Combined diffusion weighted MRI and MR spectroscopy: feasibility to improve the MRI capability in differentiation between benign and malignant neck lymphadenopathy

Combined diffusion weighted MRI and MR spectroscopy Feasibility to improve the MRI capability in differentiation between benign and malignant neck lymphadenopathy The Egyptian Journal of Radiology and[.]

Original Article

Combined diffusion-weighted MRI and MR spectroscopy: Feasibility

to improve the MRI capability in differentiation between benign and

malignant neck lymphadenopathy

Tamer F Taha Ali⇑, Mona A El Hariri

Department of Radiodiagnosis, Faculty of Medicine, Zagazig University, Sharkia, Egypt

a r t i c l e i n f o

Article history:

Received 8 November 2016

Accepted 19 December 2016

Available online xxxx

Keywords:

DWI

MRS

Lymph nodes

Cervical

Benign

Malignant

a b s t r a c t

The study aims to evaluate the additional value of MRS and DWI in differentiating malignant and benign neck lymphadenopathy

Materials and methods: Thirty-three patients with enlarged neck lymph nodes of malignant suspicious underwent DWI and MRS ADC values, presence of Cho peak and Cho/Creatine ratio of the dominant node were assessed and results were compared with histopathological results

Results: the patients were classified into benign (n = 9) and malignant (n = 24: 17 metastases and 7 lymphoma) The mean ADC values of the benign, metastasis and lymphoma patients were 1.56 ± 0.23, 1.01 ± 0.23 and 0.71 ± 0.02 10 3mm2/s respectively It was significantly higher in benign than malig-nant (p < 0.0001) and in metastatic than lymphomatous (p = 0.001) as well as in well- and moderately than poorly differentiated metastatic (p = 0.01) lymph nodes Using the receiver operating characteristic (ROC), cutoff value of 1.15 10 3mm2/s of ADC could differentiate benign from malignant nodes with sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) of 91.6%, 77.7%, 91.9%, 77.7%, KAPPA = 0.69 and p < 0.001 respectively

Malignant nodes showed a significant choline peak (n = 24, 100%) while benign nodes showed choline peak in only two cases (22%) Mean Cho/Cr ratio was significantly higher in malignant nodes than benign ones (2.64 ± 1.16 versus 1.09 ± 0.04) (p < 0.0001), furthermore it was significantly higher in lymphoma versus metastatic (4.3 ± 0.35 versus 1.94 ± 0.34, p < 0.001) as well as poor versus Well- to moderately dif-ferentiated metastases (2.3 ± 0.11 versus 1.69 ± 0.18, p < 0.01) The MRS sensitivity, specificity, PPV, NPV and Kappa in differentiating benign and malignant cervical lymph nodes were 100.0, 77.7, 92.3, 100.0% and 0.83 and p value = 0.001

Combination of DWI and MRS showed higher diagnostic value than DWI or MRS alone with sensitivity, specificity, PPV, NPV and Kappa of 100, 88.9, 96, 100% and 0.92 respectively (p < 0.0001)

Conclusion: ADC and MRS can help in the differentiation between malignant and benign neck lymph nodes Combination of both techniques achieved higher diagnostic performance

Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine Production and hosting by Elsevier This

is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/

4.0/)

1 Introduction

Discrimination of benign and malignant neck lymph nodes are

crucial especially in the presence of head and neck malignancy as it

is needed for proper staging and designing treatment plane as well

as follow up evaluation[1–5]

Conventional imaging relies on the morphological pattern of lymph nodes as short axis diameter, lymph node hilum loss and necrosis together with heterogenous pattern of post contrast enhancement and perinodal infiltrative process[6–9] Ultrasound (US), computed tomography (CT) and conventional magnetic reso-nance (MR) can detect cervical lymphadenopathy, yet their ability

in the discrimination between benign and malignant lymph nodes

is not highly accurate Positron emission tomography (PET) and PET/CT depend on the metabolic imaging and can aid in this differ-entiation but it is limited by low spatial resolution, as well as false physiological and inflammatory uptake Fine needle aspiration

http://dx.doi.org/10.1016/j.ejrnm.2016.12.008

0378-603X/Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine Production and hosting by Elsevier.

This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

Peer review under responsibility of The Egyptian Society of Radiology and Nuclear

Medicine.

⇑ Corresponding author.

E-mail address: Drtamerfathi@yahoo.com (T.F Taha Ali).

Contents lists available atScienceDirect

The Egyptian Journal of Radiology and Nuclear Medicine

j o u r n a l h o m e p a g e : w w w s c i e n c e d i r e c t c o m / l o c a t e / e j r n m

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-cytology (FNAC) is invasive with a risk of false results as its

opera-tor dependent[10–13]

Proton (1H) MR spectroscopy (1H MRS) is a noninvasive

tech-nique that depends on the evaluation of the metabolism at cell

level and measures the chemicals and metabolites in the body

1H MRS has been shown as a valuable tool in the cancer evaluation

with accumulating literatures validated its usage in variable types

of body cancers[14–19]

The large water and lipid resonances at 4.7 and 1.3 ppm (ppm)

has to be suppressed for proper assessment of the metabolites of

interest Shimming is needed to make the magnetic field is as

homogeneous as possible The spectrum of MRS has horizontal axis

corresponds to the metabolite resonance frequency in respect to

the water resonance peak at 4.7 ppm On the other hand the

verti-cal axis corresponds to the relative metabolite signal amplitude in

arbitrary units MRS shows quantitative data (presence or absence

of metabolites) and semi-quantitative (amplitude of metabolites or

its ratio relative to control)[20–24]

On the other hand, diffusion-weighted MRI (DWI) is considered

also as a non-invasive technique that can analyze the water

mole-cules motion to characterize the tissues These signal changes can

be quantified and reflected by apparent diffusion coefficient (ADC)

[25–30]

The aim of our study was to evaluate the additional value of

using MRS and DWI in the differentiation between malignant

and benign neck lymphadenopathy

2 Patients and methods

2.1 Patients

This prospective study had included 33 patients (17 male and

16 female patients, their mean age 53.1 ± 8.7 years) with enlarged

neck lymph nodes clinically suspicious of malignancy before going

to histopathological examination either by neck dissection (n = 11),

core biopsy (n = 10) or surgical (n = 12) Approval from our

institu-tional review board was achieved and patient informed written

consents were taken

2.2 MR imaging protocol

2.2.1 Conventional MRI

MR examination was done using a 1.5-T MR (Achieva, Philips

Medical Systems, Netherland B.V.) A standard head and neck coil

was used

Axial, coronal and sagittal localizer images were obtained first

then conventional study was completed including T1-weighted

images (repetition time (TR)/echo time (TE) = 600–600/8–10 ms)

and T2-weighted fast spin echo images (TR/TE = 3000–4000/80–

100)

Planes included axial and coronal planes with a slice thickness

of 3–4 mm, an inter-slice gap of 2 mm, a field of view (FOV) of

250–300 mm, an acquisition matrix 256 224 and a flip angle of

90 degree T1-weighted images were performed with and without

fat saturation after IV injection of 15 ml of gadopentetate

dimeglumine

2.2.2 Diffusion weighted MR

Acquisition of diffusion-weighted images using single shot echo

planar imaging (EPI) sequence was achieved in the axial plane

before the administration of contrast with 3–4 mm slice thickness,

1 mm intersection gap, FOV 250–300 mm, TR/TE = 2000–

2600/70 ms Application of the diffusion sensitizing gradient was

done in the three orthogonal planes (X, Y, Z) The b values used

were 0 and 1000 s/mm2

2.2.3 1-H MR spectroscopy Magnetic resonance spectroscopy was done for all patients Homogeneity of the magnetic field before recording the spectrum was achieved by application of automatic shimming, when the automatic shimming was difficult due to significant susceptibility differences the manual shimming was applied with a linewidth

of 12–14 Hz

Point resolved spectroscopic sequence (PRESS) – single voxel technique was obtained with the following parameters; TR/

TE = 2000/135 ms, signal acquisition 64, spectral bandwidth

1000 Hz, and number of points 512 Water suppression using chemical shift-selective suppression was done The acquired data was processed automatically with an average scan time of 4:48 min

The dominant solid node was selected for the analysis The vol-ume of interest (VOI) was positioned on the solid part of the node

in three planes (axial, sagittal and coronal planes) to limit inclusion

of surrounding fat as much as possible

Spectra are assessed for the presence of choline (Cho) and Crea-tine (Cr) peak at 3.22 and 3.03 ppm respectively and Cho/Cr ratio was calculated automatically

2.2.4 Calculation of the ADC value

A region of interest (ROI) was positioned on the ADC map on the same location as the VOI

Automatic calculation of apparent diffusion coefficient (ADC) maps was achieved by MRI machine software incorporated in same sequence of diffusion weighted image

2.3 Image analysis and data interpretation

MR images were evaluated independently by the two radiolo-gists sharing this study without any previous knowledge of its pathology Conventional MRI image were analyzed to evaluate of the lesion

MRI results were compared to the results of pathological exam-ination of the biopsy

Topographic correlation was applied by recording the maxi-mum lymph node short axis and its exact location and anatomical relations to ensure that the surgically removed node is the same subjected to analysis

2.4 Statistical analysis According to the histopathological results of lymph nodes, the patients in this study were categorized into benign and malignant (metastasis either well/moderately or poorly differentiated and lymphoma) Statistical analysis was done using SPSS version 17 The mean and standard deviations for ADC and Cho/Cr ratio were calculated for each group

One way analysis of variance (ANOVA) and post hoc analysis were applied to analyze the difference in ADC and Cho/Cr ratio among the different histolopathological types T-test was used to assess the difference in ADC values and Cho/Cr ratio between well and moderately differentiated and poorly differentiated metastasis

Receiver operating characteristic (ROC) curve was applied to evaluate the diagnostic ability of the ADC value Multiple thresh-olds of ADC values were analyzed to rule out the cutoff value deter-mined by Kappa test to differentiate benign from malignant nodes The sensitivity, specificity, positive predictive value (PPV) nega-tive predicnega-tive value (NPV) and Kappa test of DWI, MRS and com-bined DWI + MRS in differentiating benign and malignant cervical lymph nodes were calculated

The probability (p value) of <0.05 was considered significant

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-3 Results

33 patients with neck lymphadenopathy of malignant

suspi-cious were subjected to MRI including diffusion-weighted and

MR spectroscopy techniques before the pathological examination

by neck dissection, core or surgical excision Depending on the

pathological results the patients were classified into benign and

malignant group Benign group included (9 patients) with

lymphadenitis Malignant group included 24 patients of whom

17 had metastases from head and neck malignancy and 7 patients with lymphoma Metastases (Figs 1–3) were either well- to mod-erately differentiated (n = 10) or poorly differentiated (n = 7) while lymphoma (Fig 5) was either Hodgkin (n = 2) or Non-Hodgkin (n = 5)

The short axis of examined nodes ranged from 1.2 to 7.8 cm with mean was 2.9 cm

Fig 1 Metastatic lymphadenopathy: axial T2WI (a) shows left side cervical lymph node It shows high signal at DWI b 1000 (b) and low signal at ADC map (c) with ADC value

of 1.05  10 3

mm 2

/s MRS (d) shows Choline peak (Cho) (yellow arrow) with Cho/Cr ratio = 1.64.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-The mean ADC values of the benign, metastasis and lymphoma

patients were 1.56 ± 0.23, 1.01 ± 0.23 and 0.71 ± 0.02 10 3mm2/

s respectively

The ADC values of the benign neck lymph nodes (Fig 4) were

significantly higher than those of the metastatic and

lymphoma-tous nodes (p < 0.0001) The ADC values of the metastatic lymph

nodes were significantly higher than lymphoma (p = 0.001) The

mean ADC of well- and moderately differentiated metastasis

(1.11 ± 0.26 10 3mm2/s) was higher than that of poorly

differ-entiated metastasis (0.87 ± 0.02 10 3mm2/s) (p = 0.01)

Receiver operating characteristic (ROC) curve was applied to

assess the diagnostic power of the ADC in differentiating benign

from malignant nodes The area under the curve (AUC) was

(0.98) with CI (0.0–1.0) and std error = 0.01 (Table 1,Fig 6)

Cutoff value of 1.15 10 3mm2/s was used for the ADC to

dif-ferentiate benign from malignant lymph nodes with sensitivity,

specificity, positive predictive value (PPV) and negative predictive

value (NPV) of 91.6%, 77.7%, 91.9% and 77.7% respectively while

KAPPA test = 0.69 and p < 0.001

Malignant nodes showed a significant choline peak in all

malig-nant cases (n = 24, 100%) at MR spectroscopy while benign nodes

showed choline peak in only two cases (22%) The mean Cho/Cr

ratio was significantly higher in malignant nodes in comparison

to benign ones (2.64 ± 1.16 versus 1.09 ± 0.04) (p < 0.0001),

fur-thermore it was significantly higher in lymphoma compared to

metastatic lymph nodes (4.3 ± 0.35 versus 1.94 ± 0.34, p < 0.001)

as well as between poor versus well- to moderately differentiated

metastases (2.3 ± 0.11 versus 1.69 ± 0.18, p < 0.01) (Tables 1 and 2)

The MRS sensitivity, specificity, PPV, NPV and Kappa in differen-tiating benign and malignant cervical lymph nodes were 100%, 77.7%, 92.3%, 100.0% and 0.83 and p value = 0.001

Combination of both DWI and MRS showed a higher diagnostic value than the DWI or MRS alone with sensitivity, specificity, PPV, NPV and Kappa of 100%, 88.9%, 96%, 100% and 0.92 respectively and p value < 0.0001 (Table 3)

4 Discussion Treatment strategy and outcome in head and neck cancer are markedly influenced by the presence of metastatic lymphadenopa-thy While conventional imaging can detect neck lymph nodes morphology, still the functional information is needed[18,19] The motion of water molecules extracellular, through the cell membranes, and intracellular affect the diffusion coefficient of the tissue so DWI can give information about the biology and phys-iology criteria of the tumor These motions of water molecules are impacted by hypercellularity, fibers, enlarged nuclei, intracellular organelles, and macromolecules in the tissues with subsequent decreases in the ADC value in malignancy[19,25–28]

Our study included 33 patients with enlarged neck lymph nodes They were 9 patients with benign lymphadenopathy, 17 patients with metastasis from head and neck cancer and 7 patients with nodal lymphoma

The mean ADC values of the benign, metastasis and lymphoma groups were 1.56 ± 0.23, 1.01 ± 0.23 and 0.71 ± 0.02 10 3mm2/s,

Fig 2 Metastatic lymphadenopathy: axial T2WI (a) LT side cervical lymph nodes It shows high signal at DWI b 1000 (b) and while low signal at ADC map (c) with ADC value

of 0.85  10 3

mm 2

/s MRS (d) shows Choline peak (Cho) (yellow arrow) with Cho/Cr ratio = 2.2.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-respectively Benign neck lymph nodes showed a significantly

higher ADC values in comparison to those of the malignant lymph

nodes (metastatic nodes and nodal lymphoma) Furthermore, the

metastatic lymph nodes showed ADC of significantly higher values

than those of nodal lymphoma This coincides with multiple earlier

reports[29–33] This can be explained by higher cellularity and

lower extracellular space in lymphoma[31]

Razek et al.[29]reported that the mean ADC value of metastatic lymph nodes was 1.09 ± 0.11 10 3mm2/s while that of lym-phomatous nodes was 0.97 ± 0.27 10 3mm2/s and both was sig-nificantly lower than mean ADC of benign cervical lymph nodes (1.64 ± 0.16 10 3mm2/s) (p < 0.04)

Perrone et al.[31]reported also a significantly higher mean ADC value in benign nodes (1.448 10 3

mm2/s) in comparison to

Fig 3 Metastatic lymphadenopathy: axial T2WI (a) show multiple bilateral cervical lymph nodes It shows high signal at DWI b 1000 (b) while low signal at ADC map (c) with ADC value of 1.05  10 3 mm 2 /s MRS (d) showed Choline peak (yellow arrow) with Cho/Cr ratio = 2.3.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-those of malignant nodes (0.85 10 3mm2/s) Similar results was

achieved by Bondt et al.[33]and they reported mean ADC values

malignant lymph nodes of 0.85 ±10 3mm2/s versus

1.2 ± 0.24 10 3mm2/s for benign nodes

Moreover, earlier studies showed that the ADC of metastatic

lymph nodes was significantly higher than that of lymphoma

[32,34]

While in the study of Sumi et al [34] reported a significant higher ADC in metastatic nodes (0.410 ± 0.105 10 3mm2/s) in comparison to inflammatory nodes 0.302 ± 0.062 10 3mm2/s) and attributed that to central necrosis in examined metastatic lymph nodes (48%) which altered the ADC values of these nodes

In the current study, the mean ADC in well- and moderately differentiated metastasis (1.11 ± 0.26 10 3

mm2/s) was

signifi-Fig 4 Lymphadenitis (yellow arrow): axial T2WI (a) The lymph node shows intermediate signal at DWI b 1000 (b) while high signal is noted at ADC map (c) with ADC value

of 1.71  10 3 mm 2 /s MRS (d) shows no Choline peak.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-cantly higher than that of poorly differentiated metastasis

(0.87 ± 0.02 10 3mm2/s) This coincides with an earlier report

[24]which showed that the ADC of well- and moderately versus

poorly differentiated metastatic lymph nodes are

1.13 ± 0.1110 3mm2/s versus 0.89 ± 0.1210 3mm2/s respectively

(p < 0.02)

This is also in agreement with another report [34] which

explained that by the increased nucleus to- cytoplasm ratio and

hypercellularity in poor differentiated carcinoma leading to reduc-tion of the diffusion space for water protons On the other hands, King et al [32] reported no such significant difference between poorly and well/moderately differentiated squamous cell carci-noma (SCC)

In our study we tries to find the best ADC threshold value for discriminating benign from malignant lymph nodes, it was 1.15 10 3

mm2/s with sensitivity, specificity, PPV and NPV of

Fig 5 Lymphoma: axial T2WI (a) show multiple bilateral cervical lymph nodes It shows high signal at DWI b 1000 (b) and low signal at ADC map (c) with ADC value of 0.73  10 3

mm 2

/s MRS (d) shows Choline peak (Cho) (yellow arrow) with Cho/Cr ratio = 4.6.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-91.6, 77.7, 91.9 and 77.7 respectively and KAPPA test = 0.69 with

and p < 0.001

By using the receiver operating characteristic (ROC) curve of the

ADC value to differentiate benign from malignant lymph nodes, the

area under the curve was The area under the curve (AUC) was

(0.98) with CI (0.0–1.0) and std error, p < 0.001

In an earlier report[31]the corresponding best threshold value was 1.03 10 3mm2/s, gaining a sensitivity of 100% and a speci-ficity of 92.9% The area under (ROC) was 0.983 The corresponding value in the report of Razek et al.[29]was 1.38 10 3mm2/s with

an accuracy of 96%, sensitivity of 98%, and specificity of 88% while PPV and NPV were 98.5% and 83.7% respectively with the area under the curve of 0.955

The best ADC threshold value for Bondt et al.[33]for the diag-nosis of malignant cervical lymph nodes was 1.0 10 3mm2/s with sensitivity of 92.3% and specificity 83.9%

The MRS spectra acquired from head and neck lesions are usu-ally of lower spectral resolution in comparison to that obtained from brain and this can be attributed to shimming difficulties due to susceptibility differences of tissues in cervical regions [2,17,18]

Many reports described the typical pattern of cancer MRS to have high total choline (Cho, 3.2 ppm) signal intensity (SI) relative

to creatine (Cr, 3.0 ppm), usually accompanied by the presence of other metabolites as lactate (Lac, 1.3 ppm)[35–37]

Choline is thought to originate from the metabolism of cell membrane phospholipid so increased Cho/Cr levels can be a mar-ker of active cellular proliferation and indicates a high cellular membrane turnover in malignant lesions[20,38,39]

The reduction of the creatine level can be explained by the increased metabolism rate and more energy consumption which takes place in the highly aggressive tumors[5]

Razek et al.[19]reported a significant higher Cho/Cr and lower ADC values for poorly to un-differentiated in comparison to mod-erate to well-differentiated HNSCC tumors (p = 0.003 and

p = 0.001, respectively) and suggested Cho/Cr and ADC as a new imaging parameter for estimation of HNSCC prognosis

In the study of Star-Lack et al.[40], they compared Cho/Cr ratio for metastatic lymph nodes to muscle tissue and found a signifi-cant higher ratio in former (2.9 ± 1.6 versus 0.55 ± 0.21 respec-tively, P = 0.0006) Another earlier study [24] also showed a significantly higher Cho/Cr ratio in tumor than in normal tissue They concluded the potential value of MRS to help in differentia-tion between primary squamous cell carcinoma and nodal metas-tases containing squamous cell carcinoma from normal tissue King et al.[2]did not detect Cho or Cr in the tuberculous nodes while frequently detected it in malignant nodes and they sug-gested that Cho and Cr presence in the spectra may exclude infec-tious lymphadenopathy In their case of Castleman’s disease, despite of absence of any signs of malignant changes, there was elevated Cho, and they attributed that to the hypercellularity of this disease

Our results are matched with these previous reports as we detected a significant choline peak in all malignant nodes in cur-rent study (n = 24, 100%) at MR spectroscopy while only two cases (22%) of benign nodes showed choline peak

Furthermore, King et al.[2]showed that Cho is highest in NHL, followed by UDC and SCC Cho/Cr ratios were significantly lower in SCC compared to UDC

Razek et al.[19]reported negative correlation between Cho/Cr level and the ADC value in HNSCC (r = 0.662, p = 0.001) with a significant difference in the Cho/Cr and ADC values at different degrees of tumor differentiation (p = 0.003 and p = 0.001) They

Table 1

Mean ADC and Cho/Cr values in the different types of neck lymphadenopathy.

Number

of cases ADC values (mean ± SD) 

10 3 mm 2 /s

Cho/Cr

-Well and moderately

differentiated

10 1.11 ± 0.26 1.69 ± 0.18 -Poorly differentiated 7 0.87 ± 0.02 2.3 ± 0.11

Fig 6 Receiver operating characteristic (ROC) curve of the ADC value for

differentiation between benign and malignant The area under the curve is 0.98.

Table 2

Correlation of choline peak with pathological results.

Table 3

comparison between the diagnostic values of DWI, MRS and combined DWI + MRS.

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-proposed a cut-off value for Cho/Cr and ADC for each category as

1.83, 0.95 and 1.94, 0.99, respectively, and estimated the areas

under the curve to be 0.771, 0.967 and 0.726, 0.795, respectively,

for each category

Our results are in matching with these studies, as in current

study the mean Cho/Cr was significantly higher in malignant nodes

in comparison to benign ones (2.64 ± 1.16 versus 1.09 ± 0.04)

(p < 0.0001), furthermore it was significantly higher in lymphoma

compared to metastatic lymph nodes (4.3 ± 0.35 versus

1.94 ± 0.34, p < 0.0001) also the mean Cho/Cr was significantly

lower in well- to moderately versus poorly differentiated

(1.69 ± 0.18 versus 2.3 ± 0.11, p < 0.0001)

Similar results were achieved by King et al.[41]who obtained

MRS spectra in seven (78%) of nine primary tumors and 16 (89%)

of 18 metastatic nodes They showed a significantly higher Cho/

Cr ratio for metastatic nodes (5.3 ± 1.6) than that for primary

(2.6 ± 0.5) NPC lesions (P < 0.02) at TE 136 They concluded also

that Cho/Cr ratios for NPC were high compared with those for

nor-mal neck muscle

Yu et al.[14]classified their lesions into 3 types: type 1

(with-out Cho signals); type 2 (with Cho signals and Cho/noise ratio < 3)

and type 3 (with Cho signals and Cho/noise ratio > 3 and concluded

that the differences of Cho signals and Cho/noise ratios can help in

differentiating malignant tumors and chronic infections

Combined DWI and MRS were tried in some previous reports

[15,19] El-Hariri et al.[15]used this combination in differentiation

between benign and malignant thyroid nodules and confirmed

higher diagnostic accuracy than using each of them separately

Also in the study of Razek et al.[19], they showed that the

combi-nation of higher Cho/Cr values and lower ADC values characterized

the poorly differentiated and considered this to be promising tool

to rule out highly aggressive HNSCC

In current study the combination of both DWI and MRS showed

a higher diagnostic value (sensitivity, specificity, PPV, NPV and

Kappa test of 100, 88.9, 96, 100% and 0.92 respectively and p

value < 0.0001) in comparison to the using any of DWI or MRS

alone in differentiating benign and malignant cervical lymph

nodes The sensitivity, specificity, PPV, NPV and Kappa test of

DWI were 91.6, 77.7, 91.9, 77.7% and 0.69 with p value = 0.001

while with MRS the sensitivity, specificity, PPV, NPV and Kappa

test were 100.0, 77.7, 92.3, 100.0% and 0.83 and p value = 0.001

We had some limitations in this study, 1st was that the analysis

was done on the number of patients not the number of lymph

nodes to avoid the impact of multiple nodes per patient on the

result 2nd limitation, trial to improve the diffusion sensitivity by

increasing the b value causes reduction of the signal-to-noise ratio

which limits the ADC measurement on the smaller lymph nodes

Small study number is a further limitation

In conclusion, in this preliminary study we have shown that

ADC and MRS are valuable techniques that can help in the

differen-tiation between malignant and benign cervical lymph nodes

Com-bination of both techniques increased the achieved higher

diagnostic performance

Conflict of interest

None declared

References

[1] Eida S, Sumi M, Yonetsu K, Kimura Y, Nakamura T Combination of helical CT

and Doppler sonography in the follow-up of patients with clinical N0 stage

neck disease and oral cancer AJNR 2003;24:312–8

[2] King AD, Yeung DK, Ahuja AT, Yuen EHY, Ho SFM, Tse GMK, et al Human

cervical lymphadenopathy: evaluation with in vivo 1H-MRS at 1.5 T Clin

Radiol 2005;60:592–8

[3] Fischbein N, Noworolski S, Herny R, Kaplan M, Dillon W, Nelson S Assessment

of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging AJNR Am J Neuroradiol 2003;24:301–11

[4] Sumi M, Ohki M, Nakamura T Comparison of sonography and CT for differentiating benign from malignant cervical lymph nodes in patients with squamous cell carcinoma of the head and neck AJR 2001;176:1019–24 [5] Bezabeh T, Odlum O, Patel R, Ahing S, Nason R, Sutherland D, et al Predicting prognosis and clinical outcome in head and neck tumors: a spectroscopic approach Proc Int Soc Magn Reson Med 2002;10:2050

[6] King AD, Tse GM, Ahuja AT, Yuen EH, Vlantis AC, To EW, et al Necrosis in metastatic neck nodes: diagnostic accuracy of CT, MR imaging, and US Radiology 2004;230:720–6

[7] Sigal R, Vogl T, Casselman J, Moulin G, Veillon F, Hermans R, et al Lymph node metastases from head and neck squamous cell carcinoma: MR imaging with ultrasmall superparamagnetic iron oxide particles (Sinerem MR)–results of a phase-III multicenter clinical trial Eur Radiol 2002;12:1104–13

[8] Ferreira T Comments on Castelijns and van den Brekel: imaging of lymphadenopathy in the neck Eur Radiol 2003;13:2236

[9] Ali TF Neck lymph nodes: characterization with diffusion-weighted MRI EJRNM 2012;43:173–81

[10] Yamazaki Y, Saitoh M, Notani K, Tei K, Totsuka Y, Takinami S, et al Assessment

of cervical lymph node metastases using FDG-PET in patients with head and neck cancer Am Nucl Med 2008;22:177–84

[11] Jeong HS, Baek CH, Son YI, Ki Chung M, Kyung Lee D, Young Choi J, et al Use of integrated (18)F-FDG PET/CT to improve the accuracy of initial cervical nodal evaluation in patients with head and neck squamous cell carcinoma Head Neck 2007;29(3):203–10

[12] Ishikawa M, Anzai Y MR imaging of lymph nodes in the head and neck Neuroimag Clin North Am 2004;14:679–94

[13] Castelijns JA, van den Brekel MW Imaging of lymphadenopathy in the neck Eur Radiol 2002;12:727–38

[14] Yu Q, Yang J, Wang P Malignant tumors and chronic infections in the masticator space: preliminary assessment with in vivo single-voxel 1H-MR spectroscopy AJNR Am J Neuroradiol 2008;29:716–9

[15] El-Hariri M, Gouhar GK, Said NS, Riad MM Role of diffusion-weighted imaging with ADC mapping and in vivo 1H-MR spectroscopy in thyroid nodules EJRNM 2012;43:183–92

[16] Thomas MA, Binesh N, Yue K, Bruhl ND Volume localized two dimensional correlated magnetic resonance spectroscopy of human breast cancer J Magn Reson Imaging 2001;14:181–6

[17] Huang W, Roche P, Shindo M, Madoff D, Geronimo C, Button T Evaluation of head and neck tumor response to therapy using in vivo 1H MR spectroscopy: correlation with pathology In: Proceedings of the 8th annual meeting of ISMRM, Denver, USA; 2000 p 552.

[18] Huang W, Roche P, Shindo M, Button T Evaluation of thyroid masses using

in vivo 1H MR spectroscopy: correlation with pathology In: Proceedings of the 9th annual meeting of ISMRM Glasgow, UK; 2001 p 2347.

[19] Razek AA, Nada N Correlation of choline/creatine and apparent diffusion coefficient values with the prognostic parameters of head and neck squamous cell carcinoma NMR Biomed 2016;29:483–9

[20] Shah GV, Gandhi D, Mukherji S Magnetic resonance spectroscopy of head and neck neoplasms Top Magn Reson Imaging 2004;15:87–94

[21] Razek AA, Poptani H MR spectroscopy of head and neck cancer Eur J Radiol 2013;82(6):982–9

[22] Tran T, Ross B, Lin A Magnetic resonance spectroscopy in neurological diagnosis Neurologic Clinic 2008;27:21–60

[23] Bisdas S, Baghi M, Huebner F, et al In vivo proton MR spectroscopy of primary tumours, nodal and recurrent disease of the extracranial head and neck Eur Radiol 2007;17:251–7

[24] Mukherji SK, Gerstle RJ In vitro 1H MR spectroscopy of squamous cell carcinoma of the extracranial head and neck Neuroimaging Clin North Am 1998;8:835–47

[25] Schafer J, Srinivasan A, Mukherji S Diffusion magnetic resonance imaging in the head and neck Magn Reson Imaging Clin North Am 2011;19:55–67 [26] Wang J, Takashima S, Takayama F, Kawakami S, Saito A, Matsushita T, et al Head and neck lesions: characterization with diffusion-weighted echo-planar

MR imaging Radiology 2001;220:621–30 [27] Hwang I, Choi SH, Kim YJ, Kim KG, Lee AL, Yun TJ, et al Differentiation of recurrent tumor and posttreatment changes in head and neck squamous cell carcinoma: application of high b-value diffusion-weighted imaging Am J Neuroradiol 2013;34:2343–8

[28] Thoeny HC, De Keyzer F, King AD Diffusion-weighted MR imaging in the head and neck Radiology 2012;263:19–32

[29] Razek AA, Soliman NY, Elkhamary S, Alsharaway MK, Tawfik A Role of diffusion-weighted MR imaging in cervical lymphadenopathy Eur Radiol 2006;16:1468–77

[30] Sumi M, Cauteren MV, Nakamura T MR microimaging of benign and malignant nodes in the neck AJR 2006;186:749–57

[31] Perrone A, Guerrisi P, Izzo L, D’Angeli I, Sassi S, Mele LL, et al Diffusion-weighted MRI in cervical lymph nodes: differentiation between benign and malignant lesions Eur J Radiol 2011:281–6

[32] King AD, Ahuja AT, Yeung DKW, Fong DKY, Lee YYP, Lei KIK, et al Malignant cervical lymphadenopathy: diagnostic accuracy of diffusion-weighted MR imaging Radiology 2007;245(3):806–13

[33] Bondt RBJ, Hoeberigs MC, Nelemans PJ, Deserno WM, Kootstra CP, Kremer B,

et al Diagnostic accuracy and additional value of diffusion-weighted imaging

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI

capa-for discrimination of malignant cervical lymph nodes in head and neck

squamous cell carcinoma Neuroradiology 2009;51:183–92

[34] Sumi M, Sakihama N, Sumi T, Morikawa M, Uetani M, Kabasawa H, et al.

Discrimination of metastatic cervical lymph nodes with diffusion-weighted

MR imaging in patients with head and neck cancer AJNR 2003;24:1627–34

[35] Kwock L, Smith JK, Castillo M, Ewend MG, Collichio F, Morris DE, et al Clinical

role of proton magnetic resonance spectroscopy in oncology: brain, breast, and

prostate cancer Lancet Oncol 2006;7:859–68

[36] King AD, Yeung DKW, Ahuja AT, Leung SF, Tse GMK, van Hasselt AC In vivo

proton MR spectroscopy of primary and nodal nasopharyngeal carcinoma.

AJNR Am J Neuroradiol 2004;25:484–90

[37] Castelijnsa JA, Leemansb CR Closing the uncertainty gap in the diagnosis of

parotid tumors AJNR 2009;30:442–3

[38] Chawla S, Kim S, Loevner LA, et al Proton and phosphorous MR spectroscopy

in squamous cell carcinomas of the head and neck Academic Radiol 2009;16:1366–72

[39] Mukherji SK, Gerstle RJ In vitro 1H MR spectroscopy of squamous cell carcinoma of the extracranial head and neck Neuroimaging Clin North Am 1998;8:835–47

[40] Star-Lack JM, Adalsteinsson E, Adam MF, Terris DJ, Pinto HA, Brown JM, et al In vivo 1H MR spectroscopy of human head and neck lymph node metastasis and comparison with oxygen tension measurements AJNR Am J Neuroradiol 2000;21:183–93

[41] King AD, Yeung DK, Ahuja AT, Leung SF, Gary MK, Van Hasselt AC In vivo proton MR spectroscopy of primary and nodal nasopharyngeal carcinoma Am

J Neuroradiol 2004;25:484–90

Please cite this article in press as: Taha Ali TF, El Hariri MA Combined diffusion-weighted MRI and MR spectroscopy: Feasibility to improve the MRI