The Biomarker Hunt

Predicting Who Will Respond to Lenvatinib in Endometrial Cancer

The key to personalized cancer therapy might lie in our own blood and tissue.

Imagine a future where, before starting a powerful cancer treatment, a doctor can know with high certainty whether it will work for you. This is the promise of predictive biomarkers—biological clues that can forecast treatment response. For patients with advanced endometrial cancer, the combination therapy of lenvatinib and pembrolizumab has emerged as a significant treatment option. However, with only a subset of patients experiencing long-term benefits, the race is on to find reliable biomarkers that can guide treatment decisions and maximize outcomes while minimizing unnecessary side effects.

The Treatment and The Challenge

Endometrial cancer, which originates in the lining of the uterus, is the most common gynecological malignancy in developed nations. While often detected early, advanced or recurrent disease has a much poorer prognosis, with five-year survival rates dropping below 20% 5 7 .

The combination of lenvatinib, a multi-kinase inhibitor that blocks tumor growth signals and blood vessels, and pembrolizumab, an immunotherapy that reactivates the immune system against cancer, has shown meaningful clinical benefits. In a pivotal clinical trial, this combination demonstrated a 38% objective response rate in previously treated patients 9 .

However, this also means that a majority of patients did not respond as favorably, underscoring a critical problem: without predictive biomarkers, clinicians cannot reliably identify which patients will benefit. This diagnostic challenge fuels extensive research into biomarkers that can serve as a treatment compass 7 .

Treatment Response

In clinical trials, only 38% of patients responded to lenvatinib-pembrolizumab combination therapy.

The Contenders: Promising Biomarker Candidates

Researchers are exploring biomarkers from various sources, including peripheral blood and the tumor microenvironment itself. Two categories show particular promise: systemic inflammatory markers and specific immune cell populations.

The Blood Test: Neutrophil-to-Lymphocyte Ratio (NLR)

One of the most accessible biomarkers comes from a simple complete blood count. The neutrophil-to-lymphocyte ratio (NLR) measures the balance between different types of white blood cells and reflects the body's systemic inflammatory state.

A 2025 study investigated NLR in 25 patients with advanced endometrial cancer receiving lenvatinib/pembrolizumab. The findings were striking:

  • The most accurate predictor of progression-free status was the NLR score in peripheral blood measured before starting lenvatinib/pembrolizumab treatment.
  • Patients with a lower NLR (below 5.39) had significantly longer progression-free survival compared to those with a higher NLR—13.5 months versus 3.0 months 1 .

This simple blood marker potentially offers an effective, non-invasive way to stratify patients before beginning treatment.

NLR as a Predictor for Lenvatinib/Pembrolizumab Response
NLR Group Progression-Free Survival Statistical Significance
NLR < 5.39 13.5 months (median) p = 0.023
NLR ≥ 5.39 3.0 months (median)
NLR Impact on Progression-Free Survival

The Tumor Microenvironment: CD20+ B Cells and CD8/CD20 Ratio

While blood-based biomarkers are convenient, the tumor microenvironment—the ecosystem surrounding cancer cells—holds crucial information. A 2024 study analyzed tumor samples from 28 patients with advanced endometrial cancer before they started lenvatinib/pembrolizumab treatment, focusing on specific immune cells 2 .

The research revealed that the composition of immune cells within tumors, particularly B lymphocytes, strongly correlates with treatment response:

  • Patients who responded to treatment had a significantly higher proportion of CD20+ B lymphocytes in their tumor stroma compared to non-responders.
  • The ratio of CD8+ T cells to CD20+ B cells was also a powerful differentiator, with responders showing a lower ratio 2 .

These findings highlight that the immune contexture of the tumor, especially the presence of B cells, plays a critical role in determining the success of immunotargeted therapy.

Tumor Microenvironment Biomarkers
Biomarker Responders Non-Responders Predictive Power
CD20+ B lymphocytes 0.24% 0.08% Sensitivity: 85.71%, Specificity: 70.59%
CD8/CD20 lymphocyte ratio 1.44 arbitrary units 19.00 arbitrary units Sensitivity: 85.71%, Specificity: 85.71%
CD20+ B Cell Comparison
CD8/CD20 Ratio Comparison

A Deeper Dive: The KEYNOTE-146 Biomarker Analysis

While the aforementioned studies identified positive biomarkers, a crucial 2024 analysis from the KEYNOTE-146 clinical trial asked a different question: Do established immunotherapy biomarkers predict response to the lenvatinib-pembrolizumab combination? This experiment represents one of the most comprehensive biomarker studies for this therapy to date 9 .

The Methodology

  • Patient Population: The study included 93 and 79 patients for RNA-sequencing and whole-exome sequencing analyses, respectively. All had previously treated advanced endometrial cancer.
  • Treatment Regimen: Patients received oral lenvatinib (20 mg daily) plus intravenous pembrolizumab (200 mg every 3 weeks).
  • Biomarker Assessment: Researchers analyzed tumor samples using multiple advanced techniques:
    • RNA-sequencing to evaluate 12 different gene expression signatures, including the T-cell-inflamed gene expression profile (GEP).
    • Whole-exome sequencing to assess tumor mutational burden (TMB) and mutations in key genes like PIK3CA, PTEN, and TP53.
  • Statistical Analysis: The team examined associations between these biomarker scores and clinical outcomes, primarily objective response rate and progression-free survival.

The Surprising Results

Contrary to expectations, this rigorous analysis found that none of the biomarkers showed a statistically significant association with treatment response:

  • The T-cell-inflamed GEP, which predicts response to pembrolizumab alone, was not associated with better outcomes for the combination therapy.
  • Tumor mutational burden (TMB), another established biomarker for immunotherapy, did not correlate with response.
  • Mutational status of individual genes (PIK3CA, PTEN, TP53) did not predict benefit.
  • Responses were observed regardless of the biomarker status 9 .
Biomarkers That Did NOT Predict Response in KEYNOTE-146
Biomarker Category Specific Examples Finding
Gene Expression Profiles T-cell-inflamed GEP, Angiogenesis, Hypoxia No significant association with ORR or PFS
Tumor Mutational Burden TMB (≥175 or <175 mutations/exome) Response seen across all TMB levels
Specific Gene Mutations PIK3CA, PTEN, TP53 No correlation with treatment outcome

The Significance

These negative findings are scientifically crucial. They suggest that the combination of lenvatinib and pembrolizumab may work through broader mechanisms than pembrolizumab alone, potentially overcoming the immunological limitations of "cold" tumors that would not respond to immunotherapy alone.

The authors concluded that the established biomarkers for immunotherapy lack clinical utility for this specific combination, highlighting the need to discover entirely new predictive models 9 .

The Scientist's Toolkit: Essential Research Reagents

To conduct the critical research highlighted above, scientists rely on specialized reagents and tools. The following table details key components of the biomarker researcher's toolkit.

Essential Research Reagents for Biomarker Discovery
Reagent/Tool Primary Function Application in Research
Multiplex Immunofluorescence Labels multiple protein markers on a single tissue section Simultaneous visualization of CD8+ T-cells, CD20+ B-cells, FoxP3+ T-regs in tumor microenvironment 2
RNA-sequencing Measures gene expression levels across the entire genome Analysis of gene signatures like T-cell-inflamed GEP or angiogenesis pathways 9
Whole-Exome Sequencing Identifies mutations across all protein-coding genes Evaluation of tumor mutational burden and specific mutations in genes like PIK3CA and PTEN 9
CustomMAP Immunoassay Panels Quantifies multiple protein biomarkers in serum Measurement of cytokine and angiogenic factor levels in blood samples
Flow Cytometry Analyzes physical and chemical characteristics of cells Enumeration and characterization of lymphocyte populations in peripheral blood 1
Research Method Applications
Biomarker Discovery Timeline
NLR Discovery (2025)

Blood-based biomarker showing predictive value for lenvatinib response

Tumor Microenvironment (2024)

CD20+ B cells and CD8/CD20 ratio identified as predictors

KEYNOTE-146 Analysis (2024)

Established immunotherapy biomarkers found ineffective for combination therapy

Future Research

Composite scores, spatial transcriptomics, longitudinal monitoring

The Future of Biomarker Research

The search for reliable biomarkers continues to evolve. While simple measures like NLR show promise and complex gene signatures have so far disappointed, research is exploring other avenues:

Composite Biomarker Scores

Combining multiple biomarkers into a single score may improve predictive power. This approach has shown promise in renal cell carcinoma treated with lenvatinib-everolimus .

Spatial Transcriptomics

This emerging technology allows scientists to see which genes are being turned on and where exactly they're active within the tumor microenvironment, providing unprecedented detail.

Longitudinal Monitoring

Tracking biomarker changes during treatment rather than relying solely on baseline measurements may provide dynamic indicators of response.

AI-Powered Analysis

Machine learning algorithms can integrate diverse data types to identify complex patterns that may predict treatment response more accurately than individual biomarkers.

The journey to precision medicine for endometrial cancer patients continues, with each study bringing us closer to the goal of prescribing the right treatment, to the right patient, at the right time.

Clinical Implications

The discovery of reliable biomarkers for lenvatinib response could:

  • Improve patient outcomes by targeting therapy to those most likely to benefit
  • Reduce unnecessary side effects for patients unlikely to respond
  • Lower healthcare costs by avoiding ineffective treatments
  • Accelerate drug development through better patient stratification in clinical trials
Research Progress
NLR Validation
75%
TME Biomarkers
60%
Composite Scores
40%
Clinical Implementation
25%

References