median age of the young group was 41 (25.8-45.0) years. The young age group had a higher frequency of lymph node involvement (25% vs. 22% and 19%), poorly differentiated tumors (42% vs. 26% and 27%), ER-negative tumors (20% vs. 11% and 11%) and triple negative molecular IHC subtype (16% vs. 9% ad 8%). Median tumor size was the same across the 3 age groups (17mm). Of the 1100 young patients, 61% were C-high while the 70-GS assessed 48% as G-high. Overall, 31% were CL/GL (vs. 43% in other age groups), 9% CL/GH, 21% CH/GL and 40% CH/GH (vs. 24% and 25%).

In the discordant risk groups, chemotherapy (CT) allocation when randomized to no chemo occurred in 5% of young women as compared to 3% and 1% in the older age groups. Reason for non-compliance was 50/50 between patient refusal and PI decision.

Overall, the 5-year DMFS was 94.1% (95% CI 92.4-95.4) in <45 age group, 95.3% (95% CI 94.2-96.1) in 45-55 and 94.9% (95% CI 94.0-95.6) in >55. For the young patients, 5-year DMFS was 98.3% for the CL/GL (96.0-99.3), 97.4% in CL/GH (90.0-99.4), 95.5% in CH/GL (91.6-97.7) and 89.2% in CH/GH (85.6-92.0). In the older two age groups (45-55 and >55), the 5-year DMFS rates were 97.8% (96.5.98.6) and 97.2% (96.2-98.0) for CL/GL, 93.9% (88.8-96.7) and 94.5% (91.0-96.7) for CL/GH, 94.5% (92.0-96.3) and 95.4% (93.5-96.8) for CH/GL and 92.0% (89.2-94.1) and 90.4% (88.0-92.4) for CH/GH, respectively. With 9 events in the <45 group at a CH/GL risk, numbers were too small to evaluate chemotherapy effect in this population.

Conclusion: The use of the 70-GS reduces the proportion of patients characterized as high risk as compared to traditional clinical risk assessment (48% vs. 61%). Outcome was comparable for the 3 age categories with a very good 5-year DMFS of 95-98% in all GL groups. Performing the 70-GS provides clinically relevant information concerning the prognosis for young early-stage BC patients categorized as CH. These results add important new data to the limited available evidence on genomic expression in young BC patients.

Citation Format: Aalders K, Genbrugge E, Poncet C, Kuijer A, Pistilli B, Piccart M, Tryfonidis K, van Dalen T, Cardoso F, van 't Veer L, Rutgers E. Young age and the risk of disease recurrence as assessed by the 70-gene signature – an analysis from the EORTC 10041/BIG 03-04 MINDACT trial [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-08.

Abstract P5-14-01: Chemotherapy randomization of the EORTC 10041/ BIG 3-04 MINDACT (microarray in node-negative and 1 to 3 positive lymph node disease may avoid chemotherapy) trial

This abstract was withdrawn by the authors.

Abstract P1-03-04: Concordance of local immunohistochemistry with TargetPrint microarray based assessment of ER, PR and Her2 and BluePrint molecular subtyping in the Symphony Triple A study

PURPOSE: A decade ago intrinsic biological breast cancer subtypes have been identified which have proven to be of clinical importance in terms of outcome and response to systemic treatment. The aim of the current study is to assess concordance between breast cancer subtypes determined by local immunohistochemistry (IHC) assessment of estrogen receptor (ER), progesterone receptor (PR) and Her2-receptor status and microarray based molecular subtyping in a subset of ER+ early stage breast cancer patients.

PATIENTS AND METHODS: In this prospective observational multicenter study information on local pathology assessment and BluePrint/TargetPrint results were obtained in ER+ Dutch early stage breast cancer patients in whom a 70-gene profile (MammaPrint) was used as they were enrolled in clinical trial based on the existence of controversy regarding the additional value of adjuvant CT. Local IHC assessment of ER, PR and Her2 status were compared with microarray based assessment (TargetPrint/BluePrint) of these characteristics. Reclassification of ER and PR overexpression was assessed by a McNemars test and by Spearman correlation. Furthermore, concordance between the clinical subtypes based on local pathology (Luminal-type: ER+/PR+/Her2-; Her2-type: Her2+ disease) and molecular subtyping was assessed.

RESULTS: Between January 2013 And December 2015 660 patients, treated in 31 hospitals, were enrolled. In 564 (85%) BluePrint and/or TargetPrint was performed in addition to the 70-GS. The majority of patients had ER+/Her2- disease and TargetPrint reclassified 1% (n = 7) of patients as ER-negative (r = 0,250, p <0,001). TargetPrint reclassified 7% (n = 40) and 2% (n = 11) of patients for PR and Her2 status respectively (table 1, r = 0,580, p <0,001 for PR

Table 1. Concordance between immunohistochemistry and TargetPrint. TargetPrint result (ER, PR and Her2 resp.)  ImmunohistochemistryPositiveNegativeOverall discordance (%)p-value*Estrogenreceptor status    Positive557 (99%)6 (1%)  Negativen.a.n.a.1%n.a.Progesterone receptor status    Positive474 (96%)18 (4%)  Negative22 (31%)49 (69%)7%0,636Her2 receptor status    Positive3 (30%)7 (70%)  Negative4 (3%)546 (97%)2%0,549Equivocal0 (0%)3 (1%)  * P-value represents results of the McNemar test.). Based on IHC 545 (98%) patients were regarded as luminal-type and the remaining 2% as Her2-type. BluePrint reclassified 2% of the clinical luminal-type patients: 4 (1%) patients were reclassified as basal-type and 3 (0%) patients as Her2-type. Of the clinical Her2-type patients 80% (n=8) was reclassified by BluePrint as molecular luminal-type.

Table 2. Concordance between clinical subtyping and molecular subtyping according to BluePrint.  BluePrint resultClinical SubtypeNo. ptsLuminalBasalHer2Luminal545539 (99%)4 (1%)3 (0%)Her2108 (80%)02 (20%)Note. Overall discordance 3%.

Conclusion: In the current study we observe a high concordance between microarray-based assessment of ER, PR and Her2 and local pathology in Dutch ER+ early stage breast cancer patients. In the small subset of ER+ patients who are considered candidates for 70 GS use and who have HER2+ tumors by IHC molecular typing of HER2 status is of additional value.

Citation Format: Kuijer A, Straver M, Elias S, Smorenburg C, Wesseling J, Linn S, Rutgers E, Siesling S, van Dalen T. Concordance of local immunohistochemistry with TargetPrint microarray based assessment of ER, PR and Her2 and BluePrint molecular subtyping in the Symphony Triple A study [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-03-04.

Abstract P5-17-09: Biomarkers to distinguish hazardous from harmless ductal carcinoma in situ (DCIS) of the breast

Background. The incidence of DCIS has increased since the introduction of population-based screening. This has not resulted in a decrease in invasive breast cancer incidence, implying overdiagnosis exists. All women with DCIS are still intensively treated, by surgery, radiotherapy, and/or hormonal treatment, although only a minority will develop a subsequent invasive breast cancer. As we cannot discriminate such hazardous from harmless DCIS lesions, accurate prognostic biomarkers are urgently needed. In the current study we aim to identify molecular markers for DCIS aggressiveness, using a large population-based cohort.

Patients and methods. We used a population-based, nation-wide cohort consisting of 10,090 women treated for primary DCIS between 1989 and 2004 with a median follow-up time of 10.7 years. Within this cohort, a case-control study was set up to analyse which markers are associated with progression to invasive breast cancer. Formalin-fixed paraffin embedded (FFPE) tissue blocks were retrieved from 1580 DCIS patients who were treated by breast conserving surgery without radiotherapy (316 DCIS patients with a subsequent ipsilateral invasive breast cancer (iiBC): i.e. the "cases"; and 1264 DCIS patients without subsequent invasive breast cancer: i.e. the "controls"). A first study using this population-based cohort will involve immunohistochemistry (IHC) on 200 "cases" and 500 "controls" for an 8-marker IHC panel (ER, PR, HER2, Ki67, p16, p53, COX-2, and Annexin A1). Molecular subtypes of the DCIS and invasive breast cancer lesions will be determined and intra-individual heterogeneity will be assessed. IHC marker expression will be both compared between "cases" and " controls" as well as between DCIS lesions and its subsequent invasive breast cancer. In a second study, DNA and RNA will be isolated from these specimens, using laser microdissection, and extensive molecular profiling will be performed.

Results. We have collected FFPE tissue blocks of 287 "cases" and 1149 "controls" (86% of requested material) from 56 participating hospitals. At present, the specimens of 223 "cases" (matched DCIS and iiBC specimen) and 103 "controls" have been centrally revised for extensive morphological characteristics. Only a small part (14%) of the specimens had to be excluded from the study population. IHC staining of the tissue specimens, using the 8-marker IHC panel is ongoing.

Conclusion. Within a nation-wide cohort of 10,090 patients diagnosed with primary DCIS, we were able to collect tissue material of a representative case-control series of 200 "cases" with subsequent invasive breast cancer and 500 invasive breast cancer-free "controls". This is the first time such a large unique, unbiased DCIS series, with long-term follow-up is analysed integrating clinical, histological, and immunohistochemical data. The results will be presented at SABCS 2015.

Citation Format: Visser L, Elshof L, Groen E, van de Vijver K, Lips E, de Maaker M, Nieboer F, Schaapveld M, Rutgers E, Wesseling J. Biomarkers to distinguish hazardous from harmless ductal carcinoma in situ (DCIS) of the breast. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-17-09.

Abstract P5-17-06: Prognostic value of method of detection in primary pure DCIS

Background

Population-based mammographic screening programs have led to a substantial increase in incidence of ductal carcinoma in situ (DCIS). We assessed whether the method of detection provides prognostic information among women with DCIS detected through the Dutch screening program (screen-detected DCIS) and those with DCIS not detected within the national screening program (non-screen-detected DCIS). This could have impact on the treatment strategy of screen-detected DCIS as compared to symptomatic DCIS.

Methods

We studied a population-based retrospective cohort comprising 7,106 women aged 49-76 years with primary pure DCIS, who were treated by mastectomy or breast conserving surgery with or without radiotherapy between 1989 and 2004 in the Netherlands. Risk of subsequent ipsilateral and contralateral invasive breast cancer and overall survival among women with screen-detected (n=4,905) and non-screen-detected (n=2,201) DCIS were compared using Cox regression, adjusting for treatment (time-dependent), age (time-scale), diagnosis period and follow-up duration. Because of gradual implementation of the screening program in the Netherlands, we defined two periods based on year of DCIS diagnosis: 1989-1998 (gradual implementation of screening) and 1999-2004 (full coverage of screening).

Results

With a median follow-up of 10.5 years (interquartile range 7.7-14.0 years) 366 ipsilateral (screen-detected DCIS n=234, non-screen-detected DCIS n=132) and 380 contralateral (screen-detected DCIS n=245, non-screen-detected DCIS n=135) invasive breast cancers were diagnosed, and 1,088 of 7,106 women died (screen-detected DCIS n=603, non-screen-detected DCIS n=485). From 1989 to 2004 the number of non-screen-detected DCIS remained stable (mean 140, range 110-187 per year), whereas the number of screen-detected primary pure DCIS increased from 8 in 1989 to 596 in 2004. Ipsilateral invasive breast cancer risk was lower for screen-detected DCIS compared to DCIS not detected within the national screening program, irrespective of DCIS treatment, period of diagnosis, and follow-up duration (adjusted hazard ratio [HR] 0.74, 95% confidence interval [CI] 0.59-0.92, p < 0.01). The prognostic value of method of detection was similar across categories of treatment, period of diagnosis, and follow-up duration. The risk of contralateral invasive breast cancer did not differ between screen-detected DCIS and non-screen-detected DCIS (adjusted HR 0.89, 95% CI 0.71-1.11, p = 0.3) and neither did all-cause mortality (adjusted HR 0.91, 95% CI 0.79-1.04, p = 0.2).

Conclusion

Women with primary pure DCIS detected through the Dutch screening program had lower risk of subsequent ipsilateral invasive breast cancer, irrespective of DCIS treatment, compared to women whose DCIS was not detected within the national screening program. However, the magnitude of this risk difference does not warrant a different treatment strategy of screen-detected DCIS as compared to non-screen-detected DCIS. Having a screen-detected DCIS was not associated with risk of subsequent contralateral invasive breast cancer and all-cause mortality.

Citation Format: Elshof LE, Schaapveld M, Schmidt MK, van Leeuwen FE, Rutgers EJTh, Wesseling J. Prognostic value of method of detection in primary pure DCIS. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-17-06.

High concordance of protein (by IHC), gene (by FISH; HER2 only), and microarray readout (by TargetPrint) of ER, PgR, and HER2: results from the EORTC 10041/BIG 03-04 MINDACT trial

BACKGROUND

To investigate the correlation of TargetPrint with local and central immunohistochemistry/fluorescence in situ hybridization assessment of estrogen (ER), progesterone (PgR), and human epidermal growth factor receptor 2 (HER2) in the first 800 patients enrolled in the MINDACT trial.

PATIENTS AND METHODS

Data from local (N = 800) and central (N = 626) assessments of receptor status were collected and compared with TargetPrint results.

RESULTS

For ER, the positive agreement (the percentage of central pathology positive assessments that were also TargetPrint/local laboratory positive) for TargetPrint in comparison to centralized assessment was 98% with a negative agreement (the percentage of central pathology negative assessments that were also TargetPrint/local laboratory negative) of 96%. For PgR, the positive agreement was 83% with a negative agreement of 92%. For HER2, the positive agreement was 75% with a negative agreement of 99%. Even though the local assessment showed higher positive agreement for PgR (89%) and higher positive agreement for HER2 (85%), the range of discordant local versus central assessments were as high as 6.7% for ER, 12.9% for PgR, and 4.3% for HER2.

CONCLUSION

TargetPrint and local assessment of ER, PgR, and HER2 show high concordance with central assessment in the first 800 MINDACT patients. However, there are concerns about the higher discordance rates for some local sites. TargetPrint can improve the reliability of hormone receptor and HER2 testing for those centers with a lower rate of concordance with the reference laboratory, with the limitation of a positive agreement of 75% for HER2. TargetPrint consequently has important implications for treatment decisions in clinical practice and is a reliable alternative to local assessment for ER.

CLINICAL TRIALS NUMBER

NCT00433589.

Abstract OT2-1-03: Preoperative accelerated partial breast irradiation trial (PAPBI); defining radiosensitivity

Background and aim of the study:

A.The ongoing Preoperative Accelerated Partial Breast Irradiation (PAPBI) trial (NCT01024582) is based on the rationale that three-dimensional conformal external-beam radiation (3D-CRT) leads to more dose homogeneity compared with brachy-or intraoperative radiotherapy (RT). By irradiating preoperatively this can lead to more accurate tumor delineation and smaller irradiated volumes. As the tumor remains in situ during irradiation, more precise delivery of the radiation dose is guaranteed with CT cone beam linear accelerators, avoiding the uncertainties of the original tumor position in the operation cavity as is the case in postoperative RT. Tumor excision 6 weeks after RT removes the high dose volume tissue and can lead to better cosmesis.

B. By assessing tumor response to radiotherapy, an additional goal of the study is to develop a gene expression profile that predicts breast cancer radiosensitivity. This gene signature of breast radiosensitivity would further design optimal treatment strategies for individual breast cancer patients treated with BCT.

Inclusion citeria:

Patients 60 years or older with a cT< = 3cm, ductal carcinoma (no in situ component), unifocal on mammogram and MRI, pN0(sn) (sentinel node procedure before RT), will be treated by preoperative RT (CTV = GTV + 2 cm, 10 × 4 Gy IMRT/VMAT over two weeks). Six weeks after pre-operative RT, a wide local excision will be performed. Skin toxicity and fibrosis is scored using EORTC/RTOG criteria. Patients are followed during RT and on a 3-monthly basis. Cosmesis is scored and photographs are taken for analysis (BCCT.core project score).

To study radiosensit ivity, gene expression profiling from RNA and DNA isolated from biopsies (mRNA gene expression profiles, the miRNA expression profiles and the DNA copy number changes) taken of the tumor before radiotherapy and at time of surgery will be correlated with response to radiotherapy, defined as pathologic response at the time of the lumpectomy. Response of the tumor will be evaluated by MRI scan and PET (before radiotherapy and before surgery) and classical pathology.

Endpoint :

The main objective is to investigate the impact of a short fractionated schedule given preoperatively on cosmesis and breast fibrosis. Therefore, it is anticipated that the percentage of moderate or severe fibrosis will decrease from 27% as found in the boost arm of the EORTC boost-no boost trial to 15% (Collette et al EJC 2008). The total sample size of 120 patients will provide in excess of 80% power to detect the difference between the null hypotheses (a rate of fibrosis of 27%) and the alternative hypothesis (a rate of fibrosis of 15%) with an exact binomial test at 0.05 2-sided significance level. In addition, the 2-sided 95% confidence interval for the proportion of patients without local recurrence will extend 0.035 from the observed proportion for an expected proportion of 96%. An additional objective is to build a classifier (genomic or proteomic or any kind of molecular signature) to identify responders and non-responders. A total of 120 patients will be included in the study. The main analysis will include 60 patients in the training set and 60 in the validation set.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr OT2-1-03.

Abstract P3-05-02: Pathological assessment of discordant cases for molecular (BluePrint and MammaPrint) versus clinical subtypes for breast cancer among 621 patients from the EORTC 10041/BIG 3–04 (MINDACT) trial

Background: Biology has become the main driver of breast cancer therapy. Intrinsic biological subtypes by gene expression profiling have been identified. Pathology can be used to define surrogates of these subtypes but these are not always concordant, which may lead to different treatment plans. We investigated the concordance between BluePrint and MammaPrint (microarray-based) breast cancer subtypes and pathological surrogates (based on ER, PR, HER2 & Ki67).

Methods: Using available data (centrally assessed pathology and genomics) from the MINDACT pilot phase [Rutgers et al. EJC 2011] 621 tumors were analyzed. Patients were classified according to 4-category based pathology (ER, PR, HER2 and Ki67); additionally, classification was performed adhering to the recent St. Gallen recommendations [Goldhirsch et al. 2011], which recognizes an additional category (Luminal B HER2+). Based on BluePrint and MammaPrint 4 subtypes are formed: Luminal A (Luminal-type/MammaPrint Low Risk); Luminal B (Luminal-type/MammaPrint High Risk); HER2-type; and Basal-type. This study is an analysis of discordant patient groups (i.e. clinical HER2+/BluePrint Luminal-type; clinical Hormone Receptor (HR)-positive/BluePrint Basal-type) providing comparison of centrally assessed tumor heterogeneity as well as comparison of quantified ER, PR and HER2 results.

Results: Ki67 is often used as biomarker to distinguish Luminal A from Luminal B subgroups. The concordance between MammaPrint and centrally assessed Ki67 in Luminal-type patients is 71%, with a κ score of 0.35 (95% CI 0.26–0.45) indicating that Ki67 and MammaPrint cannot reliably substitute for each other. There is a relatively large group of clinical HER2+ cases that are BluePrint Luminal-type (29 out of 76; 38%) indicating that tumor expression of the Luminal profile is dominant compared with expression of the HER2 profile. These patients have high IHC ER results and fall into the group that St Gallen separately defines as Luminal B HER2-type. These patients may have lower response to trastuzumab [von Minckwitz et al. JCO 2012]. 12 out of 76 BluePrint Basal-type patients are clinical HR+. These patients have low centrally assessed IHC ER and PR expression (≥1% and <10%).

Conclusions: Marked differences are observed between BluePrint and MammaPrint (microarray based) breast cancer subtypes and centrally re-assessed pathological surrogates (based on ER, PR, HER2 & Ki67). The greatest discordance is seen in the sub-stratification of Luminal patients, and in the HR+/HER2+ patients. The observed subtype discrepancies may have an important impact on treatment decision making.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-05-02.

P1-07-06: High Concordance of Protein (by IHC), Gene (by FISH; HER-2 Only) and Microarray Readout (by TargetPrint) of ER/PR/HER2: Results from the MINDACT Trial

Background

Previously, the micro-array readout of ER, PR and HER2 by TargetPrint was shown to be strongly correlated with high quality immunohistochemistry (IHC)/FISH assessment, especially for ER and HER2. Concordance rates were 93% (k=0.79) for ER; 83% (k=0.65) for PR and 96% for HER2 (k=0.88) in 636 patients (Roepman et al., Clin Cancer Res, 2009).

This study analysis was undertaken to further determine the correlation of microarray readout with IHC/FISH assessment both locally and centrally determined in the 1st 800 pts enrolled in the MINDACT trial. This work is essential to determine the quality of biological data in the two risk assessment methods used in MINDACT based upon which adjuvant chemotherapy decision is made, in order to exclude bias.

Methods: ER/PR/HER2 IHC assessment was performed on the 1st 800 primary breast cancers (BC) of pts enrolled in the MINDACT study. The assessment was performed locally at each center (n=800) and by central review at the laboratory of the European Institute of Oncology (n=626). A tumor was classified positive for ER and PR when 1% of tumor cells showed positive staining. HER2 IHC status was scored as 0, 1+, 2+ or 3+; a score of 3+ was considered positive. In 2+ cases FISH was performed to assess final HER2 status. Gene expression data for ER, PR and HER2 were obtained by TargetPrint stratified as receptor positive or negative using previously determined and validated thresholds for ER, PR and HER2 mRNA levels (n=800).

Results: Comparison of local assessment (IHC & FISH for HER2) with central review indicated highly similar results for receptor readout with a concordance of 98% (k=0.90) for ER; and 96% for HER2 (k=0.80) and slightly lower for PR (90% (k=0.72)).

Comparison of central assessment (IHC & FISH for HER2) with micro array readout by TargetPrint indicated highly similar results for receptor readout with a concordance of 97% (k=0.88) for ER and 95% for HER2 (k=0.76). For PR the concordance was lower but still quite acceptable (85% (k=0.62)).

Conclusion: Local and centrally assessed ER, PR and HER2 status in the first 800 MINDACT patient samples indicate a high level of quality for pathology in the local participating hospitals. These results exclude any bias induced by a lower quality of “traditional” pathology results as compared to the centrally assessed MammaPrint, both used for risk assessment and adjuvant chemotherapy decision in the MINDACT trial. The microarray-based assessment of ER, PR and HER2 gives results comparable to IHC & FISH and provides an objective and quantitative assessment of tumor receptor status. These results indicate that TargetPrint can serve as a second pathology assessment for locally assessed parameters, especially since TargetPrint is part of a multi-profile platform for breast cancer treatment management. This work was funded by the Breast Cancer Research Foundation and the EU Framework Program VI.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-07-06.

The Role of FDG PET/CT in Patients with Locoregional Breast Cancer Recurrence and a Comparison to Conventional Imaging

Background: The prognosis and clinical management of patients with locoregional recurrence of breast cancer depends on the extent of the disease. Conventional work-up, consisting of bone scintigraphy, liver ultrasound, X-ray or CT of the thorax and MRI of the thoracic wall, may underestimate the local extent or miss the presence of distant metastases. The aim of this study was to evaluate the impact of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) on clinical management in patients with locoregional breast cancer recurrence amenable for locoregional treatment and compare the results with conventional imaging.Material and Methods: We performed a retrospective analysis of all patients referred for FDG PET/CT with locoregional breast cancer recurrence between January 2006 and August 2008. PET/CT findings were compared with results of additional imaging, pathology and follow-up. The impact of FDG PET/CT findings on clinical management of patients, was evaluated on the basis of clinical decisions obtained from patient files.Results: 56 patients were included. In 93% of the patients FDG PET/CT confirmed the locoregional recurrence. In 32 patients (57%) FDG PET/CT revealed additional tumour localisations. Distant metastases were detected in 11 patients on conventional imaging and in 23 patients on FDG PET/CT images (p<0,01). In 25 patients (45%), FDG PET/CT detected additional lesions not visible on conventional imaging. FDG PET/CT had an impact on clinical management in 27 patients (48%) by detecting more extensive locoregional disease or distant metastases. In 20 patients (36%) unnecessary mutilating surgery was prevented and treatment was changed to palliative systemic treatment. The sensitivity, specificity, accuracy, positive and negative predictive values of FDG PET/CT were respectively 97%, 92%, 95%, 94% and 96%.Discussion: FDG PET/CT, in addition to conventional imaging, plays an important role in staging patients with locoregional breast cancer recurrence since its result changed the clinical management in almost half of the patients. FDG PET/CT could potentially replace conventional staging imaging in patients with a locoregional breast cancer recurrence.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5011.

Prediction of Prognosis and Treatment Benefit by the Adjuvant! Web-Based Tool: Calibration and Discriminatory Accuracy in 5380 Dutch Breast Cancer Patients

Adjuvant! online (AOL) is a web-based tool that calculates individualized 10-year survival probabilities and predicted benefit of adjuvant systemic therapy. The AOL model was constructed using data from breast cancer patients recorded in the US population-based SEER registry and has not yet been validated in any large European series. European patients may differ from those in the US with regard to prognostic factors and intrinsic tumor characteristics. Therefore the aim of our study was to validate AOL in Dutch patients, investigating both its calibration and discriminatory accuracy.Patients who were treated at the Netherlands Cancer Institute for breast cancer between 1987 and 1998 were selected according to the following criteria: pT1-3, pN0-3, M0, definitive primary surgery, and complete axillary staging. Median follow-up was 11.7 years; >96% of patients had more than 10 years follow-up. Clinicopathological characteristics and adjuvant treatment data were entered into the AOL version 8.0 batch processer, with blinding to outcome.For all 5380 patients included, the 10-year observed overall survival (69.0%) and breast cancer-specific survival (78.6%) rates as compared to the 10-year rates predicted by AOL (69.1% and 77.8%, respectively) were within 1% and not statistically different. Moreover, differences between predicted and observed outcomes were within 2% for most relevant clinicopathological subgroups. Subsets of patients for which there was a discrepancy between the predicted outcomes by AOL and the actual observed outcomes included patients under 40 years, for which both the predicted overall and breast cancer-specific survival were overly optimistic (Pred – Obs: by 4.2% and 4.7% respectively; p<0.05).The concordance(c)-index, which indicates discriminatory accuracy at the individual level, was 0.71 for BCSS in the entire cohort, indicating that AOL performs better than chance. The c-index for a multivariate Cox regression model fitted to the 5380 patients using a backward approach including clinicopathological variables was also 0.71 for BCSS, suggesting AOL optimally incorporates the information in these variables.In order to validate the AOL model, we evaluated both its calibration (goodness of fit) and discriminatory accuracy. Although the latter is rarely examined, it is of paramount importance to justify the use of prognostic models for clinical outcome prediction. We showed that in addition to good calibration, the model was capable of separating individuals with a poor outcome from those with a good outcome with moderate power. In conclusion, AOL accurately predicted 10-year outcomes in this first large scale European validation study and is of use for adjuvant treatment-decision making. Additional inclusion of biological markers in AOL in the future is expected to lead to increased discriminatory accuracy.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4031.

Evaluation of the effect of screening on the detection of good and poor prognosis breast cancers

1525

Background: Since the advent of screening, breast cancer incidence has significantly increased in the United States as well as other countries. SEER data from the US shows that the increase is largely in localized disease without a concomitant or significant decrease in regionalized disease. In addition, the increase has occurred essentially in women over 50. We asked whether molecular characterization of tumors would shed light on the types of tumors detected in screening and in locally advanced breast cancers (LABC). Methods: We identified two groups of patients from European studies before and after the introduction of population based screened (in 1995) where the 70-gene prognosis test (MammaPrint) results were available. The first source is the European Validation Study (EVS) (Buyse et al, J Natl Cancer Inst. 2006). The second was a prospective implementation trial of MammaPrint in the Netherlands (RASTER study) (Bueno de Mesquita et al, Lancet Oncol. 2007). The source of LABC patients comes from the multicenter I-SPY TRIAL (CALGB 150007/150012) who underwent routine screening (prior mammogram within 2 years of diagnosis). Results: In women age 50–60, the fraction of cancers that were MammaPrint good prognosis were 40% and 60% prior to and after the introduction of screening, respectively. For women 50–60 who were actually undergoing screening, the fraction of tumors that were MammaPrint good prognosis was 67%. For patients aged 30–40 with stage I and II lymph node-negative disease, who did not undergo screening in either period, the fraction of MammaPrint good prognosis tumors did not change (29% and 31%, for EVS and Raster, respectively). For patients with LABC, from the I-SPY TRIAL, the fraction of tumors that are MammaPrint good prognosis is 7% and 29% for women aged 30–40 and 50–60, respectively. Of the LABC patients undergoing screening, 80% presented clinically during the interval between routine mammograms. Conclusions: Current screening programs are increasing the burden of low-risk cancers. Screening programs should consider including molecular profiles at the time of diagnosis to reduce overtreatment. The majority of LABCs have high-risk molecular profiles and do not present as screen detected cancers. New strategies are needed for early detection of LABC.

[Table: see text]

Benefit of the 70-gene profile for widely used guidelines: an answer to increased selection for adjuvant chemotherapy in breast cancer

Abstract #1084

Introduction:
 Risk assessment for breast cancer patients differs substantially among treatment guidelines. The NCCN, St.Gallen, Adjuvant!Online, and Dutch 2008 guidelines are less restrictive in comparison to the 2004 Dutch guidelines and Nottingham Prognostic Index, when selecting patients for adjuvant systemic treatment. The Dutch Institute for Healthcare Improvement (CBO) has introduced slight changes in the concept 2008 guidelines. Adjuvant systemic treatment is only advised when the absolute 10-years survival benefit is 5% or more. The new recommendations for adjuvant systemic treatment are based on survival tables used in the Adjuvant! software. The changes affect only node-negative patients aged 35 years: since 2008 patients with G2/G3 tumors >1cm or every tumor >2cm are advised to undergo adjuvant systemic therapy, while in 2004 this was the case for tumors G3>1cm, G2>2cm or every tumor >3cm.
 Patients and Methods:
 Risk was assessed for 427 lymph-node negative (LNN) patients in the prospective RASTER-study (Bueno-de-Mesquita, 2007) and 151 LNN patients in the NEJM-series (van de Vijver, 2002). Clinical risk was calculated using the 2004 and 2008 Dutch guidelines. Genetic risk was assigned according to the result of the Amsterdam 70-gene signature. Survival analyses were done according to the univariate Kaplan-Meier-method.
 Results:
 Instead of 57% in 2004, in 2008 only 24% of patients were assigned to clinical low risk in the RASTER study, and 141 patients (33%) changed from low to high risk (p<0.001). The rate of discordant findings between clinical assessment and 70-gene signature increased from 30% in 2004 to 41% in 2008 (p<0.001, table 1). Similar results were obtained for the 151 patients of the NEJM-series (30% and 35% discordance for the 2004 and 2008 CBO guidelines, respectively). At 10 years follow-up, differences between the high and low risk categories according to the 2004, 2008 guidelines and the 70-gene profile, were best predicted by the latter: distant-disease-free survival (DDFS) log-rank p=0.002, p=0.11 and p<0.001 respectively.
 
 Conclusion:
 As adjuvant systemic treatments become more effective, guidelines become less restrictive, resulting in more patients being selected for adjuvant therapy. Since newer treatment guidelines do not better predict survival, the magnitude of the 70-gene profiles benefit, which is defined as proportion of patients in whom over- and undertreatment can be avoided, depends on present guidelines for risk assessment and shows a remarkable increase.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 1084.

Guidelines on the standards for the training of specialised health professionals dealing with breast cancer

According to EUSOMA position paper 'The requirements of a specialist breast unit', each breast unit should have a core team made up of health professionals who have undergone specialist training in breast cancer. In this paper, on behalf of EUSOMA, authors have identified the standards of training in breast cancer, to harmonise and foster breast care training in Europe. The aim of this paper is to contribute to the increase in the level of care in a breast unit, as the input of qualified health professionals increases the quality of breast cancer patient care.

The management of lobular carcinoma in situ (LCIS). Is LCIS the same as ductal carcinoma in situ (DCIS)?

Lobular carcinoma in situ was first described over 60 years ago. Despite the long history, it continues to pose significant difficulties in screening, diagnosis, management and treatment. This is partly due its multi-focal and bilateral presentation, an incomplete understanding of its biology and natural history and perpetuation of misconceptions gathered over the last decades. In this review, the working group on behalf of EUSOMA has attempted to summarise the current thinking and management of this interesting lesion.

Prediction of response to neoadjuvant chemotherapy in breast cancer

20005

Background: Neoadjuvant chemotherapy is increasingly employed in operable breast cancer. Our initial studies on a cDNA array platform failed to identify gene expression patterns predicting response to neoadjuvant chemotherapy in breast cancer patients (J Clin Oncol 23:3331, 2005). Now we included more patients and used oligo microarrays. Methods: Patients with operable or locally advanced breast cancer were included in a randomized phase II study or received neoadjuvant chemotherapy off protocol. All except 7 patients began chemotherapy with 3 courses of dose-dense adriamycin and cyclophosphamide (ddAC) and response was evaluated by MRI. Patients with a response and a HER2-positive tumor were then randomized between either 3 additional courses of ddAC or six weekly courses of carboplatin, paclitaxel and trastuzumab (CPT). Patients without response were switched to CPT. Patients with HER2-negative tumors were randomized between 3 courses of either ddAC or capecitabine and docetaxel (CD). After evaluation, patients without response were switched to the alternative treatment arm. From all patients 14G core needle biopsies were taken before treatment and total RNA was isolated. Amplified mRNA was labeled and hybridized to 35k human oligo microarrays from our microarray facility. Results: So far, 77 patients have been included into the study. From 48 of these, good quality RNA from tissue with >50% tumor cells was isolated. 43 patients had received ddAC as initial chemotherapy; 32 of these had not been switched to another regimen. In a training set containing 11 pathological complete remissions (pCR) and 9 non-responders (NR) we could separate these groups by using 20 genes in a supervised classification and a 9-step cross validation. These results could be validated in an independent set of 11 samples (6 pCR, 5 NR). From 10 out of 11 samples, response status could be predicted correctly, independent from the treatment regimen. Although ER-positive tumors have a lower pCR rate than ER-negative ones, the steroid hormone receptors were not present in the classifier. Conclusions: We conclude that it should be possible to identify a reliable gene expression profile associated with response to adriamycin based neoadjuvant chemotherapy in breast cancer.

No significant financial relationships to disclose.

Early detection of breast and ovarian cancer in families with BRCA mutations

Women at risk of breast and ovarian cancer due to a genetic predisposition may opt for preventive surgery or surveillance. The aim of this study was to determine the effectiveness of surveillance in families with a BRCA mutation. Sixty-eight BRCA-families underwent surveillance using annual mammography, transvaginal ultrasound, and estimation of CA125. Two hundred and two women had at least one breast examination, and 138 at least one examination of the ovaries. After a mean follow-up of 33 months, breast cancer was detected in 21 women, four with lymph node metastases. After a mean follow-up of 37 months, six advanced ovarian cancers were detected. The percentage of metastatic breast cancers in the current study appeared to be acceptable. However, because these women have a high-risk of developing breast cancer, they still have a substantial risk of developing metastatic disease under surveillance. Surveillance for ovarian cancer was not effective.

Predictors of patients' choices for breast-conserving therapy or mastectomy: a prospective study

A study was undertaken to describe the treatment preferences and choices of patients with breast cancer, and to identify predictors of undergoing breast-conserving therapy (BCT) or mastectomy (MT). Consecutive patients with stage I/II breast cancer were eligible. Information about predictor variables, including socio-demographics, quality of life, patients' concerns, decision style, decisional conflict and perceived preference of the surgeon was collected at baseline, before decision making and surgery. Patients received standard information (n=88) or a decision aid (n=92) as a supplement to support decision making. A total of 180 patients participated in the study. In all, 72% decided to have BCT (n=123); 28% chose MT (n=49). Multivariate analysis showed that what patients perceived to be their surgeons' preference and the patients' concerns regarding breast loss and local tumour recurrence were the strongest predictors of treatment preference. Treatment preferences in itself were highly predictive of the treatment decision. The decision aid did not influence treatment choice. The results of this study demonstrate that patients' concerns and their perceptions of the treatment preferences of the physicians are important factors in patients' decision making. Adequate information and communication are essential to base treatment decisions on realistic concerns, and the treatment preferences of patients.

Treatment of operable breast cancer in the elderly: a randomised clinical trial EORTC 10851 comparing tamoxifen alone with modified radical mastectomy

For treatment of early breast cancer in older women, little evidence is available from randomised trials. We conducted a randomised trial comparing modified radical mastectomy (MRM) with tamoxifen (TAM) as the sole initial therapy in 164 patients aged >/=70 years with operable breast cancer. 82 were treated by MRM and 82 with TAM. Survival curves were estimated using the Kaplan-Meier method: multivariate analyses were performed using the Cox's proportional hazards model. Endpoints included survival, time to first relapse or progression, loco-regional progression, time to distant progression and progression-free survival. After a median follow-up of approximately 10 years, there was a significantly decreased time to progression in the TAM only group (logrank P<0.0001) and significantly shorter time to local progression within the TAM group (logrank P<0.0001). Overall survival of the two groups was similar. The results indicate that tamoxifen alone leads to an unacceptably high rate of local progression or relapse.

Randomised trial of high-dose chemotherapy and haemopoietic progenitor-cell support in operable breast cancer with extensive axillary lymph-node involvement

BACKGROUND

Uncontrolled studies suggest that high-dose chemotherapy is beneficial in patients with breast cancer and multiple metastases to the axillary lymph nodes. Many physicians accept this treatment as standard care. We aimed to assess adjuvant high-dose chemotherapy in breast cancer in a phase II randomised trial.

METHODS

97 women aged younger than 60 years, who had breast cancer with extensive axillary-node metastases (confirmed by a tumour-positive infraclavicular lymph-node biopsy), received three courses of up-front chemotherapy (FE120C). This regimen consisted of cyclophosphamide 500 mg/m2, epirubicin 120 mg/m2, and 5-fluorouracil 500 mg/m2 once weekly for 3 weeks. After surgery, stable patients or those who responded to chemotherapy were randomly assigned conventional therapy (fourth course of FE120C, followed by radiation therapy and 2 years of tamoxifen [40 patients]) or high-dose therapy (identical treatment but an additional high-dose regimen and peripheral-blood progenitor-cell [PBPC] support after the fourth FE120C course [41 patients]). This high-dose regimen comprised cyclophosphamide 6 g/m2, thiotepa 480 mg/m2, and carboplatin 1600 mg/m2. The primary endpoint was overall and disease-free survival. All analyses were by intention to treat.

FINDINGS

No patients died from toxic effects of chemotherapy. With a median follow-up of 49 (range 21-76) months, the 4-year overall and relapse-free survivals for all 97 patients were 75% and 54%, respectively. There was no significant difference in survival between the patients on conventional therapy and those on high-dose therapy.

INTERPRETATION

High-dose therapy is associated with substantial cost and acute toxic effects, but also has potentially irreversible long-term effects. Until the benefit of this therapy is substantiated by large-scale phase III trials, high-dose chemotherapy should not be used in the adjuvant treatment of breast cancer, apart from in randomised studies.

Comparison of the surgical procedures for breast conserving treatment of early breast cancer in seven EORTC centres

The aim of this study was to develop a standardised surgical report for breast-conserving procedures, supporting the systematic documentation of the different aspects of the surgery. The surgical procedure for tumourectomy and axillary clearance was translated into a series of steps that could be quantitatively documented. This description was submitted twice to a group of surgeons from different departments to ensure that all steps that are considered to have relevance for outcome were included and that no superfluous data were collected. After two corrective phases, a first test format was developed. Between February 1993 and May 1994, seven surgical departments, participating in EORTC trials, completed this questionnaire for a number of their patients. The data collected related to general information on the department, the tumour excision itself, the axillary dissection and, in a later phase, on pathology. 269 questionnaires (264 tumour excisions, 259 axillary dissections and 189 pathology reports) were collected and analysed. Even though the participating departments were involved in a single trial on breast-conserving surgery and had previously developed regular contacts about the practical aspects of treatment, many differences were detected. In general, variations were found in the waiting time between treatment prescription and execution, experience of the surgeon, duration of the procedure, and the use of prophylactic antibiotics. Also, in the practical execution of the procedure, major variations in the type of incision, width of tumour excision, closure of the breast tissue and skin, the use of frozen sections and the extent of the axillary dissection were found. The most relevant differences and their possible consequences are discussed. It has been proven possible and feasible to document quantitatively a surgical procedure. The fact that within a group of surgeons participating in the same clinical trials, many differences in the surgical techniques are observed, stresses the need to reach a consensus on a stricter set of guidelines for breast-conserving procedures and their documentation, especially when conducting clinical trials.