PET/CT based cross-modal deep learning signature to predict occult nodal metastasis in lung cancer

Abstract

Occult nodal metastasis (ONM) plays a significant role in comprehensive treatments of non-small cell lung cancer (NSCLC). This study aims to develop a deep learning signature based on positron emission tomography/computed tomography to predict ONM of clinical stage N0 NSCLC. An internal cohort (n = 1911) is included to construct the deep learning nodal metastasis signature (DLNMS). Subsequently, an external cohort (n = 355) and a prospective cohort (n = 999) are utilized to fully validate the predictive performances of the DLNMS. Here, we show areas under the receiver operating characteristic curve of the DLNMS for occult N1 prediction are 0.958, 0.879 and 0.914 in the validation set, external cohort and prospective cohort, respectively, and for occult N2 prediction are 0.942, 0.875 and 0.919, respectively, which are significantly better than the single-modal deep learning models, clinical model and physicians. This study demonstrates that the DLNMS harbors the potential to predict ONM of clinical stage N0 NSCLC.

Fig. 1. Flow chart illustrating study design.

PET/CT, positron emission tomography-computed tomography; ROI, region of interest; DLNMS, deep learning nodal metastasis signature; SND, systematic nodal dissection; LND, limited nodal dissection; ROC, receiver operating characteristic curve.

Fig. 2. PET and CT texture features related to the DLNMS.

A Top 10 PET and (B) top 10 CT texture features related to the DLNMS N1 prediction in the training set. C Top 10 PET and (D) top 10 CT texture features related to the DLNMS N2 prediction in the training set. n = 1528 biologically independent samples were examined. Source data are provided as a Source Data file. DLNMS, deep learning nodal metastasis signature; PET, positron emission tomography; CT, computed tomography.

Fig. 3. Predictive performances of the DLNMS for occult nodal metastasis in clinical stage N0 non-small cell lung cancer.

ROC curves and performance metrics of models to predict occult N1 and N2 in the (A, B) validation set, C, D External cohort and (E, F) prospective cohort. ROC curves and performance metrics of the DLNMS to predict occult nodal metastasis in (G) adenocarcinoma and (H) squamous cell carcinoma for patients in validation set, external cohort and prospective cohort. n = 383, 355, and 999 biologically independent samples were examined for the validation set, external cohort, and prospective cohort, respectively. p values from Delong’s tests were adjusted by the Benjamini and Hochberg corrections for 5 multiple comparisons. Source data are provided as a Source Data file. ROC, Receiver operating characteristic curve; DLNMS, deep learning nodal metastasis signature; PPV, positive predictive value; NPV, negative predictive value; PET, positron emission tomography; CT, computed tomography.

Fig. 4. Predictive performances of the DLNMS for occult N2 metastasis diagnosed by nodal biopsy in clinical stage N0 non-small cell lung cancer.

A ROC curves and performance metrics of models to predict occult N2 diagnosed by nodal biopsy. B Scatter graphs illustrating the DLNMS score distributions. C, D Scatter graphs describing the DLNMS correct cases falsely predicted by the PET and CT models. n = 366 biologically independent samples were examined. p values from Delong’s tests were adjusted by the Benjamini and Hochberg corrections for 5 multiple comparisons. Source data are provided as a Source Data file. ROC, Receiver operating characteristic curve; DLNMS, deep learning nodal metastasis signature; PPV, positive predictive value; NPV, negative predictive value; PET, positron emission tomography; CT, computed tomography.

Fig. 5. Prognosis of clinical stage I non-small cell lung cancer treated with different surgical strategies for low-risk and high-risk patients in the validation set and external cohort.

Survival comparisons between (A, B) sublobectomy versus lobectomy and (C, D) LND versus SND in low-risk patients. Survival comparisons between (E, F) sublobectomy versus lobectomy and (G, H) LND versus SND in high-risk patients. n = 1324 biologically independent samples were examined. Survival data were compared by the log-rank test. Source data are provided as a Source Data file. SND, systematic nodal dissection; LND, limited nodal dissection; OS, overall survival; RFS, recurrence-free survival.

Fig. 6. Prognosis of pathological stage I non-small cell lung cancer with adjuvant therapy and without adjuvant therapy for low-risk and high-risk patients in the validation set and external cohort.

Survival comparisons between with adjuvant therapy versus without adjuvant therapy in (A) and (B) low-risk and (C) and (D) high-risk patients. n = 1182 biologically independent samples were examined. Survival data were compared by the log-rank test. Source data are provided as a Source Data file. POAT, postoperative adjuvant therapy; OS, overall survival; RFS, recurrence-free survival.

Fig. 7. Biologic basis of the DLNMS.

A, B Radar charts illustrating histologic patterns between low-score and high-score patients. C, D Bar charts showing frequency of gene alternations between patients with low scores and high scores. E, F Dot plots showing the top 20 upregulated molecular pathways in patients with high scores, p values were adjusted by the Benjamini and Hochberg corrections. G, H Boxplots comparing proportions of infiltrated immune cells between low-score and high-score patients. The centre of box denotes the 50th percentile, the bounds of box contain the 25th to 75th percentiles, the whiskers mark the maximum and minimum values, values beyond these upper and lower whiskers are considered outliers and marked with dots. n = 144 biologically independent samples were examined. Source data are provided as a Source Data file. DLNMS, deep learning nodal metastasis signature; LVI, lymphovascular invasion; VPI, visceral pleural invasion; STAS, tumor spread through air space; NES, normal enrichment score; EGFR, epidermal growth factor receptor; KRAS, kirsten ratsarcoma viral oncogene homolog; BRAF, v-raf murine sarcoma viral oncogene homolog B1; ALK, anaplastic lymphoma kinase; ROS1, c-ros oncogene 1; MDSC, myeloid-derived suppressor cells.