Introduction

Lung cancer represents a significant global public health concern. The improvement of diagnostic techniques and surgical procedures has led to a significant extension in the survival of non-small cell lung cancer patients. Nevertheless, lung cancer continues to represent the primary cause of mortality in China and developed countries1,2. Adenocarcinoma represents the predominant histological subtype of lung cancer, accounting for approximately 50% of all cases. In clinical practice, the diagnosis of lung adenocarcinoma typically relies on pathological puncture biopsy. However, puncture biopsy is an invasive procedure that can cause certain harm to patients, and a few patients may experience post-procedure complications3. Additionally, puncture biopsy involves small sample collection, which has limitations in providing a comprehensive evaluation of larger lung lesions, thus exhibiting certain constraints4. Furthermore, as research on positron emission tomography/computed tomography (PET/CT) continues to deepen, it shows significant potential in distinguishing between primary and metastatic lung lesions, as well as in determining different pathological subtypes of lung cancer5,6. The nuclear protein Ki-67 is present in all proliferating animal cells. Accumulation of Ki-67 occurs exclusively during the S, G2, and M phases of cell proliferation, with continued degradation occurring during the G1 and G0 phases. Ki-67 is a hierarchical marker of the cell cycle and does not participate in the cellular proliferation process7,8. However, the removal of Ki-67 has the effect of deregulating cancer-related pathways, which in turn impedes tumour growth. Conversely, elevated levels of Ki-67 expression have been linked to the promotion of cellular transformation, which in turn contributes to tumour growth and an elevated risk of distant metastasis9. It can be concluded that Ki-67 plays a pivotal role in the process of tumour formation and advancement, and serves as a valuable marker of the extent of cellular proliferation10.

Low-dose computed tomography (CT) has emerged as a powerful tool for routine lung cancer screening11,12. In recent years, the development and clinical application of 2-deoxy-2-fluorine-18-fluoro-D-glucose positron emission tomography/computed tomography (18F-FDG PET/CT) have been of great significance in the assessment of tumour proliferation. The maximum standardised uptake value (SUVmax) is a semi-quantitative metabolic index commonly used in PET/CT, which can accurately reflect the distribution of glucose metabolism in tumour tissues, and thus visually display the most active part of the tumour. It has been demonstrated that the more malignant a tumour is, the faster it proliferates, and that its glucose metabolism is significantly higher than that of normal tissue, with a significant correlation between the two13. A number of studies have demonstrated that measuring the standardised uptake value (SUV) of tumour tissue can effectively reflect the degree of proliferation, which can be used to characterise a range of tumours, including breast cancer, oesophageal cancer, non-small cell lung cancer, lymphoma, and others. This approach can also be employed to predict the prognosis of patients14,15,16,17,18.

Despite its potential, the application of PET/CT-derived parameters in reflecting the degree of lung tumor proliferation remains limited. Furthermore, the literature on the responsiveness of PET/CT-related parameters, including SUVmax, mean standardised uptake value (SUVmean), tumour metabolic volume (MTV), and total lesion glycolysis (TLG), to tumor lesions has not reached a definitive conclusion14,19,20,21. The present study sought to evaluate the correlation between PET/CT-related parameters, specifically SUVmax, SUVmean, MTV, TLG, and maximum tumour mediastinal background ratio (SUVtmr), and Ki-67 expression levels in lung adenocarcinoma patients. Additionally, the study aimed to explore the limitations of PET/CT-related parameters in this context.

Materials and methods

This retrospective study was approved by the Institutional Review Board of Rizhao People’s Hospital and was conducted in accordance with the relevant guidelines and regulations(NO.2025- Ethical Opinions-MR-22-01). This study follows the principles of the Declaration of Helsinki and its subsequent amendments. Due to the retrospective nature of the study, the Institutional Review Board of Rizhao People’s Hospitalwaived the need of obtaining informed consent.

Patients

A retrospective analysis was conducted on patients with lung adenocarcinoma who underwent 18F-FDG PET/CT scanning at Rizhao People’s Hospital between 1 January 2019 and 1 June 2024. In order to be eligible for enrolment, patients were required to meet the following criteria: (1) all patients were initially identified as having a lung-occupying lesion; (2) all patients underwent 18F-FDG scanning prior to surgical intervention; (3) all patients had not undergone any prior intervention on the tumour; and (4) all patients fulfilled the indications for direct surgical treatment, with no distant metastases or multiple intrapulmonary metastases present; (5) all patients were free from other significant comorbidities and neoplastic disease; (6) All patients underwent a postoperative pathological examination.

18F-FDG PET/CT imaging

The production and synthesis of 18F-FDG were conducted by Unity, and the administered dose was calculated to be 3.7 MBq/kg. All patients were scanned on a GE PET/CT Discovery 710 scanner in our hospital, under identical scanning conditions. Before the PET/CT scan, patients were required to meet the following criteria: (1) All patients were required to fast for at least eight hours; (2) Their blood glucose level must be below 8 mmol/L before tracer injection; (3) Patients’ body temperature should remain at a normal level, and strenuous exercise was prohibited within 24 h prior to the PET/CT scan. The scanning range extends from the vertex of the skull to the mid-femur. Some scan data included additional scanning ranges based on the patients’ clinical manifestations and other diagnostic considerations. The CT scans were performed using a 64-slice spiral CT integrated into the machine, with a tube voltage of 120 KV and a scan slice thickness of 3.75 mm. Immediately after the CT scan, a PET scan was conducted, with a scanning time of 2 min per bed position. The PET images were attenuation-corrected using CT data and reconstructed using the TrueX + TOF method following image acquisition.

Imaging analysis

The PET/CT images were analysed using medical image processing software from Shanghai Xingxiang Medical Devices Co Ltd (Shanghai, China). All data acquisition was conducted by two senior nuclear medicine physicians, each of whom had accumulated over three years of experience in PET/CT diagnostics. In the coronal, sagittal, and axial positions, the largest FDG uptake layer of the tumor lesion was selected. Using the post-processing plug-in in the software, a circular region of interest (ROI) that completely encompassed the lesion was manually drawn. Afterward, the software automatically delineated the volume of interest (VOI) of the tumor tissue and analyzed the SUVmax, SUVmean, and MTV of the tumor lesion. TLG was calculated as the product of MTV and SUVmean. The software’s automatic drawing results from the three different observation levels were carefully checked to ensure that the VOI accurately encompassed the lung lesion. The average value of each data point was calculated to reduce errors.

Statistical analysis

The initial data collection and counting of all relevant patient information was conducted using Excel 2021. This data was then imported into the SPSS 25.0 software, where it was used to establish a database for the statistical analysis of the experimental results. For two-by-two comparisons between the three groups, the qualitative data were tested by ², and then α-separated for two-by-two comparisons of multiple subgroups. Three simultaneous comparisons were made, with P ≤ 0.017 indicating that they were statistically different. Quantitative data should initially be tested for normality using the Shapiro-Wilk (SW) test, followed by a chi-square test. If the data meet the normality criteria, an F-test can then be applied. However, if the data do not meet the normality requirements, two-by-two comparisons between multiple groups can be made using the Bonferroni method. Results were expressed using the median M (P25, P75), with the P value corrected using Bonferroni. P ≤ 0.05 indicates that a statistical difference has been identified.

In order to make comparisons between two groups, it is first necessary to test the quantitative data for normality. If the data are normally distributed, a t-test can be used to analyse them; if the data are not normally distributed, a non-parametric Wilcoxon’s rank-sum test should be employed. In both cases, the results are expressed as mean ± standard deviation (̅±S). Finally, if the P<0.05, it can be concluded that there is a statistically significant difference between the two groups. The receiver operating characteristic curve (ROC) was plotted using the statistical software package SPSS.

Results

Clinical characteristics of patients and PET/CT-related parameters between the three groups

A total of 50 patients with lung adenocarcinoma were included in this study, provided that the requisite screening conditions were available. The cohort comprised 30 female and 20 male patients, aged 37–80 years, with a mean age of 64.27 ± 7.97 years. The patients were divided into three groups according to the Ki-67 level as determined by postoperative pathology. The patients were divided into three groups according to the Ki-67 level shown by postoperative pathology: Ki-67 (+) (Ki-67 ≤ 10%), Ki-67 (++) (10% < Ki-67 ≤ 50%), and Ki-67 (+++) (Ki-67 > 50%). Among the three groups, the expression level of Ki-67 exhibited a skewed distribution, with specific results as follows: Ki-67(+) group: 5.0% (3%, 8.5%); Ki-67(++) group: 25% (15%, 40%); Ki-67(+++) group: 70% (60%, 70%). The gender of the patients was analysed using the chi-squared test. The PET/CT parameters SUVmax, SUVmean, MTV, TLG, and SUVtmr were tested for normality using the Shapiro-Wilk test, and all were found to be skewed. Two-by-two comparisons between multiple groups were performed using the Bonferroni method. The statistical results are presented in Table 1.

Table 1 Comparison of patient general characteristics and PET/CT related parameters among the three groups
Full size table

The above statistical results indicated that there were no statistical differences in age, MTV, and TLG between the three groups of data. Between the three groups of data, there was a statistical difference in gender (χ2 = 9.85, P = 0.007 ≤ 0.017), a statistical difference between the Ki-67(++) group and the Ki-67(+++) group (P < 0.017), and no statistical difference between Ki-67(+) and Ki-67(++), and Ki-67(+++) ( P ≥ 0.017). Between the three groups, SUVmax, SUVmean, and SUVtmr were statistically different (P = 0.001). There was no statistical difference in SUVmax, SUVmean, and SUVtmr between Ki-67 (++) and Ki-67 (+++) (P > 0.05). There was a statistical difference in SUVmax, SUVmean, and SUVtmr between Ki-67 (++) and Ki-67 (+++), and vs. Ki-67 (+) (P ≤ 0.05).

Diagnostic performance evaluation of PET/CT-related parameters

\Given the absence of a statistically significant difference in PET/CT-related parameters (SUVmax, SUVmean, SUVtmr) between the Ki-67 (++) and Ki-67 (+++) groups, Subsequently, the patients were regrouped according to the level of Ki-67 expression, with those exhibiting a value of ≤ 10% classified as the low-level Ki-67 group and those with a value of > 10% as the high-level Ki-67 group. The statistical results are presented in Table 2. The diagnostic performance of PET/CT-related parameters, namely SUVmax, SUVmean, and SUVtmr, was evaluated using receiver operating characteristic (ROC) curves. As illustrated in Fig. 1; Table 3,

Table 2 Comparison of PET/CT related parameters.
Full size table
Fig. 1
figure 1

The ROC curves of SUVmax, SUVmean, SUVtmr.

Full size image

The above statistics indicated that the median SUVmax, SUVmean, and SUVtmr were 2.40 (1.68–4.53), 1.24 (0.84–2.70), and 1.48 (0.94–3.42), respectively, in the low level Ki-67 group (Ki-67 ≤ 10%), and SUVmax in the high level Ki-67 group (Ki-67 > 10%), SUVmean, and median SUVtmr were 7.26 (5.01–9.63), 3.63 (2.78–5.76), and 4.47 (2.83–6.18), respectively. There was a statistical difference in the distribution of SUVmax, SUVmean, and SUVtmr between the two groups (P < 0.001).

Table 3 Independent ROC curve analysis of SUVmax, SUVmean, and SUVtmr.
Full size table

The receiver operating characteristic (ROC) curves demonstrated that AUC for the SUVmax(A), SUVmean(B), and SUVtmr(C) were 0.865, 0.861, and 0.847, respectively. Among these, SUVmax exhibited the greatest capacity to discriminate between lung adenocarcinoma with a low Ki-67 level and lung adenocarcinoma with a high Ki-67 level. The sensitivity, specificity, and Youden’s index were 0.844, 0.778, and 0.622, respectively.

Discussion

The present study revealed the correlation between Ki-67 expression levels in lung adenocarcinomas and patients’ clinical characteristics and PET/CT-related parameters through statistical analyses in different subgroups. The findings indicated that there was no statistically significant discrepancy in the age of patients with lung adenocarcinomas exhibiting disparate Ki-67 expression levels. It has been demonstrated that male sex and advanced age are independent risk factors for the development of lung nodules into malignancy22 and for the proliferation of lung adenocarcinoma23. In our study, the proportion of males was significantly higher than that of females in the Ki-67(+++) group (Ki-67 > 50%), in contrast to the findings observed in the Ki-67(+) and Ki-67(++) groups.

MTV and TLG are functional parameters related to PET/CT functionality. It has been demonstrated that MTV and TLG can accurately predict the treatment prognosis24,25,26 and overall survival of lung adenocarcinoma, particularly in patients who have not undergone surgical treatment27. Furthermore, there are notable distinctions in MTV and TLG between lung adenocarcinomas with varying degrees of proliferation21,28. Nevertheless, our study revealed no significant correlation between MTV and TLG and the Ki-67 expression level in lung adenocarcinoma. MTV and TLG demonstrated limited capacity to differentiate lung adenocarcinomas with disparate Ki-67 expression levels. It is postulated that this may be attributed to the considerable inter-patient variability in the proliferative activity of lung adenocarcinoma observed in the enrolled cohort. Nevertheless, it is clear that MTV and TLG are valuable for assessing the relevant indicators of lung adenocarcinoma29.

In the present study, parameters related to PET/CT demonstrated limited capacity to discriminate lung adenocarcinomas with higher Ki-67 expression levels. In various forms of lung cancer, elevated Ki-67 expression has frequently been associated with poorer differentiation, larger tumour volume, higher pathological stage and increased risk of lymph node metastasis30,31. Prior research has demonstrated that the expression level of Ki-67 is progressively elevated during the late G1, S, and G2 phases of cell proliferation. At the final stage of cell division, the expression of the Ki-67 antigen is rapidly lost32. Furthermore, no expression of the Ki-67 antigen was observed in resting cells33. Nevertheless, this does not appear to be a satisfactory explanation of the experimental results.

Our experimental results showed that there was no statistically significant difference in SUVmax, SUVmean, and SUVtmr between the Ki-67 (++) and Ki-67 (+++) groups. This difference only exists between the Ki-67 (+) group (Ki-67 ≤ 10%) and the Ki-67(++) group (10% < Ki-67 ≤ 50%), as well as between the Ki-67(+) group and the Ki-67(+++) group (Ki-67 > 50%). In addition to the difference in the proliferation period, we speculate that this may be due to a strange ‘saturation phenomenon’ of lung adenocarcinoma cells with respect to the uptake of 18F-FDG, or due to the limitation of the dose of 18F-FDG injected into the lung adenocarcinoma cells. The reason for this is unknown, and this is the limitation of PET/CT parameters for identifying lung adenocarcinoma with high Ki-67 expression. We evaluated the diagnostic efficacy of each of the three statistically different indicators, SUVmax, SUVmean, and SUVtmr, by ROC curves and compared their AUC values. The results showed that the diagnostic efficacy of SUVmax > SUVmean > SUVtmr (0.865 > 0.861 > 0.847), and SUVmax ≥ 4.44 was the optimal threshold for determining the level of highly expressed Ki-67 in lung adenocarcinoma. There is a significant correlation between SUVmax and the degree of tumour differentiation in patients with lung adenocarcinoma34. In clinical work, the SUVmax value by measuring the VOI of a tumour is a simple and intuitive indicator to determine the degree of tumour proliferation.

Limitations

This study is a retrospective study. In addition, due to the gradual updating and advancement of therapeutic modalities for lung adenocarcinoma, most of the patients will undergo relevant adjuvant therapies before surgery. As a result, there are fewer lung adenocarcinoma patients without intervention before surgery, which in turn leads to a smaller sample size of the study. Therefore, the experimental results may have some selection bias.

Conclusion

PET/CT-related parameters SUVmax, SUVmean, and SUVtmr are important factors in responding to the Ki-67 expression level of lung adenocarcinoma. However, the above parameters have limitations in identifying lung adenocarcinomas with high Ki-67 expression levels.