Correlation between lung cancer probability and number of pulmonary nodules in baseline computed tomography lung cancer screening: A retrospective study based on the Chinese population

Quanyang Wu, Shijun Zhao, Yao Huang, Jianwei Wang, Wei Tang, Lina Zhou, Linlin Qi, Zewei Zhang, Yuting Xie, Jiaxing Zhang, Hongjia Li, Ning WuFrontiers in Oncology 2023 · DOI

Abstract

Lung cancer (LC) probability does not significantly change with the total number of nodules detected during baseline screening. However, when stratified by nodule size, the effect of nodule number on LC probability becomes size-dependent. Specifically, solitary nodules of intermediate size (6–15 mm) require greater clinical attention and active follow-up compared to multiple nodules of the same size. Assessing nodule number in conjunction with size in baseline low-dose computed tomography (LDCT) screening is considered beneficial for accurate risk assessment.

Introduction

Low-dose computed tomography (LDCT) has revolutionized lung cancer screening, but its high sensitivity presents a clinical dilemma: the frequent detection of multiple small pulmonary nodules. The vast majority of these nodules are benign, often resulting from past infections or inflammation.

Current clinical guidelines primarily focus on the characteristics of the largest or most suspicious "dominant" nodule to determine cancer risk. However, there is ongoing debate regarding whether the total number of nodules should influence a clinician's suspicion. Does a higher nodule count imply a higher risk of malignancy, or does it suggest a systemic, non-cancerous process? This study clarifies this relationship within the Chinese population, where the prevalence of non-smoking-related lung cancer and prior granulomatous diseases (such as tuberculosis) differs from Western cohorts.

Methods

This retrospective study analyzed 10,167 asymptomatic participants who underwent LDCT screening at the National Cancer Center in China between 2010 and 2018. The selection process, including age-based exclusion criteria and data accessibility filters, is detailed in Figure 1.

Flowchart of participant selection and exclusion criteria.

Researchers analyzed noncalcified nodules $\ge$ 4 mm and categorized participants based on their nodule count (1, 2, 3, 4, and >4). Crucially, nodules were stratified into three risk groups based on size and type:

  • Group A (Low Risk): Solid nodules (SNs) < 6 mm or non-solid nodules (NSNs) < 8 mm.
  • Group B (Intermediate Risk): SNs or the solid component of partial-solid nodules (PSNs) measuring 6–15 mm, or NSNs measuring 8–15 mm.
  • Group C (High Risk): SNs/solid components or NSNs $\ge$ 15 mm.

Results and Distribution

Of the 10,167 participants, 3,076 (30.2%) had at least one nodule detected. A total of 5,875 positive nodules were identified, leading to 119 confirmed malignant nodules in 97 participants (an overall cancer rate of 0.95%).

The distribution of nodules among participants is shown in Figure 2. Over half (55.9%) of participants with nodules presented with a solitary finding. As the number of nodules per person increased, the proportion of solid nodules also tended to increase, yet the cancer risk did not follow a simple upward trajectory.

Distribution of participants by the number of pulmonary nodules detected.

Nodule Count and Cancer Probability

On a general level, the total number of nodules did not significantly predict cancer risk ($p > 0.05$). LC probabilities were 2.9% for one nodule, 3.2% for two, 3.4% for three, 3.6% for four, and 5.1% for more than four nodules.

However, a critical "Solitary Nodule Paradox" emerged in Group B (intermediate risk). In this size range (6–15 mm), the probability of lung cancer was significantly higher for solitary nodules than for participants with more than four nodules ($p < 0.05$). This suggests that "multinodularity" in this size range is often a marker for benign inflammatory or infectious processes. For very small (Group A) or very large (Group C) nodules, the nodule count provided little additional predictive value.

Discussion and Clinical Utility

The study suggests that nodule count should not be viewed in isolation but rather as a "modifier" for intermediate-sized findings. A solitary 10 mm nodule requires more aggressive follow-up than one of five 10 mm nodules. Among the 119 malignant nodules identified, 41.2% were non-solid, 31.1% were part-solid, and 27.7% were solid, reinforcing that nodule type and size remain the primary drivers of risk.

Limitations and Future Work

The authors acknowledge several limitations, including the retrospective nature of the study and the lack of smoking status for all participants, which prevented a more nuanced multivariate analysis. Furthermore, as a single-center study in China, the results may reflect regional environmental exposures and disease prevalence that differ from other global populations. Future risk-prediction models, including AI-based tools, should integrate nodule count as a weighted variable, particularly for intermediate-sized findings, rather than treating each nodule as an independent risk event.

Summary updated

Abstract

LC probability does not significantly change with the number of nodules. However, when stratified by the nodule size, the effect of nodule number on LC probability was nodule-size dependent, and greater attention and active follow-up are required for solitary nodules especially SNs/solid component of PSNs measuring 6-15 mm or NSNs measuring 8-15 mm. Assessing the nodule number in conjunction with nodule size in baseline LDCT LC screening is considered beneficial.
BibTeX
@inproceedings{wu2023correlation,
  author = {Quanyang Wu and Shijun Zhao and Yao Huang and Jianwei Wang and Wei Tang and Lina Zhou and Linlin Qi and Zewei Zhang and Yuting Xie and Jiaxing Zhang and Hongjia Li and Ning Wu},
  title = {{Correlation between lung cancer probability and number of pulmonary nodules in baseline computed tomography lung cancer screening: A retrospective study based on the Chinese population}},
  booktitle = {Frontiers in Oncology},
  year = {2023},
  doi = {10.3389/fonc.2022.1061242},
}
Figure