Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

PREDICT underestimates survival of patients with HER2-positive early-stage breast cancer

PREDICT underestimates survival of patients with HER2-positive early-stage breast cancer www.nature.com/npjbcancer ARTICLE OPEN PREDICT underestimates survival of patients with HER2-positive early-stage breast cancer 1,2 3 4 5 6 7 8 Elisa Agostinetto , Lieveke Ameye , Samuel Martel , Philippe Aftimos , Noam Pondé , Christian Maurer , Sarra El-Abed , 9 10 11 12 13 14 15 Yingbo Wang , Malou Vicente , Saranya Chumsri , Judith Bliss , Judith Kroep , Marco Colleoni , Fausto Petrelli , 16,17 11 10 3 1 Lucia Del Mastro , Alvaro Moreno-Aspitia , Martine Piccart , Marianne Paesmans , Evandro de Azambuja and 16,17 Matteo Lambertini The prognostic performance of PREDICT in patients with HER2-positive early breast cancer (EBC) treated in the modern era with effective chemotherapy and anti-HER2 targeted therapies is unclear. Therefore, we investigated its prognostic performance using data extracted from ALTTO, a phase III trial evaluating adjuvant lapatinib ± trastuzumab vs. trastuzumab alone in patients with HER2-positive EBC. Our analysis included 2794 patients. After a median follow-up of 6.0 years (IQR, 5.8–6.7), 182 deaths were observed. Overall, PREDICT underestimated 5-year OS by 6.7% (95% CI, 5.8–7.6): observed 5-year OS was 94.7% vs. predicted 88.0%. The underestimation was consistent across all subgroups, including those according to the type of anti HER2-therapy. The highest absolute differences were observed for patients with hormone receptor negative-disease, nodal involvement, and large tumor size (13.0%, 15.8%, and 15.3%, respectively). AUC under the ROC curve was 73.7% (95% CI 69.7–77.8) in the overall population, ranging between 61.7% and 77.7% across the analyzed subgroups. In conclusion, our analysis showed that PREDICT highly underestimated OS in HER2-positive EBC. Hence, it should be used with caution to give prognostic estimation to HER2-positive EBC patients treated in the modern era with effective chemotherapy and anti-HER2 targeted therapies. npj Breast Cancer (2022) 8:87 ; https://doi.org/10.1038/s41523-022-00452-8 INTRODUCTION diagnosed after 2004, to ensure that patients did not receive trastuzumab . A subsequent validation was done in 2012 in a The addition of trastuzumab to adjuvant chemotherapy has British Columbia Canadian cohort . This study demonstrated that dramatically improved the outcomes of patients with HER2- the inclusion of the HER2 status allowed the model to perform positive early breast cancer, reducing the risk of mortality by more 1 better than the previous PREDICT version and Adjuvant! Online in than 30% . Despite the undoubted benefit of adjuvant therapy, estimating overall and breast-cancer-specific survival . several clinical questions remain open. Approximately 25% of Although the use of PREDICT is recommended to aid decision patients still experience recurrence up to 10 years from diagnosis, making in the adjuvant setting , its prognostic role in HER2- and further research efforts are needed to better refine patient positive early breast cancer patients treated with modern selection for adopting escalation or de-escalation treatment chemotherapy and anti-HER2 therapies remains unclear. We 2,3 strategies . aimed to investigate the prognostic performance of PREDICT in PREDICT (www.predict.nhs.uk) is a publicly available, online tool patients with HER2-positive early breast cancer who received that helps to predict the individual prognosis of patients with early trastuzumab-based therapy started concurrently with chemother- breast cancer and to show the impact of adjuvant treatments apy within the ALTTO trial. The ALTTO trial is the largest adjuvant administered after breast cancer surgery. It uses traditional study ever conducted in the field of HER2-positive early breast clinical-pathological factors, and it is aimed to support clinical cancer and, including at least 5-year follow-up data from all decision making in the adjuvant setting. The original version of patients , represented a unique opportunity to investigate the PREDICT (v.1.0) was derived from cancer registry information on reliability and prognostic performance of PREDICT in women with 5,694 women treated in East Anglia from 1999–2003, and was HER2-positive disease. subsequently validated in several datasets of patients with breast 4,5 cancer . In 2011, the model was updated to include HER2 status. RESULTS Estimates for the prognostic effect of HER2 status were based on an analysis of 10,179 cases collected by the Breast Cancer Out of 8381 patients included in the ALTTO trial, 2836 were Association Consortium (BCAC), none of which had been treated with chemotherapy and concurrent trastuzumab-based 1 2 Academic Trials Promoting Team, Institut Jules Bordet and l’Université Libre de Bruxelles (U.L.B), Brussels, Belgium. Humanitas University, Department of Biomedical 3 4 Sciences, via Rita Levi Montalcini 4, 20090 Pieve Emanuele, Milan, Italy. Data Center, Institut Jules Bordet, Brussels, Belgium. Department of Hemato-Oncology, CISSS Montérégie-Centre/Hôpital Charles-Le Moyne, Université de Sherbrooke, Greenfield Park, QC, Canada. Clinical Trials Conduct Unit, Institut Jules Bordet – Université Libre de 6 7 Bruxelles, Brussels, Belgium. Clinical Oncology Department, AC Camargo Cancer Center, São Paulo, Brazil. University of Cologne, Department I of Internal Medicine, Center for 8 9 Integrated Oncology Aachen Bonn Cologne Dusseldorf, Cologne, Germany. Breast International Group (BIG), Brussels, Belgium. Novartis Pharma AG, Basel, Switzerland. 10 11 12 Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium. Robert and Monica Jacoby Center for Breast Health, Mayo Clinic, Jacksonville, FL, USA. The Institute of Cancer Research, Clinical Trials & Statistics Unit, London, UK. Department of Medical Oncology, Leiden University Medical Center, P.O. Box 9600, 2300 RC, Leiden, 14 15 16 The Netherlands. IEO European Institute of Oncology, IRCCS, Milan, Italy. Oncology Unit, ASST Bergamo Ovest, Treviglio (BG), Italy. Department of Medical Oncology, U.O. Clinica di Oncologia medica, IRCCS Ospedale Policlinico San Martino, Genova, Italy. Department of Internal Medicine and Medical Specialties (DiMI), School of Medicine, University of Genova, Genova, Italy. email: matteo.lambertini@unige.it Published in partnership with the Breast Cancer Research Foundation 1234567890():,; E. Agostinetto et al. (>50 mm) (15.3%), and high number of nodes (≥4 positive lymph Patients enrolled in ALTTO nodes) (15.8%). (n=8381) Discrimination AUC under the ROC curve was 73.7% (95% CI 69.7–77.8) in the Patients excluded (n=5545) overall population (Fig. 3). because assigned to lapatinib This finding of suboptimal discriminatory accuracy was only arm, and/or to design 1 of consistent across all subgroups, ranging from 61.7% (in patients chemotherapy administration with ≥4 positive lymph nodes) to 77.7% (in patients receiving trastuzumab alone as anti-HER2 therapy) (Table 3). The lowest discriminatory accuracy was observed for patients with high number of nodes (≥4 and 1–3 positive lymph nodes) (Supple- Patients potentially eligible for the present ALTTO sub- mentary Figs 1 and 2), and for patients receiving a non- analysis anthracycline-based chemotherapy (61.7%, 64.8%, and 65.2%, (n=2836) respectively). The highest discriminatory accuracy was observed for patients with negative lymph nodes (Supplementary Fig. 3) and for patients receiving trastuzumab alone as anti-HER2 therapy Patients excluded (n=42) because: (77.3% and 77.7%, respectively). - Age < 25 years old (n=7) - No tumor size available (n=13) - No nodal status available DISCUSSION (n=22) To the best of our knowledge, PREDICT is the only publicly available, free, online tool developed to predict individual prognosis in the specific population of patients with HER2- positive early breast cancer based on traditional and easily Patients included in the present retrieved clinical-pathological factors including HER2. In our analysis ALTTO analysis, PREDICT highly underestimated patients’ OS; this (n=2794) finding was consistent across all patient subgroups, with highest absolute differences for patients with hormone receptor-negative Fig. 1 STROBE flow-chart. This figure illustrates the patient disease, nodal involvement, and large tumor size. In terms of selection process. discrimination, the accuracy of PREDICT was overall low, with the lowest discriminatory accuracy observed in patients with nodal involvement (≥4 and 1–3 positive lymph nodes), and in patients therapy and were potentially eligible for the present analysis. In 42 receiving non-anthracycline-based chemotherapy. patients, the PREDICT algorithm was not evaluable (due to age of The low performance of this tool raises several questions about the patient <25 years old [n = 7], missing tumor size [n = 13], or the reliability of PREDICT to give prognostic estimation in HER2- missing lymph nodes status [n = 22]). Therefore, 2794 patients positive early breast cancer patients. To potentially explain the were included in the present analysis (Fig. 1). reasons for the underestimation of patients’ OS, we can speculate Most patients (71%) were aged between 41 and 64 years (Table whether the population used to validate this prognostic tool 1). Twenty-five percent of patients had negative nodal status, 45% accurately mirrors the real-world population of patients with had a tumor size ≤2 cm and 58% had hormone receptor-positive HER2-positive disease treated in the modern era with effective disease. Regarding administered treatments, 88% underwent an chemotherapy and anti-HER2 targeted therapies. The prognostic anthracycline-based chemotherapy regimen (design 2). The effect of HER2 status was evaluated and incorporated in the majority of patients with hormone receptor-positive disease PREDICT tool for the first time in October 2011, based on data (45%) received a selective estrogen receptor modulator (SERM) from the Breast Cancer Association Consortium (BCAC) consisting (tamoxifen). in 10,179 cases not exposed to anti-HER2 treatment (Supplemen- Median follow-up of included patients was 6.0 years (inter- tary Table 1). The subsequently developed model (called PREDICT quartile range: 5.8–6.7). Overall, 182 deaths were observed. Plus) was then validated in the original British Columbia dataset, a cohort including 203 HER2-positive breast cancer patients . In this latter cohort, PREDICT demonstrated an improved ability to Calibration estimate breast cancer-specific and overall survival in HER2- Median predicted and observed 5-year overall survival (OS) were positive patients, compared to other prognostication tools such as 88.0% and 94.7%, respectively (standard error 0.0044, difference PREDICT and Adjuvant! Online . In the HER2-positive cohort of the −6.7%, 95% Confidence Intervals [CI] −7.5 to −5.8), thus British Columbia dataset, the observed 10-year OS was 44.3%, and indicating an underestimation of OS by PREDICT score (Table 2, none of the included patients had received trastuzumab .A Fig. 2). further step forward, was the inclusion in PREDICT of the estimates This finding was consistent across all subgroups, with a of benefit from adjuvant trastuzumab, with its proportional difference ranging from 2.7% (in the hormone receptor-positive reduction of 31% in the mortality rate up to five years. These subgroup) to 15.8% (in patients with ≥4 positive lymph nodes) estimates were based on the results of four clinical trials: FinHER , (Table 2). The underestimation of survival by PREDICT was 11 12,13 14 HERA , B31/N9831 , and BCIRG006 (Supplementary Table 2). consistent and similar in all analyzed subgroups, including among Patients with HER2-positive early breast cancer are experiencing patients treated with lapatinib and trastuzumab (Predicted— a consistent shift towards better survival across the years, mainly observed 5-year OS: −6.98), trastuzumab alone (Predicted – due to the increasingly effective local and systemic therapies observed 5-year OS: −6.28), or trastuzumab followed by lapatinib available in this setting. This change might not be reflected by a (Predicted—observed 5-year OS: −6.82). prognostic tool developed and validated 10 years ago. In The highest absolute differences were observed for patients particular, newer drugs like pertuzumab and T-DM1 have become with hormone receptor-negative disease (13.0%), larger tumor size available for many patients developing disease progression after npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation 1234567890():,; E. Agostinetto et al. Table 1. Characteristics of the patients (overall and per randomization arm). All patients Trastuzumab+ lapatinib Trastuzumab alone Trastuzumab followed by lapatinib N (%) 2794 (100.0) 925 (100.0) 936 (100.0) 933 (100.0) Age at randomization ≤40 495 (17.7) 161 (17.4) 160 (17.1) 174 (18.7) 41–64 1989 (71.2) 667 (72.1) 667 (71.3) 655 (70.2) ≥65 310 (11.1) 97 (10.5) 109 (11.7) 104 (11.2) Ethnicity Asian 606 (21.7) 201 (21.7) 200 (21.4) 205 (22.0) Black 59 (2.1) 28 (3.0) 14 (1.5) 17 (1.8) White 2001 (71.6) 657 (71.0) 677 (72.3) 667 (71.5) Other/missing 128 (4.6) 39 (4.2) 45 (4.8) 44 (4.7) Histology Ductal 2605 (93.2) 868 (93.8) 867 (92.6) 870 (93.2) Lobular 101 (3.6) 34 (3.7) 38 (4.1) 29 (3.1) Others 119 (4.3) 35 (3.8) 43 (4.6) 41 (4.4) Central HR status Negative 1185 (42.4) 393 (42.5) 398 (42.5) 394 (42.2) Positive 1609 (57.6) 532 (57.5) 538 (57.5) 539 (57.8) Number of positive lymph nodes 0 567 (25.5) 180 (24.7) 200 (26.5) 187 (25.3) 1–3 945 (42.6) 319 (43.8) 314 (41.6) 312 (42.3) ≥4 709 (31.9) 230 (31.6) 240 (31.8) 239 (32.4) Tumor size (mm) ≤20 mm 1248 (44.7) 397 (42.9) 436 (46.6) 415 (44.5) 21–50 mm 1356 (48.5) 466 (50.4) 439 (46.9) 451 (48.3) >50 mm 190 (6.8) 62 (6.7) 61 (6.5) 67 (7.2) Tumor grade 1 79 (2.8) 26 (2.8) 22 (2.4) 31 (3.3) 2 936 (33.6) 310 (33.6) 295 (31.6) 331 (35.6) 3 1698 (60.9) 561 (60.9) 589 (63.0) 548 (58.9) X (differentiation cannot be assessed) 75 (2.7) 25 (2.7) 29 (3.1) 21 (2.3) Surgery BCS 1226 (43.9) 399 (43.1) 408 (43.6) 419 (44.9) Mastecomy 1538 (56.1) 526 (56.9) 529 (56.4) 514 (55.1) Type of CT Non-anthracycline based 322 (11.5) 103 (11.1) 109 (11.7) 110 (11.8) Anthracycline-based 2472 (88.5) 822 (88.9) 827 (88.4) 823 (88.2) Type of endocrine therapy AI 581 (39.4) 192 (39.4) 197 (40.3) 192 (38.5) AI & SERM 212 (14.4) 68 (14.0) 77 (15.8) 67 (13.4) LHRH 11 (0.8) 4 (0.8) 5 (1.0) 2 (0.4) SERM 671 (45.5) 223 (45.8) 210 (42.9) 238 (47.7) HR hormone receptors, BCS breast conserving surgery, CT chemotherapy, AI aromatase inhibitors, SERM selective estrogen receptor modulators, LHRH luteinizing hormone-releasing hormone. treatment in the ALTTO trial. These two drugs improve OS in positive disease. As such, the discordance between OS estimated metastatic patients and may contribute to the “better-than- by PREDICT and the current real-world OS is expected to be even 15,16 predicted” OS . Moreover, the current standard of care for early higher. Therefore, our results suggest that the current version of breast cancer is even superior to the treatment received by many PREDICT should be used with caution for prognostication in HER2- patients in the ALTTO study, including neoadjuvant therapy with positive early breast cancer patients treated in the modern era pertuzumab, adjustment of adjuvant therapy based on patholo- with effective chemotherapy and anti-HER2 targeted therapies. gical response to neoadjuvant therapy (i.e., T-DM1 for patients It should be also considered that at least part of the discordance who do not achieve pathological complete response) and observed between the observed and predicted 5-yr OS by PREDICT considering extended adjuvant anti-HER2 therapy with neratinib could be due to the differences existing between a highly selected and endocrine therapy for patients with hormone receptor- population enrolled in a clinical trial and the real-world patient Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. Table 2. Median predicted probability of 5-year overall survival and observed 5-year overall survival rate in the study population. % 5 years OS Predicted Observed s. e. Difference (95% CI) All patients 2794 88.00 94.69 0.44 −6.69 (−7.55 to–5.83) Type of anti-HER2 therapy Lapatinib + trastuzumab 925 87.90 94.88 0.75 −6.98 (−8.45–5.51) Trastuzumab alone 936 87.90 94.18 0.79 −6.28 (−7.83–4.73) Trastuzumab followed by lapatinib 933 88.20 95.02 0.73 −6.82 (−8.25–5.39) Type of chemotherapy Non-anthracycline-based 322 88.15 96.22 1.12 −8.07 (−10.27–5.87) Anthracycline-based 2472 87.95 94.51 0.47 −6.56 (−7.48–5.64) Age at randomization ≤40 495 90.40 95.64 0.95 −5.24 (−7.10–3.38) 41–64 1989 88.20 94.91 0.51 −6.71 (−7.71–5.71) ≥65 310 82.05 91.78 1.61 −9.73 (−12.89–6.57) Central HR status Negative 1185 80.20 93.15 0.76 −12.95 (−14.44–11.46) Positive 1609 93.10 95.82 0.52 −2.72 (−3.74–1.70) Number of positive lymph nodes 0 567 91.80 97.93 0.62 −6.13 (−7.35–4.91) 1–3 945 87.40 96.40 0.63 −9.00 (−10.23–7.77) ≥4 709 71.80 87.61 1.27 −15.81 (−18.30–13.32) Tumor size (mm) ≤20 mm 1248 91.25 97.43 0.46 −6.18 (−7.08–5.28) 21–50 mm 1356 86.00 93.33 0.70 −7.33 (−8.70–5.96) >50 mm 190 71.05 86.37 2.59 −15.32 (−20.40–10.24) OS overall survival, s.e. standard error, CI confidence interval, CT chemotherapy, HR hormone receptors. 50 60 70 80 90 100 Predicted 5-year OS Fig. 2 Calibration plot showing observed versus predicted 5-year Fig. 3 Discriminatory accuracy of PREDICT represented by the overall survival: for each decile of the predicted 5-year overall area under the receiver-operator characteristic (ROC) curve at the survival, the mean observed 5-year overall survival is presented, 5-year timepoint in the overall population. ROC receiver-operator with error bars presenting the standard error. OS overall survival. characteristic, AUC area under curve. 17,18 population, which might have slightly different prognosis .Clinical of age, comorbidity status, and overall health, differences between trials have a strong internal validity, but their external validity could be predicted and observed OS in a clinical trial sample as compared to weaker, particularly in the case of narrow inclusion criteria. For this real-world data are expected. Consistently with our findings, an reason, findings from clinical trials might overestimate outcomes as independent validation of PREDICT on data from real-world patients compared to real-world practice. Due to differences in the distribution led by Gray and colleagues showed a general pattern of npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation Observed 5-year OS E. Agostinetto et al. administered or not, but also in HER2-positive disease. Indeed, Table 3. Discriminatory accuracy of PREDICT in the overall population although in HER2-positive breast cancer almost all patients and in subgroups. deserve chemotherapy as per standard of care, a reliable prognostic estimation has several implications, from the planning AUC for time-point 5 years of premenopausal patients’ reproductive life (e.g. affecting the (95% CI) choice of having or not a pregnancy later on ), to a therapeutic All patients 2794 73.75 (69.73–77.76) perspective (adoption of escalation or de-escalation treatment strategies, including type of chemotherapy to be administered Type of anti-HER2 therapy together with anti-HER2 treatment and use of extended adjuvant Lapatinib + trastuzumab 925 72.37 (64.31–80.42) endocrine therapy in hormone receptor-positive disease ). Trastuzumab alone 936 77.67 (72.02–83.32) Several molecular assays are now available for hormone Trastuzumab followed by 933 70.64 (63.51–77.78) receptor-positive/HER2-negative breast cancer , and, recently, lapatinib some molecular assays have been also developed for HER2- Type of chemotherapy positive disease . Non-anthracycline based 322 65.18 (50.36–80.00) It is likely that these assays will refine prognostication beyond what can be provided by clinical prognostic models like Anthracycline based 2472 74.44 (70.32–78.57) 27,28 PREDICT , and their increasing use, as a consequence, will Age at randomization reduce reliance on tools like PREDICT. Nevertheless, one strength ≤40 495 76.09 (66.20–85.97) of PREDICT is the fact that it is “free” and easy to use in everyday 41–64 1989 73.69 (68.75–78.62) clinical practice, and its integration with molecular assay could ≥65 310 67.42 (56.95–77.89) provide a more complete prognostic evaluation of each single patient. Recently, Prat et al. developed a new prognostic score, HR status HER2DX, based on the combination of clinical-pathological and Negative 1185 71.87 (65.79–77.96) molecular characteristics of the tumor (nodal and tumor stage, the Positive 1609 76.81 (71.58–82.04) number of stromal tumor-infiltrating lymphocytes, Number of positive lymph nodes PAM50 subtypes, and expression of 13 genes relating to 26,29 0 567 77.25 (65.5–89.01) proliferation and underlying subtype-related biology) . This was the first attempt to build a combined prognostic score based 1–3 945 64.76 (54.58–74.96) on clinicopathological and genomic variables in early-stage HER2- ≥4 709 61.74 (55.05–68.43) positive breast cancer, using tumor samples from the phase 3 Tumor size (mm) 30 Short-HER trial . However, the HER2DX prognostic model is still ≤20 mm 1248 70.63 (61.83–79.44) immature to be used as biomarker, and future clinical validations 21–50 mm 1356 68.61 (63.27–73.94) are warranted in order to establish its use in different scenarios, especially in the neoadjuvant setting. >50 mm 190 72.97 (63.09–82.84) Our study has some limitations that should be acknowledged. AUC area under the curve, CI confidence interval, CT chemotherapy, HR First, this is an unplanned exploratory analysis. Second, some hormone receptors. information (including prognostic factors like the proliferation index Ki67 and type of method for breast cancer detection) were not available in the ALTTO database and could not be included in overestimation of mortality (expected and observed 5-year mortality: the model. Third, PREDICT did not allow for estimates of dual- 15.3% and 14.5%, respectively), although not focusing specifically on targeted anti-HER2 therapy efficacy, and, in particular, does not HER2-positive disease . provide estimates for lapatinib use. However, our subgroup Additionally, prognostication estimates of PREDICT are provided analysis confirmed that PREDICT still underperforms for patients as OS rates. Although OS is an important endpoint, being free treated with trastuzumab alone. Additionally, PREDICT tool does from any ambiguity in its definition, it could be influenced by not consider the presence of comorbidities and/or the patient several variables (competing risks) not strictly related to breast performance status, thus further limiting the possibility to cancer and not considered in PREDICT, such as comorbidities and compare predicted vs. observed outcomes using a clinical trial performance status . Non-cancer deaths may not entirely reflect sample. Finally, only the point estimates by PREDICT, without its tumor biology, aggressiveness, and responsiveness to therapy . range, were included in the present analysis. On the other hand, the more aggressive the disease, the higher On the other hand, our study has several strengths. Our results the relevance of OS. Indeed, HER2-positive breast cancer tend to derive from a large cohort (n = 2794) of patients enrolled in the develop more early recurrences compared to hormone receptor largest, randomized adjuvant trial ever conducted in the field of positive/HER2-negative disease, thus having an undoubtedly more HER2-positive breast cancer. We included only patients receiving relevant impact on OS . adjuvant trastuzumab-based therapy started concurrently with In our analysis, the highest absolute differences between modern chemotherapy. Trial sample size allowed the exploration observed and predicted OS were observed for patients with of relevant patient subgroups. All data used for the analyses were hormone-receptor negative disease, larger tumor size, and high prospectively collected during the trial conduction, as detailed in number of nodes (≥4 positive lymph nodes), namely those the study protocol. patients traditionally considered at higher risk of relapse. Further In conclusion, in patients with HER2-positive early breast cancer investigations are urgently needed to better predict prognosis of enrolled in the ALTTO trial and treated with modern chemother- these patients. Of note, despite the traditional stigma of poor apy and trastuzumab-based therapies, the PREDICT score highly prognosis for patients with high-risk HER2-positive breast cancer, underestimated OS. The suboptimal performance of this prog- recent clinical trials have shown good outcomes also for this high- nostic tool was observed irrespective of type of anti-HER2 risk subset of patients . treatment, type of chemotherapy regimen, age of the patients The prediction of prognosis in patients with early breast cancer at the time of diagnosis, central hormone receptor status, is an issue of paramount importance, not only in hormone pathological nodal status, and pathological tumor size. Our results receptor-positive/HER2-negative disease, where prognostication suggest that the current version of PREDICT should be used with may settle whether adjuvant chemotherapy should be caution to give prognostic estimation in HER2-positive early breast Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. cancer patients treated in the modern era with effective The most updated ALTTO database was used for this analysis , which corresponds to at least 5-year follow-up for every single patient. chemotherapy and anti-HER2 targeted therapies. The further improvement of therapeutic strategies expected in the next future will likely increase the survival of patients with HER2-positive early Statistical analysis breast cancer, thus requiring the current version of PREDICT to be The present analysis should be considered as exploratory, since it was not updated to provide reliable prognostic estimation in these preplanned in the study protocol and the power of the statistical analyses patients. performed was not pre-specified. The prognostic performance of PREDICT was evaluated by assessing the following endpoints: i) calibration, defined as the agreement between the METHODS predicted and observed survival rates, and ii) discriminatory accuracy, defined as the ability of distinguishing individuals who will survive 5 years Study design and patients compared to those who will not (i.e. the ability to discern patients with Details of the ALTTO trial study design were previously published . good outcomes from those with poor outcomes at the individual patient Shortly, the ALTTO trial (Breast International Group [BIG] 2-06/EGF106708 level). and North Central Cancer Treatment Group [Alliance] N063D) was an The observation time for each patient was defined as the time between international, open-label, randomized phase III study testing the use of the date of diagnosis and an event. OS event was defined as death from trastuzumab and/or lapatinib as adjuvant anti-HER2 therapy in patients any cause. with HER2-positive early breast cancer. The median predicted 5-year OS was calculated from individual Primary tumor samples from all patients were centrally tested to assess predicted outcomes by PREDICT v. 2.2. 32 33 HER2 and hormone receptor status . For assessing calibration, the median predicted 5-year survival Eligible patients were randomized to one of four anti-HER2 treatment probabilities (by PREDICT) were compared with the observed 5-year arms: trastuzumab alone, lapatinib alone, sequential treatment with survival rates (as obtained by Kaplan-Meier curves). We had to use the trastuzumab for 12 weeks followed by a 6-week washout period before median 5-year prediction instead of the mean 5-year prediction, due to the other 34 weeks of lapatinib, and dual anti-HER2 blockade with trastuzumab skewness in the distribution, i.e. mean 5-year prediction was 83.6% while plus lapatinib. The CONSORT diagram of the ALTTO study is reported in the median 5-year prediction was 88.0%, and thus the mean predicted 5-year ALTTO primary analysis paper. survival probability underestimated the center of the distribution. There- Anti-HER2 treatment could be administered as per physician’s choice fore, we used the median as a robust estimator of the center of the following chemotherapy completion (design 1), or concomitantly, either distribution. Using the standard error as obtained by the Kaplan-Meier with a taxane after anthracycline-based chemotherapy (design 2) or with 6 curve, we calculated 95% CI for the difference in predicted vs. observed cycles of docetaxel and carboplatin in an anthracycline-free regimen 5-year survival. Calibration plots for PREDICT were constructed by (design 2B). In all treatment arms, adjuvant anti-HER2 therapy was visualizing mean predicted vs. observed survival outcomes by deciles of administered for 1 year. predicted outcomes. In 2011, after the first interim analysis, the lapatinib arm was closed and For assessing discriminatory accuracy, the area under the receiver- patients were offered adjuvant commercial trastuzumab . operator characteristic curve (AUC under the ROC) and corresponding 95% In the present analysis, in order to reflect current clinical practice in this CI for 5-year predicted OS were calculated. The AUC translates into the setting, only patients who received concurrent chemotherapy (design 2 probability that the predicted outcome of a randomly selected patient who and design 2B) and who received trastuzumab-based anti-HER2 therapy indeed had that outcome is higher than that of a patient who did not; the (i.e. trastuzumab alone arm, trastuzumab followed by lapatinib arm and higher the AUC, the better the tool is at identifying patients with a better trastuzumab plus lapatinib arm) were included. All patients originally survival. assigned to the lapatinib alone arm, and those who received anti-HER2 Subgroup analyses were performed to investigate the prognostic therapies at the completion of all chemotherapy (sequential treatment, performance of PREDICT according to the type of anti-HER2 treatment and chemotherapy received, age at the time of diagnosis, central hormone design 1) were excluded. receptor status, pathological nodal status, and pathological tumor size. Statistical analysis was performed by L.A. using SAS 9.4 statistical Ethics section software (SAS Institute, Cary, NC) and R. All patients signed a written informed consent prior to enrollment in ALTTO. The project proposal of the present exploratory analysis was Reporting summary submitted and approved by the ALTTO Steering Committee. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Study objectives The primary objective of the current analysis was to investigate the prognostic performance of PREDICT in breast cancer patients with early- DATA AVAILABILITY stage HER2-positive disease treated with modern chemotherapy and Data can be made available upon reasonable request. Data and results from the Data concurrent trastuzumab-based anti-HER2 therapy. Centre at Institut Jules Bordet in Brussels (Belgium) can be made available upon Secondary objectives were to investigate the prognostic performance of approval of a research proposal. PREDICT according to the type of anti-HER2 treatment received (trastuzumab alone, trastuzumab followed by lapatinib and, trastuzumab Received: 19 January 2022; Accepted: 1 July 2022; plus lapatinib), type of chemotherapy regimen received (anthracycline- based chemotherapy regimens vs. non-anthracycline-based chemotherapy regimens), age of patients at the time of diagnosis (age ≤ 40 years vs. age 41–64 vs. age ≥65 years), central hormone receptor status (hormone receptor -positive vs. negative), pathological nodal status (node-negative vs. node-positive disease [1–3 positive nodes] vs. node-positive disease [≥4 REFERENCES positive nodes]), and pathological tumor size (small [≤2 cm] vs. medium 1. Bradley, R. et al. Trastuzumab for early-stage, HER2-positive breast cancer: a [2–5 cm] vs. large [>5 cm] disease). meta-analysis of 13,864 women in seven randomised trials. Lancet Oncol. 22, 1139–1150 (2021). 2. Jackisch, C. et al. Risk-based decision-making in the treatment of HER2-positive Data extraction early breast cancer: recommendations based on the current state of knowledge. PREDICT estimates for each patient were calculated by one investigator Cancer Treat. Rev. 99, 10222 (2021). blinded to patient outcomes. Patient and tumor characteristics, as well as 3. Pernas, S. & Tolaney, S. M. Management of early-stage human epidermal growth administered adjuvant anticancer treatments, were entered in the PREDICT factor receptor 2-positive breast cancer. JCO Oncol. Pract. 17, 320–330 (2021). v.2.2 program to calculate the predicted 5-year OS for each patient. 4. Wishart, G. C. et al. PREDICT: a new UK prognostic model that predicts survival Detection modality and Ki67 status were considered “unknown” for all following surgery for invasive breast cancer. Breast Cancer Res. 12, R1 (2010). patients (as these variables were not collected as part of the ALTTO trial). npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation E. Agostinetto et al. 5. Wishart, G. C. et al. A population-based validation of the prognostic model 32. Wolff, A. C. et al. American Society of Clinical Oncology/College of American PREDICT for early breast cancer. Eur. J. Surg. Oncol. 37, 411–417 (2011). Pathologists guideline recommendations for human epidermal growth factor 6. The Breast Cancer Association Consortium (BCAC). http://bcac.ccge.medschl. receptor 2 testing in breast cancer. J. Clin. Oncol. 25, 118–145 (2007). cam.ac.uk/. 33. Lambertini, M. et al. Dissecting the effect of hormone receptor status in patients 7. Wishart, G. C. et al. PREDICT Plus: development and validation of a prognostic with HER2-positive early breast cancer: exploratory analysis from the ALTTO (BIG model for early breast cancer that includes HER2. Br. J. Cancer 107, 800–807 2-06) randomized clinical trial. Breast Cancer Res. Treat. 177, 103–114 (2019). (2012). 8. NICE. Tumour profiling tests to guide adjuvant chemotherapy decisions in early breast cancer. https://www.nice.org.uk/guidance/dg34/ (2018). ACKNOWLEDGEMENTS 9. Moreno-Aspitia, A. et al. Updated results from the international phase III ALTTO This work was supported by Les Amis de l’Institut Bordet (Grant No. 2018-09). The trial (BIG 2-06/Alliance N063D). Eur. J. Cancer 148, 287–296 (2021). ALTTO trial received financial support from GlaxoSmithKline (until January 2015), 10. Joensuu, H. et al. Fluorouracil, epirubicin, and cyclophosphamide with either Novartis Pharma AG (as of January 2015) and the National Cancer Institute of the docetaxel or vinorelbine, with or without trastuzumab, as adjuvant treatments of National Institutes of Health (NCI-NIH; Grant No. U10CA180821 and U10CA180882 to breast cancer: final results of the FinHer Trial. J. Clin. Oncol. 27, 5685–5692 (2009). the Alliance for Clinical Trials in Oncology and Grant No. CA025224 to the legacy 11. Smith, I. et al. 2-year follow-up of trastuzumab after adjuvant chemotherapy in North Central Cancer Treatment Group). Matteo Lambertini acknowledges the HER2-positive breast cancer: a randomised controlled trial. Lancet 369,29–36 support from the Italian Association for Cancer Research (“Associazione Italiana per la (2007). Ricerca sul Cancro”, AIRC; MFAG 2020 ID 24698) and the Italian Ministry of Health (5 × 12. Romond, E. H. et al. Trastuzumab plus adjuvant chemotherapy for operable HER2- 1000 funds 2017). The funders of the study had no role in study design, data positive breast cancer. N. Engl. J. Med. 353, 1673–1684 (2005). collection, data analysis, data interpretation, or writing of the present manuscript. All 13. Perez, E. A. et al. Four-year follow-up of trastuzumab plus adjuvant chemotherapy authors agreed and share the final responsibility for the provided interpretation of for operable human epidermal growth factor receptor 2-positive breast cancer: the study results and for the decision to submit for publication. The abstract of the joint analysis of data from NCCTG N9831 and NSABP B-31. J. Clin. Oncol. 29, present work was presented at the American Society of Clinical Oncology (ASCO) 3366–3373 (2011). Annual Meeting 2021 and was awarded an ASCO Merit Award. 14. Slamon, D. et al. Adjuvant trastuzumab in HER2-positive breast cancer. N. Engl. J. Med. 365, 1273–1283 (2011). 15. Swain, S. M. et al. Pertuzumab, trastuzumab, and docetaxel for HER2-positive AUTHOR CONTRIBUTIONS metastatic breast cancer (CLEOPATRA): end-of-study results from a double-blind, M.L. conceived and designed the study. E.A. and E.d.A. were involved in data randomised, placebo-controlled, phase 3 study. Lancet Oncol. 21, 519–530 (2020). acquisition. L.A. analyzed the data. All authors were involved in data interpretation. 16. Diéras, V. et al. Trastuzumab emtansine versus capecitabine plus lapatinib in E.A., E.d.A., and M.L. wrote the first draft of the paper. LA and EA produced the patients with previously treated HER2-positive advanced breast cancer (EMILIA): a manuscript figures and tables. All authors were involved in reviewing and editing descriptive analysis of final overall survival results from a randomised, open-label, drafts of the paper and approving the manuscript. phase 3 trial. Lancet Oncol. 18, 732–742 (2017). 17. Saturni, S. et al. Randomized controlled trials and real life studies. Approaches and methodologies: a clinical point of view. Pulm. Pharmacol. Ther. 27, 129–138 COMPETING INTERESTS (2014). 18. Rothwell, P. M. External validity of randomised controlled trials: “To whom do the The authors have no known competing financial interests or personal relationships results of this trial apply?” Lancet 365,82–93 (2005). that could have appeared to influence the work reported in this paper. Elisa 19. Gray, E. et al. Independent validation of the PREDICT breast cancer prognosis Agostinetto: Consultancy/honoraria from Eli Lilly, Sandoz. Support to attend medical prediction tool in 45,789 patients using Scottish Cancer Registry data. Br. J. Cancer conferences (travel accommodations/expenses) from Novartis, Roche, Eli Lilly, 119, 808–814 (2018). Genetic, Istituto Gentili (outside the submitted work). Samuel Martel: Honoraria 20. Hudis, C. A. et al. Proposal for standardized definitions for efficacy end points in and advisory board for Novartis, Pfizer, Knight Therapeutics, Exact Science, Taiho, adjuvant breast cancer trials: the STEEP system. J. Clin. Oncol. 25, 2127–2132 Apobiologix (outside the submitted work). Philippe Aftimos: Consulting: Boehringer (2007). Ingelheim, Macrogenics, Roche, Novartis, Amcure, Servier, G1 Therapeutics, Radius, 21. van Maaren, M. C. et al. Ten-year recurrence rates for breast cancer subtypes in Deloitte, Menarini; Honoraria: Synthon, Amgen, Novartis, Gilead; Travel grants: the Netherlands: a large population-based study. Int. J. Cancer 144, 263–272 Amgen, MSD, Pfizer, Roche; Research funding to my institution: Roche (outside the (2019). submitted work). Noam Pondé: Honoraria and advisory board for Lilly, Novartis, 22. Piccart, M. et al. Adjuvant pertuzumab and trastuzumab in early HER2-positive Roche/Genentech, AstraZeneca and Pfizer (outside the submitted work). Christian breast cancer in the APHINITY trial: 6 years’ follow-up. J. Clin. Oncol. 39, Maurer: Travel grants: Mundipharma, Amgen, Servier Deutschland GmbH, Abbvie. 1448–1457 (2021). Honoraria: Abbvie. Advisory board: Celgene/ BMS and Pfizer (outside the submitted 23. Lambertini, M. et al. Pregnancies during and after trastuzumab and/or lapatinib in work). Sarra El-Abed: grants from Novartis during the conduct of the ALTTO study, patients with human epidermal growth factor receptor 2–positive early breast and grants from Roche/Genentech and Pfizer (outside of the submitted work). Yingbo cancer: analysis from the NeoALTTO (BIG 1-06) and ALTTO (BIG 2-06) trials. Cancer Wang: Employee of Novartis Pharmaceuticals and holds stock in the company 125, 307–316 (2019). (outside the submitted work). Saranya Chumsri: Research funding from Merck and 24. Burstein, H. J. et al. Adjuvant endocrine therapy for women with hormone Pfizer (outside the submitted work). Judith Bliss: Research Funding to her Institution: receptor–positive breast cancer: asco clinical practice guideline focused update. AstraZeneca, Merck Sharp & Dohme, Medivation, Puma Biotechnology, Clovis J. Clin. Oncol. 37, 423–438 (2018). Oncology, Pfizer, Janssen-Cilag, Roche, Novartis (previously GSK), Eli Lilly (outside 25. Puppe, J. et al. Genomic signatures in luminal breast cancer. Breast Care 15, the submitted work). Judith Kroep: Honoraria and advisory board: AstraZeneca, 355–365 (2020). Daiichi Sankyo, Eisai, Lilly, MSD, Novartis, Pfizer, GSK. Research grant: Amgen, Astra 26. Prat, A. et al. A multivariable prognostic score to guide systemic therapy in early- Zeneca, Novartis, Philips (outside the submitted work). Marco Colleoni: Research stage HER2-positive breast cancer: a retrospective study with an external eva- grant to his Institution: Roche (outside the submitted work). Lucia Del Mastro: luation. Lancet Oncol. 21, 1455–1464 (2020). Honoraria: Roche, Novartis, Eli Lilly, MSD, Pfizer, Ipsen. Honoraria to her Institution: 27. Cardoso, F. et al. 70-gene signature as an aid to treatment decisions in early-stage Novartis. Consulting or Advisory role: Roche, Eli Lilly, Novartis, MSD, Genomic Health, breast cancer. N. Engl. J. Med. 375, 717–729 (2016). Pierre Fabre, Daiichi Sankyo, Seagen, Astrazeneca, Eisai, Gilead, Exact Science. Travel, 28. Cardoso, F. et al. MINDACT: Long-term results of the large prospective trial testing accomodation, expenses: Celgene, Roche, Pfizer, Daiichi Sankyo (outside the the 70-gene signature MammaPrint as guidance for adjuvant chemotherapy in submitted work). Alvaro Moreno Aspitia: Institutional research funding from breast cancer patients. J. Clin. Oncol. 38, 506 (2020). GlaxoSmithKline, Genentech, Daiichi Sankyo, Novartis, Merck, Pfizer, AbbVie and Eli 29. Hurvitz, S. A. HER2DX: a tool that might inform treatment choices for HER2- Lilly (outside the submitted work). Martine Piccart: Board Member (Scientific Board): positive breast cancer. Lancet Oncol. 21, 1392–1393 (2020). Oncolytics. Consultant (honoraria): AstraZeneca, Camel-IDS, Immunomedics, Lilly, 30. Conte, P. et al. Nine weeks versus 1 year adjuvant trastuzumab in combination Menarini, MSD, Novartis, Odonate, Pfizer, Roche-Genentech, Seattle Genetics, with chemotherapy: final results of the phase III randomized Short-HER study‡. Immutep, Seagen, NBE Therapeutics. Research grants to her Institute: AstraZeneca, Ann. Oncol. 29, 2328–2333 (2018). Immunomedics, Lilly, Menarini, MSD, Novartis, Pfizer, Radius, Roche-Genentech, 31. Piccart-Gebhart, M. et al. Adjuvant Lapatinib and Trastuzumab for early human Servier, Synthon (outside the submitted work). Evandro de Azambuja: honoraria and epidermal growth factor receptor 2-positive breast cancer: results from the advisory board: Roche/GNE, Novartis, Seattle Genetics, Zodiacs, Libbs and Pierre randomized phase III adjuvant lapatinib and/or trastuzumab treatment optimi- Fabre; travel grants: Roche/GNE, GSK/Novartis. Research grant for his Institute: Roche/ zation trial. J. Clin. Oncol. 34, 1034–1042 (2016). GNE, AstraZeneca, Novartis, and Servier (outside the submitted work). Matteo Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. Lambertini acted as a consultant for Roche, Lilly, AstraZeneca, Exact Sciences, Seagen, Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims Gilead, Pfizer, MSD and Novartis, and received honoraria from Sandoz, Takeda, Roche, in published maps and institutional affiliations. Lilly, Pfizer, Ipsen, Libbs, Knight and Novartis (outside the submitted work). Lieveke Ameye, Marianne Paesmans, Malou Vicente, and Fausto Petrelli have no conflict of interests to declare. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give ADDITIONAL INFORMATION appropriate credit to the original author(s) and the source, provide a link to the Creative Supplementary information The online version contains supplementary material Commons license, and indicate if changes were made. The images or other third party available at https://doi.org/10.1038/s41523-022-00452-8. material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the Correspondence and requests for materials should be addressed to Matteo article’s Creative Commons license and your intended use is not permitted by statutory Lambertini. regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http:// Reprints and permission information is available at http://www.nature.com/ creativecommons.org/licenses/by/4.0/. reprints © The Author(s) 2022 npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png npj Breast Cancer Springer Journals

Loading next page...
 
/lp/springer-journals/predict-underestimates-survival-of-patients-with-her2-positive-early-nSCunlIget

References (38)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
2374-4677
DOI
10.1038/s41523-022-00452-8
Publisher site
See Article on Publisher Site

Abstract

www.nature.com/npjbcancer ARTICLE OPEN PREDICT underestimates survival of patients with HER2-positive early-stage breast cancer 1,2 3 4 5 6 7 8 Elisa Agostinetto , Lieveke Ameye , Samuel Martel , Philippe Aftimos , Noam Pondé , Christian Maurer , Sarra El-Abed , 9 10 11 12 13 14 15 Yingbo Wang , Malou Vicente , Saranya Chumsri , Judith Bliss , Judith Kroep , Marco Colleoni , Fausto Petrelli , 16,17 11 10 3 1 Lucia Del Mastro , Alvaro Moreno-Aspitia , Martine Piccart , Marianne Paesmans , Evandro de Azambuja and 16,17 Matteo Lambertini The prognostic performance of PREDICT in patients with HER2-positive early breast cancer (EBC) treated in the modern era with effective chemotherapy and anti-HER2 targeted therapies is unclear. Therefore, we investigated its prognostic performance using data extracted from ALTTO, a phase III trial evaluating adjuvant lapatinib ± trastuzumab vs. trastuzumab alone in patients with HER2-positive EBC. Our analysis included 2794 patients. After a median follow-up of 6.0 years (IQR, 5.8–6.7), 182 deaths were observed. Overall, PREDICT underestimated 5-year OS by 6.7% (95% CI, 5.8–7.6): observed 5-year OS was 94.7% vs. predicted 88.0%. The underestimation was consistent across all subgroups, including those according to the type of anti HER2-therapy. The highest absolute differences were observed for patients with hormone receptor negative-disease, nodal involvement, and large tumor size (13.0%, 15.8%, and 15.3%, respectively). AUC under the ROC curve was 73.7% (95% CI 69.7–77.8) in the overall population, ranging between 61.7% and 77.7% across the analyzed subgroups. In conclusion, our analysis showed that PREDICT highly underestimated OS in HER2-positive EBC. Hence, it should be used with caution to give prognostic estimation to HER2-positive EBC patients treated in the modern era with effective chemotherapy and anti-HER2 targeted therapies. npj Breast Cancer (2022) 8:87 ; https://doi.org/10.1038/s41523-022-00452-8 INTRODUCTION diagnosed after 2004, to ensure that patients did not receive trastuzumab . A subsequent validation was done in 2012 in a The addition of trastuzumab to adjuvant chemotherapy has British Columbia Canadian cohort . This study demonstrated that dramatically improved the outcomes of patients with HER2- the inclusion of the HER2 status allowed the model to perform positive early breast cancer, reducing the risk of mortality by more 1 better than the previous PREDICT version and Adjuvant! Online in than 30% . Despite the undoubted benefit of adjuvant therapy, estimating overall and breast-cancer-specific survival . several clinical questions remain open. Approximately 25% of Although the use of PREDICT is recommended to aid decision patients still experience recurrence up to 10 years from diagnosis, making in the adjuvant setting , its prognostic role in HER2- and further research efforts are needed to better refine patient positive early breast cancer patients treated with modern selection for adopting escalation or de-escalation treatment chemotherapy and anti-HER2 therapies remains unclear. We 2,3 strategies . aimed to investigate the prognostic performance of PREDICT in PREDICT (www.predict.nhs.uk) is a publicly available, online tool patients with HER2-positive early breast cancer who received that helps to predict the individual prognosis of patients with early trastuzumab-based therapy started concurrently with chemother- breast cancer and to show the impact of adjuvant treatments apy within the ALTTO trial. The ALTTO trial is the largest adjuvant administered after breast cancer surgery. It uses traditional study ever conducted in the field of HER2-positive early breast clinical-pathological factors, and it is aimed to support clinical cancer and, including at least 5-year follow-up data from all decision making in the adjuvant setting. The original version of patients , represented a unique opportunity to investigate the PREDICT (v.1.0) was derived from cancer registry information on reliability and prognostic performance of PREDICT in women with 5,694 women treated in East Anglia from 1999–2003, and was HER2-positive disease. subsequently validated in several datasets of patients with breast 4,5 cancer . In 2011, the model was updated to include HER2 status. RESULTS Estimates for the prognostic effect of HER2 status were based on an analysis of 10,179 cases collected by the Breast Cancer Out of 8381 patients included in the ALTTO trial, 2836 were Association Consortium (BCAC), none of which had been treated with chemotherapy and concurrent trastuzumab-based 1 2 Academic Trials Promoting Team, Institut Jules Bordet and l’Université Libre de Bruxelles (U.L.B), Brussels, Belgium. Humanitas University, Department of Biomedical 3 4 Sciences, via Rita Levi Montalcini 4, 20090 Pieve Emanuele, Milan, Italy. Data Center, Institut Jules Bordet, Brussels, Belgium. Department of Hemato-Oncology, CISSS Montérégie-Centre/Hôpital Charles-Le Moyne, Université de Sherbrooke, Greenfield Park, QC, Canada. Clinical Trials Conduct Unit, Institut Jules Bordet – Université Libre de 6 7 Bruxelles, Brussels, Belgium. Clinical Oncology Department, AC Camargo Cancer Center, São Paulo, Brazil. University of Cologne, Department I of Internal Medicine, Center for 8 9 Integrated Oncology Aachen Bonn Cologne Dusseldorf, Cologne, Germany. Breast International Group (BIG), Brussels, Belgium. Novartis Pharma AG, Basel, Switzerland. 10 11 12 Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium. Robert and Monica Jacoby Center for Breast Health, Mayo Clinic, Jacksonville, FL, USA. The Institute of Cancer Research, Clinical Trials & Statistics Unit, London, UK. Department of Medical Oncology, Leiden University Medical Center, P.O. Box 9600, 2300 RC, Leiden, 14 15 16 The Netherlands. IEO European Institute of Oncology, IRCCS, Milan, Italy. Oncology Unit, ASST Bergamo Ovest, Treviglio (BG), Italy. Department of Medical Oncology, U.O. Clinica di Oncologia medica, IRCCS Ospedale Policlinico San Martino, Genova, Italy. Department of Internal Medicine and Medical Specialties (DiMI), School of Medicine, University of Genova, Genova, Italy. email: matteo.lambertini@unige.it Published in partnership with the Breast Cancer Research Foundation 1234567890():,; E. Agostinetto et al. (>50 mm) (15.3%), and high number of nodes (≥4 positive lymph Patients enrolled in ALTTO nodes) (15.8%). (n=8381) Discrimination AUC under the ROC curve was 73.7% (95% CI 69.7–77.8) in the Patients excluded (n=5545) overall population (Fig. 3). because assigned to lapatinib This finding of suboptimal discriminatory accuracy was only arm, and/or to design 1 of consistent across all subgroups, ranging from 61.7% (in patients chemotherapy administration with ≥4 positive lymph nodes) to 77.7% (in patients receiving trastuzumab alone as anti-HER2 therapy) (Table 3). The lowest discriminatory accuracy was observed for patients with high number of nodes (≥4 and 1–3 positive lymph nodes) (Supple- Patients potentially eligible for the present ALTTO sub- mentary Figs 1 and 2), and for patients receiving a non- analysis anthracycline-based chemotherapy (61.7%, 64.8%, and 65.2%, (n=2836) respectively). The highest discriminatory accuracy was observed for patients with negative lymph nodes (Supplementary Fig. 3) and for patients receiving trastuzumab alone as anti-HER2 therapy Patients excluded (n=42) because: (77.3% and 77.7%, respectively). - Age < 25 years old (n=7) - No tumor size available (n=13) - No nodal status available DISCUSSION (n=22) To the best of our knowledge, PREDICT is the only publicly available, free, online tool developed to predict individual prognosis in the specific population of patients with HER2- positive early breast cancer based on traditional and easily Patients included in the present retrieved clinical-pathological factors including HER2. In our analysis ALTTO analysis, PREDICT highly underestimated patients’ OS; this (n=2794) finding was consistent across all patient subgroups, with highest absolute differences for patients with hormone receptor-negative Fig. 1 STROBE flow-chart. This figure illustrates the patient disease, nodal involvement, and large tumor size. In terms of selection process. discrimination, the accuracy of PREDICT was overall low, with the lowest discriminatory accuracy observed in patients with nodal involvement (≥4 and 1–3 positive lymph nodes), and in patients therapy and were potentially eligible for the present analysis. In 42 receiving non-anthracycline-based chemotherapy. patients, the PREDICT algorithm was not evaluable (due to age of The low performance of this tool raises several questions about the patient <25 years old [n = 7], missing tumor size [n = 13], or the reliability of PREDICT to give prognostic estimation in HER2- missing lymph nodes status [n = 22]). Therefore, 2794 patients positive early breast cancer patients. To potentially explain the were included in the present analysis (Fig. 1). reasons for the underestimation of patients’ OS, we can speculate Most patients (71%) were aged between 41 and 64 years (Table whether the population used to validate this prognostic tool 1). Twenty-five percent of patients had negative nodal status, 45% accurately mirrors the real-world population of patients with had a tumor size ≤2 cm and 58% had hormone receptor-positive HER2-positive disease treated in the modern era with effective disease. Regarding administered treatments, 88% underwent an chemotherapy and anti-HER2 targeted therapies. The prognostic anthracycline-based chemotherapy regimen (design 2). The effect of HER2 status was evaluated and incorporated in the majority of patients with hormone receptor-positive disease PREDICT tool for the first time in October 2011, based on data (45%) received a selective estrogen receptor modulator (SERM) from the Breast Cancer Association Consortium (BCAC) consisting (tamoxifen). in 10,179 cases not exposed to anti-HER2 treatment (Supplemen- Median follow-up of included patients was 6.0 years (inter- tary Table 1). The subsequently developed model (called PREDICT quartile range: 5.8–6.7). Overall, 182 deaths were observed. Plus) was then validated in the original British Columbia dataset, a cohort including 203 HER2-positive breast cancer patients . In this latter cohort, PREDICT demonstrated an improved ability to Calibration estimate breast cancer-specific and overall survival in HER2- Median predicted and observed 5-year overall survival (OS) were positive patients, compared to other prognostication tools such as 88.0% and 94.7%, respectively (standard error 0.0044, difference PREDICT and Adjuvant! Online . In the HER2-positive cohort of the −6.7%, 95% Confidence Intervals [CI] −7.5 to −5.8), thus British Columbia dataset, the observed 10-year OS was 44.3%, and indicating an underestimation of OS by PREDICT score (Table 2, none of the included patients had received trastuzumab .A Fig. 2). further step forward, was the inclusion in PREDICT of the estimates This finding was consistent across all subgroups, with a of benefit from adjuvant trastuzumab, with its proportional difference ranging from 2.7% (in the hormone receptor-positive reduction of 31% in the mortality rate up to five years. These subgroup) to 15.8% (in patients with ≥4 positive lymph nodes) estimates were based on the results of four clinical trials: FinHER , (Table 2). The underestimation of survival by PREDICT was 11 12,13 14 HERA , B31/N9831 , and BCIRG006 (Supplementary Table 2). consistent and similar in all analyzed subgroups, including among Patients with HER2-positive early breast cancer are experiencing patients treated with lapatinib and trastuzumab (Predicted— a consistent shift towards better survival across the years, mainly observed 5-year OS: −6.98), trastuzumab alone (Predicted – due to the increasingly effective local and systemic therapies observed 5-year OS: −6.28), or trastuzumab followed by lapatinib available in this setting. This change might not be reflected by a (Predicted—observed 5-year OS: −6.82). prognostic tool developed and validated 10 years ago. In The highest absolute differences were observed for patients particular, newer drugs like pertuzumab and T-DM1 have become with hormone receptor-negative disease (13.0%), larger tumor size available for many patients developing disease progression after npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation 1234567890():,; E. Agostinetto et al. Table 1. Characteristics of the patients (overall and per randomization arm). All patients Trastuzumab+ lapatinib Trastuzumab alone Trastuzumab followed by lapatinib N (%) 2794 (100.0) 925 (100.0) 936 (100.0) 933 (100.0) Age at randomization ≤40 495 (17.7) 161 (17.4) 160 (17.1) 174 (18.7) 41–64 1989 (71.2) 667 (72.1) 667 (71.3) 655 (70.2) ≥65 310 (11.1) 97 (10.5) 109 (11.7) 104 (11.2) Ethnicity Asian 606 (21.7) 201 (21.7) 200 (21.4) 205 (22.0) Black 59 (2.1) 28 (3.0) 14 (1.5) 17 (1.8) White 2001 (71.6) 657 (71.0) 677 (72.3) 667 (71.5) Other/missing 128 (4.6) 39 (4.2) 45 (4.8) 44 (4.7) Histology Ductal 2605 (93.2) 868 (93.8) 867 (92.6) 870 (93.2) Lobular 101 (3.6) 34 (3.7) 38 (4.1) 29 (3.1) Others 119 (4.3) 35 (3.8) 43 (4.6) 41 (4.4) Central HR status Negative 1185 (42.4) 393 (42.5) 398 (42.5) 394 (42.2) Positive 1609 (57.6) 532 (57.5) 538 (57.5) 539 (57.8) Number of positive lymph nodes 0 567 (25.5) 180 (24.7) 200 (26.5) 187 (25.3) 1–3 945 (42.6) 319 (43.8) 314 (41.6) 312 (42.3) ≥4 709 (31.9) 230 (31.6) 240 (31.8) 239 (32.4) Tumor size (mm) ≤20 mm 1248 (44.7) 397 (42.9) 436 (46.6) 415 (44.5) 21–50 mm 1356 (48.5) 466 (50.4) 439 (46.9) 451 (48.3) >50 mm 190 (6.8) 62 (6.7) 61 (6.5) 67 (7.2) Tumor grade 1 79 (2.8) 26 (2.8) 22 (2.4) 31 (3.3) 2 936 (33.6) 310 (33.6) 295 (31.6) 331 (35.6) 3 1698 (60.9) 561 (60.9) 589 (63.0) 548 (58.9) X (differentiation cannot be assessed) 75 (2.7) 25 (2.7) 29 (3.1) 21 (2.3) Surgery BCS 1226 (43.9) 399 (43.1) 408 (43.6) 419 (44.9) Mastecomy 1538 (56.1) 526 (56.9) 529 (56.4) 514 (55.1) Type of CT Non-anthracycline based 322 (11.5) 103 (11.1) 109 (11.7) 110 (11.8) Anthracycline-based 2472 (88.5) 822 (88.9) 827 (88.4) 823 (88.2) Type of endocrine therapy AI 581 (39.4) 192 (39.4) 197 (40.3) 192 (38.5) AI & SERM 212 (14.4) 68 (14.0) 77 (15.8) 67 (13.4) LHRH 11 (0.8) 4 (0.8) 5 (1.0) 2 (0.4) SERM 671 (45.5) 223 (45.8) 210 (42.9) 238 (47.7) HR hormone receptors, BCS breast conserving surgery, CT chemotherapy, AI aromatase inhibitors, SERM selective estrogen receptor modulators, LHRH luteinizing hormone-releasing hormone. treatment in the ALTTO trial. These two drugs improve OS in positive disease. As such, the discordance between OS estimated metastatic patients and may contribute to the “better-than- by PREDICT and the current real-world OS is expected to be even 15,16 predicted” OS . Moreover, the current standard of care for early higher. Therefore, our results suggest that the current version of breast cancer is even superior to the treatment received by many PREDICT should be used with caution for prognostication in HER2- patients in the ALTTO study, including neoadjuvant therapy with positive early breast cancer patients treated in the modern era pertuzumab, adjustment of adjuvant therapy based on patholo- with effective chemotherapy and anti-HER2 targeted therapies. gical response to neoadjuvant therapy (i.e., T-DM1 for patients It should be also considered that at least part of the discordance who do not achieve pathological complete response) and observed between the observed and predicted 5-yr OS by PREDICT considering extended adjuvant anti-HER2 therapy with neratinib could be due to the differences existing between a highly selected and endocrine therapy for patients with hormone receptor- population enrolled in a clinical trial and the real-world patient Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. Table 2. Median predicted probability of 5-year overall survival and observed 5-year overall survival rate in the study population. % 5 years OS Predicted Observed s. e. Difference (95% CI) All patients 2794 88.00 94.69 0.44 −6.69 (−7.55 to–5.83) Type of anti-HER2 therapy Lapatinib + trastuzumab 925 87.90 94.88 0.75 −6.98 (−8.45–5.51) Trastuzumab alone 936 87.90 94.18 0.79 −6.28 (−7.83–4.73) Trastuzumab followed by lapatinib 933 88.20 95.02 0.73 −6.82 (−8.25–5.39) Type of chemotherapy Non-anthracycline-based 322 88.15 96.22 1.12 −8.07 (−10.27–5.87) Anthracycline-based 2472 87.95 94.51 0.47 −6.56 (−7.48–5.64) Age at randomization ≤40 495 90.40 95.64 0.95 −5.24 (−7.10–3.38) 41–64 1989 88.20 94.91 0.51 −6.71 (−7.71–5.71) ≥65 310 82.05 91.78 1.61 −9.73 (−12.89–6.57) Central HR status Negative 1185 80.20 93.15 0.76 −12.95 (−14.44–11.46) Positive 1609 93.10 95.82 0.52 −2.72 (−3.74–1.70) Number of positive lymph nodes 0 567 91.80 97.93 0.62 −6.13 (−7.35–4.91) 1–3 945 87.40 96.40 0.63 −9.00 (−10.23–7.77) ≥4 709 71.80 87.61 1.27 −15.81 (−18.30–13.32) Tumor size (mm) ≤20 mm 1248 91.25 97.43 0.46 −6.18 (−7.08–5.28) 21–50 mm 1356 86.00 93.33 0.70 −7.33 (−8.70–5.96) >50 mm 190 71.05 86.37 2.59 −15.32 (−20.40–10.24) OS overall survival, s.e. standard error, CI confidence interval, CT chemotherapy, HR hormone receptors. 50 60 70 80 90 100 Predicted 5-year OS Fig. 2 Calibration plot showing observed versus predicted 5-year Fig. 3 Discriminatory accuracy of PREDICT represented by the overall survival: for each decile of the predicted 5-year overall area under the receiver-operator characteristic (ROC) curve at the survival, the mean observed 5-year overall survival is presented, 5-year timepoint in the overall population. ROC receiver-operator with error bars presenting the standard error. OS overall survival. characteristic, AUC area under curve. 17,18 population, which might have slightly different prognosis .Clinical of age, comorbidity status, and overall health, differences between trials have a strong internal validity, but their external validity could be predicted and observed OS in a clinical trial sample as compared to weaker, particularly in the case of narrow inclusion criteria. For this real-world data are expected. Consistently with our findings, an reason, findings from clinical trials might overestimate outcomes as independent validation of PREDICT on data from real-world patients compared to real-world practice. Due to differences in the distribution led by Gray and colleagues showed a general pattern of npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation Observed 5-year OS E. Agostinetto et al. administered or not, but also in HER2-positive disease. Indeed, Table 3. Discriminatory accuracy of PREDICT in the overall population although in HER2-positive breast cancer almost all patients and in subgroups. deserve chemotherapy as per standard of care, a reliable prognostic estimation has several implications, from the planning AUC for time-point 5 years of premenopausal patients’ reproductive life (e.g. affecting the (95% CI) choice of having or not a pregnancy later on ), to a therapeutic All patients 2794 73.75 (69.73–77.76) perspective (adoption of escalation or de-escalation treatment strategies, including type of chemotherapy to be administered Type of anti-HER2 therapy together with anti-HER2 treatment and use of extended adjuvant Lapatinib + trastuzumab 925 72.37 (64.31–80.42) endocrine therapy in hormone receptor-positive disease ). Trastuzumab alone 936 77.67 (72.02–83.32) Several molecular assays are now available for hormone Trastuzumab followed by 933 70.64 (63.51–77.78) receptor-positive/HER2-negative breast cancer , and, recently, lapatinib some molecular assays have been also developed for HER2- Type of chemotherapy positive disease . Non-anthracycline based 322 65.18 (50.36–80.00) It is likely that these assays will refine prognostication beyond what can be provided by clinical prognostic models like Anthracycline based 2472 74.44 (70.32–78.57) 27,28 PREDICT , and their increasing use, as a consequence, will Age at randomization reduce reliance on tools like PREDICT. Nevertheless, one strength ≤40 495 76.09 (66.20–85.97) of PREDICT is the fact that it is “free” and easy to use in everyday 41–64 1989 73.69 (68.75–78.62) clinical practice, and its integration with molecular assay could ≥65 310 67.42 (56.95–77.89) provide a more complete prognostic evaluation of each single patient. Recently, Prat et al. developed a new prognostic score, HR status HER2DX, based on the combination of clinical-pathological and Negative 1185 71.87 (65.79–77.96) molecular characteristics of the tumor (nodal and tumor stage, the Positive 1609 76.81 (71.58–82.04) number of stromal tumor-infiltrating lymphocytes, Number of positive lymph nodes PAM50 subtypes, and expression of 13 genes relating to 26,29 0 567 77.25 (65.5–89.01) proliferation and underlying subtype-related biology) . This was the first attempt to build a combined prognostic score based 1–3 945 64.76 (54.58–74.96) on clinicopathological and genomic variables in early-stage HER2- ≥4 709 61.74 (55.05–68.43) positive breast cancer, using tumor samples from the phase 3 Tumor size (mm) 30 Short-HER trial . However, the HER2DX prognostic model is still ≤20 mm 1248 70.63 (61.83–79.44) immature to be used as biomarker, and future clinical validations 21–50 mm 1356 68.61 (63.27–73.94) are warranted in order to establish its use in different scenarios, especially in the neoadjuvant setting. >50 mm 190 72.97 (63.09–82.84) Our study has some limitations that should be acknowledged. AUC area under the curve, CI confidence interval, CT chemotherapy, HR First, this is an unplanned exploratory analysis. Second, some hormone receptors. information (including prognostic factors like the proliferation index Ki67 and type of method for breast cancer detection) were not available in the ALTTO database and could not be included in overestimation of mortality (expected and observed 5-year mortality: the model. Third, PREDICT did not allow for estimates of dual- 15.3% and 14.5%, respectively), although not focusing specifically on targeted anti-HER2 therapy efficacy, and, in particular, does not HER2-positive disease . provide estimates for lapatinib use. However, our subgroup Additionally, prognostication estimates of PREDICT are provided analysis confirmed that PREDICT still underperforms for patients as OS rates. Although OS is an important endpoint, being free treated with trastuzumab alone. Additionally, PREDICT tool does from any ambiguity in its definition, it could be influenced by not consider the presence of comorbidities and/or the patient several variables (competing risks) not strictly related to breast performance status, thus further limiting the possibility to cancer and not considered in PREDICT, such as comorbidities and compare predicted vs. observed outcomes using a clinical trial performance status . Non-cancer deaths may not entirely reflect sample. Finally, only the point estimates by PREDICT, without its tumor biology, aggressiveness, and responsiveness to therapy . range, were included in the present analysis. On the other hand, the more aggressive the disease, the higher On the other hand, our study has several strengths. Our results the relevance of OS. Indeed, HER2-positive breast cancer tend to derive from a large cohort (n = 2794) of patients enrolled in the develop more early recurrences compared to hormone receptor largest, randomized adjuvant trial ever conducted in the field of positive/HER2-negative disease, thus having an undoubtedly more HER2-positive breast cancer. We included only patients receiving relevant impact on OS . adjuvant trastuzumab-based therapy started concurrently with In our analysis, the highest absolute differences between modern chemotherapy. Trial sample size allowed the exploration observed and predicted OS were observed for patients with of relevant patient subgroups. All data used for the analyses were hormone-receptor negative disease, larger tumor size, and high prospectively collected during the trial conduction, as detailed in number of nodes (≥4 positive lymph nodes), namely those the study protocol. patients traditionally considered at higher risk of relapse. Further In conclusion, in patients with HER2-positive early breast cancer investigations are urgently needed to better predict prognosis of enrolled in the ALTTO trial and treated with modern chemother- these patients. Of note, despite the traditional stigma of poor apy and trastuzumab-based therapies, the PREDICT score highly prognosis for patients with high-risk HER2-positive breast cancer, underestimated OS. The suboptimal performance of this prog- recent clinical trials have shown good outcomes also for this high- nostic tool was observed irrespective of type of anti-HER2 risk subset of patients . treatment, type of chemotherapy regimen, age of the patients The prediction of prognosis in patients with early breast cancer at the time of diagnosis, central hormone receptor status, is an issue of paramount importance, not only in hormone pathological nodal status, and pathological tumor size. Our results receptor-positive/HER2-negative disease, where prognostication suggest that the current version of PREDICT should be used with may settle whether adjuvant chemotherapy should be caution to give prognostic estimation in HER2-positive early breast Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. cancer patients treated in the modern era with effective The most updated ALTTO database was used for this analysis , which corresponds to at least 5-year follow-up for every single patient. chemotherapy and anti-HER2 targeted therapies. The further improvement of therapeutic strategies expected in the next future will likely increase the survival of patients with HER2-positive early Statistical analysis breast cancer, thus requiring the current version of PREDICT to be The present analysis should be considered as exploratory, since it was not updated to provide reliable prognostic estimation in these preplanned in the study protocol and the power of the statistical analyses patients. performed was not pre-specified. The prognostic performance of PREDICT was evaluated by assessing the following endpoints: i) calibration, defined as the agreement between the METHODS predicted and observed survival rates, and ii) discriminatory accuracy, defined as the ability of distinguishing individuals who will survive 5 years Study design and patients compared to those who will not (i.e. the ability to discern patients with Details of the ALTTO trial study design were previously published . good outcomes from those with poor outcomes at the individual patient Shortly, the ALTTO trial (Breast International Group [BIG] 2-06/EGF106708 level). and North Central Cancer Treatment Group [Alliance] N063D) was an The observation time for each patient was defined as the time between international, open-label, randomized phase III study testing the use of the date of diagnosis and an event. OS event was defined as death from trastuzumab and/or lapatinib as adjuvant anti-HER2 therapy in patients any cause. with HER2-positive early breast cancer. The median predicted 5-year OS was calculated from individual Primary tumor samples from all patients were centrally tested to assess predicted outcomes by PREDICT v. 2.2. 32 33 HER2 and hormone receptor status . For assessing calibration, the median predicted 5-year survival Eligible patients were randomized to one of four anti-HER2 treatment probabilities (by PREDICT) were compared with the observed 5-year arms: trastuzumab alone, lapatinib alone, sequential treatment with survival rates (as obtained by Kaplan-Meier curves). We had to use the trastuzumab for 12 weeks followed by a 6-week washout period before median 5-year prediction instead of the mean 5-year prediction, due to the other 34 weeks of lapatinib, and dual anti-HER2 blockade with trastuzumab skewness in the distribution, i.e. mean 5-year prediction was 83.6% while plus lapatinib. The CONSORT diagram of the ALTTO study is reported in the median 5-year prediction was 88.0%, and thus the mean predicted 5-year ALTTO primary analysis paper. survival probability underestimated the center of the distribution. There- Anti-HER2 treatment could be administered as per physician’s choice fore, we used the median as a robust estimator of the center of the following chemotherapy completion (design 1), or concomitantly, either distribution. Using the standard error as obtained by the Kaplan-Meier with a taxane after anthracycline-based chemotherapy (design 2) or with 6 curve, we calculated 95% CI for the difference in predicted vs. observed cycles of docetaxel and carboplatin in an anthracycline-free regimen 5-year survival. Calibration plots for PREDICT were constructed by (design 2B). In all treatment arms, adjuvant anti-HER2 therapy was visualizing mean predicted vs. observed survival outcomes by deciles of administered for 1 year. predicted outcomes. In 2011, after the first interim analysis, the lapatinib arm was closed and For assessing discriminatory accuracy, the area under the receiver- patients were offered adjuvant commercial trastuzumab . operator characteristic curve (AUC under the ROC) and corresponding 95% In the present analysis, in order to reflect current clinical practice in this CI for 5-year predicted OS were calculated. The AUC translates into the setting, only patients who received concurrent chemotherapy (design 2 probability that the predicted outcome of a randomly selected patient who and design 2B) and who received trastuzumab-based anti-HER2 therapy indeed had that outcome is higher than that of a patient who did not; the (i.e. trastuzumab alone arm, trastuzumab followed by lapatinib arm and higher the AUC, the better the tool is at identifying patients with a better trastuzumab plus lapatinib arm) were included. All patients originally survival. assigned to the lapatinib alone arm, and those who received anti-HER2 Subgroup analyses were performed to investigate the prognostic therapies at the completion of all chemotherapy (sequential treatment, performance of PREDICT according to the type of anti-HER2 treatment and chemotherapy received, age at the time of diagnosis, central hormone design 1) were excluded. receptor status, pathological nodal status, and pathological tumor size. Statistical analysis was performed by L.A. using SAS 9.4 statistical Ethics section software (SAS Institute, Cary, NC) and R. All patients signed a written informed consent prior to enrollment in ALTTO. The project proposal of the present exploratory analysis was Reporting summary submitted and approved by the ALTTO Steering Committee. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Study objectives The primary objective of the current analysis was to investigate the prognostic performance of PREDICT in breast cancer patients with early- DATA AVAILABILITY stage HER2-positive disease treated with modern chemotherapy and Data can be made available upon reasonable request. Data and results from the Data concurrent trastuzumab-based anti-HER2 therapy. Centre at Institut Jules Bordet in Brussels (Belgium) can be made available upon Secondary objectives were to investigate the prognostic performance of approval of a research proposal. PREDICT according to the type of anti-HER2 treatment received (trastuzumab alone, trastuzumab followed by lapatinib and, trastuzumab Received: 19 January 2022; Accepted: 1 July 2022; plus lapatinib), type of chemotherapy regimen received (anthracycline- based chemotherapy regimens vs. non-anthracycline-based chemotherapy regimens), age of patients at the time of diagnosis (age ≤ 40 years vs. age 41–64 vs. age ≥65 years), central hormone receptor status (hormone receptor -positive vs. negative), pathological nodal status (node-negative vs. node-positive disease [1–3 positive nodes] vs. node-positive disease [≥4 REFERENCES positive nodes]), and pathological tumor size (small [≤2 cm] vs. medium 1. Bradley, R. et al. Trastuzumab for early-stage, HER2-positive breast cancer: a [2–5 cm] vs. large [>5 cm] disease). meta-analysis of 13,864 women in seven randomised trials. Lancet Oncol. 22, 1139–1150 (2021). 2. Jackisch, C. et al. Risk-based decision-making in the treatment of HER2-positive Data extraction early breast cancer: recommendations based on the current state of knowledge. PREDICT estimates for each patient were calculated by one investigator Cancer Treat. Rev. 99, 10222 (2021). blinded to patient outcomes. Patient and tumor characteristics, as well as 3. Pernas, S. & Tolaney, S. M. Management of early-stage human epidermal growth administered adjuvant anticancer treatments, were entered in the PREDICT factor receptor 2-positive breast cancer. JCO Oncol. Pract. 17, 320–330 (2021). v.2.2 program to calculate the predicted 5-year OS for each patient. 4. Wishart, G. C. et al. PREDICT: a new UK prognostic model that predicts survival Detection modality and Ki67 status were considered “unknown” for all following surgery for invasive breast cancer. Breast Cancer Res. 12, R1 (2010). patients (as these variables were not collected as part of the ALTTO trial). npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation E. Agostinetto et al. 5. Wishart, G. C. et al. A population-based validation of the prognostic model 32. Wolff, A. C. et al. American Society of Clinical Oncology/College of American PREDICT for early breast cancer. Eur. J. Surg. Oncol. 37, 411–417 (2011). Pathologists guideline recommendations for human epidermal growth factor 6. The Breast Cancer Association Consortium (BCAC). http://bcac.ccge.medschl. receptor 2 testing in breast cancer. J. Clin. Oncol. 25, 118–145 (2007). cam.ac.uk/. 33. Lambertini, M. et al. Dissecting the effect of hormone receptor status in patients 7. Wishart, G. C. et al. PREDICT Plus: development and validation of a prognostic with HER2-positive early breast cancer: exploratory analysis from the ALTTO (BIG model for early breast cancer that includes HER2. Br. J. Cancer 107, 800–807 2-06) randomized clinical trial. Breast Cancer Res. Treat. 177, 103–114 (2019). (2012). 8. NICE. Tumour profiling tests to guide adjuvant chemotherapy decisions in early breast cancer. https://www.nice.org.uk/guidance/dg34/ (2018). ACKNOWLEDGEMENTS 9. Moreno-Aspitia, A. et al. Updated results from the international phase III ALTTO This work was supported by Les Amis de l’Institut Bordet (Grant No. 2018-09). The trial (BIG 2-06/Alliance N063D). Eur. J. Cancer 148, 287–296 (2021). ALTTO trial received financial support from GlaxoSmithKline (until January 2015), 10. Joensuu, H. et al. Fluorouracil, epirubicin, and cyclophosphamide with either Novartis Pharma AG (as of January 2015) and the National Cancer Institute of the docetaxel or vinorelbine, with or without trastuzumab, as adjuvant treatments of National Institutes of Health (NCI-NIH; Grant No. U10CA180821 and U10CA180882 to breast cancer: final results of the FinHer Trial. J. Clin. Oncol. 27, 5685–5692 (2009). the Alliance for Clinical Trials in Oncology and Grant No. CA025224 to the legacy 11. Smith, I. et al. 2-year follow-up of trastuzumab after adjuvant chemotherapy in North Central Cancer Treatment Group). Matteo Lambertini acknowledges the HER2-positive breast cancer: a randomised controlled trial. Lancet 369,29–36 support from the Italian Association for Cancer Research (“Associazione Italiana per la (2007). Ricerca sul Cancro”, AIRC; MFAG 2020 ID 24698) and the Italian Ministry of Health (5 × 12. Romond, E. H. et al. Trastuzumab plus adjuvant chemotherapy for operable HER2- 1000 funds 2017). The funders of the study had no role in study design, data positive breast cancer. N. Engl. J. Med. 353, 1673–1684 (2005). collection, data analysis, data interpretation, or writing of the present manuscript. All 13. Perez, E. A. et al. Four-year follow-up of trastuzumab plus adjuvant chemotherapy authors agreed and share the final responsibility for the provided interpretation of for operable human epidermal growth factor receptor 2-positive breast cancer: the study results and for the decision to submit for publication. The abstract of the joint analysis of data from NCCTG N9831 and NSABP B-31. J. Clin. Oncol. 29, present work was presented at the American Society of Clinical Oncology (ASCO) 3366–3373 (2011). Annual Meeting 2021 and was awarded an ASCO Merit Award. 14. Slamon, D. et al. Adjuvant trastuzumab in HER2-positive breast cancer. N. Engl. J. Med. 365, 1273–1283 (2011). 15. Swain, S. M. et al. Pertuzumab, trastuzumab, and docetaxel for HER2-positive AUTHOR CONTRIBUTIONS metastatic breast cancer (CLEOPATRA): end-of-study results from a double-blind, M.L. conceived and designed the study. E.A. and E.d.A. were involved in data randomised, placebo-controlled, phase 3 study. Lancet Oncol. 21, 519–530 (2020). acquisition. L.A. analyzed the data. All authors were involved in data interpretation. 16. Diéras, V. et al. Trastuzumab emtansine versus capecitabine plus lapatinib in E.A., E.d.A., and M.L. wrote the first draft of the paper. LA and EA produced the patients with previously treated HER2-positive advanced breast cancer (EMILIA): a manuscript figures and tables. All authors were involved in reviewing and editing descriptive analysis of final overall survival results from a randomised, open-label, drafts of the paper and approving the manuscript. phase 3 trial. Lancet Oncol. 18, 732–742 (2017). 17. Saturni, S. et al. Randomized controlled trials and real life studies. Approaches and methodologies: a clinical point of view. Pulm. Pharmacol. Ther. 27, 129–138 COMPETING INTERESTS (2014). 18. Rothwell, P. M. External validity of randomised controlled trials: “To whom do the The authors have no known competing financial interests or personal relationships results of this trial apply?” Lancet 365,82–93 (2005). that could have appeared to influence the work reported in this paper. Elisa 19. Gray, E. et al. Independent validation of the PREDICT breast cancer prognosis Agostinetto: Consultancy/honoraria from Eli Lilly, Sandoz. Support to attend medical prediction tool in 45,789 patients using Scottish Cancer Registry data. Br. J. Cancer conferences (travel accommodations/expenses) from Novartis, Roche, Eli Lilly, 119, 808–814 (2018). Genetic, Istituto Gentili (outside the submitted work). Samuel Martel: Honoraria 20. Hudis, C. A. et al. Proposal for standardized definitions for efficacy end points in and advisory board for Novartis, Pfizer, Knight Therapeutics, Exact Science, Taiho, adjuvant breast cancer trials: the STEEP system. J. Clin. Oncol. 25, 2127–2132 Apobiologix (outside the submitted work). Philippe Aftimos: Consulting: Boehringer (2007). Ingelheim, Macrogenics, Roche, Novartis, Amcure, Servier, G1 Therapeutics, Radius, 21. van Maaren, M. C. et al. Ten-year recurrence rates for breast cancer subtypes in Deloitte, Menarini; Honoraria: Synthon, Amgen, Novartis, Gilead; Travel grants: the Netherlands: a large population-based study. Int. J. Cancer 144, 263–272 Amgen, MSD, Pfizer, Roche; Research funding to my institution: Roche (outside the (2019). submitted work). Noam Pondé: Honoraria and advisory board for Lilly, Novartis, 22. Piccart, M. et al. Adjuvant pertuzumab and trastuzumab in early HER2-positive Roche/Genentech, AstraZeneca and Pfizer (outside the submitted work). Christian breast cancer in the APHINITY trial: 6 years’ follow-up. J. Clin. Oncol. 39, Maurer: Travel grants: Mundipharma, Amgen, Servier Deutschland GmbH, Abbvie. 1448–1457 (2021). Honoraria: Abbvie. Advisory board: Celgene/ BMS and Pfizer (outside the submitted 23. Lambertini, M. et al. Pregnancies during and after trastuzumab and/or lapatinib in work). Sarra El-Abed: grants from Novartis during the conduct of the ALTTO study, patients with human epidermal growth factor receptor 2–positive early breast and grants from Roche/Genentech and Pfizer (outside of the submitted work). Yingbo cancer: analysis from the NeoALTTO (BIG 1-06) and ALTTO (BIG 2-06) trials. Cancer Wang: Employee of Novartis Pharmaceuticals and holds stock in the company 125, 307–316 (2019). (outside the submitted work). Saranya Chumsri: Research funding from Merck and 24. Burstein, H. J. et al. Adjuvant endocrine therapy for women with hormone Pfizer (outside the submitted work). Judith Bliss: Research Funding to her Institution: receptor–positive breast cancer: asco clinical practice guideline focused update. AstraZeneca, Merck Sharp & Dohme, Medivation, Puma Biotechnology, Clovis J. Clin. Oncol. 37, 423–438 (2018). Oncology, Pfizer, Janssen-Cilag, Roche, Novartis (previously GSK), Eli Lilly (outside 25. Puppe, J. et al. Genomic signatures in luminal breast cancer. Breast Care 15, the submitted work). Judith Kroep: Honoraria and advisory board: AstraZeneca, 355–365 (2020). Daiichi Sankyo, Eisai, Lilly, MSD, Novartis, Pfizer, GSK. Research grant: Amgen, Astra 26. Prat, A. et al. A multivariable prognostic score to guide systemic therapy in early- Zeneca, Novartis, Philips (outside the submitted work). Marco Colleoni: Research stage HER2-positive breast cancer: a retrospective study with an external eva- grant to his Institution: Roche (outside the submitted work). Lucia Del Mastro: luation. Lancet Oncol. 21, 1455–1464 (2020). Honoraria: Roche, Novartis, Eli Lilly, MSD, Pfizer, Ipsen. Honoraria to her Institution: 27. Cardoso, F. et al. 70-gene signature as an aid to treatment decisions in early-stage Novartis. Consulting or Advisory role: Roche, Eli Lilly, Novartis, MSD, Genomic Health, breast cancer. N. Engl. J. Med. 375, 717–729 (2016). Pierre Fabre, Daiichi Sankyo, Seagen, Astrazeneca, Eisai, Gilead, Exact Science. Travel, 28. Cardoso, F. et al. MINDACT: Long-term results of the large prospective trial testing accomodation, expenses: Celgene, Roche, Pfizer, Daiichi Sankyo (outside the the 70-gene signature MammaPrint as guidance for adjuvant chemotherapy in submitted work). Alvaro Moreno Aspitia: Institutional research funding from breast cancer patients. J. Clin. Oncol. 38, 506 (2020). GlaxoSmithKline, Genentech, Daiichi Sankyo, Novartis, Merck, Pfizer, AbbVie and Eli 29. Hurvitz, S. A. HER2DX: a tool that might inform treatment choices for HER2- Lilly (outside the submitted work). Martine Piccart: Board Member (Scientific Board): positive breast cancer. Lancet Oncol. 21, 1392–1393 (2020). Oncolytics. Consultant (honoraria): AstraZeneca, Camel-IDS, Immunomedics, Lilly, 30. Conte, P. et al. Nine weeks versus 1 year adjuvant trastuzumab in combination Menarini, MSD, Novartis, Odonate, Pfizer, Roche-Genentech, Seattle Genetics, with chemotherapy: final results of the phase III randomized Short-HER study‡. Immutep, Seagen, NBE Therapeutics. Research grants to her Institute: AstraZeneca, Ann. Oncol. 29, 2328–2333 (2018). Immunomedics, Lilly, Menarini, MSD, Novartis, Pfizer, Radius, Roche-Genentech, 31. Piccart-Gebhart, M. et al. Adjuvant Lapatinib and Trastuzumab for early human Servier, Synthon (outside the submitted work). Evandro de Azambuja: honoraria and epidermal growth factor receptor 2-positive breast cancer: results from the advisory board: Roche/GNE, Novartis, Seattle Genetics, Zodiacs, Libbs and Pierre randomized phase III adjuvant lapatinib and/or trastuzumab treatment optimi- Fabre; travel grants: Roche/GNE, GSK/Novartis. Research grant for his Institute: Roche/ zation trial. J. Clin. Oncol. 34, 1034–1042 (2016). GNE, AstraZeneca, Novartis, and Servier (outside the submitted work). Matteo Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2022) 87 E. Agostinetto et al. Lambertini acted as a consultant for Roche, Lilly, AstraZeneca, Exact Sciences, Seagen, Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims Gilead, Pfizer, MSD and Novartis, and received honoraria from Sandoz, Takeda, Roche, in published maps and institutional affiliations. Lilly, Pfizer, Ipsen, Libbs, Knight and Novartis (outside the submitted work). Lieveke Ameye, Marianne Paesmans, Malou Vicente, and Fausto Petrelli have no conflict of interests to declare. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give ADDITIONAL INFORMATION appropriate credit to the original author(s) and the source, provide a link to the Creative Supplementary information The online version contains supplementary material Commons license, and indicate if changes were made. The images or other third party available at https://doi.org/10.1038/s41523-022-00452-8. material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the Correspondence and requests for materials should be addressed to Matteo article’s Creative Commons license and your intended use is not permitted by statutory Lambertini. regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http:// Reprints and permission information is available at http://www.nature.com/ creativecommons.org/licenses/by/4.0/. reprints © The Author(s) 2022 npj Breast Cancer (2022) 87 Published in partnership with the Breast Cancer Research Foundation

Journal

npj Breast CancerSpringer Journals

Published: Jul 20, 2022

There are no references for this article.