Skip to main content
NCBI home page
Lippincott Open Access logoLink to Lippincott Open Access
. 2026 Apr 9;10(4):e2500609. doi: 10.1200/PO-25-00609

Clinicogenomic Landscape and Function of PIK3CA, AKT1, and PTEN Mutations in Breast Cancer

Jacqueline J Tao 1, Saumya D Sisoudiya 2, Hanna Tukachinsky 2, Alexa B Schrock 2, Ericka M Ebot 2, Smruthy Sivakumar 2, Ethan S Sokol 2, Neil Vasan 3,4,✉
PMCID: PMC13086116  PMID: 41955551

Abstract

PURPOSE

To comprehensively characterize the clinical and genomic landscapes of PIK3CA, AKT1, and PTEN alterations and examine their functional and therapeutic implications in AKT-driven breast cancer.

METHODS

Comprehensive genomic profiling of 51,767 breast tumors was performed using FoundationOne CDx or FoundationOne. We examined the genomic landscape of PIK3CA, AKT1, and PTEN alterations and their distribution across clinical variables of interest. Prior deep mutational scanning (DMS) data were used to functionally characterize clinical PTEN variants. Real-world clinical outcomes were assessed in patients treated with capivasertib plus fulvestrant.

RESULTS

A total of 29,157 variants were identified across the three genes, including pathogenic variants and variants of uncertain significance. The most frequently altered gene was PIK3CA (37.4% of cases), followed by PTEN (13.5%) and AKT1 (5.4%). The most common alterations in each gene were PIK3CA H1047R (35.6% of PIK3CA-altered cases), E545K (19.7%), and E542K (11.7%); AKT1 E17K (69.7%); and PTEN homozygous copy number deletion (37.3%). PIK3CA alterations were less prevalent in patients of African genetic ancestry (27.1% vs 38.6% in European genetic ancestry), whereas AKT1 and PTEN alterations were balanced across ancestries. DMS data on missense PTEN mutations revealed that 32.5% showed discordant effects on protein stability and phosphatase activity. A subset of patients with rare AKT pathway variants derived meaningful progression-free survival and overall survival benefit from capivasertib.

CONCLUSION

Here, we present the landscape of PIK3CA, AKT1, and PTEN alterations in, to our knowledge, the largest clinical cohort examined to date. The functional complexity of rare PTEN variants underscores the need for functional validation by tools such as DMS. Rare AKT pathway variants may predict clinical benefit from AKT inhibitors and warrant further clinical investigation.

INTRODUCTION

The AKT pathway regulates critical cellular functions including growth, proliferation, survival, and metabolism and is one of the most frequently altered pathways across cancers. PIK3CA encodes PI3Kα, which recruits and activates AKT and downstream signaling through mechanistic target of rapamycin complex 1.1 The lipid phosphatase PTEN metabolizes PIP3, antagonizing the effects of PI3K. Consequently, aberrant AKT pathway activation primarily arises because of activating mutations in PIK3CA or AKT1 and inactivating alterations in PTEN.2

CONTEXT

  • Key Objective

  • What is the clinicogenomic landscape of PIK3CA, AKT1, and PTEN alterations in breast tumors, and what are their functional and therapeutic implications?

  • Knowledge Generated

  • Comprehensive genomic profiling of 51,767 breast tumors identified 29,157 variants across the three genes and revealed that PIK3CA alterations were less prevalent in patients of African genetic ancestry. Analysis of published deep mutational scanning (DMS) showed that one third of PTEN missense variants had discordant effects on protein abundance versus activity. Clinical benefit was observed in a subset of patients with rare variants treated with capivasertib and fulvestrant.

  • Relevance

  • AKT pathway variants are common in breast tumors and include many rare variants of uncertain significance. Tools such as DMS can clarify the functional implications of rare variants, which may predict clinical benefit from AKT pathway inhibitors. Ancestry-based differences in PIK3CA prevalence should be considered in clinical trials.

The AKT pathway is altered in over 40% of estrogen receptor–positive (ER+) breast cancers.3 PI3Kα-specific inhibitors including alpelisib and inavolisib are approved for PIK3CA-mutated metastatic breast cancer (MBC),4,5 but are often associated with significant toxicities including high-grade hyperglycemia. Recently, the AKT inhibitor capivasertib was approved for patients with ER+ MBC with mutations in PIK3CA, AKT1, and/or PTEN after progression on first-line endocrine-based therapy.6 In the CAPItello-291 study, capivasertib combined with fulvestrant more than doubled PFS in patients with qualifying genomic alterations7 and demonstrated a favorable toxicity profile compared with older generations of PI3K inhibitors. Notably, clinical benefit was limited to patients harboring PIK3CA/AKT1/PTEN alterations, highlighting their importance as biomarkers of sensitivity to capivasertib.

Despite significant progress in genomic characterization of the AKT pathway,8,9 prior studies of PIK3CA/AKT1/PTEN have had limited sequencing coverage and analyzed smaller genomic data sets. Early analyses evaluated fewer than 2,000 breast tumors using limited gene panel sequencing,10,11 whereas more recent studies have examined up to 7,450 breast cancers using next-generation sequencing.12

Given the US Food and Drug Administration approval of capivasertib for patients with breast cancer harboring PIK3CA, AKT1, or PTEN alterations as identified by the companion diagnostic FoundationOne CDx, understanding the genomic, functional, and clinical landscapes of this biomarker triad is increasingly critical. Here, we assess the frequencies, distributions, clinical associations, and co-occurrence patterns of PIK3CA, AKT1, and PTEN variants in the largest clinicogenomic breast cancer data set analyzed to date (N = 51,767) and integrate genomic findings with functional data and real-world clinical outcomes.

METHODS

Comprehensive Genomic Profiling of a Breast Cancer Cohort

Comprehensive genomic profiling (CGP) of formalin-fixed, paraffin-embedded tissue sections was performed using the FoundationOne CDx assay or the laboratory-developed test predecessor, FoundationOne, as part of routine clinical care in a Clinical Laboratory Improvement Amendments–certified, College of American Pathologists–accredited laboratory (Foundation Medicine, Inc, Cambridge, MA). This study included 51,767 breast cancer samples profiled between August 2014 and September 2023. Approval for this study, including a waiver of informed consent and Health Insurance Portability and Accountability Act (HIPAA) waiver of authorization, was obtained from the Western Institutional Review Board (protocol No. 20152817).

Genomic Analysis

CGP was performed on hybrid capture-selected libraries spanning exons of at least 300 cancer-related genes and select introns of genes frequently rearranged in cancer. Processing of the sequence data and identification of different classes of genomic alterations were performed as previously described.13,14 The alterations identified included single-nucleotide variants (missense, nonsense, and splice site mutations), indels (in-frame and frameshift mutations), copy number alterations, and rearrangement events. Alterations were classified as either known or likely pathogenic using annotations from various sources, including the COSMIC database, previous scientific literature (such as truncations and deletions in known tumor suppressor genes), or mutations previously characterized as pathogenic; all other alterations were classified as variants of unknown significance (VUS).

Histopathologic and Clinical Characteristics

Human epidermal growth factor receptor (HER2) status was determined from the HER2 (ERBB2) amplification status based on the FoundationOne or FoundationOne CDx assays. ER status was derived from accompanying pathology reports, where available. Clinical data including the biopsy site were obtained from the submitted test requisition form. Genetic ancestry of each patient was predicted using a single-nucleotide polymorphism–based approach, as previously described.15

Co-Occurrence and Mutual Exclusivity Analyses

The data set was interrogated for patterns of co-occurrence and mutual exclusivity of PIK3CA, AKT1, and PTEN alterations with other cancer-associated genes sequenced. A Fisher's exact test was used to identify significantly co-occurring (odds ratio [OR] > 1) and mutually exclusive (OR < 1) patterns using a false discovery rate (FDR)-adjusted P value threshold of .05. A Benjamini-Hochberg FDR correction was applied to adjust for multiple comparisons.

Functional Characterization of Clinical PTEN Alterations

We examined two published deep mutational scanning (DMS) data sets characterizing the effects of single amino acid substitutions on PTEN function. Matreyek et al16,17 generated abundance scores for 4,547 PTEN variants by measuring intracellular abundance in human cells and then assigned an abundance class of low, possibly low, possibly wild-type (WT)–like, or WT-like to each variant. Mighell et al18 generated fitness scores for 6,904 PTEN variants by assaying lipid phosphatase activity in yeast and then assigned functional categories (truncation-like, hypomorphic, or WT-like) based on fitness score percentiles. We abstracted abundance and fitness scores corresponding to the PTEN missense and nonsense variants in our clinical cohort and assigned functional classifications for each variant. Functional designations from the original publications were consolidated into binary categories of pathogenic (abundance class = low or possibly low; fitness category = truncation-like or hypomorphic) and nonpathogenic (abundance class = WT-like or possibly WT-like; fitness category = WT-like or >WT). For each missense PTEN variant, functional classifications derived from abundance versus fitness scores were compared.

Clinical Outcome Analyses

This study used the US-based deidentified Flatiron Health-Foundation Medicine breast cancer Clinico-Genomic Database (FH-FMI CGDB). Clinical data from the Flatiron Health Research Database19 are linked to genomic data, derived from FMI's CGP tests FoundationOne CDx and FoundationOne, in the FH-FMI CGDB by deterministic matching, providing a deidentified data set.13,14,20 During the study period, the deidentified data originated from approximately 280 cancer clinics (approximately 800 sites of care) and included 12,601 patients with breast cancer who underwent FMI CGP between January 2011 and December 2024. We analyzed 102 patients who had alterations in PIK3CA, AKT1, and/or PTEN and received capivasertib + fulvestrant in any line of therapy (Data Supplement, Fig S1). Of these, 93 patients had a capivasertib-qualifying variant (CAPIm+) and nine patients had other AKT pathway variants (CAPIm–). We assessed real-world progression-free survival (rwPFS), defined as the time from therapy start to the first progression event or death, and real-world overall survival (rwOS) from therapy start. Institutional review board approval was obtained before study conduct and included a waiver of informed consent based on the observational, noninterventional nature of the study (WCG IRB, Protocol No. 420180044).

Statistical Analyses

Categorical variables were compared between groups using the Fisher's exact test, with an FDR correction for multiple testing. A P value of ≤.05 was considered statistically significant. Statistics, computation, and plotting were performed using R 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Baseline Cohort Characteristics

This analysis included 51,767 breast tumors. HER2 amplification was assessed in all samples, of which 8.3% was HER2+. ER status was available in 4,219 cases, of which 57.1% was ER+/HER2–, 33.4% was ER–/HER2–, 4.4% was ER–/HER2+, and 5% was ER+/HER2+. The most common ancestries represented were European (72%) and African (14%). Tumor samples were obtained from a metastatic site (47%), local site (33%), lymph node (12%), or unknown site (8%).

Prevalence and Spectrum of AKT Pathway Alterations

There were 29,157 variants identified across PIK3CA, AKT1, and PTEN. The most frequently altered gene was PIK3CA (37.4% of samples), followed by PTEN (13.5%) and AKT1 (5.4%), including pathogenic alterations and VUS. When restricting to pathogenic alterations only, PIK3CA was altered in 36.5% of samples, PTEN in 12.7%, and AKT1 in 4.7%. Among 1,024 distinct PIK3CA alterations identified, the most common variant type was missense mutations (67.4% of all PIK3CA variants) followed by in-frame mutations (24.6%; Fig 1, Data Supplement, Table S1). The most common mutations were H1047R (35.6% of PIK3CA-altered cases), E545K (19.7%), and E542K (11.7%), consistent with known mutational hotspots.21 Among 288 distinct AKT1 alterations identified, 71.2% were missense mutations, predominantly the hotspot mutation E17K (69.7% of AKT1-altered cases; Fig 1, Data Supplement, Table S2). Finally, among 1,723 distinct PTEN alterations identified, common variant types were distributed between frameshift mutations (41.9% of PTEN variants), missense mutations (28.7%), and splice site mutations (14.3%). The most common individual PTEN alteration was homozygous loss (37.3% of PTEN-altered cases; Fig 1, Data Supplement, Table S3).

FIG 1.

FIG 1.

Open in a new tab

Frequency, distribution, and coding type of PIK3CA, AKT1, and PTEN variants. The most frequently altered AKT pathway gene was PIK3CA (37.4% of patients), followed by PTEN (13.5%) and AKT1 (5.4%). (A) Lollipop plots of PIK3CA, AKT1, and PTEN mutations. (B) Prevalence of individual variants in each gene. The most common variants in each gene were PIK3CA H1047R (35.6% of PIK3CA-altered cases), AKT1 E17K (69.7%), and PTEN deletion (37.3%). (C) Distribution of variant coding types in each gene. The most common variant coding types were PIK3CA missense mutation (67.4%), AKT1 missense mutation (71.2%), and PTEN frameshift mutation (41.9%). DSPc, dual-specificity phosphatase catalytic domain; PH, Pleckstrin homology; VUS, variants of unknown significance.

Multiple pathogenic PIK3CA alterations were seen in 3,148 samples (16.2% of all PIK3CA-altered samples and 6.1% of overall cohort). Pathogenic coalterations across genes were most common with PIK3CA and PTEN (4% of cohort) followed by PIK3CA and AKT1 (0.6%) and then AKT1 and PTEN (0.14%).

Clinical and Demographic Associations

We next examined mutational patterns by receptor subtype, patient ancestry, and biopsy site (Figs 2A-2F, Supplementary Tables S4-S6). PIK3CA pathogenic variants were more prevalent in ER+/HER2– and HER2+ disease (40.3%, P = 6.0 × 10−36 and 37.6% of cases, P = 2.3 × 10−32) than ER–/HER2– disease (20.9%), consistent with prior reports.10,22 However, variants outside the three mutational hotspots were more common in ER–/HER2– (42%) than ER+/HER2– or HER2+ cases (33%, P = .002 and 33%, P = .002). AKT1 mutations were most prevalent in ER+/HER2– disease (6.0%) and less common in HER2+ or ER–/HER2– disease (1.7%, P = 4.7 × 10−20 and 3.0%, P = 3.5 × 10−5). By contrast, PTEN alterations were enriched in ER–/HER2– (17.9%) compared with ER+/HER2– and HER2+ disease (11.3%, P = 1.8 × 10−8 and 3.9%, P = 2 × 10−57).

FIG 2.

FIG 2.

Open in a new tab

Prevalence of gene variants by breast cancer receptor subtype, genetic ancestry, and biopsy site. (A) PIK3CA variants were enriched in ER+/HER2– and HER2+ disease (altered in 40.3% and 37.6% of cases, respectively) and AKT1 variants were enriched in ER+/HER2– disease (6.0%). By contrast, PTEN variants were most prevalent in ER–/HER2– disease (17.9%). (B and C) PIK3CA alterations were less prevalent in patients with African ancestry (27.1% v 38.6% in European ancestry) and enriched in metastases (39.6% v 33.2% in local tumors). (D) PIK3CA variants outside the 542, 545, and 1,047 codons were more common in ER–/HER2– disease (42%) compared with ER+/HER2– or HER2+ disease (33% and 33%). (E and F) PIK3CA variants by ancestry and biopsy site. ER, estrogen receptor; HER2, human epidermal growth factor receptor 2.

PIK3CA pathogenic alterations were less prevalent in patients with African ancestry, both in the full cohort (27.1% v 38.6% in Europeans, P = 9.6 × 10−81) and in the ER+/HER2– subgroup (31.2% v 41.1% in Europeans, P = .002). In addition, among ER+/HER2– cases, patients with South Asian ancestry had a higher prevalence of PIK3CA pathogenic alterations (68.9% v 41.1% in Europeans, P = .002). PIK3CA pathogenic variants were enriched in metastases compared with local tumors (39.6% v 33.2%, P = 6 × 10−41). The distribution of individual PIK3CA variants was similar across ancestries and biopsy sites. The prevalence of pathogenic AKT1 and PTEN alterations was similar across all ancestries and biopsy sites.

Co-Occurrence and Mutual Exclusivity Analyses

Next, we characterized patterns of co-occurrence between pathogenic variants in PIK3CA, AKT1, and PTEN and other sequenced genes (Fig 3, Data Supplement, Tables S7-S9). PIK3CA, AKT1, and PTEN pathogenic alterations were all strongly mutually exclusive with each other (PIK3CA/AKT1 OR, 0.25, P < 10−60; PIK3CA/PTEN OR, 0.74, P = 6.6 × 10−25; AKT1/PTEN OR, 0.20, P < 10−60), suggesting that a single driver mutation may be sufficient for pathway activation. Notably, the degree of mutual exclusivity was lowest between PIK3CA and PTEN, reflecting their higher rates of comutation compared with the other gene combinations.

FIG 3.

FIG 3.

Open in a new tab

Co-occurrence and mutual exclusivity analyses of pathogenic variants in PIK3CA, AKT1, and PTEN. Pathogenic variants in these three genes were all mutually exclusive to each other. Notable co-occurrences included PIK3CA with SOX2, PRKCI, MAP3K1, TERC, CDH1, and TBX3; AKT1 with NF1 and CDH1; and PTEN with FAS, TP53, and RB1. Notable mutual exclusivities include PIK3CA with TP53, BRCA1/2, and STK11; AKT1 with ERBB2 and MYC; and PTEN with ERBB2, GATA3, and ESR1. Y-axis is capped at a P value of 10E-60.

PIK3CA was significantly coaltered with SOX2, PRKCI, MAP3K1, TERC, CDH1, and TBX3. Co-occurrence with SOX2 is not well-described in the breast cancer literature although there are reports suggesting a role of PIK3CA and SOX2 coamplification in the pathogenesis of head and neck cancers.23 There are well-known associations of PIK3CA with CDH1 and TBX3, whose losses are hallmarks of invasive lobular carcinoma (ILC),24,25 and MAP3K1, which is associated with the luminal A phenotype.26-28 These findings are consistent with the enrichment of ILC and luminal A breast cancers for PIK3CA mutations. Several tumor suppressors were mutually exclusive with PIK3CA including TP53, BRCA1/2, and STK11. Other notable mutual exclusivities included PIK3R1, FGF23, CCND2, MYC, KDM5A, and GATA3.

AKT1 was significantly coaltered with NF1 and CDH1, an association previously described in ILC.29 There were several shared co-occurrence patterns with PIK3CA, including MAP3K1, which may reflect the increased prevalence of both PIK3CA and AKT1 in ER+/HER2– tumors. AKT1 was mutually exclusive with ERBB2 and MYC.

PTEN was significantly coaltered with FAS (both located on 10q23) and the tumor suppressors TP53 and RB1. PTEN is known to be an essential mediator of Fas-mediated apoptosis, and PTEN loss can lead to impaired apoptosis and increased cell survival.30 Combined PTEN and TP53 loss is known to be associated with poor therapeutic response and prognosis, particularly in triple-negative breast cancers.31,32 Notable mutually exclusivities included ERBB2, GATA3, and ESR1.

Co-occurrence analyses of VUS in the three genes of interest showed weakly significant results (Data Supplement, Fig S2 and Tables S7-S9).

Functional Characterization of Clinical PTEN Alterations

Among the >1,700 clinical PTEN variants identified in our cohort, a substantial proportion are VUS. To explore the functional implications of these clinical PTEN variants, we examined two published DMS data sets characterizing the effects of single amino acid substitutions on PTEN abundance (protein expression) and fitness (lipid phosphatase activity).16-18 Of the 610 single amino acid substitution PTEN variants (494 missense, 116 nonsense) in our clinical cohort, 339 (55.6%) had an available abundance score and 536 (87.9%) had an available fitness score; 298 (48.9%) variants had both abundance and fitness scores available (Data Supplement, Table S10).

Nonsense variants exhibited lower abundance and fitness scores than missense variants (Fig 4A) and were almost all appropriately classified as pathogenic by both abundance score (100%) and fitness score (95.7%). By contrast, there was significant variation in the functional classifications assigned to missense variants. By abundance score, 148 (51.0% of missense variants with available abundance score) were pathogenic and 142 (49.0%) were nonpathogenic; by fitness score, 231 (52.0%) were pathogenic and 213 (48.0%) were nonpathogenic. For missense variants with both abundance and fitness scores available (n = 261; Figs 4B and 4C), 35.6% was classified as pathogenic by both assays, whereas 31.8% was classified as nonpathogenic by both assays, highlighting that a substantial proportion of missense PTEN variants do not contribute to pathogenicity. The remaining one third of missense variants were discordantly classified (32.5%), exhibiting either loss of abundance with preserved phosphatase activity (14.9%) or loss of phosphatase activity with preserved abundance (17.6%). These discordances likely reflect the distinct phenotypic effects of mutations in different amino acid residues.

FIG 4.

FIG 4.

Open in a new tab

Deep mutational scanning data for missense and nonsense PTEN variants. (A) Distribution of abundance and fitness scores among missense versus nonsense variants in the cohort, from the studies by Matreyek et al17 and Mighell et al.18 Mean abundance and fitness scores for nonsense variants are significantly lower than those for missense variants (mean scores for each group are represented by dashed lines; P = 1.12 × 10−28 and P = 4.93 × 10−25, respectively). (B) Set diagrams depicting degree of concordance in functional classification by abundance versus fitness score, for missense variants only. (C) Scatterplot of abundance and fitness scores for missense variants only, colored by concordance in functional classification between the two assays.

Clinical Outcomes Among Patients Treated With Capivasertib and Fulvestrant

Finally, we evaluated real-world treatment outcomes among patients receiving capivasertib plus fulvestrant who had capivasertib-qualifying variants (CAPIm+, n = 93) versus other AKT pathway variants (CAPIm–, n = 9). Clinical characteristics did not differ significantly between the groups (Data Supplement, Table S11). Real-world PFS was similar between the groups (median rwPFS 170 day v 193 days, hazard ratio, 1.1 [95% CI, 0.51 to 2.6]; P = .75; Figs 5A and 5B), demonstrating that meaningful responses can be achieved even in patients with rare AKT pathway variants. Individual rwOS, time of first progression after treatment initiation, treatment line, and gene variant profiles are presented in a swimmer plot (Fig 5C). Clinical benefit was observed in a substantial proportion of patients belonging to both CAPIm+ and CAPIm– groups.

FIG 5.

FIG 5.

Open in a new tab

Clinical outcomes among patients treated with capivasertib and fulvestrant. (A) Kaplan-Meier analysis and (B) hazard ratios for multivariate analysis of rwPFS among patients with capivasertib-qualifying variants (CAPIm+) versus other AKT pathway variants (CAPIm–). rwPFS was defined as the time from therapy start to the first progression event or death. Progression events within 14 days of therapy start date were excluded. Patients with no progression event within the line of therapy were censored at the last structured activity date within the line of therapy. Median rwPFS did not differ significantly between groups. A trend toward improved rwPFS was observed in patients treated in earlier (1-3) compared with later (4+) lines of therapy. (C) Swimmer plot showing rwOS from treatment initiation for each patient in the CAPIm+ (top) and CAPIm– (bottom) groups, with annotations indicating line of therapy and genomic profile. rwOS bars are colored by gene variant; black dots indicate earliest progression event after treatment initiation. AIC, Akaike information correction; HR, hazard ratio; LoT, line of treatment; mPFS, median progression-free survival; rwOS, real-world overall survival; rwPFS, real-world progression-free survival.

DISCUSSION

Here we report the largest landscape study, to our knowledge, of PIK3CA, AKT1, and PTEN genomic alterations in breast tumors, encompassing 29,157 alterations across 51,767 samples. The frequency and distribution of PIK3CA and AKT1 alterations were consistent with prior reports.10-12,22,33-35 PTEN alterations were more common (13.5%) than previously described (≤10%),11,12 potentially reflecting broader sequencing coverage.

The most common PIK3CA variants observed were H1047R, E545K, E542K, N345K, and H1047L, consistent with previous reports.34,35 The roles of nonhotspot mutations remain variably understood although some less common variants including N345K36,37 have been associated with alpelisib sensitivity in preclinical studies. We also found that 16.2% of PIK3CA-altered samples had multiple PIK3CA alterations, similar to previous studies,38 which have been shown to result in increased PI3K activity and increased sensitivity to PI3Kα inhibitors. AKT1 E17K accounted for over two thirds of AKT1 variants, leaving a notable minority of nonhotspot alterations that remain poorly characterized. A subset of non-E17K variants have been shown to sensitize tumors to AKT inhibition or confer selective hypersensitivity to ATP-competitive AKT inhibitors,39 underscoring the need to further evaluate mutant-specific therapeutic vulnerabilities. ER–/HER2– cases had a higher proportion of nonhotspot PIK3CA variants than other receptor subtypes; however, given historically negative trials of AKT pathway inhibition in ER–/HER2– breast cancer,40-43 the therapeutic implications of this trend remain unclear.

Our ancestry analyses revealed that PIK3CA alterations were less common in patients with African ancestry, consistent with smaller prior studies.44-47 Interestingly, no such differences were observed in AKT1 and PTEN, suggesting that ancestry-based differences in PIK3CA prevalence are gene-specific rather than affecting the AKT pathway globally. This could reflect differential tumor biology that limits the early clonal acquisition of PIK3CA mutations in African patients, highlighting one potential biological explanation for lower representation of this population in PIK3CA biomarker-based trials. Conversely, PIK3CA prevalence was higher in patients with South Asian ancestry and ER+/HER2– disease, underscoring the need for representative patient enrollment in biomarker-based clinical trials.

Our gene association analyses corroborated prior findings of strong mutual exclusivity among pathogenic PIK3CA, AKT1, and PTEN alterations,10,28,35 likely reflecting redundancies in AKT pathway activation. Co-occurrence patterns involving PIK3CA and AKT1 were consistent with their increased prevalence in ER+ disease and ILC. We identified a novel co-occurrence between PIK3CA and SOX2, a combination previously implicated in the pathogenesis of head and neck cancers but not well-described in breast cancers. PTEN-altered samples exhibited known enrichment for mutations in other tumor suppressors including FAS and TP53.

The functional characterization of PTEN VUS remains an important challenge in clinical genomics. By leveraging large-scale DMS data sets, we show that nearly one third of missense PTEN variants were consistently classified as nonpathogenic by both abundance and fitness-based assays, highlighting the value of functional data in refining clinical variant interpretation. However, discordant classifications in nearly one third of missense variants suggest that a substantial proportion of mutations may not predict straightforward loss of function, warranting further mechanistic investigation as to whether they predict for AKT inhibitor sensitivity. Expanding and refining such data sets will be critical for improving precision in molecular diagnostics and therapeutic guidance.

Finally, in our real-world clinical cohort treated with capivasertib and fulvestrant, patients with rare AKT pathway variants achieved clinically meaningful benefit comparable with that seen in patients with capivasertib-qualifying gene variants. Although sample sizes were small given the recent approval of capivasertib, these findings suggest that therapeutic benefit from AKT inhibition may extend to patients beyond those who meet current biomarker eligibility criteria.

Our study has limitations including the lack of available ER status on a subset of patients although the mutational distributions observed across receptor subtypes were consistent with prior reports. Samples are restricted to US-based patients, which could limit the generalizability of the ancestry-related findings.

In conclusion, this large CGP data set of >51,000 breast tumors reveals a wide spectrum of alterations in PIK3CA/AKT1/PTEN, encompassing well-established pathogenic variants and rare variants of uncertain significance. Functional tools such as DMS and continued clinical studies are needed to clarify the roles of lesser-known variants as predictive biomarkers, and ancestry-based differences in PIK3CA prevalence should be considered in clinical trials of AKT inhibition.

Saumya D. Sisoudiya

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Hanna Tukachinsky

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Alexa B. Schrock

Employment: Foundation Medicine

Stock and Other Ownership Interests: Foundation Medicine, Roche

Smruthy Sivakumar

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Ethan S. Sokol

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Neil Vasan

Leadership: Meliora Therapeutics

Stock and Other Ownership Interests: Meliora Therapeutics, Reactive Biosciences

Honoraria: Guidepoint Global, Gerson Lehrman Group, ExpertConnect

Consulting or Advisory Role: Apertor Therapeutics, Avenzo Therapeutics, Reactive Biosciences, Scorpion Therapeutics

Research Funding: Gilead Sciences

Patents, Royalties, Other Intellectual Property: Vasan N, Baselga J. U.S. Patent Number WO2020041684A1: “Biomarkers for Determining Responsiveness of a Cancer to PI3K Inhibitors,” Patent granted February 27, 2020, Vasan N, Sanjana N “Compositions and Methods For Base Editing Kinase Genes,” U.S. Provisional Patent No. 18/824,298, filed September 4, 2024, Wolters Kluwer royalties

No other potential conflicts of interest were reported.

PRIOR PRESENTATION

Presented at the San Antonio Breast Cancer Symposium, San Antonio, TX, December 10-13, 2024.

SUPPORT

Supported by NIH K08CA245192, the Breast Cancer Alliance Young Investigator Grant, the Gilead Science Research Scholars Grant, the Conquer Cancer Foundation Young Investigator Award, and the Conquer Cancer Foundation ASCO-BCRF Advanced Clinical Research Award in Breast Cancer.

*

J.J.T. and S.D.S. contributed equally to this work.

DATA SHARING STATEMENT

A data sharing statement provided by the authors is available with this article at DOI https://doi.org/10.1200/PO-25-00609.

All relevant data are provided within the article and its accompanying Supplementary Data. Due to HIPAA requirements, we are not authorized to share individualized patient genomic data, which contain potentially identifying or sensitive patient information. Foundation Medicine is committed to collaborative data analysis, with well-established and widely utilized mechanisms by which investigators can query our core genomic database of >700,000 deidentified sequenced cancers to obtain aggregated data sets. For more information and mechanisms of access, please contact the corresponding author(s) or the Foundation Medicine, Inc Data Governance Council at data.governance.council@foundationmedicine.com. For the data analysis of the clinicogenomic database (CGDB), the data were originated by and are the property of Flatiron Health, Inc and Foundation Medicine, Inc Requests for data sharing by license or by permission for the specific purpose of replicating results in this manuscript can be submitted to PublicationsDataAccess@flatiron.com and cgdb-fmi@flatiron.com.

AUTHOR CONTRIBUTIONS

Conception and design: Jacqueline J. Tao, Ethan S. Sokol, Neil Vasan

Collection and assembly of data: Saumya D. Sisoudiya

Data analysis and interpretation: Jacqueline J. Tao, Saumya D. Sisoudiya, Hanna Tukachinsky, Alexa B. Schrock, Ericka M. Ebot, Smruthy Sivakumar, Ethan S. Sokol

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/po/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Saumya D. Sisoudiya

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Hanna Tukachinsky

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Alexa B. Schrock

Employment: Foundation Medicine

Stock and Other Ownership Interests: Foundation Medicine, Roche

Smruthy Sivakumar

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Ethan S. Sokol

Employment: Foundation Medicine

Stock and Other Ownership Interests: Roche

Neil Vasan

Leadership: Meliora Therapeutics

Stock and Other Ownership Interests: Meliora Therapeutics, Reactive Biosciences

Honoraria: Guidepoint Global, Gerson Lehrman Group, ExpertConnect

Consulting or Advisory Role: Apertor Therapeutics, Avenzo Therapeutics, Reactive Biosciences, Scorpion Therapeutics

Research Funding: Gilead Sciences

Patents, Royalties, Other Intellectual Property: Vasan N, Baselga J. U.S. Patent Number WO2020041684A1: “Biomarkers for Determining Responsiveness of a Cancer to PI3K Inhibitors,” Patent granted February 27, 2020, Vasan N, Sanjana N “Compositions and Methods For Base Editing Kinase Genes,” U.S. Provisional Patent No. 18/824,298, filed September 4, 2024, Wolters Kluwer royalties

No other potential conflicts of interest were reported.

REFERENCES

  • 1.Vasan N, Cantley LC: At a crossroads: How to translate the roles of PI3K in oncogenic and metabolic signalling into improvements in cancer therapy. Nat Rev Clin Oncol 19:471-485, 2022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Vanhaesebroeck B, Perry MWD, Brown JR, et al. : PI3K inhibitors are finally coming of age. Nat Rev Drug Discov 20:741-769, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Castel P, Toska E, Engelman JA, et al. : The present and future of PI3K inhibitors for cancer therapy. Nat Cancer 2:587-597, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Andre F, Ciruelos E, Rubovszky G, et al. : Alpelisib for PIK3CA-mutated, hormone receptor-positive advanced breast cancer. N Engl J Med 380:1929-1940, 2019 [DOI] [PubMed] [Google Scholar]
  • 5.Turner NC, Im SA, Saura C, et al. : Inavolisib-based therapy in PIK3CA-mutated advanced breast cancer. N Engl J Med 391:1584-1596, 2024 [DOI] [PubMed] [Google Scholar]
  • 6.U.S. Food and Drug Administration : FDA Approves Capivasertib with Fulvestrant for Breast Cancer. 2023. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-capivasertib-fulvestrant-breast-cancer [Google Scholar]
  • 7.Turner NC, Oliveira M, Howell SJ, et al. : Capivasertib in hormone receptor-positive advanced breast cancer. N Engl J Med 388:2058-2070, 2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Millis SZ, Ikeda S, Reddy S, et al. : Landscape of Phosphatidylinositol-3-Kinase pathway alterations across 19 784 diverse solid tumors. JAMA Oncol 2:1565-1573, 2016 [DOI] [PubMed] [Google Scholar]
  • 9.Razavi P, Chang MT, Xu G, et al. : The genomic landscape of endocrine-resistant advanced breast cancers. Cancer Cell 34:427-438 e6, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Stemke-Hale K, Gonzalez-Angulo AM, Lluch A, et al. : An integrative genomic and proteomic analysis of PIK3CA, PTEN, and AKT mutations in breast cancer. Cancer Res 68:6084-6091, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Tada H, Miyashita M, Harada-Shoji N, et al. : Clinicopathogenomic analysis of PI3K/AKT/PTEN-altered luminal metastatic breast cancer in Japan. Breast Cancer 32:208-216, 2025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Bhave MA, Quintanilha JCF, Tukachinsky H, et al. : Comprehensive genomic profiling of ESR1, PIK3CA, AKT1, and PTEN in HR(+)HER2(-) metastatic breast cancer: Prevalence along treatment course and predictive value for endocrine therapy resistance in real-world practice. Breast Cancer Res Treat 207:599-609, 2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Frampton GM, Fichtenholtz A, Otto GA, et al. : Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol 31:1023-1031, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Milbury CA, Creeden J, Yip WK, et al. : Clinical and analytical validation of FoundationOne(R)CDx, a comprehensive genomic profiling assay for solid tumors. PLoS One 17:e0264138, 2022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Jiagge E, Jin DX, Newberg JY, et al. : Tumor sequencing of African ancestry reveals differences in clinically relevant alterations across common cancers. Cancer Cell 41:1963-1971 e3, 2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Matreyek KA, Starita LM, Stephany JJ, et al. : Multiplex assessment of protein variant abundance by massively parallel sequencing. Nat Genet 50:874-882, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Matreyek KA, Stephany JJ, Ahler E, et al. : Integrating thousands of PTEN variant activity and abundance measurements reveals variant subgroups and new dominant negatives in cancers. Genome Med 13:165, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Mighell TL, Evans-Dutson S, O'Roak BJ: A saturation mutagenesis approach to understanding PTEN lipid phosphatase activity and genotype-phenotype relationships. Am J Hum Genet 102:943-955, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Flatiron Health : Database Characterization Guide. 2025. https://flatiron.com/database-characterization [Google Scholar]
  • 20.Singal G, Miller PG, Agarwala V, et al. : Association of patient characteristics and tumor genomics with clinical outcomes among patients with non-small cell lung cancer using a clinicogenomic database. JAMA 321:1391-1399, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Samuels Y, Wang Z, Bardelli A, et al. : High frequency of mutations of the PIK3CA gene in human cancers. Science 304:554, 2004 [DOI] [PubMed] [Google Scholar]
  • 22.Xiao W, Zhang G, Chen B, et al. : Mutational landscape of PI3K-AKT-mTOR pathway in breast cancer: Implications for targeted therapeutics. J Cancer 12:4408-4417, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Montoro-Jimenez I, Granda-Diaz R, Menendez ST, et al. : Combined PIK3CA and SOX2 gene amplification predicts laryngeal cancer risk beyond histopathological grading. Int J Mol Sci 25:2695, 2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.An Y, Adams JR, Hollern DP, et al. : Cdh1 and Pik3ca mutations cooperate to induce immune-related invasive lobular carcinoma of the breast. Cell Rep 25:702-714 e6, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Ciriello G, Gatza ML, Beck AH, et al. : Comprehensive molecular portraits of invasive lobular breast cancer. Cell 163:506-519, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Nixon MJ, Formisano L, Mayer IA, et al. : PIK3CA and MAP3K1 alterations imply luminal A status and are associated with clinical benefit from pan-PI3K inhibitor buparlisib and letrozole in ER+ metastatic breast cancer. NPJ Breast Cancer 5:31, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Avivar-Valderas A, McEwen R, Taheri-Ghahfarokhi A, et al. : Functional significance of co-occurring mutations in PIK3CA and MAP3K1 in breast cancer. Oncotarget 9:21444-21458, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mosele F, Stefanovska B, Lusque A, et al. : Outcome and molecular landscape of patients with PIK3CA-mutated metastatic breast cancer. Ann Oncol 31:377-386, 2020 [DOI] [PubMed] [Google Scholar]
  • 29.Sokol ES, Feng YX, Jin DX, et al. : Loss of function of NF1 is a mechanism of acquired resistance to endocrine therapy in lobular breast cancer. Ann Oncol 30:115-123, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Di Cristofano A, Kotsi P, Peng YF, et al. : Impaired Fas response and autoimmunity in Pten+/- mice. Science 285:2122-2125, 1999 [DOI] [PubMed] [Google Scholar]
  • 31.Liu JC, Voisin V, Wang S, et al. : Combined deletion of Pten and p53 in mammary epithelium accelerates triple-negative breast cancer with dependency on eEF2K. EMBO Mol Med 6:1542-1560, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Gasparyan M, Lo MC, Jiang H, et al. : Combined p53- and PTEN-deficiency activates expression of mesenchyme homeobox 1 (MEOX1) required for growth of triple-negative breast cancer. J Biol Chem 295:12188-12202, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Rudolph M, Anzeneder T, Schulz A, et al. : AKT1 (E17K) mutation profiling in breast cancer: Prevalence, concurrent oncogenic alterations, and blood-based detection. BMC Cancer 16:622, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Martinez-Saez O, Chic N, Pascual T, et al. : Frequency and spectrum of PIK3CA somatic mutations in breast cancer. Breast Cancer Res 22:45, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Rugo HS, Raskina K, Schrock AB, et al. : Biology and targetability of the extended spectrum of PIK3CA mutations detected in breast carcinoma. Clin Cancer Res 29:1056-1067, 2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Gymnopoulos M, Elsliger MA, Vogt PK: Rare cancer-specific mutations in PIK3CA show gain of function. Proc Natl Acad Sci USA 104:5569-5574, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Dogruluk T, Tsang YH, Espitia M, et al. : Identification of variant-specific functions of PIK3CA by rapid phenotyping of rare mutations. Cancer Res 75:5341-5354, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Vasan N, Razavi P, Johnson JL, et al. : Double PIK3CA mutations in cis increase oncogenicity and sensitivity to PI3Kalpha inhibitors. Science 366:714-723, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Shrestha Bhattarai T, Shamu T, Gorelick AN, et al. : AKT mutant allele-specific activation dictates pharmacologic sensitivities. Nat Commun 13:2111, 2022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Pascual J, Turner NC: Targeting the PI3-kinase pathway in triple-negative breast cancer. Ann Oncol 30:1051-1060, 2019 [DOI] [PubMed] [Google Scholar]
  • 41.Martin M, Chan A, Dirix L, et al. : A randomized adaptive phase II/III study of buparlisib, a pan-class I PI3K inhibitor, combined with paclitaxel for the treatment of HER2- advanced breast cancer (BELLE-4). Ann Oncol 28:313-320, 2017 [DOI] [PubMed] [Google Scholar]
  • 42.Dent RA, Kim SB, Oliveira M, et al. : Ipatasertib plus paclitaxel for patients with PIK3CA/AKT1/PTEN-altered locally advanced unresectable or metastatic triple-negative breast cancer in the IPATunity130 phase III trial. Clin Cancer Res 30:4329-4338, 2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.AstraZeneca : Update on the CAPItello-290 phase III trial for truqap plus chemotherapy in advanced or metastatic triple-negative breast cancer. 2024. https://www.astrazeneca.com/media-centre/press-releases/2024/update-on-capitello-290-phase-iii-trial.html
  • 44.Chen JW, Murugesan K, Newberg JY, et al. : Comparison of PIK3CA mutation prevalence in breast cancer across predicted ancestry populations. JCO Precis Oncol 10.1200/PO.22.00341 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Pitt JJ, Riester M, Zheng Y, et al. : Characterization of Nigerian breast cancer reveals prevalent homologous recombination deficiency and aggressive molecular features. Nat Commun 9:4181, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Jouali F, Marchoudi N, Talbi S, et al. : Detection of PIK3/AKT pathway in Moroccan population with triple negative breast cancer. BMC Cancer 18:900, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Omilian AR, Wei L, Hong CC, et al. : Somatic mutations of triple-negative breast cancer: A comparison between Black and White women. Breast Cancer Res Treat 182:503-509, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

A data sharing statement provided by the authors is available with this article at DOI https://doi.org/10.1200/PO-25-00609.

All relevant data are provided within the article and its accompanying Supplementary Data. Due to HIPAA requirements, we are not authorized to share individualized patient genomic data, which contain potentially identifying or sensitive patient information. Foundation Medicine is committed to collaborative data analysis, with well-established and widely utilized mechanisms by which investigators can query our core genomic database of >700,000 deidentified sequenced cancers to obtain aggregated data sets. For more information and mechanisms of access, please contact the corresponding author(s) or the Foundation Medicine, Inc Data Governance Council at data.governance.council@foundationmedicine.com. For the data analysis of the clinicogenomic database (CGDB), the data were originated by and are the property of Flatiron Health, Inc and Foundation Medicine, Inc Requests for data sharing by license or by permission for the specific purpose of replicating results in this manuscript can be submitted to PublicationsDataAccess@flatiron.com and cgdb-fmi@flatiron.com.

Articles from JCO Precision Oncology are provided here courtesy of Wolters Kluwer Health

RESOURCES