Prognostic and predictive value of radiomic signature in stage I lung adenocarcinomas following complete lobectomy

Machine Learning Based on Clinical Information and Integrated CT Radiomics to Predict Local Recurrence of Stage Ia Lung Adenocarcinoma after Microwave Ablation

PURPOSE

To develop and compare 3 different machine learning-based models of clinical information and integrated radiomics features predicting the local recurrence of Stage Ia lung adenocarcinoma after microwave ablation (MWA) for assisting clinical decision making.

MATERIALS AND METHODS

The data of 360 patients with Stage Ia lung adenocarcinoma who underwent MWA were included in the training (n = 208), internal test (n = 90), and external test (n = 62) sets based on the inclusion and exclusion criteria. The predictors associated with local recurrence were identified using univariate and multivariate analyses of clinical information. The integrated radiomics features were extracted from pre-MWA and post-MWA (scanned immediately after the ablation) computed tomography (CT) images, and 10 radiomics features were selected by the t-test and least absolute shrinkage and selection operator. The L2-logistic regression of machine learning was applied for the clinical model, CT radiomics model, and combined model including clinical predictors and radiomics features. Model performance was evaluated by the receiver operating characteristic and decision curve analysis.

RESULTS

The ablative margin was an independent clinical predictor (P = 0.001; odds ratio [OR], 0.46; 95% CI, 0.29-0.73). The combined model showed the highest area under the curve value among the 3 models (training, 0.86; 95% CI, 0.81-0.91; internal test, 0.93; 95% CI, 0.87-0.98; external test, 0.89; 95% CI, 0.79-0.96).

CONCLUSIONS

The combined model could accurately predict the local recurrence of Stage Ia lung adenocarcinoma after MWA to better support a clinical decision.

Radiomic score is prognostic in clinical stage I lung adenocarcinoma ≤2 cm undergoing surgery

Background

As sub-lobar resection becomes acceptable for lung cancer ≤2 cm, a preoperative marker of tumor aggressiveness to choose an appropriate extent of resection becomes necessary. We sought to assess the utility of Computer-Aided Nodule Assessment and Risk Yield (CANARY), a validated radiomic tool, in clinical stage I adenocarcinoma ≤2 cm.

Methods

We performed a retrospective review of resected lung cancer patients from 2016-2022. Our eligibility criteria included clinical stage I adenocarcinoma, availability of pre-operative computed tomography (CT) imaging, and a lesion size of ≤2 cm. Preoperative imaging was input into the CANARY program, and this was then used to categorize each lesion into good, intermediate, and poor. Kaplan-Meier curve was used to compare the recurrence-free survival (RFS). Descriptive statistics and log-rank tests were conducted to compare RFS between risk groups.

Results

Study population (n=134) had a median age of 68.6 and follow up of 2.9 years. By CANARY profile, 29 patients (21.6%) were good risk, 52 (38.8%) intermediate, and 53 (39.6%) poor. By procedure, 52 patients (38.8%) received wedge resections. Overall, the 3-year RFS was 96.3%, 92.0%, and 72.7% for patients with good, intermediate, and poor risks, respectively. There was a statistically significant difference in RFS between each risk group (χ2=12.6, P=0.002). Patients with poor risk were associated with a significantly increased risk of recurrence relative to those with good/intermediate risks [hazard ratio (HR) =5.7, 95% confidence interval (CI): 1.9-17.5].

Conclusions

Poor risk on CANARY analysis is significantly associated with increased risk of recurrence after resection in clinical stage I adenocarcinoma lesions ≤2 cm.

Nomogram for predicting prognosis and identifying chemotherapy beneficiaries for completely resected stage I invasive mucinous lung adenocarcinoma

Background

At present, there is a lack of studies in invasive mucinous adenocarcinoma (IMA) that combine clinicopathological and imaging features to stratify risk and select optimal treatment regimen. We aimed to develop and validate a nomogram for predicting recurrence-free survival (RFS) and identifying adjuvant chemotherapy (ACT) beneficiaries for completely resected stage I primary IMA.

Methods

This retrospective study included 750 patients from three hospitals. Patients from two hospitals were divided into training (n=424) and validating cohort (n=185), and patients from the remaining other one hospital constituted external test cohort (n=141) and preoperative computed tomography (CT) image features of each patient were consecutively evaluated. The nomogram was developed by integrating significant prognostic factors of RFS identified in the multivariate analysis. The risk score (RS) based on nomogram was calculated in the entire cohort and the optimal cut-off point for risk stratification was obtained by X-tile software. The Kaplan-Meier method, log-rank test and interaction were used to evaluate the difference in RFS and overall survival (OS) between different risk and treatment groups.

Results

Visceral pleural invasion (VPI, P<0.001), lymph-vascular invasion (LVI, P<0.001), tumor size (P<0.001), smoking history (P<0.001), lobulation (P<0.001) were identified as independent prognostic factors for RFS. The concordance index (C-index) of the nomogram was higher than that of tumor-node-metastasis (TNM) staging system (validation cohort: 0.73±0.09 vs. 0.62±0.08, P<0.001; external test cohort: 0.74±0.10 vs. 0.70±0.09, P=0.035). The patients with higher RS were associated with worse RFS [hazard ratios (HRs) ≥4.76] and OS (HRs ≥2.55) in all included cohorts. Chemotherapy benefits in terms of RFS and OS were observed for patients in higher RS group in both stage IB (interaction P=0.012 for RFS and P=0.037 for OS) and stage I IMA (interaction P<0.001 for both RFS and OS).

Conclusions

The nomogram based on CT image and clinicopathologic features showed superior performance in predicting RFS for stage I IMA and might identify ACT candidates for personalized patient treatment.

Preoperative fluorescence in situ hybridization analysis as a predictor of tumor recurrence in patients with non-muscle invasive bladder cancer: a bi-institutional study

BACKGROUND

Non-muscle invasive bladder cancer (NMIBC) is known for its elevated recurrence rate, necessitating an enhancement in the current risk stratification for recurrence. The urine-based fluorescence in situ hybridization (FISH) assay has emerged as a noninvasive auxiliary tool for detecting bladder cancer. The aim of this study was to explore the potential relationship between the preoperative FISH assay and recurrence, and to develop a FISH-clinical nomogram for predicting the recurrence-free survival (RFS) in NMIBC patients.

METHODS

In total, 332 eligible patients were enrolled from two hospitals. The SYSMH cohort was randomly assigned to the training set (n = 168) and the validation set I (n = 72) at a ratio of 7:3, while the SYSUTH cohort was allocated to the validation set II (n = 92). The correlation between the preoperative FISH assay and recurrence was determined through the Cox regression analysis. The least absolute shrinkage and selection operator (LASSO) Cox regression algorithm was used for model construction. The performance of the model was assessed by its discrimination, calibration, and clinical usefulness.

RESULTS

We uncovered that chromosome 7 aneuploidy, p16 locus loss, number of the positive FISH sites, and the FISH test result were significantly associated with tumor recurrence. Then, a FISH-clinical nomogram incorporating the FISH test result, T stage, associated CIS, tumor grade, and tumor status was developed. It showed favorable calibration and discrimination with a C-index of 0.683 (95%CI, 0.611-0.756) in the training set, which was confirmed in the validation set I and validation set II with C-indexes of 0.665 (95%CI, 0.565-0.765) and 0.778 (95%CI, 0.665-0.891), respectively. Decision curve analysis revealed the clinical usefulness of the nomogram. Moreover, our proposed nomogram significantly outperformed the guideline-recommended EORTC and CUETO scoring models.

CONCLUSION

Our study confirmed the prognostic value of the preoperative FISH assay and proposed a FISH-clinical nomogram to predict RFS in NMIBC patients. Our nomogram can serve as a more precise tool for recurrence risk stratification, which may optimize disease management in bladder cancer and improve patient prognosis.