The clinical value of spectral CT combined with targeted scanning technology in the diagnosis of invasive ground glass nodules in pulmonary adenocarcinoma

Abstract

Background

Pulmonary adenocarcinoma, a predominant form of lung cancer, is characterized by diverse histopathological subtypes, including ground-glass nodules (GGNs), which may represent different degrees of malignancy. The accurate differentiation between invasive and non-invasive GGNs is paramount, as it significantly influences clinical management and treatment strategies.

Objective

To evaluate the clinical efficacy of spectral CT combined with targeted scanning in diagnosing invasive GGNs in pulmonary adenocarcinoma.

Methods

A retrospective analysis of 120 patients with GGNs, who underwent spectral CT and targeted scanning at the Second Affiliated Hospital of Qiqihar Medical College (Nov 2021 - Dec 2022), was conducted. Patients were categorized based on postoperative pathology into non-invasive (66) and invasive (54) groups. Imaging features and values of various indicators including water concentration (WC), spectral curve slope (k value), and iodine concentration across different phases were analyzed and compared. Key independent factors for invasive GGN and the predictive value of the combined imaging technique were explored.

Results

The invasive lesion group showed a higher incidence of fissure sign, spiculation sign, and bronchial inflation sign, along with increased values of WC, WCAP, and WCVP across phases (p < 0.05). These factors were identified as main influencers of invasive GGN, with OR values >1. The combined imaging parameters achieved an AUC of 0.914, sensitivity of 0.925, and specificity of 0.880, significantly outperforming individual indicators.

Conclusion

Fissure sign, spiculation sign, bronchial inflation sign, WC, WCAP, and WCVP effectively predict invasive GGNs in pulmonary adenocarcinoma. Their combined application enhances predictive accuracy.

Keywords

spectral CT targeted scanning technology lung adenocarcinoma ground glass nodule

1 Introduction

Lung tumors, which can be classified as benign or malignant, represent a significant health concern globally due to their high mortality rates.¹ Lung cancer was characterized by uncontrolled cell growth in lung tissues and are the leading cause of cancer-related deaths worldwide.¹ According to the World Health Organization (WHO), lung cancer accounted for approximately 2.1 million new cases and 1.8 million fatalities in 2020, making it the most prevalent form of cancer.¹ Regionally, the incidence of lung cancer varies significantly, with the highest rates reported in Eastern Asia, particularly in China, where smoking prevalence is high.² This alarming trend emphasizes the necessity of investigating the pathophysiology and diagnostic strategies for lung tumors, particularly the invasive characteristics of ground-glass nodules (GGNs), which pose a unique challenge in clinical practice.

According to the 2015 WHO histological classification of lung tumors, pulmonary adenocarcinoma mainly includes four types: atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC), with AAH and AIS collectively referred to as preinvasive lesions.¹ Studies have shown² that there are differences in invasiveness among different pathological subtypes, leading to the adoption of different treatment regimens in clinical practice. Generally, patients with preinvasive lesions and MIA undergo sublobar resection, such as wedge resection or segmentectomy, while patients with IAC often require lobectomy, and sometimes even including mediastinal lymph node dissection. Therefore, the accurate preoperative diagnosis of the pathological type of pulmonary adenocarcinoma is crucial. While pathological examination has been traditionally used for the diagnosis of pulmonary adenocarcinoma with relatively high accuracy, it unavoidably causes certain trauma to patients.³

Spectral CT imaging, which includes single-energy imaging, spectral curve, effective atomic number, and images of iodine-based or water-based substances, can be used for the early clinical diagnosis of lung cancer, differentiation of lung cancer pathological types, assessment of lung perfusion, evaluation of lung lymph node metastasis, guidance for TNM staging, and assessment of lung cancer treatment efficacy.⁴ Spectral technology serves as a reliable method for evaluating and predicting tumor treatment efficacy, providing new evidence for clinical diagnosis and treatment.^5,6 Previous studies have found⁷ that the accuracy and sensitivity of spectral CT alone for predicting the invasiveness of pulmonary adenocarcinoma ground glass nodules (GGN) are not high. Therefore, it is necessary to combine spectral CT with other examination techniques to improve the predictive efficacy of invasive GGN in pulmonary adenocarcinoma. Targeted scanning involves the examination of the internal structure of pulmonary adenocarcinoma, its relationship with surrounding blood vessels, and the presence of mobile and communicating vessels using maximum density projection (MIP), MPR, and CPR reconstruction techniques.⁸

GGNs have gained significant attention in recent years due to their association with early-stage lung cancer, particularly pulmonary adenocarcinoma. Studies have shown that while many GGNs are benign, a substantial proportion can be indicative of malignancy, necessitating accurate diagnostic approaches.^6–8 A notable study have demonstrated that the likelihood of malignancy increases with the size and radiological characteristics of GGNs.⁷ Furthermore, recently studies highlighted the limitations of conventional imaging techniques, such as computed tomography (CT) and positron emission tomography (PET), in differentiating between invasive and non-invasive GGNs.^7,8 These findings underscore the need for advanced imaging modalities and diagnostic strategies to improve preoperative evaluations.

Accordingly, this study explores the clinical value of spectral CT combined with targeted scanning technology in the diagnosis of invasive GGN in pulmonary adenocarcinoma, aiming to provide potential directions for improving the treatment efficacy for early lung cancer or precancerous lesions and improving patient prognosis.

2 Methods

2.1 Study population

A retrospective analysis was conducted on the clinical data of 120 patients with pulmonary adenocarcinoma ground glass nodules (GGN) who underwent spectral CT combined with targeted scanning technology at the Second Affiliated Hospital of Qiqihar Medical College from November 2021 to December 2022. According to postoperative pathological results, the patients were divided into the invasive lesion group (including AAH, AIS, and MIA) with 66 cases, and the non-invasive lesion group with 54 cases (IAC).

The inclusion criteria were: (1) aged between 40 and 80 years; (2) underwent surgical resection of GGN lesions within one week after spectral plain scan + enhanced CT examination; (3) presence of a single GGN lesion with a maximum cross-sectional diameter of ≤3.0 cm.

The exclusion criteria were: (1) a history or treatment history of pulmonary GGN lesions; (2) a history of malignant tumors; (3) incomplete imaging, clinical, and pathological data.

2.2 Imaging and image processing

All patients underwent chest spectral CT scans with the GE Revolution scanner, followed by targeted three-dimensional image processing, as detailed below: (1) Spectral CT examination: Prior to the examination, the operating technician provided a detailed explanation to the patient about the contraindications and precautions of this examination method. Before the spectral CT plain scan and enhanced scan, a venous indwelling needle was inserted into the antecubital vein. Before the scan commenced, the examining technician thoroughly explained the examination process to the patient and patiently discussed potential physiological reactions induced by the examination to reduce patient tension and anxiety, including breath-holding training. During the examination, the patient was positioned supine with the arms extended and raised towards the head to facilitate foot-first scanning. The scan range covered from the thoracic inlet to the level of the diaphragm, encompassing the entire lung field. The scanning parameters included tube voltages of 80 kVp to 140 kVp with automatic tube current modulation, a slice thickness of 5.0 mm, a 5.0 mm interval, a 512 × 512 matrix, a pitch of 0.938:1, a rotation time of 0.6 s, and a reconstructed image interval of 1.25 mm. The spectral CT plain scan was performed first, followed by the spectral CT enhanced arterial phase scan. A high-pressure dual-barrel injector was used to inject 70 ml to 85 ml of contrast agent iohexol injection (350 mg I/ml) at a flow rate of 2.5 ml/s to 3.5 ml/s via the right antecubital vein, followed by a 20 ml injection of normal saline at the same flow rate, with monitoring at the level of the descending aorta (carina) and a trigger threshold of 110.0 HU. The venous phase scan commenced 25 s after the arterial phase scan. The scan ranges for the plain scan, arterial phase, and venous phase all remained consistent, and all original scan images were required to be transferred to the ADW4.7 workstation. The relevant phases were selected, followed by clicking on GSI Viewer, then on MD analysis to begin the quantitative analysis of spectral material parameters. An area of interest (ROI) was manually outlined within the plane displaying the maximum cross-sectional diameter of the lesion, with the ROI area covering 70% or more of the lesion. For the same patient, the positioning and size of the ROI for the plain scan, arterial phase, and venous phase lesions were consistent, and the selected areas avoided bronchi, blood vessels, calcifications, necrotic areas, and other unsuitable areas. The measured data included the following information: water concentration (WC) in the plain scan phase, spectral curve slope (k value); water concentration value in the arterial phase (WCAP), iodine concentration value (ICAP), standardized iodine concentration value (NICAP), k value; water concentration value in the venous phase (WCVP), iodine concentration value (ICVP), standardized iodine concentration value (NICVP), k value.(2) Targeted three-dimensional imaging processing: Thin-section high-resolution computed tomography (HRCT) targeted scanning was used to scan nodular tissue, and the scanned raw data and images were transmitted to the ADW4.7 workstation for volume rendering (VR), multiplanar reformation (MPR), surface-shaded display (SSD), maximum intensity projection (MIP), and minimum intensity projection (MinIP) reconstructions. The American GE Lung VCAR was used for morphological analysis of lesions on non-enhanced lung window (window width 1200, window level −600) thin-section CT images. Evaluation criteria included fissure sign (yes/no), spiculation (yes/no), bronchus inflation sign (yes/no), air cavity sign (yes/no), and pleural depression sign (yes/no).

All data were analyzed by two diagnostic physicians with over 10 years of experience in chest imaging diagnosis and proficiency in spectral image analysis and multiple quantitative parameter measurements. Single-blind analysis was conducted, independently evaluating the morphological signs of GGN, measuring quantitative parameters, calculating related parameter data, making diagnostic judgments on lesion pathology, and classifying groups. In the event of disagreement between the two observing physicians’ opinions, a consensus regarding the diagnosis of the lesion should be reached through joint discussion and analysis. Strict adherence to standards was observed during the morphological sign analysis and data measurement processes to ensure objectivity and accuracy, minimizing the impact of physician subjectivity on the results.

2.3 Statistical methods

Data were analyzed using SPSS 25.0 statistical software. Frequency data were presented as n(%) and differences between two groups were compared using the chi-square test. When the sample size was ≥40, but the theoretical frequency 1 ≤ T < 5, the chi-square test correction formula was used. If the sample size was <40 or the theoretical frequency T < 1, statistical analysis was performed using Fisher's exact probability method. For sample sizes ≥40 and the smallest theoretical frequency ≥5, the Pearson non-corrected method was utilized. Normally distributed metric data were represented as (±s) and analyzed using the t-test. Pearson's correlation coefficient was employed for collinearity analysis, with an absolute correlation coefficient >0.8 indicating potential collinearity. Further analysis involving variance inflation factor (VIF) was conducted, with a VIF value >10 suggesting collinearity, prompting reconsideration of its removal from the model for subsequent analysis. Logistic regression was applied to identify independent influencing factors for invasive GGN, followed by receiver operating characteristic (ROC) curve construction. The area under the curve (AUC) was used to evaluate the diagnostic performance of the predictive index, with AUC values of 0.5–0.7 indicating poor diagnostic performance, 0.7–0.9 indicating moderate diagnostic performance, and >0.9 indicating good diagnostic performance. A significance level of p < 0.05 was considered statistically significant.

3 Results

3.1 Comparison of baseline data between the two groups

As shown in Table 1, there were no statistically significant differences in general data such as gender, age, occupation, and medical history between the two groups (p $>$ 0.05), indicating comparability between the groups.

Table 1.
Baseline data comparison between the two groups.

Non-invasive Invasive

lesion lesion

Group group (n = 66) group (n = 54) χ^2/t p

Gender [n(%)] Male 45(68.18) 32(59.26) 0.677 0.411

Female 21(31.82) 22(40.74)

Age (years) 52.40 ± 5.62 51.96 ± 5.78 0.422 0.674

Occupation [n(%)] Painter 23(34.85) 19(35.19) 0.009 1

Miner 18(27.27) 15(27.78)

Asbestos worker 15(22.73) 12(22.22)

Other 10(15.15) 8(14.81)

Medical history [n(%)] Asthma 9(13.64) 6(11.11) 0.019 0.89

Chronic bronchitis 10(15.15) 8(14.81) 0 1

Emphysema 9(13.64) 4(7.41) 0.635 0.425

Pleurisy 3(4.55) 2(3.70) 0 1

		Non-invasive	Invasive
Gender [n(%)]	Male	45(68.18)	32(59.26)	0.677	0.411
	Female	21(31.82)	22(40.74)
Age (years)		52.40 ± 5.62	51.96 ± 5.78	0.422	0.674
Occupation [n(%)]	Painter	23(34.85)	19(35.19)	0.009	1
	Miner	18(27.27)	15(27.78)
	Asbestos worker	15(22.73)	12(22.22)
	Other	10(15.15)	8(14.81)
Medical history [n(%)]	Asthma	9(13.64)	6(11.11)	0.019	0.89
	Chronic bronchitis	10(15.15)	8(14.81)	0	1
	Emphysema	9(13.64)	4(7.41)	0.635	0.425
	Pleurisy	3(4.55)	2(3.70)	0	1

3.2 Comparison of morphological features of GGN lesions between two groups

As depicted in Table 2, the proportions of fissure sign, spiculation sign, and bronchus inflation sign were higher in the invasive lesion group compared to the non-invasive lesion group, demonstrating statistical differences between the groups (p < 0.05). The proportions of air cavity sign and pleural depression sign showed no statistically significant differences between the two groups (p $>$ 0.05), suggesting that combined targeted scanning may play a predictive role in invasive GGN in patients with lung adenocarcinoma.

Table 2.
Comparison of morphological signs of GGN lesions in the two groups [n(%)].

Non-invasive lesion Invasive lesion

Group group (n = 66) group (n = 54) χ² p

Lobulated sign Present 25(37.88) 42(77.78)

Absent 41(62.12) 12(22.22) 17.589 p < 0.001

Spiculated sign Present 22(33.33) 44(81.48)

Absent 44(66.66) 10(18.52) 25.908 p < 0.001

Bronchus inflation sign Present 17(25.76) 43(79.63)

Absent 49(74.24) 11(20.37) 32.357 p < 0.001

Air cyst sign Present 61(92.42) 50(92.59)

Absent 5(7.58) 4(7.41) 0 1

Pleural indentation sign Present 19(28.79) 16(29.63)

Absent 47(71.21) 38(70.37) 0 1

		Non-invasive lesion	Invasive lesion
Lobulated sign	Present	25(37.88)	42(77.78)
	Absent	41(62.12)	12(22.22)	17.589	p < 0.001
Spiculated sign	Present	22(33.33)	44(81.48)
	Absent	44(66.66)	10(18.52)	25.908	p < 0.001
Bronchus inflation sign	Present	17(25.76)	43(79.63)
	Absent	49(74.24)	11(20.37)	32.357	p < 0.001
Air cyst sign	Present	61(92.42)	50(92.59)
	Absent	5(7.58)	4(7.41)	0	1
Pleural indentation sign	Present	19(28.79)	16(29.63)
	Absent	47(71.21)	38(70.37)	0	1

3.3 Comparison of WC and k values between the two groups

Based on Table 3, during the plain scan phase, the WC of the invasive lesion group was found to be higher than that of the non-invasive lesion group, indicating a statistically significant difference between the groups (p < 0.05). However, there was no statistically significant difference in the k values between the two groups (p $>$ 0.05). These findings suggest that the WC during the plain scan phase may be an important predictive factor for invasive ground glass nodules in patients with lung adenocarcinoma.

Table 3.
Comparison of plain scan phase WC and k values between the two groups ( ${\bar{χ}}_{\pm s}$ ).

Non-invasive lesion Invasive lesion

Group group (n = 66) group (n = 54) t p

WC (mg/cm³) 286.35 ± 54.36 368.50 ± 58.70 7.884 p < 0.001

K Value 5.25 ± 1.02 5.14 ± 0.98 0.637 0.526

	Non-invasive lesion	Invasive lesion
WC (mg/cm³)	286.35 ± 54.36	368.50 ± 58.70	7.884	p < 0.001
K Value	5.25 ± 1.02	5.14 ± 0.98	0.637	0.526

3.4 Comparison of WCAP, ICAP, NICAP and k values in arterial phase between the two groups

According to Table 4, during the arterial phase, the WCAP of the invasive lesion group was higher than that of the non-invasive lesion group, demonstrating a statistically significant difference between the groups (p < 0.05). However, there was no statistically significant difference in the ICAP, NICAP, and k values between the two groups (p < 0.05). These results suggest that WCAP during the arterial phase may have diagnostic value for invasive ground glass nodules in patients with lung adenocarcinoma.

Table 4.
Comparison of arterial phase WCAP, ICAP, NICAP, mNICAP, and k values between the two groups ( ${\bar{χ}}_{\pm s}$ ).

Non-invasive lesion Invasive lesion

Group group (n = 66) group (n = 54) t p

WCAP (mg/cm³) 298.25 ± 46.22 374.54 ± 48.10 8.797 p < 0.001

ICAP (mg/cm³) 2.95 ± 0.45 3.10 ± 0.52 1.612 0.11

NICAP (mg/cm³) 0.25 ± 0.04 0.26 ± 0.05 1.678 0.097

K Value 5.46 ± 1.02 5.84 ± 1.10 1.933 0.056

	Non-invasive lesion	Invasive lesion
WCAP (mg/cm³)	298.25 ± 46.22	374.54 ± 48.10	8.797	p < 0.001
ICAP (mg/cm³)	2.95 ± 0.45	3.10 ± 0.52	1.612	0.11
NICAP (mg/cm³)	0.25 ± 0.04	0.26 ± 0.05	1.678	0.097
K Value	5.46 ± 1.02	5.84 ± 1.10	1.933	0.056

3.5 Comparison of WCVP, ICVP, NICVP and k value between the two groups in venous phase

As indicated in Table 5, during the venous phase, the WCVP of the invasive lesion group was higher than that of the non-invasive lesion group, revealing a statistically significant difference between the groups (p < 0.05). However, there was no statistically significant difference in the ICVP, NICVP, and k values between the two groups (p < 0.05). These findings suggest that WCVP during the venous phase may be one of the main predictive factors for invasive ground glass nodules in patients with lung adenocarcinoma.

Table 5.
Comparison of venous phase WCVP, ICVP, NICVP, mNICVP, and k values between the two groups ( ${\bar{χ}}_{\pm s}$ ).

Non-invasive lesion Invasive lesion

Group group (n = 66) group (n = 54) t p

WCVP (mg/cm³) 290.76 ± 45.40 373.32 ± 46.82 9.741 p < 0.001

ICAP (mg/cm³) 2.68 ± 0.55 2.70 ± 0.50 0.209 0.835

NICAP (mg/cm³) 0.45 ± 0.07 0.47 ± 0.08 1.66 0.1

K Value 5.10 ± 1.20 4.98 ± 1.16 0.534 0.595

	Non-invasive lesion	Invasive lesion
WCVP (mg/cm³)	290.76 ± 45.40	373.32 ± 46.82	9.741	p < 0.001
ICAP (mg/cm³)	2.68 ± 0.55	2.70 ± 0.50	0.209	0.835
NICAP (mg/cm³)	0.45 ± 0.07	0.47 ± 0.08	1.66	0.1
K Value	5.10 ± 1.20	4.98 ± 1.16	0.534	0.595

3.6 Analysis of collinearity

As shown in Tables 6 and 7, if the absolute value of the correlation coefficient between certain pairs of WC, WCAP, and WCVP is excessively high (e.g., > 0.8), it likely indicates the presence of multicollinearity issues. Further analysis revealed that the VIF values of the individual indicators of WC, WCAP, and WCVP were all <10, suggesting the absence of multicollinearity issues and enabling their inclusion in the logistic regression analysis.

Table 6.
Pearson correlation coefficients.

Standard Pathological type of

Project Mean deviation lung adenocarcinoma WC WCAP WCVP

Pathological type of lung adenocarcinoma 0.450 0.500 1

WC 323.337 43.701 0.840* 1

WCAP 332.580 41.755 0.813* 0.877* 1

WCVP 327.907 44.244 0.832* 0.870* 0.846* 1

		Standard	Pathological type of
Pathological type of lung adenocarcinoma	0.450	0.500	1
WC	323.337	43.701	0.840*	1
WCAP	332.580	41.755	0.813*	0.877*	1
WCVP	327.907	44.244	0.832*	0.870*	0.846*	1

Note: *p < 0.05.

Table 7.

Collinearity diagnosis.

Variable	VIF Value	Tolerance
Pathologic types of lung adenocarcinoma	23.647	0.042
WC	8.773	0.114
WCAP	6.144	0.163
WCVP	7.667	0.130

3.7 Multifactorial analysis of factors affecting invasive ground glass nodules

As indicated in Table 8, after including the differing indicators in the logistic regression equation, it was found that spiculated sign, lobulated sign, bronchus inflation sign, WC, WCAP, and WCVP are all major factors influencing invasive GGN. The odds ratio (OR) values for these factors were all <1, suggesting that the aforementioned target scan imaging signs and quantitative parameters from spectral CT are influential factors for invasive GGN.

Table 8.
Multifactor analysis of factors affecting invasive ground glass nodules.

Influencing Standard Odds ratio 95% Confidence

Factor β Value error p Value (OR) interval (95% CI)

Lobulated sign 0.745 0.870 0.001 1.345 0.8305 $\sim$ 3.8862

Spiculated sign 0.802 0.902 0.024 1.874 1.1032 $\sim$ 5.7245

Bronchus inflation sign 0.565 0.066 0.001 1.798 0.7262 $\sim$ 3.9482

WC 1.450 0.470 0.001 1.884 0.6250 $\sim$ 4.8876

WCAP 0.674 0.728 0.006 1.670 1.1045 $\sim$ 2.9035

WCVP 0.825 0.934 0.020 1.705 1.4562 $\sim$ 5.7850

Influencing		Standard		Odds ratio	95% Confidence
Lobulated sign	0.745	0.870	0.001	1.345	0.8305 $\sim$ 3.8862
Spiculated sign	0.802	0.902	0.024	1.874	1.1032 $\sim$ 5.7245
Bronchus inflation sign	0.565	0.066	0.001	1.798	0.7262 $\sim$ 3.9482
WC	1.450	0.470	0.001	1.884	0.6250 $\sim$ 4.8876
WCAP	0.674	0.728	0.006	1.670	1.1045 $\sim$ 2.9035
WCVP	0.825	0.934	0.020	1.705	1.4562 $\sim$ 5.7850

3.8 The predictive value of individual or combined factors for invasive GGN in lung adenocarcinoma

As demonstrated in Table 9, ROC analysis revealed that the AUC value for predicting invasive GGN in lung adenocarcinoma using a combination of spectral CT multi-parameters and target scan imaging signs was 0.914. This approach exhibited a sensitivity of 0.925, specificity of 0.880, and a Youden index of 0.501, significantly higher than the predictive value of each individual parameter. These results suggest that the combined use of spectral CT multi-parameters and target scan imaging signs can enhance the predictive value for the pathological type of lung adenocarcinoma.

Table 9.
Predictive value of individual or combined factors for invasive GGN in lung adenocarcinoma.

Feature AUC value Sensitivity Specificity Youden index

Lobulated sign 0.707 0.725 0.614 0.214

Spiculated sign 0.746 0.710 0.652 0.352

Bronchus inflation sign 0.751 0.708 0.701 0.501

WC 0.820 0.725 0.779 0.379

WCAP 0.807 0.725 0.614 0.214

WCVP 0.816 0.710 0.752 0.352

Combined 0.914 0.925 0.880 0.501

Feature	AUC value	Sensitivity	Specificity	Youden index
Lobulated sign	0.707	0.725	0.614	0.214
Spiculated sign	0.746	0.710	0.652	0.352
Bronchus inflation sign	0.751	0.708	0.701	0.501
WC	0.820	0.725	0.779	0.379
WCAP	0.807	0.725	0.614	0.214
WCVP	0.816	0.710	0.752	0.352
Combined	0.914	0.925	0.880	0.501

4 Discussion

The pathological basis of ground glass nodules (GGN) is the growth of lesion tissues along the alveolar wall without disrupting the alveolar structure and with relatively high gas content in the alveoli.⁹ Within the pathological spectrum of lung adenocarcinoma, atypical adenomatous hyperplasia (AAH) manifests as localized non-classic hyperplasia covering the alveoli and respiratory bronchioles, while adenocarcinoma in situ (AIS) is characterized by the growth of adenocarcinoma cells along the alveolar wall without interstitial, vascular, or pleural infiltration, thus being considered as non-invasive precursor lesions.^10,11 Minimally invasive adenocarcinoma (MIA) is characterized by infiltration within the lesion being less than or equal to 5 mm, while invasive adenocarcinoma (IAC) includes infiltration greater than 5 mm; therefore, the latter two are considered invasive lesions.¹² Non-invasive precursor lesions and invasive lesions exhibit different biological behaviors, with the latter being more prone to lymph node metastasis and having a poorer prognosis. Therefore, the critical clinical importance of effectively distinguishing between the two cannot be overstated.^13,14 Studies by BAI WL et al.¹⁵ found that CT features such as lobulation, spiculation, air bronchogram or air space, and pleural indentation signs are more commonly observed in invasive lesions. However, in this study, only lobulation, spiculation, and bronchus inflation signs were more common in invasive lesions. The analysis suggests that this discrepancy may be due to the larger sample size of invasive lesions in the study by BAI WL et al., which resulted in differences in the final analysis. A study by WU LY et al.¹⁶ also found a higher proportion of lobulation, spiculation, and bronchus inflation signs in invasive lesions in CT features, similar to the results of this study, indicating that individuals with these features have a higher malignant degree and are more likely to be diagnosed with invasive GGN.

Due to the small volume and limited CT features of pulmonary ground glass nodules (GGN), the assessment and differentiation of their invasiveness by traditional CT present considerable challenges. Spectral CT achieves substance separation by exploiting the differential absorption and attenuation of X-rays by different materials, yielding various material pairs in the images (such as iodine-water, water-calcium pairs).^17,18 Clinical studies have indicated that the iodine-water pair is the most stable material pair.¹⁹ Hence, this study utilized the iodine-water material pair for analysis. Research by XIAO RX et al.²⁰ found that arterial and venous phase CT values (WC) in the invasive lesion group were higher than that in the non-invasive lesion group (p < 0.05). Consistently, our study found that arterial and venous phase CT values for WCAP were higher in the invasive lesion group than in the non-invasive lesion group (p < 0.05), consistent with the above study, and plain scan phase WC was also higher in the non-invasive lesion group (p < 0.05). This may be attributed to the increasing malignancy as GGN progresses to invasive stages, leading to higher cell density, increased tumor components, and interstitial thickening, ultimately resulting in elevated WC.^21,22 The higher malignancy in the invasive lesion group leads to faster cell proliferation and metabolism, resulting in a significant increase in metabolic byproducts and interstitial fluid, thereby significantly elevating WC measured by spectral CT compared to pre-invasive lesions.²³ Research by Wu LY et al.²⁴ suggested that a combination of plain scan, arterial, and venous phase ICVP, NICVP, mNICVP, and k values can also differentiate invasive GGN. However, our study did not identify differences in the above indicators between pre-invasive and invasive lesion groups, potentially due to variations in sample size, equipment types, and post-processing software used in different studies.

Despite the individual significance of CT features and quantitative spectral parameters in predicting invasive ground glass nodules (GGN), their sensitivity, specificity, and accuracy are generally low.²⁵ However, multi-parameter combinations can effectively enhance diagnostic efficacy. In the research by Park et al.,²⁶ the combination of CT values and CT features was found to have important predictive value for invasive GGN, with sensitivity and specificity of 91.7% and 71.4%, respectively, and an AUC of 0.820. In the present study, we found that lobulated sign, spiculated sign, bronchus inflation sign, WC, WCAP, and WCVP are all significant factors influencing invasive GGN, with odds ratios exceeding 1. Combining spectral CT parameters with targeted scan imaging features resulted in an AUC value of 0.914, sensitivity of 0.925, specificity of 0.880, and a Youden index of 0.501, significantly higher than individual indicators, suggesting that this combined approach can enhance the predictive value for invasive GGN in lung adenocarcinoma, providing potential diagnostic evidence for radiologists.

Looking toward the future, the integration of artificial intelligence (AI) into the diagnostic process holds great promise for further improving accuracy and efficiency in identifying malignant nodules. AI algorithms, particularly those utilizing deep learning techniques, could potentially analyze complex imaging data more comprehensively than traditional methods, enabling more nuanced recognition patterns that human observers might overlook. Additionally, AI could assist in the real-time evaluation of imaging results, optimize protocols for spectral CT scans, and personalize patient assessments based on individual risk profiles. Future research should explore the synergistic relationship between AI and multi-parametric imaging technologies, aiming to create automated diagnostic frameworks that could alleviate the burden on radiologists while increasing precision in the diagnosis of pulmonary adenocarcinoma.

Despite the promising findings of this study, several limitations must be acknowledged. First, the sample size, while sufficient for our analyses, may limit the generalizability of the results to larger populations. A larger cohort would be necessary to validate these findings across diverse demographic groups and various clinical settings. Additionally, the study was conducted at a single center, which may introduce bias related to specific imaging protocols, patient selection criteria, and radiological interpretation standards that could differ in other institutions. Moreover, our study utilized retrospective data; therefore, inherent biases, such as selection bias and inconsistencies in data collection, may have influenced the outcomes. Prospective studies with predefined protocols would be beneficial in minimizing these biases and confirming the reliability of our results. Lastly, while we focused on specific imaging parameters and their predictive capabilities, other factors such as patient history, genetic markers, and tumor biology should also be considered in future research to provide a more comprehensive approach to diagnostics in invasive pulmonary adenocarcinoma.

5 Conclusion

This study underscored the significant clinical value of combining spectral CT and targeted scanning technology in predicting invasive ground-glass nodules (GGNs) in pulmonary adenocarcinoma. Specifically, key imaging features such as the fissure sign, spiculation sign, and bronchial inflation sign, alongside quantitative spectral parameters like water concentration (WC), water concentration in the arterial phase (WCAP), and water concentration in the venous phase (WCVP), have demonstrated robust predictive capabilities for distinguishing between invasive and non-invasive lesions. The high area under the curve (AUC) of 0.914 indicates that this approach significantly surpasses the predictive power of individual indicators, thereby enhancing diagnostic accuracy essential for the timely treatment of patients.

Footnotes

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Second Affiliated Hospital of Qiqihar Medical College, China (No. [2022] 45). Signed written informed consents were obtained from the patients and/or guardians.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Heilongjiang Provincial Health Commission Project (Project number: 20220909010622).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

Wang

Zhang

, et al. 3D Deep learning based classification of pulmonary ground glass opacity nodules with automatic segmentation. Comput Med Imaging Graph 2021; 88: 101814.

Halpenny

Das

Ziv

, et al. Percutaneous computed tomography guided biopsy of sub-solid pulmonary nodules: Differentiating solid from ground glass components at the time of biopsy. Clin Imaging 2021; 69: 332–338.

Succony

Rassl

Barker

, et al. Adenocarcinoma spectrum lesions of the lung: Detection, pathology and treatment strategies. Cancer Treat Rev 2021; 99: 102237.

Nakamura

Goto

Nara

, et al. Pure ground glass opacity(GGO)on chest CT: A rare presentation of lung metastasis of Malignant Phyllodes tumor. Breast Cancer 2020; 27: 1187–1190.

Tækker

Kristjánsdóttir

Graumann

, et al. Diagnostic accuracy of low-dose and ultra-low-dose CT in detection of chest pathology: A systematic review. Clin Imaging 2021; 74: 139–148.

Tsai

Yeh

Hsu

, et al. CT-Guided Core biopsy for peripheral sub-solid pulmonary nodules to predict predominant histological and aggressive subtypes of lung adenocarcinoma. Ann Surg Oncol 2020; 27: 4405–4412.

Yang

Tan

, et al. Preoperative prediction of the degree of differentiation of lung adenocarcinoma presenting as sub-solid or solid nodules with a radiomics nomogram. Clin Radiol 2022; 77: e680–e688.

Sugiyama

Tajiri

Fujita

, et al.

The approach of scratch-imprint cytology: Is it an alternative to frozen section for intraoperative assessment of pulmonary lesions?

Pathol Int 2020; 70: 31–39.

Wang

Liu

, et al. Predicting pathological invasiveness of lung adenocarcinoma manifesting as GGO-predominant nodules: A combined prediction model generated from DECT. Acad Radiol 2021; 28: 509–516.

10.

Yotsukura

Asamura

Motoi

, et al. Long-Term prognosis of patients with resected adenocarcinoma in situ and minimally invasive adenocarcinoma of the lung. J Thorac Oncol 2021; 16: 1312–1320.

11.

Mori

Rao

Popper

, et al. Atypical adenomatous hyperplasia of the lung: A probable forerunner in the development of adenocarcinoma of the lung. Mod Pathol 2001; 14: 72–84.

12.

Sun

, et al. Disparate genomic characteristics of patients with early-stage lung adenocarcinoma manifesting as radiological subsolid or solid lesions. Lung Cancer 2022; 166: 178–188.

13.

Deng

Wang

, et al. Ten-year follow-up of lung cancer patients with resected adenocarcinoma in situ or minimally invasive adenocarcinoma: Wedge resection is curative. J Thorac Cardiovasc Surg 2022; 164: 1614–1622.e1.

14.

Nie

Yao

Zhu

, et al. Evolutionary metabolic landscape from preneoplasia to invasive lung adenocarcinoma. Nat Commun 2021; 12: 6479.

15.

Bai

Zhang

Wang

, et al. Comparative analysis of the long-term outcomes of segmentectomy and lobectomy for stage IA1 lung adenocarcinoma in patients with or without previous malignancy of other organs: A population-based study. Expert Rev Anticancer Ther 2022; 22: 215–228.

16.

Zhang

, et al. Ground glass opacity featured lung adenocarcinoma in teenagers. J Cancer Res Clin Oncol 2021; 147: 3719–3724.

17.

Higashiyama

Kobayashi

Kashima

, et al. Invasive mucinous adenocarcinoma of the lung with a mural nodule-like lesion. Am J Surg Pathol 2022; 46: 1524–1532.

18.

Bruehl

Doxtader

Cheng

, et al. Does histological assessment accurately distinguish separate primary lung adenocarcinomas from intrapulmonary metastases?A study of paired resected lung nodules in 32 patients using a routine next-generation sequencing panel for driver mutations. J Clin Pathol 2022; 75: 390–396.

19.

Huang

Zheng

Chen

, et al. Fusion of CT images and clinical variables based on deep learning for predicting invasiveness risk of stage I lung adenocarcinoma. Med Phys 2022; 49: 6384–6394.

20.

Xiao

, et al. Lung adenocarcinoma manifesting as subsolid nodule potentially represents tumour in the equilibrium phase of immunoediting. Immunology 2023; 168: 290–301.

21.

Song

Xing

Zhu

, et al. Hybrid clinical-radiomics model for precisely predicting the invasiveness of lung adenocarcinoma manifesting as pure ground-glass nodule. Acad Radiol 2021; 28: e267–e277.

22.

Zhang

Liu

Zhang

, et al. Extracellular vesicle long RNA markers of early-stage lung adenocarcinoma. Int J Cancer 2023; 152: 1490–1500.

23.

Lin

Zhang

. Differentiation of adenocarcinoma in situ with alveolar collapse from minimally invasive adenocarcinoma or invasive adenocarcinoma appearing as part-solid ground-glass nodules (≤2 cm) using computed tomography. Jpn J Radiol 2022; 40: 29–37.

24.

Gao

, et al. The value of various peritumoral radiomic features in differentiating the invasiveness of adenocarcinoma manifesting as ground-glass nodules. Eur Radiol 2021; 31: 9030–9037.

25.

Yin

, et al. A comparative study for the evaluation of CT-based conventional,radiomic,combined conventional and radiomic,and delta-radiomic features,and the prediction of the invasiveness of lung adenocarcinoma manifesting as ground-glass nodules. Clin Radiol 2022; 77: e741–e748.

26.

Park

Doo

Sung

, et al. Computed tomography findings for predicting invasiveness of lung adenocarcinomas manifesting as pure ground-glass nodules. Can Assoc Radiol J 2023; 74: 137–146.

		Non-invasive	Invasive
		lesion	lesion
Group		group (n = 66)	group (n = 54)	χ^2/t	p
Gender [n(%)]	Male	45(68.18)	32(59.26)	0.677	0.411
	Female	21(31.82)	22(40.74)
Age (years)		52.40 ± 5.62	51.96 ± 5.78	0.422	0.674
Occupation [n(%)]	Painter	23(34.85)	19(35.19)	0.009	1
	Miner	18(27.27)	15(27.78)
	Asbestos worker	15(22.73)	12(22.22)
	Other	10(15.15)	8(14.81)
Medical history [n(%)]	Asthma	9(13.64)	6(11.11)	0.019	0.89
	Chronic bronchitis	10(15.15)	8(14.81)	0	1
	Emphysema	9(13.64)	4(7.41)	0.635	0.425
	Pleurisy	3(4.55)	2(3.70)	0	1

		Standard	Pathological type of
Project	Mean	deviation	lung adenocarcinoma	WC	WCAP	WCVP
Pathological type of lung adenocarcinoma	0.450	0.500	1
WC	323.337	43.701	0.840*	1
WCAP	332.580	41.755	0.813*	0.877*	1
WCVP	327.907	44.244	0.832*	0.870*	0.846*	1

Influencing		Standard		Odds ratio	95% Confidence
Factor	β Value	error	p Value	(OR)	interval (95% CI)
Lobulated sign	0.745	0.870	0.001	1.345	0.8305 $\sim$ 3.8862
Spiculated sign	0.802	0.902	0.024	1.874	1.1032 $\sim$ 5.7245
Bronchus inflation sign	0.565	0.066	0.001	1.798	0.7262 $\sim$ 3.9482
WC	1.450	0.470	0.001	1.884	0.6250 $\sim$ 4.8876
WCAP	0.674	0.728	0.006	1.670	1.1045 $\sim$ 2.9035
WCVP	0.825	0.934	0.020	1.705	1.4562 $\sim$ 5.7850

The clinical value of spectral CT combined with targeted scanning technology in the diagnosis of invasive ground glass nodules in pulmonary adenocarcinoma

Abstract

Background

Objective

Methods

Results

Conclusion

Keywords

1 Introduction

2 Methods

2.1 Study population

2.2 Imaging and image processing

2.3 Statistical methods

3 Results

3.1 Comparison of baseline data between the two groups

Table 3. Comparison of plain scan phase WC and k values between the two groups ( χ ¯ ± s ). Non-invasive lesion Invasive lesion Group group (n = 66) group (n = 54) t p WC (mg/cm3) 286.35 ± 54.36 368.50 ± 58.70 7.884 p < 0.001 K Value 5.25 ± 1.02 5.14 ± 0.98 0.637 0.526

Table 6. Pearson correlation coefficients. Standard Pathological type of Project Mean deviation lung adenocarcinoma WC WCAP WCVP Pathological type of lung adenocarcinoma 0.450 0.500 1 WC 323.337 43.701 0.840* 1 WCAP 332.580 41.755 0.813* 0.877* 1 WCVP 327.907 44.244 0.832* 0.870* 0.846* 1

5 Conclusion

Footnotes

Ethics approval and consent to participate

Funding

Declaration of conflicting interests

Availability of data and materials

References

Table 3.
Comparison of plain scan phase WC and k values between the two groups ( ${\bar{χ}}_{\pm s}$ ).

Non-invasive lesion Invasive lesion

Group group (n = 66) group (n = 54) t p

WC (mg/cm³) 286.35 ± 54.36 368.50 ± 58.70 7.884 p < 0.001

K Value 5.25 ± 1.02 5.14 ± 0.98 0.637 0.526

Table 6.
Pearson correlation coefficients.

Standard Pathological type of

Project Mean deviation lung adenocarcinoma WC WCAP WCVP

Pathological type of lung adenocarcinoma 0.450 0.500 1

WC 323.337 43.701 0.840* 1

WCAP 332.580 41.755 0.813* 0.877* 1

WCVP 327.907 44.244 0.832* 0.870* 0.846* 1