Neoadjuvant Therapy with Weekly Nanoparticle Albumin‐Bound Paclitaxel for Luminal Early Breast Cancer Patients: Results from the NABRAX Study (GEICAM/2011‐02), a Multicenter, Non‐Randomized, Phase II Trial, with a Companion Biomarker Analysis

The NABRAX study was designed to evaluate the antitumor activity and safety of single‐agent weekly nab‐paclitaxel as neoadjuvant treatment of estrogen receptor‐positive/HER2‐negative breast cancer patients. This article reports on the trial and efforts to define the role of this drug and biomarkers of activity in this particular subtype of breast cancer.

Abstract

Background.

Nanoparticle albumin‐bound paclitaxel (nab‐Paclitaxel) is an alternative to standard taxanes for breast cancer (BC) treatment. We evaluated nab‐Paclitaxel efficacy as neoadjuvant treatment for early estrogen receptor‐positive (ER+), human epidermal growth factor receptor 2‐negative (HER2‐) disease.

Materials and Methods.

Women with ER+, HER2‐, stage II–III BC were treated preoperatively with four cycles of weekly nab‐Paclitaxel (150 mg/m²), 3 weeks on and 1 week off. We hypothesized that poor pathological response rate (residual cancer burden [RCB] III; Symmans criteria) would be ≤16%.

Results.

Eighty‐one patients with a median age of 47 years were treated; 64.2% were premenopausal, and 69% of tumors were stage II. Residual cancer burden III rate was 28.4% (95% confidence interval [CI]: 18.6%–38.2%), RCB 0+I (good response) rate was 24.7% (95% CI: 15.3%–34.1%) and RCB 0 (complete response) rate was 7.4% (95% CI: 1.7%–13.1%). Objective response rate by magnetic resonance imaging was 76.5% and rate of conversion to breast conserving surgery was 40.0%. The most frequent grade 3 and 4 toxicity was neutropenia (12.3% and 3.7% of patients, respectively), without any febrile neutropenia. Sensory neuropathy grade 2 and 3 were seen in 25.9% and 2.5% of patients, respectively. Tumor secreted protein, acidic, cysteine‐rich (SPARC) overexpression was significantly associated with RCB 0 (odds ratio: 0.079; 95% CI: 0.009–0.689; p = .0216).

Conclusion.

Despite failing to confirm an RCB III rate ≤16% in nab‐Paclitaxel‐treated patients, the RCB 0+I rate indicates a significant drug antitumor activity with low rates of grade 3–4 toxicity. Our exploratory biomarker analysis suggests a potential predictive role of complete response for SPARC. Confirmatory analyses are warranted, adapting dose and schedule to decrease peripheral neurotoxicity. (Trial registration: European Clinical Trials Database study number: 2011‐004476‐10; ClinicalTrials.gov: NCT01565499).

Implications for Practice.

The pathological response rate (residual cancer burden [RCB]; Symmans criteria) of nanoparticle albumin‐bound paclitaxel administered as neoadjuvant treatment for early estrogen receptor‐positive, human epidermal growth factor receptor 2‐negative disease was evaluated. Whereas poor response (RCB III) was 24.7%, similar to that for docetaxel, good response (RCB 0+I) reached 23.0%, far superior to the 13% for docetaxel, while keeping toxicity low. Exploratory biomarker analysis suggests secreted protein, acidic, cysteine‐rich overexpression in tumor cells as a potential predictor of complete response (RCB 0). Findings point to an encouraging single‐agent neoadjuvant treatment with low toxicity, which warrants future research and development.

Introduction

Neoadjuvant chemotherapy of hormone receptor (HR)‐positive/human epidermal growth factor receptor 2 (HER2)‐negative breast cancer (BC) can produce tumor shrinkage in the majority of patients, allowing for a more conservative surgery (i.e., breast‐preserving surgery) in some of them [1]. In addition, a good pathological response to neoadjuvant therapy is associated with better outcomes [2]. However, the prognostic value of obtaining a pathological complete response (pCR) in this subtype is probably less marked than in others, like triple negative (TN) and HER2‐positive/HR‐negative [3].

In addition, the neoadjuvant setting is an excellent model for testing the activity of new compounds and searching for predictive biomarkers.

Taxanes, paclitaxel and docetaxel, are among the most effective treatments for BC [4] and are considered a significant component of neoadjuvant BC regimens [5]. However, the hydrophobic nature of conventional taxanes mandates the concomitant use of synthetic solvents, which may be responsible for some of their toxicity [6], [7], and may limit their efficacy by reducing the intratumoral concentrations of the active drug [8].

Nanoparticle albumin‐bound paclitaxel (nab‐Paclitaxel) is a formulation of paclitaxel that consists of nanometer‐range particles of paclitaxel, bound to human serum albumin. This union increases drug solubility, improving its delivery to tumor cells and thus avoiding the need for solvents [9].

In the pivotal phase III trial of nab‐Paclitaxel monotherapy versus solvent‐based paclitaxel, the objective response rate (ORR) and time to progression were significantly increased in all patients treated with nab‐Paclitaxel. Overall survival (OS) was also increased for patients treated in second or subsequent lines [10]. Nab‐Paclitaxel has also been evaluated in the neoadjuvant setting of BC patients [11], [12], [13], [14], [15], [16], [17], [18].

All of these neoadjuvant trials administered nab‐Paclitaxel in combination with other drugs and usually in mixed BC subtypes. We designed a trial (the NABRAX study) to evaluate the antitumor activity and safety of single‐agent weekly nab‐Paclitaxel as neoadjuvant treatment of estrogen receptor (ER)‐positive/HER2‐negative BC patients. Our main aims are to further define the role this drug plays in this particular subtype of BC and to search for biomarkers of activity.

Materials and Methods

Study Design

This is a multicenter, single‐arm, phase II trial. All eligible patients were administered intravenous nab‐Paclitaxel weekly (150 mg/m²) on days 1, 8, and 15, every 4 weeks, for four cycles. Upon completion of nab‐Paclitaxel, patients underwent mastectomy or breast‐conserving surgery (BCS), plus axillary lymph node dissection (unless previous sentinel lymph node biopsy had been negative) within a maximum of 6 weeks after the last dose of the study drug. Adjuvant treatment was left at the investigator criteria.

This trial was approved by all the participating institutions’ Ethical Review Boards and the Spanish Health Authorities and registered in European Clinical Trials Database (2011‐004476‐10) and ClinicalTrial.gov (NCT01565499). It was conducted in compliance with Good Clinical Practices and the tenets of the Declaration of Helsinki. Written informed consent was obtained from all patients before study entry.

Eligibility Criteria

Patients aged >18 years, with an Eastern Cooperative Oncology Group Performance Status (ECOG PS) <2, were eligible for the trial if they had histologically proven primary unilateral invasive early BC through a core biopsy, and their largest tumor size in the breast was ≥2 cm, or <2 cm if positive for axillary involvement. Tumors had to be ER‐positive (>1%) and HER2‐negative (immunohistochemistry [IHC] 0 or 1+, or negative gene amplification by fluorescence/chromogenic in situ hybridization [ISH]) by local determinations. Patients were required to have adequate bone marrow, renal, and liver functions and be oncology treatment‐naïve. For potentially fertile women, adequate contraception and a negative pregnancy test were required.

Patients were excluded if they had inflammatory BC (T4d), supraclavicular lymph nodes (N3), synchronous contralateral or multicentric BC, evidence of metastatic disease or a pre‐existing neurotoxicity grade ≥2 (based on the National Cancer Institute‐Common Terminology Criteria for Adverse Events [NCI‐CTCAE] version 4.0, http://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm). Finally, patients were also excluded if they had a previous history of cancer other than adequately treated carcinoma in situ of the cervix, stage I colon cancer, skin carcinoma (non‐melanoma), or other malignant tumors treated more than 5 years before the study entry with no subsequent evidence of recurrence.

Assessments and Endpoints

The primary endpoint of the study was to determine the rate of poor pathological response (residual cancer burden [RCB] III). Secondary endpoints included rate of good pathological response (RCB 0+I), ORR, rate of conversion to BCS, invasive disease‐free survival, and toxicity. Exploratory analyses of tumor and blood samples for potentially predictive biomarkers were centrally performed.

The following baseline procedures were performed before study entry: (a) breast tumor assessment by physical exam; (b) bilateral breast mammogram and magnetic resonance imaging (MRI); (c) lymph node status determination by axillary ultrasound (with pathological confirmation in case of suspicion); (d) ECOG PS evaluation; (e) hematology; (f) serum chemistry; (g) pregnancy test (potentially fertile women only); (h) electrocardiogram; and (i) local determination of ER/progesterone receptor (PgR)/HER2 expression levels. Eastern Cooperative Oncology Group Performance Status evaluation, hematology, and serum chemistry were repeated before each cycle and the breast tumor assessment was repeated before surgery.

Pathological response was centrally assessed at surgery following the Symmans classification criteria [2]. Objective response rate was evaluated according to the Response Evaluation Criteria in Solid Tumors 1.1 criteria [19] before surgery. Toxicities were assessed after each cycle and graded according to the NCI‐CTCAE version 4.0.

Putative Predictive Biomarker Assessment

We analyzed data from patients with available centralized biomarker determination and pathological response information.

Biomarker analysis by IHC and/or ISH in Formalin‐fixed paraffin‐embedded tumor samples was carried out at a central laboratory by a pathologist blinded to clinical parameters (see additional information in supplemental online Appendix 2). The assessment of ER, PgR, HER2, and Ki67 was performed following the American Society of Clinical Oncology and the College of American Pathologists guidelines [20], [21], [22]. The cut‐off considered for Ki67 expression was 20% of positively stained tumor cells [23]. Secreted protein, acidic, cysteine‐rich was evaluated for both tumor and stroma as described previously [24]. Its expression was categorized as negative when the intensity was absent‐to‐weak (+), or moderate (++)‐to‐strong (+++) with a proportion of stained cells <10%. Immunolabeling was positive if the intensity was moderate (++)‐to‐strong (+++) and the extent of staining was ≥10%. Caveolin (Cav)‐1 was evaluated in the stroma and its expression was categorized in low, moderate, or high. The high expression of Cav‐1 was considered as positive. Molecular subtypes were classified according to St. Gallen criteria 2013 [23] and Prat et al. [25] into Luminal A (ER+, PgR >20%, HER2‐, Ki67 <14%), Luminal B1 (ER+, HER2‐, PgR ≤20% and/or Ki67 ≥14%), Luminal B2 (ER+, HER2+, any PgR, any Ki67), TN (ER‐, PgR‐, HER2‐), and HER2‐enriched (ER‐, PgR‐, HER2+) subtypes.

Statistical Considerations

The sample size was calculated based on the null hypothesis of a poor pathological response (RCB III) of 33% (seen in patients with luminal BC when treated with single‐agent docetaxel [26]) and an alternative hypothesis (H1) of RCB III of 16% or less with single‐agent nab‐Paclitaxel. For significance (alpha) of 0.05 and a two‐tailed test, 70 evaluable patients had to be included in the study to yield a statistically significant result with a power of 91.8%. Assuming a dropout rate of 10%, 78 was the total number of patients to be recruited into this study.

The efficacy and safety variables were analyzed in the intent‐to‐treat population (defined as study patients receiving at least one dose of the study drug).

The number and proportion of patients experiencing the different grades of pathological response (III, 0, 0+I), and their corresponding two‐sided 95% CIs, were calculated.

Univariate analyses and multivariate logistic regression analyses were used to study the association of biomarkers with pathological response. The biomarkers were also analyzed by Fisher's exact test.

All analyses were performed using SAS Enterprise Guide 5.1 software (SAS Institute, Cary, NC, https://www.sas.com/en_us/home.html).

Results

Patient Characteristics

Between April 2012 and January 2013, 83 patients were registered in 15 Spanish sites. Two of them never received treatment and were excluded from the analysis. The median age was 47 years (range: 28–75 years), 64% of patients were premenopausal, 97.5% had an ECOG PS of 0, and 69.0% and 22.0% were diagnosed with stage II and III, respectively. Most tumors (91.4%) had histology of invasive ductal carcinoma, 48% were histological grade 2, 32% were grade 3, and 21% were PgR negative.

Dose Administration

The median relative dose‐intensity for nab‐Paclitaxel was 98.5%. Of the 81 patients analyzed, six discontinued treatment early; thus, 75 patients completed the treatment as planned. Reasons for discontinuation included adverse events (five patients, three of them due to sensory neuropathy) and consent withdrawal (one patient). Concerning dose modifications, 5.3% of the cycles were omitted, another 5% of the cycles were delayed, and 10.6% of nab‐Paclitaxel doses had to be reduced. Main reasons for dose modification were neutropenia (in 25.9% of patients) and sensory neuropathy (in 18.5% of patients).

Toxicity

Table 1 summarizes all frequent grade 2–4 toxicities per patient and per cycle. The most frequent grade 3–4 toxicities were neutropenia in 13 patients (16%) and sensory neuropathy in two patients (2.5%). Two serious cases of adverse events related with nab‐Paclitaxel were reported; both were grade 3 sensory neuropathy that descended to grade 1 following treatment discontinuation. However, these adverse effects were still present at the two‐year follow‐up.

Table 1. Grade 2–4 related adverse events (toxicity) based on the National Cancer Institute‐Common Terminology Criteria for Adverse Events version 4.0.

Abbreviation: G, grade.

Efficacy

The RCB evaluation was available for 80 of the 81 treated patients (one patient received three cycles but refused further treatment and withdrew the informed consent). The RCB III rate in the 81 treated patients was 28.4% (95% CI: 18.6%–38.2%); 24.7% (95% CI: 15.3%–34.1%) of patients achieved a good pathological response (RCB 0+I), with six patients (7.4%; 95% CI: 1.7%–13.1%) achieving pathological complete response (RCB 0).

Three patients included in the trial were reclassified as TN by the central laboratory (see Tumor Biomarkers and Clinical‐Pathological Variables Predictive of Efficacy section). Two of them achieved pCR and one achieved an RCB type I response. When excluding these three patients, the rates of RCB III and RCB 0+I were 29.7% and 23.0%, respectively.

The ORR measured by MRI was 76.5% (95% CI: 67.3%–85.7%) and 60.5% (95% CI: 49.9%–71.1%) when measured by mammography.

Breast‐conserving surgery was performed in 49 patients (60.5%). The rate of conversion to BCS in patient candidates for a mastectomy at time of diagnosis was 40%.

Tumor Biomarkers and Clinical‐Pathological Variables Predictive of Efficacy

Of the 81 patients, 77 had available pretreatment tumor sample for biomarker analysis (ER, PgR, HER2, Ki67, Cav‐1 and SPARC) in a central laboratory. The distribution by BC subtypes following St. Gallen 2013 and Prat et al. criteria [23], [25] was as follows: Luminal A (25%), Luminal B1 (71%), and TN (4%). Of these 77 patients, five were excluded from biomarker analysis (three patients [4%] due to central TN phenotype, one due to missing pathological response data, and one due to incomplete central biomarker data).

Residual cancer burden 0, 0+I, and RCB III were correlated with biomarker expression in 72 patients (Tables 2, 3, and 4, respectively). In the univariate analysis, there was no correlation between RCB III and any biomarker or subtype (data not shown). Tumor size <20 mm (p = .0029), absence of regional lymph node metastases (p = .0005), and Cav‐1 overexpression in stroma (p = .0320) seemed to be associated with RCB 0+I (supplemental online Table 1). In the multivariate analysis, only tumor size (odds ratio [OR]: 0.070; 95% CI: 0.008–0.635; p = .0181) and lymph node metastases (OR: 0.070; 95% CI: 0.012–0.394; p = .0026) were significantly associated with RCB (0+I; supplemental online Table 2). Finally, tumor SPARC overexpression was associated with RCB 0 in the univariate analyses (OR: 0.079; 95% CI: 0.009–0.689; p = .0216; supplemental online Table 3).

Table 2. Correlation between individual biomarkers and clinicopathological parameters and pathological complete response (RCB 0).

^a7th Edition AJCC Cancer Staging Atlas (2010).

Bolded values indicate statistically significant.

Abbreviations: AJCC, American Joint Committee on Cancer; Cav‐1, caveolin‐1; NA, not applicable; PgR, progesterone receptor; pN, pathological lymph nodes; pT, pathological tumor size; RCB, residual cancer burden; SPARC, tumor secreted protein, acidic, cysteine‐rich.

Table 3. Correlation between individual biomarkers and clinico‐pathological parameters and good pathological response (RCB 0+I).

^a7th Edition AJCC Cancer Staging Atlas (2010).

Bolded values indicate statistically significant.

Abbreviations: AJCC, American Joint Committee on Cancer; Cav‐1, caveolin‐1; NA, not applicable; PgR, progesterone receptor; pN, pathological lymph nodes; pT, pathological tumor size; RCB, residual cancer burden; SPARC, tumor secreted protein, acidic, cysteine‐rich.

Table 4. Correlation between individual biomarkers and clinico‐pathological parameters and poor pathological response (RCB III).

^a7th Edition AJCC Cancer Staging Atlas (2010).

Bolded values indicate statistically significant.

Abbreviations: AJCC, American Joint Committee on Cancer; Cav‐1, caveolin‐1; NA, not applicable; PgR, progesterone receptor; pN, pathological lymph nodes; pT, pathological tumor size; RCB, residual cancer burden; SPARC, tumor secreted protein, acidic, cysteine‐rich.

Discussion

The NABRAX study failed to confirm the a priori antitumor activity hypothesis (a Symmans RCB III rate of 16% or less in ER‐positive/HER2‐negative BC) because the observed rate was 28.4% (95% CI: 18.6%–38.2%). We have no evidence of any difference in the poor response rate experienced by patients receiving nab‐Paclitaxel and the 33.0% poor response observed in historical controls treated with docetaxel (100 mg/m2) [26].

However, we found a remarkable rate of good pathological response with nab‐Paclitaxel (Symmans RCB 0+I). Twenty of the 81 treated patients, (24.7%, 95% CI: 15.3%–34.1%) achieved RCB 0+I and six patients (7.4%, 95% CI: 1.7%–13.1%) had a pathological complete response (RCB 0). Excluding those three patients reclassified as having TN phenotype BC in the central review, the rate of RCB 0+I was 23.0%. In the historical control series treated with docetaxel, upon which the statistical assumption of the NABRAX study was based [26], the rate of RCB 0+I seen in Luminal A and B patients was 13%, apparently inferior to that observed in the NABRAX trial, although inter‐study comparisons should be interpreted with extreme caution. In support of the good antitumor activity of nab‐Paclitaxel in this population, the ORR by MRI in the NABRAX trial was 76.5% (95% CI: 67.3%–85.7%) and the rate of conversion to BCS in patients who are candidates for mastectomy at diagnosis reached 40%.

Two studies have studied nab‐Paclitaxel in comparison with conventional paclitaxel in the neoadjuvant setting in early‐stage high‐risk BC patients, with inconclusive results. The GEPARSEPTO study found that nab‐Paclitaxel was superior to conventional paclitaxel in achieving pCR (p = .00065) [16]. In that trial, both nab‐Paclitaxel and conventional paclitaxel were administered in 12 weekly doses in consecutive weeks and followed by sequential anthracyclines. This trial started with a dose of 150 mg/m2 of nab‐Paclitaxel but, after a preplanned safety interim analysis, the dose had to be reduced to 125 mg/m2 due to toxicity. The Evaluating Treatment with Neoadjuvant Abraxane trial showed a trend to a superiority of nab‐Paclitaxel but it was not statistically significant (p = .127). In that trial, nab‐Paclitaxel was administered weekly (125 mg/m2; 3 weeks on and 1 week off, for four cycles) and followed by sequential anthracyclines [11]. In the NABRAX study, the schedule administered (150 mg/m2; 3 weeks on and 1 week off, for four cycles) allowed a better tolerance as shown by the median relative dose‐intensity of nab‐Paclitaxel (98.5%) and the low rates of toxicity, apart from grade 2 + 3 peripheral neuropathy that was reported by 28.5% of patients, causing early discontinuation in three patients.

Neoadjuvant hormonotherapy for HR‐positive/HER2‐negative tumors, an alternative approach to neoadjuvant chemotherapy, is gaining ground in some countries and in the research setting [27], [28], [29], [30]. In some studies comparing neoadjuvant hormones versus neoadjuvant chemotherapy in ER‐positive tumors, the clinical response rates were similar with both therapies [31], [32]. However, these results cannot rule out that some of these patients (i.e., those with high Ki67) could benefit from the use of both therapies in sequence. Therefore, neoadjuvant hormonal therapy (which is not usually followed by adjuvant chemotherapy after surgery) should be reserved for patients whose tumors are not candidates for chemotherapy based on proliferation index or genomic profiling.

Previous findings in lung cancer support that higher Cav‐1 levels in tumor‐associated stroma are associated with better ORR and OS [33]. Similarly, our biomarker analysis found that high expression of Cav‐1 was correlated with good pathological response in the univariate analysis but not in the multivariate analysis. Stromal Cav‐1 is a protein that favors biomechanical remodeling of the microenvironment and tumor invasion [34].

The role of SPARC in cancer is controversial. Most studies suggest a protumorigenic role, but others suggest an antitumorigenic activity [35]. In the NABRAX study, SPARC expression in tumor was associated with pathological complete response (RCB 0) in the univariate analyses. Similarly to other proteins involved in extracellular matrix remodeling and invasion in human BC, expression of SPARC in tumor cells has been associated with reduction of characteristic epithelial markers and acquisition of an aggressive and stem cell‐related phenotype in these tumors [36]. Similar findings have been reported in other epithelial tumor types (lung, prostate, ovarian, or endometrial), suggesting the involvement of SPARC present in tumor cells in cancer growth, apoptosis and metastasis, cell migration, and stroma formation [37], [38]. Moreover, SPARC tumor expression and stem cell‐related phenotype have been extensively correlated with response to systemic therapies. In BC, SPARC has been associated with a higher chance of achieving a pathological complete remission after Docetaxel/Doxorubicine/Cyclophosphamide or TAC + Four Cycles of Vinorelbine and Capecitabine (NX) chemotherapy [39]. It has also been hypothesized that binding of SPARC to albumin in the tumor microenvironment may enrich the concentration of nab‐Paclitaxel in head and neck tumors, enhancing its antitumor activity [40]. Furthermore, in a phase I/II trial in pancreas cancer patients treated with gemcitabine and nab‐Paclitaxel, the high expression of SPARC was correlated with improved OS. Patients with high levels of SPARC had a median survival of 17.8 months and patients with low SPARC expression had a median survival of 8.1 months [41]. However, results from the phase III MPACT trial of nab‐Paclitaxel plus gemcitabine versus gemcitabine alone for patients with metastatic pancreatic cancer showed that there was no association between stromal, tumor epithelial, or plasma SPARC levels and efficacy in terms of OS [24]. This result is in line with GEPARSEPTO trial [16]. The role of SPARC in tumor cells warrants further study. The interest of Cav‐1 and SPARC as predictors of response to nab‐Paclitaxel should be validated in other, larger, cohorts of patients.

Conclusion

Our patient response rate related to nab‐Paclitaxel failed to support its superiority as a neoadjuvant treatment, in terms of reduction of poor pathological response rate, with respect to historical controls treated with docetaxel. However, it showed an encouraging single‐agent activity in patients with ER‐positive/HER2‐negative BC that warrants future development in the neoadjuvant setting, maybe adapting dose and schedule to decrease peripheral neurotoxicity.

See http://www.TheOncologist.com for supplemental material available online.

Author Contributions

Conception/design: Miguel Martín

Provision of study material or patients: Miguel Martín, César Rodríguez‐Martín, Rosalía Caballero, María I. Casas, Federico Rojo, Eva Carrasco

Collection and/or assembly of data: José I. Chacón, Antonio Antón, Arrate Plazaola, Elena García‐Martínez, Miguel A. Seguí, Pedro Sánchez‐Rovira, José Palacios, Lourdes Calvo, Carmen Esteban, Enrique Espinosa, Agusti Barnadas, Norberto Batista, Angel Guerrero, Montserrat Muñoz, Estefania Romio, Silvia Antolín

Data analysis and interpretation: Miguel Martín, César Rodríguez‐Martín, Rosalía Caballero, Federico Rojo, Eva Carrasco

Manuscript writing: Miguel Martín, José I. Chacón, Antonio Antón, Arrate Plazaola, Elena García‐Martínez, Miguel A. Seguí, Pedro Sánchez‐Rovira, José Palacios, Lourdes Calvo, Carmen Esteban, Enrique Espinosa, Agusti Barnadas, Norberto Batista, Angel Guerrero, Montserrat Muñoz, Estefania Romio, César Rodríguez‐Martín, Rosalía Caballero, María I. Casas, Federico Rojo, Eva Carrasco, Silvia Antolín

Final approval of manuscript: Miguel Martín, José I. Chacón, Antonio Antón, Arrate Plazaola, Elena García‐Martínez, Miguel A. Seguí, Pedro Sánchez‐Rovira, José Palacios, Lourdes Calvo, Carmen Esteban, Enrique Espinosa, Agusti Barnadas, Norberto Batista, Angel Guerrero, Montserrat Muñoz, Estefania Romio, César Rodríguez‐Martín, Rosalía Caballero, María I. Casas, Federico Rojo, Eva Carrasco, Silvia Antolín

Disclosures

Montserrat Muñoz: AstraZeneca (SAB), Roche (ET). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board