Radiol med (2014) 119:20–26 DOI 10.1007/s11547-013-0303-x

CHEST RADIOLOGY

Extrapleural and cardiophrenic lymph nodes: prevalence, clinical significance and diagnostic value Beatrice Feragalli • Cesare Mantini • Nicoletta Civitareale • Roberta Polverosi • Armando Tartaro • Antonio Raffaele Cotroneo

Received: 22 February 2012 / Accepted: 4 April 2012 / Published online: 19 November 2013 Ó Italian Society of Medical Radiology 2013

Abstract Purpose The aim of this study was to evaluate the prevalence, clinical meaning and diagnostic value of extrapleural and cardiophrenic nodes occasionally observed on computed tomography (CT) scans of the chest. Materials and methods We included 750 consecutive patients who underwent CT of the chest for different clinical purposes (340 nonneoplastic patients, 270 with extrathoracic neoplasms, 120 with intrathoracic neoplasms, 20 with pleural metastasis) and 91 patients with histologically proven malignant pleural mesothelioma (MPM). For each group of patients, we analysed the presence of extrapleural and cardiophrenic nodes, their number (single or multiple) and their size. Results The prevalence of cardiophrenic nodes between 6 and 10 mm and [10 mm was significantly higher in patients with MPM (28.6 and 26.4 %, respectively) than in all other categories of patients, except for patients with pleural metastasis (30 and 25 %, respectively). The prevalence of extrapleural nodes, independently from their size, was significantly higher in patients with MPM (68 %)

B. Feragalli (&) Department of Medical, Oral and Biotechnological Sciences, ‘‘G. D’Annunzio’’ University, Via dei Vestini, 66013 Chieti, Italy e-mail: [email protected] C. Mantini  N. Civitareale  A. Tartaro  A. R. Cotroneo Section of Diagnostic Imaging and Therapy, Radiology Division, Department of Neuroscience and Imaging, ‘‘G. D’Annunzio’’ University, Chieti, Italy R. Polverosi Veneto Institute of Oncology IOV, IRCCS Padua, Via Gattamelata 64, 35128 Padua, Italy

123

compared with all other groups, including patients with pleural metastasis (5 %; p \ 0.0001). Conclusions Cardiophrenic nodes [5 mm and extrapleural nodes of any size have a significant diagnostic value in malignant pleural disease, either primary or secondary, whereas they are extremely rare in other neoplastic or nonneoplastic diseases. Keywords Lymph nodes  Multidetector-row computed tomography  Neoplastic disease  Mesothelioma

Introduction Thoracic lymph nodes are commonly named using descriptive terminology according to the blood vessels or visceral structures to which they are most closely related or by their general anatomical location. Although there are slight differences in the classification of thoracic nodes [1, 2], these can be divided into visceral and parietal lymph nodes. The visceral group receives lymphatic drainage from pulmonary and tracheobronchial structures and include the anterior and posterior mediastinal nodes as well as the tracheobronchial nodes. The parietal group receives lymphatic drainage from musculoskeletal and articular structures and include cardiophrenic, extrapleural, sternal, internal mammary and intercostal nodes [3]. The object of our study was to investigate the clinical meaning and diagnostic value of cardiophrenic and extrapleural lymph nodes. Cardiophrenic nodes are located in the fatty tissue of the most basal portion of the mediastinum, bordered by the base of the heart, the diaphragm and the chest wall (Fig. 1) [4]. Depending on their position relative to the heart, they are classified into (1) an anterior group situated between

Radiol med (2014) 119:20–26

21

Fig. 1 Axial computed tomography (CT) scan showing location of cardiophrenic lymph nodes. These nodes are classified into an anterior group (white), a middle or lateral group (grey) and a posterior group (black). ICV inferior cava vein, DA descending aorta

Fig. 3 Axial computed tomography (CT) scan showing the position of the extrapleural lymph nodes (white) in adipose paraspinal tissue near the costovertebral joints

Fig. 2 Axial computed tomography (CT) scan showing the pericardium (arrow) as a relatively hyperdense line that separates the epicardial adipose tissue (?) from the cardiophrenic space (asterisk)

They receive afferent vessels from the parietal pleura, diaphragm, liver and abdominal wall [6, 7]. Extrapleural lymph nodes are located in a triangular area at the costovertebral joint, near the head of the rib, in the adipose paraspinal tissue [7]. This region, located in the lower part of the mediastinum and called the retrocrural space, is bordered anteriorly and laterally by the diaphragmatic pillars and posteriorly by the body of the last thoracic and the first lumbar vertebrae (Fig. 3). The retrocrural space communicates with the posterior mediastinum above and with the retroperitoneal space below, without clear anatomical limitations, representing a potential conduit for the spread of pathological processes between thorax and abdomen [8]. Lymphatics of the posterior mediastinum, the posterior portion of the diaphragm, the spine and the posterior parietal pleura drain into the extrapleural lymph nodes; those of the anterior parietal pleura drain into the internal mammary chain for the upper and middle part of the thorax and into the cardiophrenic lymph nodes for the lower part [7]. Several neoplastic and nonneoplastic diseases can spread easily from thoracic organs to the abdomen along these lymphatic chains [9–12]. Thoracic lymph nodes can be evaluated with various methods, such as computed tomography (CT), magnetic resonance imaging (MRI) and fluorodeoxyglucose-positron emission tomography (FDG-PET). CT is the primary noninvasive technique for diagnostic evaluation of thoracic lymph nodes. Lymph node abnormalities are depicted by CT and MRI as an increase in nodal size and/or number, and patterns of thoracic lymph node involvement can provide important clues in the diagnosis of many pulmonary and extrapulmonary diseases. CT criteria used in distinguishing pathological from normal nodes are based

the pericardium posteriorly and the xiphoid process of the sternum and costochondral articulation of the seventh rib anteriorly; (2) a middle or lateral group located between the pericardium medially and lung hilum laterally in the region where the phrenic nerves run; and (3) a posterior group located near the oesophagus, posteromedial to the inferior vena cava, which goes down to the aortic hiatus [5]. Most diseases of the cardiophrenic space affect the anterior compartment. When evaluating this space, the observer must pay attention to visualising the pericardium, which separates the epicardial adipose tissue from the cardiophrenic space (Fig. 2). Cardiophrenic lymph nodes have a very complex drainage system, involving thoracic and abdominal organs.

123

22

on their size. Generally, lymph nodes with a short-axis diameter [1 cm are considered pathological [13, 14]. There are little data on the normal size of cardiophrenic and extrapleural lymph nodes, for which no definite cutoff value has yet been established.

Materials and methods Patient population We included in our study 750 consecutive patients (430 men, 320 women; mean age, 59 ± 6 years) referred to our institution between January and September 2009 for multidetector-row CT (MDCT) of the chest for different clinical reasons. According to their primary disease, these patients were divided into four groups: 340 nonneoplastic patients, 270 patients with extrathoracic neoplasms, 120 patients with intrathoracic neoplasms and 20 patients with pleural metastasis. We also included 91 patients (61 men, 30 women; mean age, 67 ± 9 years) with histologically proven malignant pleural mesothelioma (MPM) who had undergone MDCT of the chest at our institute between January 2001 and September 2009.

Radiol med (2014) 119:20–26

(single or multiple) and short-axis diameter (B5 mm; 6–10 mm; [10 mm) of the extrapleural and cardiophrenic nodes. In the presence of multiple lymph nodes, the largest was considered indicative. In the event of disagreement between the two readers, the final evaluation was obtained by consensus. Images were analysed using a mediastinal window (width, 400 HU; level, 40 HU). For statistical purposes, the prevalence of extrapleural and cardiophrenic nodes of each size category (B5 mm; 6–10 mm; [10 mm) was calculated in patients with MPM and in each patient group (nonneoplastic, extrathoracic neoplasm, intrathoracic neoplasm, and pleural metastasis). Comparison of the prevalence between patients with MPM and the remaining four groups was done using the Chi square test, with significance set at p \ 0.05. Interobserver agreement was analysed using kappa statistics.

Results The 750 patients in the study comprised 340 nonneoplastic, 270 with extrathoracic neoplasm, 120 with intrathoracic neoplasm and 20 with pleural metastasis. Cardiophrenic lymph nodes

Scan protocol and image acquisition CT studies were performed on a four-channel MDCT (Somatom Volume Zoom, Siemens Medical Solutions, Forchheim, Germany) using standard parameters for MDCT of the chest: 120 kV; effective 400 mA; gantry rotation time, 500 ms; table speed, 6 mm/rotation. The examination was performed with or without administration of contrast medium according to the clinical question and the presence of contraindications. In all patients, acquisition was extended from the pulmonary apex to the adrenal glands in a single breath-hold using thin collimations (4 9 1.25 mm). Images were reconstructed with a slice width of 1 mm and reconstruction interval of 1 mm. Multiplanar reconstructions were obtained in selected cases, especially in MPM patients, for staging purposes. A dose of 90–100 ml of nonionic contrast medium (Iomeron 300 mg iodine/ml, Bracco, Milan, Italy) was injected via an antecubital vein at 2–3 ml/s. A standard scan delay of 25 s was selected in all patients; in patients with MPM, we performed another acquisition with a 45- to 50-s scan delay to obtain better enhancement of pleural surfaces and disease. Imaging evaluation and statistical analysis Axial CT images were independently evaluated by two chest radiologists who assessed the presence, number

123

Results regarding cardiophrenic lymph nodes are summarised in Table 1. Cardiophrenic nodes, independent of their size, were found in 66 (8.8 %) of 750 patients assessed; in particular, they were B5 mm in 47 patients, 6–10 mm in 15 and [10 mm in the remaining four. Considering each category of patients, cardiophrenic nodes were present in 26 of 340 (7.6 %) nonneoplastic patients (B5 mm in 19, 6–10 mm in five, [10 mm in two), in 22 of 270 (8.1 %) patients with extrathoracic neoplasm (B5 mm in 15, 6–10 mm in six, [10 mm in one), in 16 of 120 (13.3 %) patients with intrathoracic neoplasm (B5 mm in 12, 6–10 mm in three, [10 mm in one) and in 13 of 20 (65 %) patients with pleural metastasis (B5 mm in two, 6–10 mm in six, [10 mm in five). Among the 91 patients with MPM, cardiophrenic nodes were present in 60 (66 %): in particular, ten (11 %) were B5 mm, 26 (28.6 %) 6–10 mm and 24 (26.4 %) [ 10 mm. Regarding the prevalence of nodes B5 mm, no significant differences (p [ 0.05) were found between patients with MPM (11 %) and the other four groups (5.5 % in patients with extrathoracic neoplasm, 10 % in patients with intrathoracic neoplasm and 10 % in patients with pleural metastasis), including patients without neoplasm (5.6 %) (Fig. 4). The prevalence of cardiophrenic nodes measuring 6–10 mm was significantly higher in patients with MPM (28.6 %) (Fig. 5) than in any other category (1.5 % in nonneoplastic, 2.2 % in extrathoracic, 2.5 % in intrathoracic neoplasm)

Radiol med (2014) 119:20–26

23

Table 1 Results regarding cardiophrenic nodes Any size Nonneoplastic pts. (n = 340)

B5 mm

6–10 mm

[10 mm

26 (7.6 %)

19 (5.6 %)

5 (1.5 %)

2 (0.6 %)

p \ 0.0001

p = 0.1116

p \ 0.0001

p \ 0.0001

22 (8.1 %)

15 (5.5 %)

6 (2.2 %)

1 (0.4 %)

p \ 0.0001

p = 0.0932

p \ 0.0001

p \ 0.0001

16 (13.3 %)

12 (10 %)

3 (2.5 %)

1 (0.8 %)

p \ 0.0001

p = 0.9957

p \ 0.0001

p \ 0.0001

13 (65 %)

2 (10 %)

6 (30 %)

5 (25 %)

p = 0.8567

p = 0.7882

p = 0.8848

p = 0.8772

Total [ pts. (n = 750)

66 (8.8 %)

47 (6.3 %)

15 (2 %)

4 (0.5 %)

Pts. with MPM (n = 91)

60 (66 %)

10 (11 %)

26 (28.6 %)

24 (26.4 %)

Pts. with extrathoracic neoplasm (n = 270) Pts. with intrathoracic neoplasm (n = 120) Pts. with pleural metastasis (n = 20)

Pts patients, MPM malignant pleural mesothelioma

except for patients with pleural metastasis, in whom prevalence was 30 %. Finally, the prevalence of cardiophrenic nodes [10 mm was very low in all patient categories (0.6 % in nonneoplastic, 0.4 % in extrathoracic, 0.8 % in intrathoracic neoplasm) except for those with pleural metastasis (25 %) and MPM (26.4 %). Extrapleural lymph nodes

Fig. 4 Axial computed tomography (CT) scan showing three lymph nodes \5 mm in the right cardiophrenic angle (arrows) in a 52-yearold woman without known disease

Results regarding extrapleural lymph nodes are summarised in Table 2. Extrapleural nodes, independent of their size, were present in only eight (1.3 %) of 750 patients assessed; in particular, they were B5 mm in four patients, 6–10 mm in two and [10 mm in the remaining two. Considering each category of patients, extrapleural lymph nodes were present in three of 340 (0.9 %) nonneoplastic patients (B5 mm in one, 6–10 mm in one, [10 mm in one), in two of 270 (0.7 %) patients with extrathoracic neoplasm (both B5 mm), in two of 120 (1.6 %) patients

Fig. 5 Axial computed tomography (CT) scans showing the presence of nodes [5 mm in the right cardiophrenic space in a patient with right pleural effusion related to malignant pleural mesothelioma (MPM)

123

24

Radiol med (2014) 119:20–26

Table 2 Results regarding extrapleural nodes B5 mm

6–10 mm

[10 mm

3 (0.9 %)

1 (0.3 %)

1 (0.3 %)

1 (0.3 %)

p \ 0.0001

p \ 0.0001

p \ 0.0001

p \ 0.0001

Pts. with extrathoracic neoplasm (n = 270)

2 (0.7 %)

2 (0.7 %)

0

0

p \ 0.0001

p \ 0.0001

p \ 0.0001

p \ 0.0001

Pts. with intrathoracic neoplasm (n = 120)

2 (1.6 %)

1 (0.83 %)

1 (0.83 %)

0

p \ 0.0001

p \ 0.0001

p \ 0.0001

p \ 0.0001

1 (5 %)

0

1 (5 %)

0

p \ 0.0001

p = 0.0661

p = 0.0633

p = 0.0553

Total pts. (n = 750)

8 (1.3 %)

4 (0.5 %)

2 (0.3 %)

2 (0.3 %)

Pts. with MPM (n = 91)

62 (68 %)

18 (20 %)

25 (27 %)

19 (21 %)

Any size Nonneoplastic pts. (n = 340)

Pts. with pleural metastasis (n = 20)

Pts patients, MPM malignant pleural mesothelioma

Fig. 6 Axial and coronal computed tomography (CT) images show typical extrapleural nodes located in the paraspinal extrapleural fat tissue adjacent to the rib heads in a patient with circumferential pleural involvement due to malignant pleural mesothelioma (MPM)

with intrathoracic neoplasm (B5 mm in one, 6–10 mm in one) and in one of 20 (5 %) patients with pleural metastasis (6–10 mm). Among 91 patients with MPM, extrapleural nodes were found in 62 (68 %); in particular, 18 (20 %) B 5 mm, 25 (27 %) 6–10 mm and 19 (21 %) [ 10 mm. These findings indicate a significantly higher prevalence of extrapleural nodes of any size in patients with MPM (Fig. 6) compared with all other groups (p \ 0.01), including patients with pleural metastasis.

Discussion Intrathoracic lymph nodes, particularly hilar and mediastinal, have been assessed in several scientific papers evaluating the diagnostic accuracy of CT in defining nodal involvement by different diseases, in particular, intrathoracic neoplasms. Despite the well-known limits, node size is the main criterion used to evaluate the possible involvement of these nodes [13]: hilar and mediastinal nodes are considered involved if the short-axis diameter is C10 mm [15, 16]. Whereas this dimensional criterion is

123

well accepted for hilar and mediastinal nodes, to our knowledge, no cutoff value has been established to date for cardiophrenic and extrapleural nodes. On the basis of the study by Dorfman et al., cardiophrenic nodes with a short-axis diameter [8 mm should be considered pathological [17], but other studies indicate a different cutoff of 5 mm [18]. No standard dimensional criteria are available for extrapleural and internal mammary lymph nodes, which are always considered pathological when present, even when the short-axis diameter is \5 mm [9]. Several papers about nodal involvement in MPM have recently been published. In particular, Abdel Rahman et al. [19] assessed the prevalence and pattern of nodal metastases in 53 patients with MPM. They concluded that hilar node involvement, in contrast with the TNM staging system of MPM, is secondary to parenchymal infiltration and not due to direct spread from the pleura, whereas extrapleural and cardiophrenic nodes are primarily involved in the presence of MPM. Seely et al. [20] conducted a retrospective study on 92 patients with a histological diagnosis of MPM and found a prevalence of cardiophrenic

Radiol med (2014) 119:20–26

nodes of 46 % (42/92). This study did not, however, evaluate extrapleural nodes. Our department conducted a study on 91 patients with histologically proven MPM to verify the frequency of early CT signs of this neoplasm. The most frequent CT signs in patients with MPM were diffuse or focal thickening of the mediastinal pleura (96 %), thickening of the costal pleura (87 %), unilateral pleural effusion (85 %), volume loss of the pathological hemithorax (80 %) and presence of extrapleural (68 %) and cardiophrenic (67 %) lymphadenopathy. Involvement of hilar and mediastinal nodes was very uncommon. We concluded that the presence of extrapleural and cardiophrenic nodes, associated with other signs of MPM and especially involvement of the mediastinal pleura and volume loss of the affected hemithorax, could have an important diagnostic role in the early diagnosis of this neoplasm. The few published reports concerning involvement of cardiophrenic and extrapleural nodes led us to investigate their prevalence in patients with different neoplastic and nonneoplastic diseases and to define a dimensional cutoff value for these nodes. Our results demonstrate that extrapleural nodes of any size are extremely rare in nonneoplastic patients and patients with intrathoracic or extrathoracic neoplasm, including those with pleural metastasis. The overall prevalence of these nodes was 1.3 %, with a minimum value of 0.7 % in patients with extrathoracic neoplasm and a maximum value of 5 % in patients with pleural metastasis. By contrast, the prevalence of these nodes was significantly higher in patients with MPM; in particular, 62 of 91 patients with MPM (68 %) showed involvement of extrapleural nodes, 18 (20 %) with a diameter B5 mm, 25 (27 %), with a diameter 6–10 mm and 19 (21 %) with a diameter [10 mm. This means that extrapleural nodes should be considered pathological independently of their size, even when \5 mm; considering that they are extremely rare in all categories of patients except those with MPM, these nodes could have an important diagnostic value in MPM patients when associated with other findings of this disease, such as unilateral pleural effusion, hemithorax volume loss and mediastinal pleura thickening. Regarding the prevalence of cardiophrenic nodes B5 mm, we found no significant differences between patients with MPM and the other four groups, including patients without neoplasm. When we consider nodes [5 mm, the prevalence of these nodes in MPM patients was significantly higher than in any other group, except for patients with pleural metastasis. This means that cardiophrenic nodes, unlike extrapleural nodes, should be considered pathological only when they are [5 mm in size. Because they are enlarged both in MPM and in pleural metastasis, these nodes could have an important role in the

25

diagnosis of malignant pleural diseases, whether primary or secondary.

Conclusion Extrapleural nodes of B5 mm and cardiophrenic nodes of [5 mm have significant diagnostic value in both primary and secondary malignant pleural disease, whereas they are extremely rare in other diseases, whether neoplastic or not. Conflict of interest B Feragalli, C. Mantini, N. Civitareale, R. Polverosi, A. Tartaro, A.R. Cotroneo declare no conflict of interest.

References 1. Johnson D, Ellis H (2005) Pectoral girdle and upper limb. In: Standring S (ed) Gray’s anatomy, 39th edn. Churchill Livingstone, Edinburgh, pp 817–849 2. Fraser RS, Mu¨ller NL, Colman N, Pare0 PD (1999) The lymphatic system of the lungs, pleura, and mediastinum. In: Fraser RS, Mu¨ller NL, Colman N, Pare0 PD (eds) Fraser and Pare0 ’s diagnosis of diseases of the chest, 4th edn. W.B. Saunders Company, Philadelphia, pp 172–195 3. Baldoni GC, Bastianini A (2000) Anatomia umana (Vol. 1). EdiHermes, Edmond 4. Pineda V, Andreau J, Ca`ceres J et al (2007) Lesions of cardiophrenic space: findings at cross-sectional imaging. Radiographics 27:19–32 5. Aromberg DJ, Peterson RR, Glazer HS, Sagel SS (1986) Superior diaphragmatic lymph nodes: CT assessment. JCAT 10(6): 937–941 6. Vock P, Hodler J (1986) Cardiophrenic angle adenopathy: update of causes and significance. Radiology 159:395–399 7. Light RW (2001) Pleural disease, 4th edn. Lippincott Williams & Wilkins, Philadelphia, pp 1–7 8. Restrepo C, Eraso A, Ocaziones D, Lemos J, Martinez S, Lemos DF (2008) The diaphragmatic crura and retrocrural space: normal imaging appearance, variants, and pathologic conditions. Radiographics 28:1289–1305 9. Sharma A, Fidias P, Hayman LA, Loomis LS, Taber KH, Aquino SD (2004) Patterns of lymphadenopathy in thoracic malignancies. Radiographics 24:419–434 10. Antman KH, Pass HI, Schiff PB (2001) Malignant mesothelioma. In: DeVita VT, Hellman S, Rosenberg SA (eds) Cancer: principles and practice of oncology. Lippincott Williams & Wilkins, New York, pp 1943–1969 11. Einstein DM, Singer AA, Chilcote WA, Desai RK (1991) Abdominal lymphadenopathy: spectrum of CT findings. Radiographics 11:457–472 12. Komaya T, Ueda H, Togashi K, Umeoka S, Kataoka M, Nagai S (2004) Radiologic manifestation of sarcoidosis in various organs. Radiographics 24:87–104 13. McLoud TC, Bourgouin PM, Greenberg RW et al (1992) Bronchogenic carcinoma: analysis of staging in the mediastinum with CT by correlative lymph node mapping and sampling. Radiology 182:319–323 14. Pieterman RM, van Putten JWG, Meuzelaar JJ et al (2000) Preoperative staging of non-small cell lung cancer with positron emission tomography. N Engl J Med 343:254–261

123

26 15. Glazer GM, Gross BH, Quint LE et al (1985) Normal mediastinal lymph nodes: number and size according to American Thoracic Society mapping. AJR Am J Roentgenol 144:261–265 16. Toloza EM, Harpole L, McCrory DC (2003) Non invasive staging of non-small cell lung cancer: a review of the current evidence. Chest 123:137S–146S 17. Dorfman RE, Alpern MB, Gross BH, Sandler MA (1991) Upper abdominal lymph nodes: criteria for normal size determined with CT. Radiology 180:319–322

123

Radiol med (2014) 119:20–26 18. Graham NJ, Libshitz HI (1995) Cascade of metastatic colorectal carcinoma from liver to the anterior diaphragmatic lymph nodes. Acad Radiol 2:282–285 19. Abdel Rahman AR, Gaafar RM, Baki HA et al (2008) Prevalence and pattern of lymph node metastasis in malignant pleural mesothelioma. Ann Thorac Surg 86(2):391–395 20. Seely JM, Nguyen ET, Churg AM, Mu¨ller NL (2009) Malignant pleural mesothelioma: computed tomography and correlation with histology. Eur J Radiol 70(3):485–491