Impact of lung-RADS classification system on the accurate diagnosis of pulmonary nodular lesions in oncology patients

Lung assessment is highly recommended in the management of oncology patients as it is the commonest affected site in metastatic dissemination. The low-dose CT with nodule reporting system based on Lung Reporting and Data System (lung-RADS) is a promising non-invasive tool for the characterization of incidentally detected pulmonary nodules. The authors aimed to assess the accuracy of the “lung-RADS” classification system as a non-invasive tool for the characterization of any newly developed pulmonary nodules among oncology patients. Ethics committee approval and informed written consent were obtained from the studied patients. A non-contrast LDCT study was performed on all patients with a nodule reporting system based on the lung-RADS classification system applied for evaluation of each detected pulmonary nodule. Diagnoses were established using the help of either histopathology or follow-up clinical results as a gold standard. In this prospective study, we enrolled 187 known malignancy patients with 200 suspicious newly developed pulmonary nodules. Their mean patient age was 48.4 ± 9.7 years. The studied 200 pulmonary nodular lesions were categorized using a nodule reporting system based on the lung-RADS into 6 sub-groups with 122 lesions found to be malignant and 78 lesions were of benign etiology, which showed a high sensitivity of 92.08%, specificity of 78.79%, and accuracy of 85.50% with 81.58% positive predictive value and 90.70% negative predictive value in the diagnosis of pulmonary nodules in cancer patients. Low-density CT with a nodule reporting system based on the lung-RADS classification system was found to be an accurate non-invasive tool to characterize and to risk stratify pulmonary nodules in oncology patients.


Background
In daily clinical practice, radiologists and pulmonologists are faced with incidental findings of pulmonary nodules that require an optimal decision in management and, meanwhile, to avoid morbidity and increased costs which are caused by invasive and unwarranted procedures for benign disease [1]. The differential diagnosis for pulmonary nodules is extremely broad, including both benign and malignant causes [2]. There is no consensus or guidelines regarding the investigation of pulmonary nodules in patients with extra-pulmonary malignancies. Retrospective data in populations of patients with cancer shows a high frequency of benign lesions (up to 58%) and primary lung cancers (up to 50%) found after biopsies or surgery [3]. Computed tomography (CT) is the most widely used modality for detecting lung metastases [4]. While the interpretation of CT scans for lung screening in pulmonary nodules remains challenging, various protocols and scoring systems have been developed [5]. The Reporting and Data System (lung-RADS) was introduced to standardize the reporting and management of patients undergoing screening to facilitate data collection and monitoring of patient outcomes [6]. In this system, a specific lung-RADS category is assigned to each low-dose computed tomography (LDCT) examination. Each category measures the likelihood of malignancy for the dominant finding and offers specific recommendations regarding management [7]. The primary goal of lung-RADS is to minimize variation in the management of detected lung nodules that can be implemented effectively in radiology practices outside the purview of a clinical trial [8].
The purpose of the work was to assess the accuracy of the "lung-RADS" classification system as a non-invasive tool for characterization of any newly developed pulmonary nodules among oncology patients using the help of either histopathology or follow-up clinical results as a gold standard.

Patients' demographic data
This prospective cross-sectional analytic study included 187 patients (109 females and 78 males) with a known history of malignancy who are under management and followed up with 200 newly developed pulmonary nodules that are suspicious of a metastatic etiology. The patient's age ranged from 21 to 78 years old with the mean age 48.4± 9.7 years.
Ethics committee approvals in addition to informed written consent were obtained from all patients.

Inclusion criteria
Patients with a known history of treated malignancy (either surgical, radio, or chemo-therapy) with newly developed pulmonary nodule/s on a follow-up CT chest study were included in the study.
Patients who had general contraindications for CT examination or respiratory distress that made them unable to tolerate breath-hold were excluded.

CT protocol and technique
The non-contrast LDCT study was performed using a 16-Multislice CT scanner (GE) Aquilion machine. All patients were trained to achieve a full end-inspiratory breath-hold before the start of the LDCT exam. They were scanned at the end of the full inspiration CT protocol: 120-140 kVp, 40-100 mA, and reconstructed section width 1.0 to 3.2 mm, with reconstruction scan intervals of 1-2.5 mm and a pitch of 1-2.

Post-processing
All CT images were magnified and reviewed by two radiologists who had prior experience with chest imaging and were informed of the patient's history, using GE PACS in axial images and reconstructed with soft tissue and lung windows. Axial maximum-intensity projections (16-2.5 mm). Coronal and sagittal multiplanar reformatted images were reconstructed and used for interpretation. The window setting for lung parenchyma used was between − 560 and − 1135 HU by the visual method to detect any pulmonary nodular lesions as well as the associated parenchymal abnormalities. According to the lung-RADS classification system (Table 1), nodules can be categorized into 1-4A, B, and X subgroups [9]. CT image interpretation was performed by two radiologists in consensus with adequate experience in CT lung screening using lung-RADS guidelines and nodule

Assessment of nodule size and density
On CT images, each pulmonary nodule was studied for: The lesion size is given in millimeters Leurion texture, which is expressed as soft tissue attenuation "solid," sub-solid, which could be divided into either pure ground-glass attenuation with pulmonary vessels and bronchi that may be seen within, or a semi-solid/part-solid that contains both ground glass and soft-tissue attenuation components. The categorization of a pulmonary nodule is important in that it predicts the likelihood of malignancy.

Histo-pathological and final clinical diagnosis as a gold standard
CT findings with the lung-RADS classification system were correlated by either histopathological results for very suspicious large malignant lesions or follow-up of final clinical diagnosis for benign featuring small nodular lesions in all patients. The nodules were classified as benign or malignant based on the final diagnosis.

Statistical analysis
The statistical calculations were done using the Statistical Package for Social Science, version 19, in which data is described in terms of range, mean, standard deviation, frequencies & percentages. Sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for CT results with a lung-RADS classification system for pulmonary nodules were calculated.

Results
This prospective study included 187 patients who had a positive history of treated cancer (83 breast cancer, 68 lung cancer, 19 prostate cancer, 12 GIT cancer, and 5 thyroid cancer) and were under management and follow-up with 200 newly developed pulmonary nodules that were suspicious of metastatic etiology. The patients' ages ranged from 21 to 78 years (M = 43.4 ± 9.7 years) ( Table 2).

Gold standard
The histopathological results for the biopsed103 nodules were as follows: 72 metastatic nodules, 9 Seq. cell Ca, 7 adeno-Ca, 6 alveller Ca, 3 small cell Ca, and 6 inflammatory lesions, while the final clinical diagnoses based on the clinical and follow-up (for 18-24 months) results of the remaining 97 inaccessible small/benign featuring lesions were as follows: 25 metastatic lesions that showed either increase in size or number in follow-up CT images. However, 72 lesions were categorized as benign nodules which were stationary on CT follow-up studies up to 2 years duration. Table 3 shows the classification of all pulmonary nodular lesions studied (no = 200) into two groups: group I: benign nodular lesions (78 lesions; 39%) and group II: malignant nodular lesions (122 lesions; 61%).

The diagnosis of lung RADS necessitates the examination of 200 nodules
Eighty-six lesions were diagnosed as benign (score 1-3), although 8 of them were found to be malignant  negative predictive value in the diagnosis of pulmonary nodules in cancer patients (Table 6).

Discussion
Pulmonary metastasis is seen in 20-54% of extrathoracic malignancies [9]. Lung-RADS reporting is standardized per CT lung screening, which accounts for pulmonary nodules and other findings concerning carcinoma, as well as any significant incidental findings [10,11]. The present study included 200 newly developed pulmonary nodules in 187 known cured cancer patients that were categorized by the lung-RADS calcification system into 6 sub-groups, 1-4A, 4B, and 4X, and we considered scores of 4A, 4B, and 4X as malignant. We found 86 lesions to be benign that were categorized as  , who conducted a similar study) on 228 patients and metastatic disease was found in 146 patients (64%). On biopsy, 60 patients (26.3%) were found to have a second primary lung tumor, while 22 patients (9.6%) were found to be cancer-free. Other series have also found different rates of malignant nodules on histologic analysis. In a large study with 1104 patients undergoing PNs resection, Mery et al. [12] observed a 63% malignancy rate in 337 patients with a previous cancer history. A lower rate by Khokhar et al. [13] discovered malignant pulmonary nodules in oncologic patients; 28% of small pulmonary  nodules detected on the initial CT will increase in size, suggesting metastasis. However, Reginald et al. [14] concluded that in the majority of clinical settings, an interval increase in the size of a single pulmonary nodule is considered evidence of metastasis, this size increase tends to occur early, and follow-up CT in 3 and 6 months would be appropriate for further evaluation. On the opposite hand, Meng et al. [15] divided ground-glass nodules into three groups: benign disease (inflammations), pre-invasive group, and invasive group (adenocarcinoma) in line with the pathologic findings of type IV, strongly suggesting a high likelihood of malignancy, while Kim et al. [16] stated that pulmonary subsolid nodules (SSNs) are frequently encountered within the screening CT with their main concern being lung adenocarcinoma and its precursors. Results from the National Lung Screening Trial showed that carcinoma screened by CT significantly reduced carcinoma mortality, and benign nodules were primarily confirmed by stability or shrinkage on repeat CT scans over a 2-year follow-up period [17]. Chung et al. [18] found that the typical false-positive rate was 7% for category 3 nodules, 7% for category 4A nodules, and 19% for category 4B nodules. Lung-RADS may be a tool that facilitates standardized reporting and management of abnormal findings at LCS CT [8].
Our findings revealed that the lung-RADS arrangement had a high sensitivity of 92.08%, specificity of 78.79%, and accuracy of 85.50% in the diagnosis of pulmonary nodules in oncology patients, with 81.58% positive predictive value and 90.70% negative predictive value. These results were in agreement with McKee et al. [19], who said that ACR lung-RADS reduced the general positive rate from 27.6 to 10.6%. No false negatives were present within the 152 patients with > 12-month follow-up reclassified as benign. Applying ACR lung-RADS increased the positive predictive value for diagnosed malignancy in 1603 patients with follow-up from 6.9 to 17.3% and by Kessle et al. [20], who concluded that lung-RADS produced 9 false-positives and 16 false-negative findings, whereas VRC with a 5% threshold resulted in 29 false-positives and 10 false-negative findings. Overall sensitivity and specificity for lung-RADS was 58.0% and 98.0%, and for VRC with a 5% threshold was 73.7% and 93.5%, respectively (P = .313, P = .001, respectively).

Limitations
Due to the relatively low number of lesions and the small size of most nodules, which made biopsy and histopathology difficult, we relied on a follow-up clinical diagnosis of up to 24 months, which is acceptable in oncology patients.

Conclusion
Low-density CT with a nodule reporting system based on the lung-RADS classification system was found to be an accurate non-invasive tool to characterize and to risk stratify pulmonary nodules in oncology patients.