TL;DR
This study investigated the predictive power of immunohistochemical (IHC) markers and radiomic features from CT scans for survival in Small Cell Lung Cancer (SCLC). While synaptophysin expression was linked to better overall survival and TTF-1/GLUT-1 to better progression-free survival, radiomic features showed only moderate correlations and did not independently predict survival.
What Problem Does This Paper Address?
Small Cell Lung Cancer (SCLC) is an aggressive cancer with a poor prognosis, characterized by rapid growth and early metastasis. Identifying reliable biomarkers for predicting treatment response and patient survival is crucial for personalized treatment strategies. While immunohistochemistry (IHC) is a standard diagnostic tool, its prognostic value in SCLC is not fully understood. Radiomics, which extracts quantitative features from medical images, offers a potential complementary approach to capture tumor heterogeneity and phenotype. However, the utility of radiomics in SCLC, particularly in conjunction with IHC, for survival prediction remains largely unexplored. This paper aims to bridge this gap by evaluating both IHC markers and radiomic features for their prognostic significance in SCLC patients.
Methods
The researchers conducted a retrospective analysis of 101 SCLC patients treated with radiotherapy and chemotherapy. They assessed the expression of several IHC markers from tumor biopsies, including neuroendocrine markers (CD56, chromogranin A, synaptophysin), proliferation markers (MIB-1/KI-67), and markers related to tumor microenvironment and metabolism (TTF-1, GLUT-1, HIF-1a, PD-1, PD-L1). For CD44, HIF-1a, and GLUT-1, semi-quantitative H-scores were calculated.
Additionally, a subgroup of 31 patients with homogenous CT scans underwent radiomic analysis. Seventy-two texture features were extracted from CT images using the PyRadiomics package. The study then correlated these IHC markers and radiomic features with overall survival (OS) and progression-free survival (PFS) using Kaplan-Meier analysis and Cox proportional hazards regression models.
Datasets
The study utilized data from 101 consecutive patients with histologically proven SCLC treated at the authors' institution. A subset of 31 patients from this cohort was used for the radiomics analysis, ensuring homogenous CT scan acquisition parameters. No new benchmark datasets were introduced.
Key Insights and Ideas
- Synaptophysin as a Prognostic Indicator: The study highlights synaptophysin, a well-known neuroendocrine marker, as a significant predictor of better overall survival in SCLC. This suggests that its presence might indicate a less aggressive tumor phenotype or a better response to treatment.
- TTF-1 and GLUT-1 in Progression-Free Survival: TTF-1 expression and a lower GLUT-1 H-score were associated with improved progression-free survival, indicating their potential role in predicting how long patients remain disease-free after treatment.
- Radiomics' Limited Predictive Power in SCLC: Despite extracting a comprehensive set of radiomic features, the study found no single texture feature that strongly correlated with OS or PFS in SCLC. This suggests that, in this specific context and with the current radiomic feature set, imaging alone may not be sufficient for robust survival prediction.
- Chromogranin A and PD-1 in Metastasis Risk: The absence of chromogranin A expression was linked to a higher risk of developing cerebral metastases, and PD-1 expression was associated with an increased risk of developing metastases in general. This points to potential roles for these markers in predicting metastatic spread.
Key Results
- Overall Survival (OS): The median OS was 21 months for limited-stage SCLC and 13 months for extensive-stage SCLC. Synaptophysin expression correlated with a better OS (HR 0.546, p=0.03).
- Progression-Free Survival (PFS): TTF-1 expression (HR 0.286, p=0.006) and a lower GLUT-1 H-score (HR 0.511, p=0.05) correlated with a better PFS.
- Metastasis Risk: Patients without chromogranin A expression had a higher risk of cerebral metastases (p=0.02), and PD-1 expression was associated with an increased risk of developing metastases (p=0.02).
- Radiomics: Correlation coefficients for radiomic features with OS ranged from -0.48 to 0.39, and for PFS from -0.46 to 0.38, indicating only moderate associations. No single radiomic feature independently predicted survival.
Limitations and Future Work
The authors acknowledge several limitations, including the retrospective nature of the study, its single-institution setting, a relatively small sample size, and potential selection bias in treatment delivery. They suggest that synaptophysin-negative patients might benefit from treatment intensification, warranting further evaluation. The limited predictive power of radiomics in this study suggests that future research could explore more advanced radiomic approaches, potentially integrating them with other data modalities like genomics or proteomics, or investigating different CT acquisition protocols. Further validation of the prognostic roles of TTF-1, GLUT-1, chromogranin A, and PD-1 in larger, prospective cohorts is also recommended.
Summary updated