What’s New in Cardiothoracic Imaging – July 2021

12 months ago


Application of N Descriptors Proposed by the International Association for the Study of Lung Cancer in Clinical Staging

Hwayoung Song, Sung Hyun Yoon, Junghoon Kim, Jihang Kim, Kyoung Won Lee, Woojoo Lee, Seungjae Lee, Kwhanmien Kim, Choon-Taek Lee, Jin-Haeng Chung, Kyung Hee Lee.

Radiology; 300:1–9, May 2021

Researchers from Seongam Citizen Medical Center in Korea sought to validate the newly proposed N stages as proposed by the International Association for the Study of Lung Cancer.  N classification for lung cancer staging was originally based on node location, while excluding tumor burden as a consideration.  The new proposed N system subdivides N1 and N2 into N1a, N1b, N2a1, N2a2, and N2b based on involved nodal locations and the presence of skip metastasis.  The proposed system is as follows: the cN1 stage was subdivided into cN1a (involvement of one N1 nodal station) and cN1b (multiple N1 nodal stations). The cN2 stage was divided into cN2a1 (one N2 nodal station without N1 involvement [skip metastasis]), cN2a2 (single N2 nodal station with N1 involvement), and cN2b (multiple N2 nodal stations).

This study is a retrospective observational study that drew subjects from a population of patients diagnosed with non-small cell lung cancer from January 2010 to December 2014.  2337 patients were initially discovered to have non-small cell lung cancer. Patients were excluded if there was no staging workup, distant metastasis, or previous malignancy with evidence of disease in the past 5 years. Ultimately, 1271 patients were included in the study.   Thoracic trained radiologists reviewed the available imaging studies for each patient and assigned a T and N grade based on the 8th edition of the TNM staging system.  Two clinical N stages were assigned, the first based on imaging alone, and a final based on all available evidence (PET-CT, EBUS, percutaneous biopsy, etc).  The final clinical stage was used for the primary analysis.  Nodal metastasis was considered if the short axis measured 1.0 cm.  Peripheral N1 LN (zone 12-14) was also considered metastatic 0.8 – 1.0 cm if it was located in the lymphatic pathway or if it was round, contrast-enhanced, or necrotic.  The study outcome was determined as overall survival.  Survival curves and 5-year overall survival rates were estimated using the Kaplan-Meier method. The distribution of patients into the N stages are as follows: N0 940 patients (74% of all patients), N1 86 (7%), N2 162 (13%), and N3 83 (7%).

Overall survival rates in accordance with the current staging system were 77.3%, 48.3%, 27.1%, and 12.4% for stages cN0, cN1, cN2, and cN3, respectively. The differences in survival between adjacent groups (for example N0 and N1, N1 and N2, etc.) were statistically significant P value <0.001.  The overall survival rates in accordance with the proposed system were  77.3%, 53.7%, 36.0%, 29.2%, 34.4%, 18.0%, and 12.4% for stages cN0, cN1a, cN1b, cN2a1, cN2a2, cN2b, and cN3, respectively. Differences between adjacent stages were not statistically significant for survival, except for N0 and N1a (P=0.002).  When the N2 stage was subdivided; rates were 77.3%, 48.3%, 32.0%, 18.0%, and 12.4% for stages cN0, cN1, cN2a, cN2b, and cN3, respectively.  All were statistically significant except for cN2b and cN3.  Within the N1 stage, lymph node size >2 cm was shown to be associated with a worse prognosis, with no difference with size in the N2 stage.

Ultimately, It appears that subdividing N2 stages in regards to the clinical staging of N2a and N2b can show a significant prognostic effect, but further dividing in the N2a1 and N2a2 is not significant for prognosis.  Additionally, in patients with N1 disease, lymph node size is important as it has prognostic implications.  A notable important finding was that the N2b prognosis was closer to N3 disease than N2a.

Limitations for this study include lack of generalizability across an ethnically diverse population as the study population was an ethnically homogenous Asian cohort.  Additionally, this was a single intuition study with a smaller number of subjects per variable N stage.  A potential confounder in this population is also a higher prevalence of tuberculosis potentially affecting PET/CT results.


Lung-RADS Version 1.0 versus Lung-RADS Version 1.1: Comparison of Categories Using Nodules from the National Lung Screening Trial

Julia Kastner, Rydhwana Hossain, Jean Jeudy, Farouk Dako, Varun Mehta, Sandeep Dalal, Ekta Dharaiya, Charles White.

Radiology; 300:1, May 2021

Researchers from the University of Maryland investigated the classification changes seen from Lung-RADS 1.0 to 1.1, to determine the frequency of perifissural and ground glass nodules that were declassified from categories 3/4a to the more benign category 2. Lung-RADS is a useful tool that is used to standardize results from low-dose lung cancer screening CT across the United States.  Perifissural nodules threshold size increased from 6 mm to 10 mm, and ground glass nodules from 20 mm to 30 mm between Lung-RADS v1.0 and v1.1.  Subjects were obtained from the NLST clinical trial and 2813 patients were included within the cohort as they had an examination that included a perifissural nodule that measured >6 mm but <10 mm. Nodules were classified using the de Hoop classification system; typical PFN (intrapulmonary lymph node), atypical PFN (PFN likely, meets most of criteria), or a non-PFN (parenchymal or pleural).  For this study, typical and atypical PFN was considered to meet the criteria for PFN.  Nodules were assessed by a 4th-year medical student with supervision from a cardiothoracic radiologist and each nodule was organized by its characteristics.  NLST data sheets were reviewed to determine which nodules proved to be malignant, which required pathologic proof.  Benign nodules required stability over a follow-up period or pathologic proof.

1092 solid nodules across all studies were used for analysis that fell within the 6-10 mm range. 772 (70.7%) were considered category 3, and 320 (29.3%) were considered category 4a.  216 of these nodules were considered PFNs when Lung-RADS v 1.1 was applied and downgraded to category 2.  11 total solid nodules within the 6-10 mm range were found to be malignant, however, none judged to be a PFN and down-categorized to category 2 were deemed to be malignant. LungRADS v1.1 was found to be statistically significantly better for detecting malignancy with solid nodules than v1.0 (p <0.01).  161 ground glass nodules were present on scans included in the study, 158 were category 2, and 3 in category 3 using v1.0 criteria.  2 nodules in category 3 were downgraded to category 2 per v1.1 criteria. 4 total GGN was found to be malignant; 2 in each category 2 and 3 using v1.0, and one of the downgraded nodules using the v1.1 criteria was found to be malignant.  No significant difference was found between the 2 versions for detecting malignant nodules (P = 0.48). Overall, the Lung-RADS v1.1 was found to perform better in detecting malignancy when compared to v1.0 (P <0.01).  The researchers also extrapolated their data set to apply it to the entire NLST trial sata and suggest that with v1.1, 3.4% of studies would be downgraded reducing the false positive rate.

Limitations for this study include the fact that it is a subset of the total data set of the NLST, although with the overwhelming number of patients in the NLST, an analysis of each patient would be inefficient.  Technique issues between scans could also lead to problems as different CT scanners can perform studies differently even with similar base parameters.

Central Tumor Location at Chest CT Is an Adverse Prognostic Factor for Disease-Free Survival of Node-Negative Early-Stage Lung Adenocarcinomas

Hyewon Choi, Hyungjin Kim, Chang Min Park, Young Tae Kim, Jin Mo Goo.

Radiology 2021; 299:438–447

Researchers for Seoul National University Hospital in Korea sought to investigate the role of the two proposed definitions of central lung cancers in regards to their prognostic value and prognostic implications of a quantitative definition of central lung cancer (CLC). Central lung cancer is associated with mediastinal node metastasis and occult N2 disease, however, a uniform definition for CLC is lacking.  The 2 proposals are as follows: 1) inner ⅓ of the lung as defined as concentric lines from the hilum; or 2) inner ⅓ as defined by concentric lines from the midline based on axial CT images.

This particular study was a retrospective study that examined patients from 2009-2015 at Seoul University Hospital with a pathologic stage of T1a-bN0M0.  Exclusion criteria included: sublobar resection, no mediastinal lymph node dissection, non-adenocarcinoma histologic type or adenocarcinoma in situ, synchronous or metachronous lung cancer, no survivable data, or missing clinical variables.  The pathologic stage was based on the 7th edition of the American Joint Commission on Cancer staging, however, the T stage was determined using the 8th edition with retrospective measurement of the cancer lesions by the researchers.  The primary endpoint was defined as disease-free survival-DFS (disease defined as date from surgery to first clinical recurrence and/or distant metastasis) as confirmed with imaging.  The researchers used the 2 above discussed proposals, as well as developing a third definition using a quantitative method to reduce the potential of subjectivity with the first 2 proposals. The quantitative method used an intersecting point between the horizontal midline of the thorax and mediastinum as the central point and deriving the thirds of the lung from this point.  Definition 4 is a subset of definition 3 and measurements included the inner third of the lung, definition 5 was similar but used the inner two-thirds of the lung as central. Tumors were assessed for each of the three definitions by thoracic trained radiologists. 436 total patients were used for this study.  Clinical history and demographics were also obtained and included in multivariate analysis. Interobserver agreement used the Cohen K test.

CLC was found in 34 patients using definition 1, and 3-year DFS was 87.4% for central; 93.7% for peripheral tumors.  CLC was found in 55 patients using definition 2, and 3-year DFS was 92.1% for central, 93.4% for peripheral tumors.  Definition 4 (using the quantitative inner third of the lung) had a total of 10 patients with no events.  Definition 5 had a total of 130 patients with a 3-year DFS of 89%.  Multivariate analysis showed that definition 1 and 5 CLCs had an almost threefold higher risk of recurrence or death when compared to the peripheral counterpart.  Interaction terms between each definition and other variables were not significant and thus not included in the models. Interobserver agreement for definition 1 was 0.52 consistent with moderate agreement, and 0.86 for definition 5 consistent with almost perfect agreement.  Overall, this study validates the prognostic value for disease-free survival using both qualitative and quantitative measurements in patients with resected, node-negative, early-stage adenocarcinomas.  CLC was shown to be adversely related to DFS compared to peripheral counterparts.

Limitations for this study are mostly revolved around the data set.  This was a retrospective study from a single institution with a small number of patients with limited events.  A larger multicenter study could validate these initial findings and provide more concrete evidence showing the results and help form a universal definition for CLC.


Long-Term Prognostic Value of Coronary CTA in Orthotopic Heart Transplant Recipients

Alejandra Garcia-Baizan, Meylin Caballeros, Ana Ezponda, Rebeca Manrique, Juan J. Gavira, Gregorio Rabago, Gorka Bastarrika.

American Journal of Roentgenology; 216(5): 1216-1221, May 2021

A group of researchers from Pamplona and Madrid sought to investigate the prognostic value of coronary CT angiography (CCTA) in patients that have undergone orthotopic heart transplants.  Cardiac allograft vasculopathy (CAV) is the leading cause of graft loss between 1-3 years after transplant and is a major factor in long-term survival. Conventional coronary angiography is the accepted standard for CAV, but limited as you can not see the vessel lumen.  CCTA has a potential to be a noninvasive alternative to traditional angiography.

This was a retrospective study that included 114 patients that underwent CCTA for CAV.  Subjects were included if there was no clinical evidence of CAV at the time of CCTA, and prior to the CCTA had negative echocardiography or conventional angiography.  Retrospectively ECG-gated CCTA examinations were acquired with the following parameters: tube voltage, 120 kV; gantry rotation time, 330 ms; detector collimation, 32 × 0.6 mm; slice acquisition, 64 × 0.6 mm; and variable pitch, 0.2–0.45, automatically adapted to the heart rate.  All studies were performed on a Siemens somatom scanner.CCTA examinations were performed after IV injection of 70 mL of contrast medium (400 mg I/mL, Iomeron 400, Bracco) using a dual-head power injector (CT Stellant, Medrad). Contrast material was injected at a flow rate of 5 mL/s, followed by a 50-mL saline flush. The bolus-tracking technique with the ROI placed in the ascending aorta, a 100-HU predefined trigger threshold, and a fixed delay of 6 seconds was used for data acquisition.  Patient demographics, risk factors for CAD, infections, and immunosuppression data were collected for analysis.  Studies were rated semiquantitatively as 0%, 1–24%, 25–49%, 50– 69%, 70–99%, and 100% stenosis or occlusion. CCTA studies were classified in two groups according to the presence of nonobstructive CAV (defined as lumen reduction < 50%) or obstructive disease (defined as a lumen reduction ≥ 50%). The primary endpoint of this study was the composite endpoint of major adverse cardiovascular events (MACEs), including cardiac death, nonfatal myocardial infarction, cardiac arrhythmia, unstable angina requiring hospitalization, congestive heart failure, revascularization (percutaneous coronary intervention or coronary artery bypass grafting), stroke, or retransplant. The time from CCTA examination to MACE was recorded.

The nonobstructive disease was visualized in 102 pts (89.5%) and obstructive disease was detected in 12 pts (10.5%).  55 patients had 0% stenosis (48.2%), 29 (25.4%) presented 1–24% stenosis, 18 (15.8%) presented 25–49% stenosis, 10 (8.8%) presented 50–69% stenosis, one (0.9%) presented 70–99% stenosis, and one (0.9%) presented total occlusion. A total of 21 patients developed a MACE during the follow-up period; significantly more in the obstructive group (50% vs 14.7% in nonobstructive; p <0.05).  The presence of 4 or more stenotic coronary segments was associated with an increased risk of MACE (12/21 vs 9/21; p <0.008).  CCTA showed high specificity (93.5%) and NPV (85.2%) to predict long-term MACEs.

This study of course has its limitations.  It is a retrospective, single-center study with a relatively small data cohort. Further evaluation with a larger population could produce different results.  But overall, this study is a good start for risk stratification in heart transplant patients and the use of CCTA can be a great non-invasive alternative to conventional angiography.

  • Share