Interplay between ESR1/PIK3CA codon variants, oncogenic pathway alterations and clinical phenotype in patients with metastatic breast cancer (MBC): comprehensive circulating tumor DNA (ctDNA) analysis

Background although being central for the biology and druggability of hormone-receptor positive, HER2 negative metastatic breast cancer (MBC), ESR1 and PIK3CA mutations are simplistically dichotomized as mutated or wild type in current clinical practice. Methods The study analyzed a multi-institutional cohort comprising 703 patients with luminal-like MBC characterized for circulating tumor DNA through next generation sequencing (NGS). Pathway classification was defined based on previous work (i.e., RTK, RAS, RAF, MEK, NRF2, ER, WNT, MYC, P53, cell cycle, notch, PI3K). Single nucleotide variations (SNVs) were annotated for their oncogenicity through OncoKB. Only pathogenic variants were included in the models. Associations among clinical characteristics, pathway classification, and ESR1/PIK3CA codon variants were explored. Results The results showed a differential pattern of associations for ESR1 and PIK3CA codon variants in terms of co-occurring pathway alterations patterns of metastatic dissemination, and prognosis. ESR1 537 was associated with SNVs in the ER and RAF pathways, CNVs in the MYC pathway and bone metastases, while ESR1 538 with SNVs in the cell cycle pathway and liver metastases. PIK3CA 1047 and 542 were associated with CNVs in the PI3K pathway and with bone metastases. Conclusions The study demonstrated how ESR1 and PIK3CA codon variants, together with alterations in specific oncogenic pathways, can differentially impact the biology and clinical phenotype of luminal-like MBC. As novel endocrine therapy agents such as selective estrogen receptor degraders (SERDS) and PI3K inhibitors are being developed, these results highlight the pivotal role of ctDNA NGS to describe tumor evolution and optimize clinical decision making. Supplementary Information The online version contains supplementary material available at 10.1186/s13058-023-01718-0.


Background
Metastatic breast cancer (MBC) is a treatable yet virtually incurable disease, and most deaths from breast cancer occur due to metastasis.[1,2].MBC often evolves via acquisition of new resistant mutations usually under the pressure of anticancer treatments.ESR1 and PIK3CA mutations have been the most thoroughly studied and have been implicated in the biology and druggability of hormone-receptor positive (HR +), HER2-negative (HER2−) MBC, and have important implications for therapeutic selection [3,4].
Alterations in the activity and expression of estrogen receptor α (ER) are often involved in MBC resistance and progression.Resistance may occur through loss of ER expression, increased expression of ER or related cofactors, post-translational modifications of ER, and/or delocalization of ER to the cellular membrane [5].Alterations of genes involved in other pathways or epigenetic alterations of ESR1 promoters can deregulate expression due to pathway crosstalk and modified ER activity [6][7][8].Point mutations are the most common ESR1 genetic alterations and generally arise in the ligand binding domain, most commonly in codons 538, 537, 380, and 536 [5,8,9].In some situations, gene amplifications, deletions, or translocations resulting in fusions can occur [10][11][12].
PIK3CA mutations can occur both as single and multiple concomitant mutations.The frequency of multiple co-occurring PIK3CA mutations has been estimated to be approximately 8-13% with the vast majority being double mutations (88-96% among patients with multiple PIK3CA mutations) [13].Double-PIK3CA-mutant breast cancer generally consists of a combination of a major-mutant hotspot (either E542, E545, or H1047) and a minor-mutant site (either E453, E726, or M1043) [13] Liquid biopsies that include circulating tumor DNA (ctDNA), circulating tumor cells (CTCs) and exosomes are non-invasive diagnostic tools that are being explored in real-time early cancer detection, monitoring for minimal residual disease, and longitudinal tracking of clonal evolution in the peripheral blood [14,15].Due to their increasing sensitivity and decreasing cost, high throughput genomic technologies such as next-generation sequencing (NGS) are becoming widely available [16,17].Together with its longitudinal application and its increasing deployment in clinic, ctDNA has become a promising tool for the development of further insight related to MBC's biological evolution [18][19][20].
The aim of this retrospective study was to evaluate the interplay between oncogenic pathway alterations and ESR1/PIK3CA codon variants as these findings relate to their impact on the biological and clinical behavior of HR + HER2− MBC.

Study population and design
This study retrospectively analyzed a multi-institutional cohort of 703 HR + HER2− MBC patients with ctDNA NGS sampling before starting a new treatment.Samples were collected from patients who underwent standardof-care ctDNA testing at Northwestern University (Chicago, IL), Massachusetts General Hospital (Boston, MA) and Washington University in St. Louis (St. Louis, MO) between 2015 and 2020.No selection was made based on current or prior lines of therapy.
Baseline imaging was performed prior to ctDNA collection and start of therapy according to the treating physician's choice [e.g., Computed Tomography (CT), Positron Emission Tomography (PET)].Sites of metastasis were categorized based on the presence of specific organ involvement (e.g., liver involvement, yes vs. no) independently from other metastatic sites.

ctDNA sample collection and analysis
Two 10-mL samples of whole blood were collected for each patient using standard stabilizing tubes (Streck, NE) at baseline before treatment start.Samples were analyzed using the commercial Guardant360 ™ NGS platform (Guardant Health, CA), a 72-gene panel based on single-molecule digital sequencing was utilized to detect somatic single nucleotide variants (SNVs), insertions/deletions (indels), gene fusions/rearrangements and copy number variations (CNVs) [21][22][23].Mutations were annotated through the OncoKB database according to their effect (loss of function, gain of function) and pathogenicity [24].Only pathogenic mutations based on OncoKB were included in the logistic and Cox regression models.
Mutant allele frequency was analyzed both for each ESR1 and PIK3CA codon variant (codon MAF) and these results highlight the pivotal role of ctDNA NGS to describe tumor evolution and optimize clinical decision making.
based on the highest frequency across all mutated gene detected in the patient's blood sample (Overall MAF).CNVs were dichotically considered as present/absent.

Statistical analysis
Clinical and pathologic variables were reported using descriptive analyses.Categorical variables were reported as frequency distributions, whereas continuous variables were described through median and interquartile ranges (IQRs).
Associations between clinical characteristics, pathway classification, and ESR1/PIK3CA codon variants were explored through uni-and multivariable logistic regression, inclusive of odds ratio (OR) and 95% confidence interval (95% CI) computation.
Overall survival (OS) was defined as the time from the baseline ctDNA blood draw to death from any cause.Patients without an end point event at the last follow-up visit were censored.Differences in survival were tested by log-rank test and uni-and multivariable Cox regression with 95% CI and represented by Kaplan-Meier estimator plot.Correction for significant clinical variables after univariable testing was applied to the multivariable model (i.e., previous treatment with CDK4/6i, number of lines, lung, liver, bone, and soft tissue involvement). Statistical
Across the tested genes, PIK3CA, TP53 and ESR1 were the most altered.As expected, the most likely effect of the detected SNVs was gain of function (GOF) for PIK3CA and ESR1 and loss of function (LOF) for TP53 (Fig. 1A,  B).
The association between the main ESR1 and PIK3CA codon variants (Fig. 1C, D) and concomitant gene alterations was investigated according to oncogenic pathway classification.
No concomitant alterations were confirmed after multivariable analysis for ESR1 380, although a numerical difference was highlighted for SNVs in the P53 pathway (Fig. 2, Additional file 1: Table S3).
No associations were observed for ESR1 536.

Alterations in oncogenic pathways and ESR1 / PIK3CA codon variants are differentially associated with sites of metastasis
The association between ESR1/PIK3CA codon variants and alterations in oncogenic pathways were then investigated across the main metastatic sites.Correction for number of lines was applied.
The number of concomitant ESR1 and PIK3CA mutations was then analyzed across metastatic sites.
Patients with lung, liver and bone metastases had a significantly higher number of ESR1 mutations (respectively P = 0.0244, P < 0.0001 and P < 0.0001) (Fig. 3), while PIK3CA mutations were significantly higher in patients with lung, CNS, and bone metastases (respectively P = 0.0167, P = 0.0011 and P < 0.0001) and significantly lower in patients with soft tissue involvement (P = 0.0020) (Fig. 3).

ESR1 mutations impact on overall survival together with alterations in PI3K, MYC, RAS and P53 pathways
The prognostic impact of ESR1 and PIK3CA codon variants and number of concomitant mutations was investigated in terms of OS (Fig. 4).Although the detection of ESR1 or PIK3CA mutations had a significantly unfavourable impact on OS (respectively P < 0.0001 and P = 0.0410) (Fig. 4A, B), no differential impact was observed across codon variants (P = 0.3108 and P = 0.3450) (Fig. 4A, B) or the number of concomitant ESR1 or PIK3CA SNVs (P = 0.9414 and P = 0.1301) (Fig. 4C, D).Similar results were observed for patients previously treated with CDK4/6i both for the overall impact of the ESR1 and PIK3CA mutational status (respectively P = 0.0310 and P = 0.0009) and multiple concomitant SNVs (respectively P = 0.8074 and P = 0.3443) (Fig. 4E, F).On the other hand, in patients not previously exposed to CDK4/6i only ESR1 had a significant impact on OS (P = 0.0206) (Fig. 4G, H).Similar results were observed in terms of PFS in patients treated with endocrine therapy (Additional file 1: Fig. S2A-D), apart from a numerical difference according to number of PIK3CA alterations within patients with PIK3CA-mutated MBC (P = 0.0753) (Additional file 1: Fig. S2D).

Discussion
Although pivotal in breast cancer biology and evolution, ESR1 and PIK3CA mutations clinically are simplistically dichotomized as mutated or wild type.On the other hand, preclinical data suggests that a more precise characterization of ESR1 and PIK3CA mutated variants may help improve our understanding of how these alterations and associated co-mutations drive resistance and clonal evolution [13,[26][27][28].In this study we retrospectively analyzed ctDNA samples from a cohort of 703 patients to more specifically investigate the role of ESR1 and PIK3CA codon variants and other oncogenic pathway alterations and their impact on the clinical phenotype of HR + HER2− MBC.
Similar to previous studies, alterations in ESR1 and PIK3CA were commonly observed in our cohort [29][30][31][32].We investigated differences in clinical phenotypes based on specific codon variants for each of these genes, together with other gene alterations on a pathway level.This approach enabled us to detect potential associations from a biological standpoint that would have been otherwise diluted by the low incidence of the single gene alterations within different pathways.Our results showed a differential pattern of associations for ESR1 and PIK3CA codon variants in terms of co-occurring pathway alterations patterns of metastatic dissemination, and prognosis.
The association among ESR1, PIK3CA codon variants, and gene alterations was then explored according to oncogenic pathways classification.
PIK3CA 1047 and 542 were associated with CNVs in the PI3K pathway.Co-occurring SNVs in the PI3K pathway are well documented and are often associated with exceptional responses.On the other hand, PIK3CA CNVs are less well described, since clinical trials focused on PI3K inhibitors are usually focused on PCR-based diagnostic companions and therefore typically report PIK3CA mutations but not PIK3CA CNVs.Our data suggest the need for a more specific characterization since patients with co-occurring SNVs and CNVs could have different patterns of response [33].
ESR1 537 was significantly associated with SNVs in the ER and RAF pathways and CNVs in the MYC pathway, while ESR1 538 was associated with SNVs in the cell cycle pathway.The selection of such co-occurring alterations could be the result of prior exposure to ET, suggesting that endocrine resistance may be an emerging property of cellular-wide genetic, epigenetic, and transcriptional phenomena, rather than the result of a single hit aberration [26,34].Multiparametric characterizations are therefore needed to better describe this phenomenon and select new therapeutic strategies that could target alterations on a pathway level [35].
It has been previously suggested that gene alterations and expression can influence the development of metastases in different sites [14,36,37].Therefore, in our current study, we combined oncogenic pathways classification and codon variants to better describe how these alterations may impact site of metastasis (e.g., organotropism).Similarly to previous works, ESR1 mutations were found to be associated with liver and bone metastases and PIK3CA with bone.On the other hand, when considered on a codon basis, only ESR1 538 was associated with liver metastases, while ESR1 537 and PIK3CA 1047 were significantly associated with bone metastases.A more complex role of ESR1 in the biology of HR + HER2− MBC has been suggested by previous data, not only on a genetic but also on an epigenetic standpoint where liver metastases were associated with low methylation levels of the ESR1 promoter [8,[38][39][40][41].In addition to specific codon variants, we showed that alterations in the RTK pathway were associated with liver metastases, supporting the importance of this pathway in driving therapeutic resistance.Prior studies have shown that alterations in ESR1, AKT1, ERBB2, FGFR4 and NF1 are linked to the RTK-RAS axis in patients who develop liver metastases [38].On one hand therapeutic targeting of these alterations could have implications for future treatment strategies, potentially impacting specific sites of metastatic spread [42].On the other hand, ctDNA shedding is influenced by tumor burden and may vary across metastatic sites, introducing a potential bias in the interpretation of data linked to organotropism [43].
As expected, our study confirmed the prognostic impact of alterations in ESR1 and PIK3CA on OS in univariable models [44].This significant impact on prognosis was consistent across codon variants and also in patients with polyclonal ESR1 or PIK3CA mutations (Fig. 4) [13].
As new treatment options specifically targeting ER are gaining momentum, the prognostic and predictive role of ESR1 mutations codon variants may become central in a significant number of patients with MBC [45].In terms of oncogenic pathways, our multivariable models highlighted a significant prognostic role for RAS pathway SNVs, MYC pathway CNVs, PI3K pathway CNVs, and P53 pathway SNVs, while only ESR1 380 codon variant retained its prognostic impact.The prognostic impact of gene alterations according to oncogenic pathways is an emerging paradigm that has been previously explored in the translational analysis of MONALEESA 7 [20].RTK gene alterations (defined as either SNVs and CNV in the FGFR1, ERBB2, IGF1R, EGFR, ERBB3, KDR, KIT, PDG-FRB, PDGFRA, ERBB4, VEGFA and IGF1 genes) were identified in 17% of patients, and similarly to or study, these patients experienced shorter PFS, especially in the ET-only arm [20].Of note, the incidence of RTK alterations was lower in MONALEESA 7 with respect to our cohort (24%), possibly due to less exposure to prior treatments [20].
Our study has several potential limitations.First, the NGS technology used in our study was not able to define whether multiple detected mutations where concomitantly harbored by the same cell population or if they originated from different subclones.This distinction may play an important role as cis-PIK3CA codon variants (i.e., present on the same allele) have been observed to significantly impact treatment response and MBC biology [13].Similarly, this limitation applies to associations between specific alterations and metastatic organotropism.If on one hand the real-world design of the study may have increased its clinical transferability, on the other it may have added potential bias due to not uniform treatment strategies across institutions, historical cohorts, and treatment lines.Moreover, timing from ctDNA and treatment start was decided on a clinical basis and not per protocol, introducing variability across patients.Finally, current clinical NGS platforms rely on targeted gene panels, potentially underestimating the tumor's mutational load and leaving out other less represented key genes across the considered oncogenic pathways.

Fig. 1
Fig. 1 Landscape plot of all detectable aberrations in ctDNA samples.A Incidence of the single aberrations [copy number variations (CNV), Fusions (Fus), deletions (Del), insertions (Ins), frameshift (FS), splicing variants (Spl), premature termination codons (PTC) and single nucleotide variation (SNV)] is represented on the left.The mutant allele frequency (MAF) of each mutation is shown in the middle.Effect [gain of function (GOF), loss of function (LOF) and switch of function (SOF)] and pathogenicity [yes, no, unknown (Ukn) and inconclusive (Inc)] of all the detected aberrations are show on the right.Histogram representing different frequency distribution of gene mutations across oncogenic pathways (B).Lollipop plot showing the distribution of ESR1 (C) and PIK3CA (D) mutations by their amino acid coordinates across ER and PI3K domains and across ESR1 and PIK3CA exon sequencies.Oncogenic mutations are highlighted CNVs, Copy Number Variations; SNVs, Single Nucleotide Variations

Table 1
Clinicopathologic characteristics of the luminal-like MBC cohort.

Table 2
Association of ESR1 and PIK3CA codon variants and alterations in oncogenic pathways with overall survival (OS).

Table 2
(continued)ConclusionsESR1 and PIK3CA codon variants, together with alterations in specific oncogenic pathways, can differentially impact the biology, and survival of patients with HR + HER2-negative metastatic breast cancer.As novel therapies such as selective estrogen receptor degraders and PI3K inhibitors are being developed and are entering clinical practice, our results highlight the pivotal role of ctDNA NGS in describing tumor evolution under treatment pressure and optimizing both clinical decision making and future drug development.