A Three-subtype Molecular model of Cervical Cancer: Multiple PI3K Pathway inhibitors suppress growth and cooperate with HPV-directed immunotherapy

Technical Summary: A Three-subtype Molecular Model of Cervical Cancer and PI3K Pathway Inhibition

Problem Statement
Cervical cancer (CC) is a globally prevalent malignancy driven by persistent high-risk Human Papillomavirus (HPV) infections. Despite the well-established viral etiology, CC lacks defined molecular subtypes to guide targeted therapy, and approved molecularly targeted treatments are scarce. While immune checkpoint inhibitors (e.g., anti-PD-1/PD-L1) have shown feasibility, resistance and heterogeneity remain challenges. The study identifies that PIK3CA mutations and YAP1 gene amplification are the most frequent oncogenic alterations in CC, yet their specific roles in defining molecular subtypes and their potential as therapeutic targets, particularly in combination with immunotherapy, require further elucidation.

Methodology
The authors employed a multi-faceted approach combining bioinformatic analysis, in vitro pharmacological testing, and immunological assays:

1. Molecular Subtyping: The researchers analyzed public datasets (TCGA-CESC, AACR-GENIE, MSKCC, Guatemala cohort) and the Caris Life Sciences real-world dataset (n=2,284) to categorize CC based on PIK3CA mutation status and YAP1 amplification.
2. In Vitro Drug Screening: A panel of cervical cancer cell lines (including CaSki, SiHa, ME180, C33A, SNU-17, and others) with varying PIK3CA and YAP1 statuses were treated with three targeted agents:
   • Alpelisib (BYL-719): A selective PI3Kα inhibitor.
   • Inavolisib (GDC-0077): A selective PI3Kα inhibitor.
   • Capivasertib (AZD-5363): A pan-AKT inhibitor.
     Cell proliferation was monitored using impedance-based real-time cell analysis (xCELLigence).
3. Molecular Mechanism Analysis: Western blotting and transcriptome analysis (Long-read RNA sequencing and qPCR) were performed on responsive cell lines (specifically CaSki) to assess changes in protein expression (PD-L1, HPV E7, YAP1, EGFR) and gene expression following drug treatment.
4. Immunotherapy Combination: The study utilized donor-derived, HPV16-specific CD8+ T cells (NEXI-003, targeting E6/E7 and BIRC5 peptides restricted by HLA-A*02:01) in combination with BYL-719. Cytotoxicity was measured via tumor cytolysis assays, including washout experiments to determine the optimal timing of drug and T-cell administration.

Key Contributions and Results

• Definition of Three Molecular Subtypes:
  The study establishes a three-subtype model for cervical cancer:

  • Subtype I: PIK3CA Wild Type (WT) without YAP1 amplification (approx. 65% of cases).
  • Subtype II: PIK3CA mutated without YAP1 amplification (approx. 29% of cases).
  • Subtype III: PIK3CA WT with YAP1 amplification (approx. 5.3% of cases).
  • Finding: PIK3CA mutations and YAP1 amplification are significantly mutually exclusive (p<0.00001). Patients with Subtype III (YAP1-amplified) exhibit significantly poorer overall survival compared to other subtypes, even after adjusting for clinical stage.

• Selective Efficacy of PI3K Inhibitors:

  • Alpelisib (BYL-719) and Inavolisib (GDC-0077): These selective PI3Kα inhibitors significantly suppressed proliferation in PIK3CA-mutated cell lines (IC50 values ranging from 1.3 to 15 μM) but showed no significant effect on PIK3CA WT lines (e.g., SiHa).
  • Capivasertib (AZD-5363): This pan-AKT inhibitor showed broader activity, suppressing some PIK3CA-mutated lines and, surprisingly, the PIK3CA WT SiHa line, suggesting PIK3CA mutation is not the sole determinant of sensitivity for AKT inhibition in all contexts.

• Modulation of Immune Checkpoints and Viral Oncoproteins:
  Treatment with BYL-719 in PIK3CA-mutated cells (CaSki, ME180, SNU-17) resulted in:

  • A 50–70% reduction in HPV16 E6 and E7 mRNA and protein levels.
  • Downregulation of the immune checkpoint molecule CD274 (PD-L1).
  • Reduced expression of YAP1, EGFR, and YAP1 targets (CTGF, Integrin).
  • Transcriptome analysis confirmed the downregulation of genes involved in antigen presentation (B2M, TAP1) and PD-L1 upregulation (STAT1, IRF1), suggesting the drug alters the tumor immune microenvironment.

• Synergy with Immunotherapy:

  • Combining BYL-719 with HPV-specific CD8+ T cells enhanced cytotoxicity against CaSki cells compared to either treatment alone.
  • Critical Timing: While continuous high-dose BYL-719 could inhibit T-cell activity, a "pre-conditioning" strategy yielded the best results. Pretreating tumor cells with BYL-719 (5 μM) for 24 hours, followed by drug washout and subsequent addition of T cells (with or without low-dose BYL-719), resulted in maximal tumor cell killing. This suggests that transient PI3K inhibition sensitizes tumor cells to T-cell mediated lysis without permanently suppressing T-cell function.

Significance and Claims
The paper claims to provide the first molecular classification of cervical cancer into three distinct subtypes based on PIK3CA and YAP1 status, addressing a gap in the field where no such subtypes currently exist for targeted therapy. The authors posit that:

1. Subtype II (PIK3CA-mutated) represents a distinct group (approx. 30% of CC) that is sensitive to FDA-approved PI3Kα inhibitors (Alpelisib, Inavolisib) and may benefit from these agents as monotherapy or in combination.
2. Subtype III (YAP1-amplified) is an aggressive, poor-prognosis subtype requiring different therapeutic strategies, potentially involving EGFR or YAP1 inhibitors.
3. Combination Strategy: PI3K inhibitors can serve as effective partners for HPV-directed immunotherapy. By reducing viral oncoprotein expression (E7) and immune checkpoint levels (PD-L1), PI3K inhibitors may "prime" the tumor for T-cell attack. The study suggests that optimizing the timing of administration (pre-treatment followed by washout) is crucial to avoid T-cell suppression while maximizing tumor cell vulnerability.

The authors conclude that these preclinical findings support the rationale for clinical trials evaluating PI3K inhibitors, alone or in combination with immunotherapy, specifically for patients with PIK3CA-mutated cervical cancer. They acknowledge limitations, including the small number of cell lines tested (particularly for T-cell assays) and the need for in vivo validation, but emphasize the potential for improved therapeutic outcomes through molecularly guided treatment.