Abstract
Purpose
Ovarian clear cell carcinoma (OCCC) is frequently characterized by dysregulation of the PI3K/AKT/mTOR (PAM) pathway; however, the expression patterns and clinical significance of class I PI3K catalytic isoforms remain incompletely understood. This study aimed to comprehensively evaluate PI3K catalytic subunit expression, co-expression patterns, and their prognostic relevance in OCCC.
Methods
A retrospective cohort of 100 patients with OCCC was analyzed. Immunohistochemistry was performed to assess the expression of p110α, p110β, p110δ, p110γ, and pAKT. PIK3CA exon 9 and exon 20 mutations were evaluated in available formalin-fixed paraffin-embedded tissues. Associations between PI3K isoforms, pAKT activation, clinicopathological characteristics, and survival outcomes were investigated.
Results
The expression positivity rates of the four class I PI3K catalytic subunits in OCCC were 86% for p110γ, 77% for p110α, 58% for p110δ, and 56% for p110β. p110α and p110β were positively associated with pAKT activation, whereas p110δ and p110γ showed no significant association. p110δ expression was significantly correlated with PI3KCA exon9 mutation. Frequent co-expression of multiple PI3K catalytic subunits was observed, with simultaneous expression of all four isoforms representing the predominant pattern and significantly correlating with pAKT positivity. In advanced-stage disease, p110β expression was significantly associated with poorer progression-free survival (PFS), while pAKT positivity correlated with worse overall survival (OS) and PFS.
Conclusions
OCCC displays substantial heterogeneity in PI3K catalytic isoform expression and co-expression. p110α and p110β were associated with pAKT activation. In advanced-stage OCCC, high p110β expression was associated with poorer PFS, whereas pAKT activation was associated with both poorer PFS and OS, supporting further investigation of PI3K-targeted therapeutic strategies.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s00404-026-08525-w.
Keywords: Ovarian clear cell carcinoma, PIK3CA mutations, PI3K subunits, Prognosis
Introduction
Ovarian clear cell carcinoma (OCCC) is characterized by a variety of molecular abnormalities. Among them, the mutation rate of the ARID1A gene is approximately 50%, while that of the PIK3CA gene ranges from 30 to 40% [1, 2]. Notably, PIK3CA mutations often co-occur with ARID1A mutations and are considered early molecular events in the disease process. Studies have demonstrated that coexistence of ARID1A and PIK3CA mutations can drive tumorigenesis [1], with the loss of ARID1A may lead to chromatin remodeling defects that could enhance the activation of the PI3K-AKT-mTOR (PAM) pathway by PIK3CA mutations [2]. In addition, other low-frequency mutations, including PTEN, TP53, TERT, KRAS, and BRCA2 have also been observed [2, 3]. These genetic alterations ultimately trigger the abnormal activation of the PAM pathway as a key mechanism in OCCC.
The PAM pathway is critical for regulating cell proliferation, survival, and metabolism, and its dysregulation is common in many cancers. This pathway is frequently activated by mechanisms such as PIK3CA mutations, PTEN loss, and PIK3CA amplifications [4]. The catalytic subunits of PI3K, especially the Class I isoforms (p110α, p110β, p110δ, and p110γ), play a crucial role in this activation. However, the role of different PI3K catalytic subunits in OCCC remain unclear.
This study aims to explore the expression and prognostic value of PI3K catalytic subunits in OCCC, and their associations with clinical outcomes, thereby providing a theoretical basis for targeted therapeutic strategies.
Method
Patients and tissue samples
Patients were included in this study based on the following criteria: (1) patients diagnosed with primary OCCC who underwent surgical resection at Fudan University Shanghai Cancer Center (FUSCC) between 2000 and 2015; and (2) availability of tumor specimens suitable for tissue microarray (TMA) construction. Clinical data extracted from medical records included age, tumor size, laterality, presence of ascites, preoperative serum CA125 levels, and International Federation of Gynecology and Obstetrics (FIGO) stage, which was reclassified according to the 2014 FIGO guidelines [5]. We also evaluated the presence of endometriosis (ovarian and/or extra-ovarian) in the specimens. Follow-up data were collected through medical records or telephone interviews until May 2023.
Tissue microarray (TMA) construction and immunohistochemistry
TMAs were constructed using 1-mm triplicate core samples from 108 OCCC specimens, while the remaining 8 cases were excluded due to tissue loss from the tissue microarray. Immunohistochemical staining was performed on the TMAs following previously described protocols [6], using antibodies against p110α, p110β, p110δ, p110γ, and pAKT (Supplementary Table 2). All H&E and IHC slides were reviewed by Dr. Bi. PI3K catalytic subunits and pAKT expression were semi-quantitatively scored (0–3 +) according to staining intensity. Cytoplasmic positivity was defined as immunoreactivity in ≥ 10% of tumor cells within the tissue cores. Considering the intratumoral heterogeneity within TMAs, the average score of 3 TMA cores was used as the final score. Immunohistochemical scores of 2 + and 3 + were considered positive [7].
DNA extraction and Sanger sequencing
Genomic DNA was extracted from all formalin-fixed paraffin-embedded (FFPE) tissue samples using the QIAGEN DNA FFPE Tissue Kit (Qiagen, Shanghai, China). DNA quality and yield were assessed using a NanoDrop Spectrophotometer (Thermo Fisher Scientific, Shanghai, China). Sanger DNA sequencing was performed by Sangon Biotech Co., Ltd. (Shanghai, China), and the sequencing data was analyzed using SnapGene version 6.0.2 software. All detected mutations were confirmed by resequencing. The primers used for PIK3CA exons 9 and 20 are listed in Supplementary Table 3.
Statistical analysis
Statistical analyses were performed to identify correlations between p110α, p110β, p110δ, p110γ, pAKT and other clinicopathologic factors using the χ2 test or Fisher’s exact test, as appropriate. To account for multiple testing in the association analyses, P-values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) procedure, and p-value < 0.05 was considered statistically significant. OS and PFS were analyzed using the Kaplan–Meier method, and differences between survival curves were evaluated using the log-rank test. The Cox proportional hazards regression model was used to investigate the prognostic significance of clinicopathologic factors. To identify independent prognostic factors, all significant variables in the univariate Cox regression analysis (P < 0.05) were subjected to multivariate Cox regression analysis. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. For survival analyses, a two-sided p-value < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS v25 and R v4.2.2.
Results
PI3K catalytic subunit expression and PIK3CA mutational profiles in OCCC
The median age was 54 years (range, 28–83 years), and the median maximum tumor diameter was 11 cm (range, 2–27 cm). A total of 69% (69/100) of patients were classified as FIGO stage I or II, and 79% (79/100) had unilateral disease. Endometriosis was observed in 36% (36/100) of ovarian and/or extra-ovarian specimens, adenofibroma was present in 5% (5/97) of patients, and ascites was observed in 47% (47/100) of cases. Elevated preoperative serum CA125 levels were observed in 78% (64/82) of patients. The clinicopathological characteristics of the study cohort are provided in Supplementary Table 1.
Immunohistochemical analysis of class I PI3K catalytic subunits and pAKT was successfully performed in all cases. Among the four PI3K catalytic isoforms, p110γ was the most frequently expressed subunit, detected in 86% of tumors, followed by p110α (77%), p110δ (58%), and p110β (56%). In addition, pAKT positivity was observed in 58% of tumors, indicating frequent activation of downstream PI3K signaling in OCCC. Representative immunohistochemical staining images are shown in Fig. 1a–d.
Fig. 1.
a Immunohistochemical staining of p110α; b Immunohistochemical stain for p110β; c Immunohistochemical stain for p110δ; d Immunohistochemical stain for p110γ; e Immunohistochemical stain for pAKT. The images were captured at 40 × magnification
Due to prolonged storage of FFPE tissue blocks, sequencing of PIK3CA exon 9 was successfully completed in 75 cases, and exon 20 in 77 cases. To evaluate the potential for selection bias resulting from incomplete PIK3CA sequencing, clinicopathological characteristics were compared between the sequenced (n = 77) and non-sequenced (n = 23) cohorts. No significant differences were observed between the two groups (Supplementary Table 4), suggesting that unsuccessful sequencing due to FFPE degradation was unlikely to introduce substantial selection bias.
DNA sequence analysis of exons 9 and 20 of the PIK3CA gene revealed mutations in 52% (40/77) of OCCC patients. Specifically, mutations in exon 9 were present in 39% (29/75) of tumors, while exon 20 mutations were detected in 21% (16/77). Co-mutations in both exons were observed in 6.5% (5/77) of patients. A heatmap integrating patient survival statuses with clinicopathological features, PI3K catalytic subunit expression, and PIK3CA mutational profiles is shown in Fig. 2.
Fig. 2.
Heatmap of Clinicopathological Factors, PI3K catalytic subunits and pAKT expression
Association between class I PI3K catalytic subunits expression and clinicopathological characteristics
To investigate the biological relevance of PI3K signaling activation in OCCC, we analyzed the relationship between PI3K catalytic subunit expression and pAKT activation. pAKT positivity was significantly associated with increased expression of p110α and p110β, whereas no significant associations were observed with p110δ or p110γ expression. In addition, p110δ expression was significantly correlated with PIK3CA exon 9 mutations (p = 0.035, Table 1), suggesting a potential functional link between genomic alterations and downstream pathway activation.
Table 1.
The association between the catalytic subunits of the class I PI3Ks and clinicopathological factors
| p110α | P | p110β | P | p110δ | P | p110γ | P | pAKT | P | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Negative(%) | Positive(%) | Negative(%) | Positive(%) | Negative(%) | Positive(%) | Negative(%) | Positive(%) | Negative(%) | Positive(%) | ||||||
| Age | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| < 60y | 18 (23) | 60 (77) | 36 (46) | 42 (54) | 32 (41) | 46 (59) | 12 (15) | 66 (85) | 33 (42) | 45 (58) | |||||
| ≥ 60y | 5 (23) | 17 (77) | 8 (36) | 14 (64) | 10 (45) | 12 (55) | 2 (9) | 20 (91) | 9 (41) | 13 (59) | |||||
| Tumor size | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| < 15 cm | 19 (26) | 55 (74) | 33 (45) | 41 (55) | 32 (43) | 42 (57) | 10 (13) | 64 (87) | 31 (42) | 43 (58) | |||||
| ≥ 15 cm | 4 (15) | 22 (85) | 11 (42) | 15 (58) | 10 (38) | 16 (62) | 4 (15) | 22 (85) | 11 (42) | 15 (58) | |||||
| Laterality | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Unilaterality | 17 (22) | 62 (78) | 35 (44) | 44 (56) | 32 (40) | 47 (60) | 12 (15) | 67 (85) | 34 (43) | 45 (57) | |||||
| Bilaterality | 6 (29) | 15 (71) | 9 (43) | 12 (57) | 10 (48) | 11 (52) | 2 (9) | 19 (91) | 8 (38) | 13 (62) | |||||
| CA125 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Normal | 4 (22) | 14 (78) | 6 (33) | 12 (67) | 6 (33) | 12 (67) | 3 (17) | 15 (83) | 8 (44) | 10 (56) | |||||
| Abnormal | 16 (25) | 48 (75) | 28 (44) | 36 (56) | 28 (44) | 36 (56) | 10 (16) | 54 (84) | 24 (38) | 40 (62) | |||||
| Ascites | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Absent | 11 (21) | 42 (79) | 25 (47) | 28 (53) | 20 (38) | 33 (62) | 9 (17) | 44 (83) | 26 (49) | 27 (51) | |||||
| Present | 12 (25) | 35 (75) | 19 (40) | 28 (60) | 22 (47) | 25 (53) | 5 (11) | 42 (89) | 16 (34) | 31 (66) | |||||
| Endometriosis | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Absent | 16 (25) | 48 (75) | 30 (47) | 34 (53) | 30 (47) | 34 (53) | 8 (12) | 56 (88) | 27 (42) | 37 (58) | |||||
| Present | 7 (19) | 29 (81) | 14 (39) | 22 (61) | 12 (33) | 24 (67) | 6 (17) | 30 (83) | 15 (42) | 21 (58) | |||||
| Stage | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| I-II | 13 (19) | 56 (81) | 28 (41) | 41 (59) | 27 (39) | 42 (61) | 8 (12) | 61 (88) | 28 (41) | 41 (59) | |||||
| III-IV | 10 (32) | 21 (68) | 16 (52) | 15 (48) | 15 (48) | 16 (52) | 6 (19) | 25 (81) | 14 (45) | 17 (55) | |||||
| Disease Progression | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| No | 12 (22) | 42 (78) | 24 (44) | 30 (56) | 22 (41) | 32 (59) | 6 (11) | 48 (89) | 22 (41) | 32 (59) | |||||
| Yes | 11 (24) | 35 (76) | 20 (44) | 26 (56) | 20 (43) | 26 (57) | 8 (17) | 38 (83) | 20 (44) | 26 (56) | |||||
| Survival Status | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Alive | 13 (21) | 49 (79) | 28 (45) | 34 (55) | 27 (43) | 35 (57) | 7 (11) | 55 (89) | 27 (44) | 35 (56) | |||||
| Dead | 10 (26) | 28 (74) | 16 (42) | 22 (58) | 15 (39) | 23 (61) | 7 (18) | 31 (82) | 15 (40) | 23 (60) | |||||
| pAKT | 0.020 | < 0.001 | 0.552 | 1.000 | |||||||||||
| Negative | 17 (41) | 25 (59) | 32 (76) | 10 (24) | 23 (55) | 19 (45) | 8 (19) | 34 (81) | |||||||
| Positive | 6 (10) | 52 (90) | 12 (21) | 46 (79) | 19 (33) | 39 (67) | 6 (10) | 52 (90) | |||||||
| PIK3CA Exon9 | 1.000 | 1.000 | 0.035 | 1.000 | 1.000 | ||||||||||
| Wild-type | 14 (30) | 32 (70) | 19 (41) | 27 (59) | 29 (63) | 17 (37) | 7 (15) | 39 (85) | 17 (37) | 29 (63) | |||||
| Mutant | 5 (17) | 24 (83) | 18 (62) | 11 (38) | 7 (24) | 22 (76) | 6 (21) | 23 (79) | 15 (52) | 14 (48) | |||||
| PIK3CA Exon20 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||||||
| Wild-type | 16 (26) | 45 (74) | 31 (51) | 30 (49) | 29 (47) | 32 (53) | 11 (18) | 50 (82) | 28 (46) | 33 (54) | |||||
| Mutant | 3 (19) | 13 (81) | 7 (44) | 9 (56) | 8 (50) | 8 (50) | 2 (12) | 14 (88) | 5 (31) | 11 (69) | |||||
Survival analyses demonstrated that patients with early-stage disease (FIGO I–II), unilateral tumors and absence of ascites had significantly better OS and PFS than those with advanced-stage disease, bilateral involvement and ascites (Supplementary Fig. 1 and 2a,b,d). Elevated preoperative CA125 levels were associated with poorer PFS, although no significant association with OS was observed (Supplementary Fig. 1c and 2c). Notably, analysis of individual PI3K catalytic subunits in the overall cohort did not reveal significant associations with OS or PFS (Supplementary Fig. 1 and 2e–o), suggesting that the prognostic impact of PI3K signaling may be context-dependent.
Co-expression architecture of PI3K catalytic isoforms reveals heterogeneous signaling patterns
Considering the context-dependent redundancy and compensatory interactions among PI3K catalytic isoforms [8–10], we next investigated the combinatorial expression patterns of p110α, p110β, p110δ and p110γ rather than evaluating each isoform independently.
Among the 100 evaluable tumors, the most common expression pattern was simultaneous positivity for all four catalytic subunits, observed in 35% of cases. Isolated expression of individual isoforms was relatively uncommon and was identified in only 1% of tumors for p110α and 12% for p110γ. Additional co-expression patterns included p110α + p110β + p110γ (12%), p110α + p110δ + p110γ (11%), p110α + p110γ (9%), p110α + p110δ (5%), p110β + p110γ (3%), p110α + p110β + p110δ (3%), p110δ + p110γ (2%), p110β + p110δ + p110γ (2%), and p110α + p110β (1%). Approximately 4% of tumors showed no detectable expression of PI3K catalytic subunits.
These findings revealed marked inter-patient heterogeneity in PI3K isoform co-expression patterns. Importantly, several low-frequency isoforms rarely occurred alone but preferentially co-expressed with dominant isoforms, suggesting potential cooperative or compensatory signaling interactions within the PI3K pathway.
We further analyzed the relationship between PI3K co-expression patterns and pAKT activation (Table 2). Notably, simultaneous expression of all four catalytic subunits was significantly associated with pAKT positivity (p < 0.001). No statistically significant associations were identified for other expression combinations.
Table 2.
Association of PI3K catalytic subunit expression with pAKT expression
| Parameters | pAKT | ||
|---|---|---|---|
| Negative(%) | Positive(%) | P | |
| p110α | 0.607 | ||
| Negative | 41(41) | 58(59) | |
| Positive | 1(100) | 0(0) | |
| p110γ | 0.201 | ||
| Negative | 33(38) | 55(62) | |
| Positive | 9(75) | 3(25) | |
| p110α + p110β | 1 | ||
| Negative | 42(42) | 57(58) | |
| Positive | 0(0) | 1(100) | |
| p110α + p110δ | 0.841 | ||
| Negative | 39(41) | 56(59) | |
| Positive | 3(60) | 2(40) | |
| p110α + p110γ | 0.377 | ||
| Negative | 36(40) | 55(60) | |
| Positive | 6(67) | 3(33) | |
| p110β + p110γ | 1 | ||
| Negative | 41(42) | 56(58) | |
| Positive | 1(33) | 2(67) | |
| p110δ + p110γ | 0.377 | ||
| Negative | 40(41) | 58(59) | |
| Positive | 2(100) | 0(0) | |
| p110α + p110β + p110δ | 1 | ||
| Negative | 41(42) | 56(58) | |
| Positive | 1(33) | 2(67) | |
| p110α + p110β + p110γ | 0.548 | ||
| Negative | 39(44) | 49(56) | |
| Positive | 3(25) | 9(75) | |
| p110α + p110δ + p110γ | 0.211 | ||
| Negative | 34(38) | 55(62) | |
| Positive | 8(73) | 3(27) | |
| p110β + p110δ + p110γ | 0.377 | ||
| Negative | 40(41) | 58(59) | |
| Positive | 2(100) | 0(0) | |
| p110α + p110β + p110δ + p110γ | < 0.001 | ||
| Negative | 39(60) | 26(40) | |
| Positive | 3(9) | 32(91) | |
| none | 0.548 | ||
| Negative | 39(41) | 57(59) | |
| Positive | 3(75) | 1(25) | |
*Categories with very small sample sizes (n ≤ 3) were considered exploratory, and the corresponding statistical comparisons should be interpreted with caution
Prognostic significance of PI3K pathway activation in advanced-stage OCCC
Given that advanced-stage OCCC is characterized by aggressive biological behavior and that PI3K pathway dysregulation is common in OCCC [11, 12], we further investigated the prognostic significance of PI3K signaling specifically in patients with FIGO stage III–IV disease.
In advanced-stage tumors, p110β expression was significantly associated with poorer PFS (p = 0.01). Moreover, patients with p110β-positive tumors exhibited shorter OS numerically than the p110β-negative group, although the difference did not reach statistical significance (log-rank p = 0.12) (Fig. 3a, b).
Fig. 3.
Kaplan–Meier survival curves of (a) the expression of p110β for OS in stage III-IV OCCC, b the expression of p110β for PFS in stage III-IV OCCC, c the expression of pAKT for OS in stage III-IV OCCC, d the expression of pAKT for PFS in stage III-IV OCCC
Furthermore, pAKT positivity was significantly associated with both shorter OS (p = 0.046) and poorer PFS (p = 0.039) (Fig. 3c, d). In contrast, other clinicopathological parameters did not significantly affect survival outcomes in advanced-stage cases (Supplementary Fig. 3a–i).
Univariate Cox regression analysis identified advanced-stage (HR = 8.762, p < 0.001), bilateral involvement (HR = 4.213, p < 0.001), and the presence of ascites (HR = 2.403, p = 0.009) as factors associated with poorer OS. Similarly, univariate analysis for PFS demonstrated that advanced-stage (HR = 6.887, p < 0.001), bilateral involvement (HR = 3.624, p < 0.001), ascites (HR = 1.969, p = 0.024), and elevated CA125 levels (HR = 3.065, p = 0.035) were significantly associated with unfavorable prognosis. Multivariate Cox regression analysis further confirmed advanced stage as an independent prognostic factor for both worse OS (HR = 6.687, p < 0.001) and poorer PFS (HR = 5.356, p < 0.001) (Table 3).
Table 3.
Univariable and multivariable cox regression analysis of clinicopathological characteristics
| Parameters | PFS | OS | ||||
|---|---|---|---|---|---|---|
| HR | CI(95%) | P | HR | CI(95%) | P | |
| Univariable | ||||||
| Age | ||||||
| Tumor size | ||||||
| Laterality | 3.624 | 1.980–6.634 | < 0.001 | 4.213 | 2.179–8.146 | < 0.001 |
| CA125 | 3.065 | 1.082–8.678 | 0.035 | |||
| Ascites | 1.969 | 1.095–3.543 | 0.024 | 2.403 | 1.242–4.647 | 0.009 |
| Endometriosis | ||||||
| Stage | 6.887 | 3.761–12.609 | < 0.001 | 8.762 | 4.422–17.360 | < 0.001 |
| p110α | ||||||
| p110β | ||||||
| p110δ | ||||||
| p110γ | ||||||
| pAKT | ||||||
| PIK3CA Exon9 | ||||||
| PIK3CA Exon20 | ||||||
| Multivariable | ||||||
| Laterality | 1.3 | 0.579–2.921 | 0.525 | 1.667 | 0.801–3.466 | 0.172 |
| CA125 | 2.165 | 0.722–6.488 | 0.168 | |||
| Ascites | 1.019 | 0.456–2.278 | 0.963 | 1.263 | 0.621–2.566 | 0.519 |
| Stage | 5.356 | 2.443–11.743 | < 0.001 | 6.687 | 3.070–14.567 | < 0.001 |
To further evaluate the prognostic significance of PI3K signaling in patients with advanced-stage disease (FIGO stage III–IV), we performed additional univariate Cox regression analysis. The results showed that p110β expression (HR = 2.800, 95% CI: 1.235–6.348, p = 0.014) and pAKT expression (HR = 2.268, 95% CI: 1.022–5.030, p = 0.044) were significantly associated with poorer PFS in this subgroup (Table 4).
Table 4.
Univariate cox regression analysis of prognostic factors in patients with advanced-stage disease
| Parameters | PFS | OS | ||||
|---|---|---|---|---|---|---|
| HR | CI(95%) | P | HR | CI(95%) | P | |
| Univariable | ||||||
| Age | ||||||
| Tumor size | ||||||
| Laterality | ||||||
| CA125 | ||||||
| Ascites | ||||||
| Endometriosis | ||||||
| Stage | ||||||
| p110α | ||||||
| p110β | 2.8 | 1.235–6.348 | 0.014 | |||
| p110δ | ||||||
| p110γ | ||||||
| pAKT | 2.268 | 1.022–5.030 | 0.044 | |||
| PIK3CA Exon9 | ||||||
| PIK3CA Exon20 | ||||||
Collectively, these findings indicate that although individual PI3K catalytic subunits have limited prognostic value in the overall OCCC cohort, activation of the PI3K pathway—particularly pAKT signaling—identifies a biologically aggressive subset of advanced-stage tumors associated with poor clinical outcomes. In addition, p110β expression was associated with shorter PFS in patients with advanced-stage disease, suggesting that p110β-mediated PI3K signaling may contribute to disease progression in this subgroup.
Discussion
The PI3K/AKT/mTOR (PAM) signaling pathway plays a central role in regulating cell proliferation, survival, and apoptosis, and its dysregulation is a well-established driver of tumorigenesis. As a lipid kinase, PI3K initiates intracellular signaling cascades that regulate diverse cellular processes, including proliferation, metabolism, migration, and survival [13, 14]. Among PI3K family members, class I PI3Ks are activated by growth factor signaling and function as heterodimers consisting of a regulatory subunit (p85) and a catalytic subunit (p110) [15]. Class I PI3Ks comprise four catalytic isoforms: p110α (encoded by PIK3CA), p110β (encoded by PIK3CB), p110δ (encoded by PIK3CD), and p110γ (encoded by PIK3CG) [16, 17].
Emerging evidence suggests that different PI3K isoforms exert distinct and context-dependent roles in oncogenic signaling and tumor progression [17–19]. Among them, p110α is the best-characterized isoform and is frequently altered through PIK3CA mutations across multiple cancer types, serving as a major oncogenic driver in PIK3CA-dependent tumors [20]. By contrast, p110β, p110δ, and p110γ are less frequently mutated but may still contribute to aberrant PI3K signaling through alternative mechanisms, including immune signaling and inflammatory responses [21–23]. OCCC is characterized by frequent alterations in the PAM pathway, with previous studies primarily focusing on PIK3CA mutations and p110α dysregulation [24–26]. However, the expression patterns and clinical significance of other class I PI3K catalytic subunits in OCCC remain largely unexplored. In the present study, we found that p110α and p110γ were the predominant isoforms expressed in OCCC. Notably, the class I PI3K catalytic subunits p110α and p110β were positively associated with pAKT activation, whereas p110δ and p110γ showed no significant association. These findings are consistent with prior studies suggesting that elevated class I PI3K activity is associated with enhanced AKT phosphorylation in multiple malignancies [27–29]. Beyond individual isoforms, we observed frequent co-expression of multiple PI3K catalytic subunits, with the four-subunit pattern (p110α + p110β + p110δ + p110γ) representing the predominant expression profile. Such co-expression patterns may reflect pathway redundancy and sustained PI3K signaling activity in OCCC [8, 30, 31]. The simultaneous expression of multiple isoforms may enable functional cooperation within the PI3K network, allowing tumor cells to maintain signaling robustness under heterogeneous genetic and microenvironmental conditions [32]. This signaling redundancy may partly explain the limited efficacy of single-isoform inhibition and supports the potential rationale for broader or combination PI3K-targeted therapeutic strategies [9, 30, 33].
Among non-α isoforms, p110β has been increasingly recognized as a critical mediator of cell survival, migration, and therapeutic resistance, particularly in PTEN-deficient and EGFR-driven tumors [34, 35]. Importantly, we identified p110β as a potential prognostic factor in advanced-stage OCCC, where elevated p110β expression was associated with poorer PFS. Similarly, increased pAKT activation was associated with both poorer OS and PFS in the advanced-stage subgroup, further supporting the biological relevance of PI3K pathway activation in aggressive disease. As a downstream effector of the PI3K pathway, elevated pAKT expression has been associated with poor prognosis in several hematologic and solid tumors [36]. Overexpression of p110β mRNA in colorectal cancer is associated with shorter disease-free survival [29], while high expression levels of all four isoforms (p110α, p110β, p110δ, p110γ) correlate with poor prognosis in diffuse large B-cell lymphoma [37]. Collectively, these findings suggest that p110β expression may serve as a prognostic biomarker for disease progression in advanced-stage OCCC, whereas pAKT activation may represent a broader indicator of poor survival outcomes. Furthermore, these observations support the potential therapeutic relevance of PI3K pathway inhibition in OCCC. Nevertheless, further mechanistic studies and multicenter validation studies are warranted to determine whether isoform-specific or combination PI3K inhibition could improve clinical outcomes in patients with OCCC.
This study has several limitations that should be acknowledged. First, this was a retrospective single-center study with a relatively limited sample size, and some clinicopathological variables were not systematically recorded in historical cases due to the retrospective nature of the study, which may limit the completeness of pathological characterization. Given the rarity of OCCC, validation in larger, multicenter cohorts will be necessary to confirm the prognostic significance of PI3K isoform expression. Second, due to prolonged storage of FFPE tissue blocks, PIK3CA mutational analysis could not be completed in all cases, which may have reduced the statistical power for assessing associations between genomic alterations and PI3K isoform expression. In addition, although immunohistochemistry enabled systematic evaluation of PI3K catalytic subunit expression and pAKT activation, protein expression does not necessarily reflect functional signaling activity. Therefore, mechanistic studies are required to determine whether the observed co-expression patterns directly contribute to pathway redundancy, tumor progression, or therapeutic resistance. Third, several PI3K co-expression patterns occurred at relatively low frequencies, limiting statistical robustness and increasing the possibility of overestimating prognostic associations in subgroup analyses. Moreover, the number of patients with stage III–IV disease was relatively small, and only a few patients remained at risk during long-term follow-up, which may have resulted in unstable Kaplan–Meier survival estimates at later time points. Therefore, survival estimates for the tails of the curves and subgroup-specific prognostic associations should be interpreted with caution and further validated in larger cohorts. Finally, while our findings suggest potential therapeutic implications for broader or combination PI3K inhibition, the present study was observational and did not directly evaluate treatment responses. Accordingly, the proposed therapeutic relevance of specific PI3K isoforms remains hypothesis-generating and warrants further experimental and clinical validation.
In summary, our study provides a comprehensive characterization of class I PI3K catalytic subunit expression patterns in OCCC and highlights substantial heterogeneity in isoform distribution and co-expression. We found that the class I PI3K catalytic subunits, p110α and p110β, were positively associated with pAKT activation. Importantly, high p110β expression was associated with poorer PFS in patients with advanced-stage OCCC, whereas increased pAKT activation was associated with both poorer OS and PFS in this subgroup. Moreover, both high p110β expression and increased pAKT activation were identified as adverse prognostic factors for PFS in advanced-stage disease. Together, these findings support the biological importance of PI3K signaling heterogeneity in OCCC and provide a rationale for further investigation of isoform-specific and combination PI3K-targeted strategies in larger prospective studies.
Supplementary Information
Below is the link to the electronic supplementary material.
Author contributions
Conceptualization: Xiaonan Zhou, Yifei Liu, Rui Bi Methodology: Xiaonan Zhou, Yifei Liu Investigation: Xiaonan Zhou, Yifei Liu, Jing Zhang, Min Ren, Gang Ji, Xu Cai Data curation: Jing Zhang, Min Ren, Gang Ji, Xu Cai Formal analysis: Xiaonan Zhou, Yifei Liu Funding acquisition: Rui Bi Supervision: Rui Bi Writing – original draft: Xiaonan Zhou, Yifei Liu Writing – review & editing: Rui Bi.
Funding
This work was supported by grants from the National Natural Foundation Science of China (NSFC 81802597).
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was granted by the Ethics Committee of Fudan University Shanghai Cancer Center under Approval No. 050432–4-2108* dated 10 August 2021. Written informed consent was obtained from all human participants prior to inclusion in the research.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xiaonan Zhou and Yifei Liu have equally contributed to this work.
References
- 1.Ye S, Zhou S, Wu Y, Pei X, Jiang W, Shi W, Yang W, Zhou X, Shan B, Yang H (2023) Genomic profiling of ovarian clear cell carcinoma in Chinese patients reveals potential prognostic biomarkers for survival. Ann Med 55(1):2218104 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Chao A, Huang CY, Yu W, Lin CY, Lin H, Chao AS, Lin CT, Chou HH, Huang KG, Huang HJ et al (2024) Molecular profiling reveals novel therapeutic targets and clonal evolution in ovarian clear cell carcinoma. BMC Cancer 24(1):1403 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ackroyd SA, Arguello D, Ramos P, Mahdi H, ElNaggar A, Winer I, Holloway R, Krivak T, Jones N, Turner VG et al (2023) Molecular portraits of clear cell ovarian and endometrial carcinoma with comparison to clear cell renal cell carcinoma. Gynecol Oncol 169:164–171 [DOI] [PubMed] [Google Scholar]
- 4.Glaviano A, Foo ASC, Lam HY, Yap KCH, Jacot W, Jones RH, Eng H, Nair MG, Makvandi P, Geoerger B et al (2023) PI3K/AKT/mTOR signaling transduction pathway and targeted therapies in cancer. Mol Cancer 22(1):138 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mutch DG, Prat J (2014) 2014 FIGO staging for ovarian, fallopian tube and peritoneal cancer. Gynecol Oncol 133(3):401–404 [DOI] [PubMed] [Google Scholar]
- 6.Ge H, Xiao Y, Qin G, Gu Y, Cai X, Jiang W, Tu X, Yang W, Bi R (2021) Mismatch repair deficiency is associated with specific morphologic features and frequent loss of ARID1A expression in ovarian clear cell carcinoma. Diagn Pathol 16(1):12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Firoozinia M, Zareian Jahromi M, Moghadamtousi SZ, Nikzad S, Abdul Kadir H (2014) PIK3CA gene amplification and PI3K p110alpha protein expression in breast carcinoma. Int J Med Sci 11(6):620–625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Arcucci S, Ramos-Delgado F, Cayron C, Therville N, Gratacap MP, Basset C, Thibault B, Guillermet-Guibert J (2021) Organismal roles for the PI3Kalpha and beta isoforms: their specificity, redundancy or cooperation is context-dependent. Biochem J 478(6):1199–1225 [DOI] [PubMed] [Google Scholar]
- 9.Narci K, Kahraman DC, Koyas A, Ersahin T, Tuncbag N, Atalay RC (2022) Context dependent isoform specific PI3K inhibition confers drug resistance in hepatocellular carcinoma cells. BMC Cancer 22(1):320 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cintas C, Douche T, Dantes Z, Mouton-Barbosa E, Bousquet MP, Cayron C, Therville N, Pont F, Ramos-Delgado F, Guyon C et al (2021) Phosphoproteomics identifies PI3K inhibitor-selective adaptive responses in pancreatic cancer cell therapy and resistance. Mol Cancer Ther 20(12):2433–2445 [DOI] [PubMed] [Google Scholar]
- 11.Amano T, Chano T, Yoshino F, Kimura F, Murakami T (2019) Current position of the molecular therapeutic targets for ovarian clear cell carcinoma: a literature review. Healthcare (Basel). 10.3390/healthcare7030094 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tong A, Di X, Zhao X, Liang X (2023) Review the progression of ovarian clear cell carcinoma from the perspective of genomics and epigenomics. Front Genet 14:952379 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tewari D, Patni P, Bishayee A, Sah AN, Bishayee A (2022) Natural products targeting the PI3K-Akt-mTOR signaling pathway in cancer: a novel therapeutic strategy. Semin Cancer Biol 80:1–17 [DOI] [PubMed] [Google Scholar]
- 14.Liu R, Chen Y, Liu G, Li C, Song Y, Cao Z, Li W, Hu J, Lu C, Liu Y (2020) PI3K/AKT pathway as a key link modulates the multidrug resistance of cancers. Cell Death Dis 11(9):797 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Malek M, Kielkowska A, Chessa T, Anderson KE, Barneda D, Pir P, Nakanishi H, Eguchi S, Koizumi A, Sasaki J et al (2017) PTEN regulates PI(3,4)P(2) signaling downstream of class I PI3K. Mol Cell 68(3):566-580 e510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Falasca M, Maffucci T (2014) Targeting p110gamma in gastrointestinal cancers: attack on multiple fronts. Front Physiol 5:391 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Rathinaswamy MK, Burke JE (2020) Class I phosphoinositide 3-kinase (PI3K) regulatory subunits and their roles in signaling and disease. Adv Biol Regul 75:100657 [DOI] [PubMed] [Google Scholar]
- 18.Sanchez-Castillo C, Cuartero MI, Fernandez-Rodrigo A, Briz V, Lopez-Garcia S, Jimenez-Sanchez R, Lopez JA, Graupera M, Esteban JA (2022) Functional specialization of different PI3K isoforms for the control of neuronal architecture, synaptic plasticity, and cognition. Sci Adv 8(47):eabq8109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Okkenhaug K (2013) Rules of engagement: distinct functions for the four class I PI3K catalytic isoforms in immunity. Ann N Y Acad Sci 1280:24–26 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Danyaei A, Ghanbarnasab-Behbahani R, Teimoori A, Neisi N, Chegeni N (2024) The simultaneous use of CRISPR/Cas9 to knock out the PI3Kca gene with radiation to enhance radiosensitivity and inhibit tumor growth in breast cancer. Iran J Basic Med Sci 27(12):1566–1573 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhao WJ, Zhu LL, Yang WQ, Xu SJ, Chen J, Ding XF, Liang Y, Chen G (2021) LPAR5 promotes thyroid carcinoma cell proliferation and migration by activating class IA PI3K catalytic subunit p110beta. Cancer Sci 112(4):1624–1632 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Pridham KJ, Le L, Guo S, Varghese RT, Algino S, Liang Y, Fajardin R, Rodgers CM, Simonds GR, Kelly DF et al (2018) PIK3CB/p110beta is a selective survival factor for glioblastoma. Neuro Oncol 20(4):494–505 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Skanland SS, Okkenhaug K, Davids MS (2024) PI3K inhibitors in hematology: when one door closes. Clin Cancer Res 30(17):3667–3675 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yang Q, Zhang C, Ren Y, Yi H, Luo T, Xing F, Bai X, Cui L, Zhu L, Ouyang J et al (2020) Genomic characterization of Chinese ovarian clear cell carcinoma identifies driver genes by whole exome sequencing. Neoplasia 22(9):399–430 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Driva TS, Schatz C, Haybaeck J (2023) Endometriosis-associated ovarian carcinomas: how PI3K/AKT/mTOR pathway affects their pathogenesis. Biomolecules. 10.3390/biom13081253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Iida Y, Aoki K, Asakura T, Ueda K, Yanaihara N, Takakura S, Yamada K, Okamoto A, Tanaka T, Ohkawa K (2012) Hypoxia promotes glycogen synthesis and accumulation in human ovarian clear cell carcinoma. Int J Oncol 40(6):2122–2130 [DOI] [PubMed] [Google Scholar]
- 27.Carvalho S, Milanezi F, Costa JL, Amendoeira I, Schmitt F (2010) PIKing the right isoform: the emergent role of the p110beta subunit in breast cancer. Virchows Arch 456(3):235–243 [DOI] [PubMed] [Google Scholar]
- 28.Trigka EA, Levidou G, Saetta AA, Chatziandreou I, Tomos P, Thalassinos N, Anastasiou N, Spartalis E, Kavantzas N, Patsouris E et al (2013) A detailed immunohistochemical analysis of the PI3K/AKT/mTOR pathway in lung cancer: correlation with PIK3CA, AKT1, K-RAS or PTEN mutational status and clinicopathological features. Oncol Rep 30(2):623–636 [DOI] [PubMed] [Google Scholar]
- 29.Cui B, Tao J, Yang Y (2012) Studies on the expression patterns of class I PI3K catalytic subunits and its prognostic significance in colorectal cancer. Cell Biochem Biophys 62(1):47–54 [DOI] [PubMed] [Google Scholar]
- 30.Thibault B, Ramos-Delgado F, Guillermet-Guibert J (2023) Targeting class I-II-III PI3Ks in cancer therapy: recent advances in tumor biology and preclinical research. Cancers (Basel). 10.3390/cancers15030784 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Madsen RR, Toker A (2023) PI3K signaling through a biochemical systems lens. J Biol Chem 299(10):105224 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Cirillo D, Diceglie M, Nazare M (2023) Isoform-selective targeting of PI3K: time to consider new opportunities? Trends Pharmacol Sci 44(9):601–621 [DOI] [PubMed] [Google Scholar]
- 33.Sirico M, D’Angelo A, Gianni C, Casadei C, Merloni F, De Giorgi U (2023) Current state and future challenges for PI3K inhibitors in cancer therapy. Cancers (Basel) 15(3):703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Schwartz S, Wongvipat J, Trigwell CB, Hancox U, Carver BS, Rodrik-Outmezguine V, Will M, Yellen P, de Stanchina E, Baselga J et al (2015) Feedback suppression of PI3Kalpha signaling in PTEN-mutated tumors is relieved by selective inhibition of PI3Kbeta. Cancer Cell 27(1):109–122 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zecchin D, Moore C, Michailidis F, Horswell S, Rana S, Howell M, Downward J (2020) Combined targeting of G protein-coupled receptor and EGF receptor signaling overcomes resistance to PI3K pathway inhibitors in PTEN-null triple negative breast cancer. EMBO Mol Med 12(8):e11987 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yao Z, Gao G, Yang J, Long Y, Wang Z, Hu W, Liu Y (2020) Prognostic role of the activated p-AKT molecule in various hematologic malignancies and solid tumors: a meta-analysis. Front Oncol 10:588200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Cui W, Cai Y, Wang W, Liu Z, Wei P, Bi R, Chen W, Sun M, Zhou X (2014) Frequent copy number variations of PI3K/AKT pathway and aberrant protein expressions of PI3K subunits are associated with inferior survival in diffuse large B cell lymphoma. J Transl Med 12:10 [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.
Supplementary Materials
Data Availability Statement
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.