Iodine concentration of healthy lymph nodes of neck,axilla,and groin in dual-energy computed tomography

Abstract

Background

Lymph nodes (LN) are examined in every computed tomography (CT) scan. Until now, an evaluation is only possible based on morphological criteria. With dual-energy CT (DECT) systems, iodine concentration (IC) can be measured which could conduct in an improved diagnostic evaluation of LNs.

Purpose

To define standard values for IC of cervical, axillary, and inguinal LNs in DECT.

Material and Methods

Imaging data of 297 patients who received a DECT scan of the neck, thorax, abdomen–pelvis, or a combination of those in a portal-venous phase were retrospectively collected from the institutional PACS. No present history of malignancy, inflammation, or trauma in the examined region was present. For each examined region, the data of 99 patients were used. The IC of the three largest LNs, the main artery, the main vein, and a local muscle of the examined area was measured, respectively.

Results

Normalization of the IC of LNs to the artery, vein, muscle, or a combination of those did not lead to a decreased value-range. The smallest range and confidence interval (CI) of IC was found when using absolute values of IC for each region. Hereby, mean values (95% CI) for IC of LN were found: 2.09 mg/mL (2.00–2.18 mg/mL) for neck, 1.24 mg/mL (1.16–1.33 mg/mL) for axilla, and 1.11 mg/mL (1.04–1.17 mg/mL) for groin.

Conclusion

The present study suggests standard values for IC of LNs in dual-layer CT could be used to differentiate between healthy and pathological lymph nodes, considering the used contrast injection protocol.

Keywords

Lymph nodes X-ray computed tomography iodine radiologic technology neoplasms

Introduction

Computed tomography (CT) is the imaging modality of choice in multiple clinical situations (1). The majority of all examinations for evaluation of tumor staging and tumor spread is performed with CT. Most of these examinations are performed with intravenously applied contrast media to improve detectability and interpretation of lesions (2). In conventional CT systems, perfusion of lesions cannot be quantified in a single phase. In multi-phase examinations, perfusion parameters can only be obtained indirectly via measuring differences in attenuation values.

Lately, dual-energy CT (DECT) has been introduced into clinical routine, and technical performance as well as potential clinical applications have been investigated intensively (3). In contrast to conventional CT systems, DECT acquires two energy spectra to enable different spectral applications such as material decomposition and quantification (4 –6). Different approaches exist to obtain the dual-energy dataset. Hereby, most systems use different tube voltages to generate the needed energy spectra. For example, two X-ray tubes with different peak kilovoltages (kVp) are used in dual-source CT (DSCT) or one tube with an alternating voltage is used in rapid-kVp-switching CT (7,8). With these systems, spectral data are not automatically acquired in every scan but only if the examination was designed accordingly. A different approach is realized with dual-layer CT (DLCT). Hereby, a single tube voltage (generally 120 kVp) is used and the two energy spectra are obtained by a detector which consists of two layers: the upper layer detects low-energy photons and is permeable for high-energy photons which are then detected by the lower layer (9). With this system, spectral data are automatically acquired in every scan. With all DECT systems, spectral applications become possible, with the detection and quantification of iodine being of particular interest for innovative CT imaging assessment (10). Via iodine concentration (IC), perfusion parameters can be evaluated indirectly; thus, there is the potential to differentiate between benign lesions without perfusion and malignant lesions with contrast enhancement. As a result, lesions with iodine uptake (e.g. renal cell carcinoma) can be differentiated from lesions without iodine uptake (e.g. cysts) (11,12). The differentiation between perfused malignant tissue and perfused benign/healthy tissue—both can present with a certain perfusion—is more difficult and the differentiation cannot simply be made by the discrimination between iodine uptake and no iodine uptake. To get a better understanding of the differences in iodine perfusion of different lesions, standard values of IC for healthy tissue are essentially needed as a reference standard. However, today, these standard values only exist for few tissues like myocardium (13).

Lymph nodes (LN) are evaluated in every CT scan. Enlargement of LNs is often seen and multiple reasons such as malignancy, inflammation, or idiopathic genesis are possible. In terms of malignancy, LNs are of special interest as they can be affected in nearly all malignant diseases, e.g. in terms of metastases or lymphoma. However, morphological evaluation of lymph nodes can be challenging in conventional CT as only morphological criteria such as size and shape can be evaluated. As a result, standard values of IC seem particularly important here to optimize interpretation of LNs in DECT.

The present study evaluated healthy LNs in different anatomical localizations to determine standard values of IC using a DLCT.

Material and Methods

Study cohort

The retrospective study was approved and conducted in accordance with the guidelines of the institutional review board. Informed consent was waived by the institutional review board due to the retrospective design of the study. For patient recruitment, the department’s picture archiving and communication system (PACS) was retrospectively searched for contrast-enhanced CT scans of the neck, thorax, and pelvis or a combination of the examinations. The CT scans had to be performed with the department’s DLCT in a portal-venous phase. Between January 2017 and July 2018, a total of 11,485 of the named CT scans were performed. From these patients, those with reported morphologically normal LNs in the actual examination as well as in follow-up examinations were selected. For each region (neck, axilla, groin), 100 patients were selected. One patient had to be subsequently eliminated from the study population due to severe motion artifacts. The diagnosis of morphologically normal LNs was made by two radiologists, where one radiologist was a senior physician with a minimum of six years of experience in interpreting CT scans. For confirmation, selected examinations were again evaluated by two radiologists with five and nine years of experience in CT before study enrollment, respectively. Patients with presence or history of malignancy, acute inflammation, or polytrauma were excluded. All included patients did not have any visible or known pathology in the respective anatomical area.

CT protocol

All CT examinations of the pelvis, neck, and thorax were acquired on a dual-layer DECT scanner (IQon, Philips Healthcare, Cleveland, OH, USA) according to our institutional protocol. Contrast medium (Imeron 400, Bracco Imaging, Konstanz, Germany) was injected in a standard dose of 80 mL with a flow rate of 2.5 mL/s using a dual syringe injection system (Stellant, MEDRAD, Indianola, PA, USA), followed by a 50-mL saline chaser. Portal venous phase images of the thorax, abdomen, and pelvis were obtained 70 s after intravenous administration of contrast material. CT scans of the neck were performed after the thorax–abdomen–pelvis scan to obtain images with lowered arms. Therefore, an additional 70 mL of contrast medium were injected followed by a 30-mL saline chaser. All scans were performed with the standard protocol of our institution. The scan was performed craniocaudally with a pitch of 0.9, a tube voltage of 120 kVp, and a detector configuration of 64 × 0.625 mm. All datasets were reconstructed in the axial view with a slice thickness of 3 mm and on a matrix of 512 × 512.

CT image analysis

Image analysis was performed using a commercially available spectral workstation (IntelliSpace Portal, v. 8.0.2, Philips Healthcare, USA). For measurement of IC, iodine maps showing IC (mg/mL) were used. For each measurement, a region of interest (ROI) was drawn in the area of the maximum iodine uptake of the measured LNs (Fig. 1).

Fig. 1.

Lymph nodes of the neck (a), axilla (b), and groin (c). Left: Conventional CT images; right: iodine maps.

For each patient, the three largest unobtrusive LNs of each region were used for the measurement of IC. The short-axis diameter of the lymph node had to be >5 mm to ensure a precise measurement (14). To minimize measurement errors, the largest possible circular ROI was adjusted by a SD not exceeding 0.2 mg/mL. The mean of the three measured LNs was used for further calculations and evaluation. For normalization purposes, the IC of an anatomically corresponding main artery and vein and of a local muscle was measured (Table 1). For normalization, Eq. 1 was used:

I C_{N} = \frac{I C_{I} \times N_{I}}{\sum N_{A} \div 99}

(1)

Table 1.

Examined lymph node locations with correlating arteries, veins, and muscles.

Region	Artery	Vein	Muscle
Neck	Common carotid artery	Internal jugular vein	Sternocleidomastoid muscle
Axilla	Subclavian artery	Subclavian vein	Major pectoral muscle
Groin	Abdominal aorta	Portal vein	Iliopsoas muscle

Hereby, IC_N is the normalized IC of the lymph node, IC_I is the not-normalized IC for this lymph node, N_I is the IC of the normalization tissue (i.e. artery, vein, muscle, or the mean of artery and vein), and (∑N_A ÷ 99) is the mean of the IC of the normalization tissue of all patients. Via normalization, effects of the contrast phase (e.g. relative late or early contrast phase) could be eliminated in theory.

Statistical analysis

Statistical analysis was performed by dedicated software packages (SPSS 25, IBM Corp., Armonk, NY, USA; Excel 2016, Microsoft Corp., Redmond, WA, USA; Prism 8 for Windows, GraphPad Software, San Diego, CA, USA). Unless otherwise indicated, all data are given as mean ± SD of the mean values. Data were tested for Gaussian distribution via D’Agostino–Pearson omnibus test. If Gaussian distribution was present, two-sided unpaired t-test was used between groups and paired t-test was used within groups. If Gaussian distribution was not present, Mann–Whitney test was used between groups and Wilcoxon matched-paired singed rank test was used. For correlation, Pearson’s correlation (in the case of Gaussian distribution) and Spearman’s correlation (in the case of missing Gaussian distribution) were applied.

Results

Study population

For each region (neck, axilla, groin), a total of 99 patients were included. Only patients with healthy LNs and no pathologies in the examined region were included. Gender and age information are given in Table 2.

Table 2.

Gender and age data of the participants.

	Neck		Axilla		Groin
	Male	Female	Male	Female	Male	Female
n	67	32	67	32	48	51
Age (years)	55.9	58.2	56.7	58.6	58.7	58.0
SD	17.8	14.4	17.4	14.8	19.2	18.5

Absolute iodine concentration of lymph nodes

For each anatomical localization (neck, axilla, groin), absolute IC of LNs are given in Fig. 2. Overall, the highest IC was found for LNs of the neck, followed by the axilla and groin with significant differences for each comparison. Hereby, the difference between the neck and axilla or groin was clearly greater than the difference between the axilla and groin with a mean difference of 0.85 mg/mL between the neck and axilla and 0.98 mg/mL between the neck and groin. Despite a significant difference, the LN of axilla and groin showed a difference in mean IC of only 0.135 mg/mL.

Fig. 2.

Violin plots of the absolute iodine concentration for the different anatomical localizations. The continuous line shows the mean and the dashed lines show the 25% and 75% quartiles. *P < 0.05, ****P < 0.0001.

Normalized iodine concentrations

For each region, normalization of the IC of LNs was performed to the artery, vein, and muscle, respectively. Additionally, normalization to the mean of the artery and vein was performed (Fig. 3). For each anatomical region, the normalization to the vein, the artery, to the mean of both vessels, or to the muscle did not lead to a reduced SD or confidence interval (CI). Consequently, the range was also greater for normalized values compared to absolute values. The normalization to vein, artery, or the mean of both caused a small increase of the resulting IC range whereas the normalization to muscle leads to a greater increase of the range. The IC of LNs of the neck was clearly higher than for LNs of the axilla and groin. The absolute IC of healthy LNs was 2.09 ± 0.44 mg/mL for the neck, 1.24 ± 0.43 mg/mL for the axilla, and 1.11 ± 0.34 mg/mL for the groin. The resulting 95% CIs for healthy LNs was 2.00–2.18 mg/mL for the neck, 1.16–1.33 mg/mL for the axilla, and 1.04–1.17 mg/mL for the groin.

Fig. 3.

Iodine concentration (IC) for the examined lymph nodes of the neck, axilla, and groin, shown as individual values. The horizontal line shows the mean, the whiskers represent the SD. Shown are the absolute values for IC as well as values which are normalized to the vein, artery, muscle, and the mean of the vein and artery. For the absolute values, the smallest range was found which can be clearly visualized.

Discussion

In recent years, DECT systems have found widespread use in clinical routine. With these systems, iodine differentiation and quantification became possible. It has been shown that measurements of IC can be performed in multiple clinical scenarios such as for differentiation and evaluation of lesions in different organs (11,15–17). Despite the availability of IC in all DECT scans over the last decade, standard values for IC only exist for some healthy tissues such as the myocardium (13). LNs are evaluated in every CT examination and are of significant clinical interest in many pathological conditions such as oncological or infectious diseases. For example, lymphoma is a common malignancy in children and adults (18), presenting with enlarged LNs. In CT, a differentiation between a benign and malignant genesis can only be made via morphological criteria such as size and shape (19). With positron emission tomography (PET), the metabolic activity can be evaluated via fluorine18-fludeoxyglucose which provides information about tumor vitality (20). However, PET involves an increased radiation exposure and is more time- and cost-consuming. Measurement of IC and thus an evaluation of the tissue perfusion could also enable information about metabolic activity with DECT. To differentiate healthy or benign from malignant LNs, standard values for IC would be essential.

To be able to evaluate IC in LN reliably and accurately, accurate iodine quantification is essential. The quantification of iodine became possible with DECT systems and was investigated widely. Earlier phantom studies showed that reliable and precise iodine quantification is possible even for low concentrations of 1 mg/mL and less (21,22). Quantification of iodine also works well in patient scans; therefore, applications such as virtual non-contrast imaging, where iodine content is extracted from the dataset based on the spectral information of the scan, became possible (10,23). These results demonstrate that accurate evaluation of IC is possible with DLCT so that IC of LNs should be evaluable accurately.

In the present study, LNs were examined in the neck, axilla, and groin of patients with healthy LNs. It could be seen that the absolute IC shows the smallest CI and smallest range for all regions. In particular, the normalization to the muscle did result in a significant increase of the IC range, which is most likely due to the low perfusion of the muscle. Thus, it seems not beneficial to implement a normalization of the IC of LNs to any other structure (vein, artery, muscle, or a combination of those), which makes an evaluation easier as fwer values must be measured and there are fewer sources of errors.

We found relatively large differences for IC of LN in the different regions. LNs of the neck showed a clearly higher IC (mean = 2.1 mg/mL) compared to the axilla (mean = 1.2 mg/mL) and groin (mean = 1.1 mg/mL). This is most likely due to the higher amount of contrast agent used in CT scans of the neck as, here, 70 mL of contrast medium were applied additionally as the CT was performed in two steps (first, CT of the thorax, abdomen, and pelvis followed by CT of the neck). Most CT scans of the neck are performed with the described protocol as in nearly all cases, scans of the neck are combined with an examination of other body parts. Thus, the information about IC seem representative for cervical LNs in the majority of cases but one has to keep this in mind when analyzing CT scans of the neck alone as less contrast agent is used here.

We found a 95% CI within < 0.2 mg/mL for each anatomical localization, indicating that the IC measurement works reliably and the IC of the LNs within one localization is within a small margin. The only comparable study found a mean IC for healthy LNs of the neck of 2.86 ± 0.37 mg/mL, compared to 2.09 ± 0.44 mg/mL in the present study (24). In the former study, only 16 patients with normal LNs were included, which lead to 36 measured LNs compared to 99 patients with 297 measured LNs in the present study. The most likely reason for the discrepancy in IC is the use of a different CT protocol: in the former study, 100 mL of contrast medium were used in a single bolus, whereas in the current study, 70 mL of contrast medium was used for the neck as a second bolus, making a total of 150 mL. Additionally, in the former study, DSCT was used, whereas in the present study, DLCT was used. The results of both studies indicate that the CT protocol and CT system must be considered when evaluating IC of LNs and other tissues.

The present study has some limitations. First, only LNs of three anatomical regions were examined. As the present study shows, the IC of LNs could differ between different anatomical regions. The examined regions are among the most often examined for LNs and are localized in different regions of the body. Thus, we believe that we chose a representative selection. However, the IC of LNs in other regions must be evaluated before a clinical standard value can be stated here. Second, a different quantity of contrast medium was used for the different regions (neck = total of 150 mL in two stages; axilla and groin = 80 mL). This most likely caused the different IC in the different localizations and thus, the contrast injection protocol probably has a high impact on the iodine quantification. Further, the DECT that was used could have an influence on the IC measurement. These factors must be kept in mind when IC is measured and different contrast injection protocols or other DECT systems are used, e.g. at different institutions. Third, a relatively small patient cohort was evaluated. Due to the small CI, we are convinced that a representative cohort was chosen and that the results reveal valid standard values of the examined LN. Fourth, only patients without any pathology in the examined LNs and the corresponding anatomical regions were included, without any recent history of malignancy. Thus, we did not assess values for IC of malignant or infectious LNs. Here, further studies are needed to evaluate the IC of pathological LNs and compare it to the range of healthy LNs.

In conclusion, the present study is the first to give standard values of the iodine concentration of healthy LNs of the neck, axilla, and groin. Hereby, standard mean values (95% CI) are purposed for the DLCT and contrast injection protocol used: 2.09 mg/mL (95% CI = 2.00–2.18 mg/mL) for the neck, 1.24 mg/mL (95% CI = 1.16–1.33 mg/mL) for the axilla, and 1.11 mg/mL (95% CI = 1.04–1.17 mg/mL) for the groin. These values could be used for differentiation of healthy and malignant LNs in the future.

Footnotes

Declaration of conflicting interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Andreas P Sauter

Sophie Ostmeier

References

Brenner

Hall

EJ.

Computed tomography–an increasing source of radiation exposure. N Engl J Med 2007; 357:2277–2284.

Bae

KT.

Intravenous contrast medium administration and scan timing at CT: considerations and approaches. Radiology 2010; 256:32–61.

McCollough

Leng

, et al. Dual- and multi-energy CT: principles, technical approaches, and clinical applications. Radiology 2015; 276:637–653.

Altman

Carmi

TU-E-210A-03: a double-layer detector, dual-energy CT - principles, advantages and applications. Med Phys 2009; 36:2750–2750.

Desai

Peterson

Garner

, et al. Clinical utility of dual-energy CT for evaluation of tophaceous gout. Radiographics 2011; 31:1365–1375; discussion 1376–1367.

Graser

Johnson

Bader

, et al. Dual energy CT characterization of urinary calculi: initial in vitro and clinical experience. Invest Radiol 2008; 43:112–119.

Flohr

McCollough

Bruder

, et al. First performance evaluation of a dual-source CT (DSCT) system. Eur Radiol 2006; 16:256–268.

Goodsitt

Christodoulou

Larson

SC.

Accuracies of the synthesized monochromatic CT numbers and effective atomic numbers obtained with a rapid kVp switching dual energy CT scanner. Med Phys 2011; 38:2222–2232.

Carmi

Naveh

Altman

Material separation with dual-layer CT. In: IEEE Nuclear Science Symposium Conference Record, 2005, pp.3--1878.

10.

Sommer

Schwarzwaelder

Stiller

, et al. Iodine removal in intravenous dual-energy CT-cholangiography: is virtual non-enhanced imaging effective to replace true non-enhanced imaging? Eur J Radiol 2012; 81:692–699.

11.

Ascenti

Mazziotti

Mileto

, et al. Dual-source dual-energy CT evaluation of complex cystic renal masses. AJR Am J Roentgenol 2012; 199:1026–1034.

12.

Chandarana

Megibow

Cohen

, et al. Iodine quantification with dual-energy CT: phantom study and preliminary experience with renal masses. AJR Am J Roentgenol 2011; 196:W693–700.

13.

Delgado Sánchez-Gracián

Oca Pernas

Trinidad López

, et al. Quantitative myocardial perfusion with stress dual-energy CT: iodine concentration differences between normal and ischemic or necrotic myocardium. Initial experience. Eur Radiol 2016; 26:3199–3207.

14.

Tawfik

Razek

Kerl

, et al. Comparison of dual-energy CT-derived iodine content and iodine overlay of normal, inflammatory and metastatic squamous cell carcinoma cervical lymph nodes. Eur Radiol 2014; 24:574–580.

15.

Agrawal

Pinho

Kulkarni

, et al. Oncologic applications of dual-energy CT in the abdomen. Radiographics 2014; 34:589–612.

16.

Deniffel

Sauter

Dangelmaier

, et al. Differentiating intrapulmonary metastases from different primary tumors via quantitative dual-energy CT based iodine concentration and conventional CT attenuation. Eur J Radiol 2019; 111:6–13.

17.

Iwano

Ito

Umakoshi

, et al. Evaluation of lung cancer by enhanced dual-energy CT: association between three-dimensional iodine concentration and tumour differentiation. Br J Radiol 2015; 88:20150224.

18.

Mugnaini

Ghosh

Lymphoma. Prim Care 2016; 43:661–675.

19.

Som

PM.

Detection of metastasis in cervical lymph nodes - CT and MR criteria and differential diagnosis. AJR Am J Roentgenol 1992; 158:961–969.

20.

Pieterman

van Putten

JWG

Meuzelaar

, et al. Preoperative staging of non-small-cell lung cancer with positron-emission tomography. N Engl J Med 2000; 343:254–261.

21.

Sellerer

Noel

Patino

, et al. Dual-energy CT: a phantom comparison of different platforms for abdominal imaging. Eur Radiol 2018; 28:2745–2755.

22.

Sauter

Kopp

Munzel

, et al. Accuracy of iodine quantification in dual-layer spectral CT: Influence of iterative reconstruction, patient habitus and tube parameters. Eur J Radiol 2018; 102:83–88.

23.

Sauter

Muenzel

Dangelmaier

, et al. Dual-layer spectral computed tomography: Virtual non-contrast in comparison to true non-contrast images. Eur J Radiol 2018; 104:108–114.

24.

Tawfik

Razek

Kerl

, et al. Comparison of dual-energy CT-derived iodine content and iodine overlay of normal, inflammatory and metastatic squamous cell carcinoma cervical lymph nodes. Eur Radiol 2014; 24:574–580.