Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Functional MRI for Treatment Evaluation in Patients with Head and Neck Squamous Cell Carcinoma: A Review of the Literature from a Radiologist Perspective

Functional MRI for Treatment Evaluation in Patients with Head and Neck Squamous Cell Carcinoma: A... Curr Radiol Rep (2018) 6:2 https://doi.org/10.1007/s40134-018-0262-z ENT IMAGING (A A JACOBI-POSTMA, SECTION EDITOR) Functional MRI for Treatment Evaluation in Patients with Head and Neck Squamous Cell Carcinoma: A Review of the Literature from a Radiologist Perspective 1 1 2,3 2,3 • • • Roland P. Nooij Jan J. Hof Peter Jan van Laar Anouk van der Hoorn Published online: 22 January 2018 The Author(s) 2018. This article is an open access publication Abstract Keywords MRI  Treatment evaluation  Primary tumor Purpose of review To show the role of functional MRI in Lymph nodes  Head/neck squamous cell carcinoma patients treated for head and neck squamous cell Review carcinoma. Recent findings MRI is commonly used for treatment Abbreviations evaluation in patients with head and neck tumors. How- ADC Apparent diffusion coefficient ever, anatomical MRI has its limits in differentiating ASL Arterial spin labeling between post-treatment effects and tumor recurrence. DCE Dynamic contrast enhanced Recent studies showed promising results of functional MRI DKI Diffusion kurtosis imaging for response evaluation. DSE Dynamic susceptibility enhanced DWI Diffusion weighted imaging Summary This review analyzes possibilities and limita- tions of functional MRI sequences separately to obtain HNSCC Head and neck squamous cell carcinoma insight in the post-therapy setting. Diffusion, perfusion and IVIM Intravoxel incoherent motion spectroscopy show promise, especially when utilized complimentary to each other. These functional MRI sequences aid in the early detection which might improve survival by increasing effectiveness of salvage therapy. Introduction Future multicenter longitudinal prospective studies are needed to provide standardized guidelines for the use of Head and neck cancer affects 550,000 new cases and functional MRI in daily clinical practice. 380,000 deaths worldwide annually [1–3]. Head and neck squamous cell carcinomas comprise over 90% of the head and neck carcinomas [4]. Patients frequently present with a locally advanced stage for which the current therapy is This article is part of the Topical collection on ENT Imaging. multimodal including surgery, radiation therapy and/or chemotherapy [5–8]. Many patients demonstrate unfavor- & Anouk van der Hoorn able treatment response, with locoregional recurrence seen a.van.der.hoorn@umcg.nl in about 30–60% [7]. This is in about 2/3 due to primary tumor recurrence, 1/3 due to regional nodal metastasis and Department of Radiology, Medical Spectrum Twente, Enschede, The Netherlands in 1/3 due to both primary tumor recurrence as well as regional nodal metastasis [9]. Department of Radiology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Conventional anatomical MRI techniques are commonly P. O. Box 30.001, 9700 RB Groningen, The Netherlands used for treatment evaluation, but are often not able to have Medical Imaging Center, University Medical Center reliable assess treatment response. Surgery as well as Groningen, University of Groningen, Groningen, The chemoradiotherapy induce false positives on imaging as a Netherlands 123 2 Page 2 of 15 Curr Radiol Rep (2018) 6:2 result of inducing benign changes involving architectural Higher diagnostic accuracy than 84% sensitivity and •• • • distortion, fibrosis and/or necrosis [10 , 11, 12, 13 , 14 ]. 82% specificity post-therapeutically is needed to differen- These benign treatment-induced changes should be differ- tiate treatment effects from true malignancy for the local entiated from residual and/or recurrent tumor on imaging to tumor site and the regional lymph nodes to reliable either prevent unjustified alteration in treatment plan, e.g. salvage initiate new therapy, adjust the current therapy or discon- therapy or (dis)continuation of therapy. Early detection of tinue unjustified therapy. local recurrence could lead to timely salvage therapy which can lead to an increase in overall survival [15–17]. Post-treatment surveillance can consist of ultrasound, Technical Background of Functional MRI •• • • PET-CT [18–21] and MRI [10 , 11, 12, 13 , 14 ]. Several Techniques recent studies have shown the potential usefulness of functional MRI techniques for treatment evaluation in Diffusion Weighted Imaging •• • • patients with head and neck tumors [10 , 11, 12, 13 , 14 , •• • • 22 , 23–28, 29 , 30–33 ]. Diffusion-weighted imaging is DWI measures cellular density and cytoarchitecture using used to image changes in cytoarchitecture and measure the measurement of water diffusivity. Random diffusion cellular density. Perfusion-weighted MRI techniques can results from the Brownian motion of water molecules. identify tumor-induced neovascularization. Changes in Motion of water molecules is hindered, restricted, by concentrations of metabolites are shown with magnetic interactions with other molecules and cellular barriers such resonance spectroscopy (MRS). as fibers, cell membranes and macromolecules. Diffusion This review will analyze the functional MR imaging abnormalities of water molecules thus reflect changes of sequences with regards to their possibilities and limitations tissue organization at a cellular level affecting the MR in head and neck squamous cell carcinoma. Clinical signal of a DWI sequence as can be seen in a number of •• • •• implications, applicability and possibilities of these processes including malignancy [10 , 11, 12, 13 , 14, 22 , •• •• • • sequences for treatment evaluation will be addressed. 24 , 25 , 26–28, 29 , 30, 32, 33 ]. DWI sequences are based on a T2-weighted sequence. At least two b values are needed to analyze motion of Role of Conventional Anatomical MRI in Head water. DWI is done at different b values (in s/mm ), which and Neck SCC represent the duration between the gradient pulses used. Simplified, it is the time that water is allowed to diffuse Conventional anatomical MRI techniques are used for before the distance is measured. Most commonly, a b0 and treatment evaluation. MRI is superior to CT yielding higher b800 or b1000 value are used for head and neck imaging. • • anatomical detail [11, 12, 13 , 14 , 34–37]. Anatomical Diffusion is quantified using ADC in mm /s. Having MRI to assess HNSCC should include a T1 without fat measured at least two different b values (e.g. b0 and b800), suppression, T2 with and without fat suppression and T1 the logarithm of relative signal intensity of a tissue is post-contrast with fat suppression. These sequences are plotted on the y axis against the b values on the x axis. The used to analyze certain characteristics of the primary tumor slope of the line fitted through the plots describes the ADC. • • and possible nodal involvement [11, 12, 13 , 14 , 34–37]. This mono-exponential fitting represents a rough approxi- However, anatomical MRI techniques are often unable mation of ADC and is most often used in clinical routine. to accurately identify treatment response showing a pooled This parameter is independent of the magnetic field sensitivity and specificity for local treatment response strength. Lower values indicate more restricted diffusion. •• evaluation in HNSCC of 84 and 82%, respectively [22 ]. However, mono-exponential fitted ADC values cannot This is due to benign treatment effects such as inflamma- separate the pure molecular diffusion from the motion of tion, fibrosis and necrosis as a result of surgery and water molecules in the capillary network [32]. Low b val- chemoradiotherapy. These post-therapy changes show ues are most influenced by the capillary component which overlapping signal characteristics with tumor. Most prob- influences the ADC values. Multi-exponential models lematic for the primary site is that inflammation and tumor using several b values are more suitable for accurate both show high T2 signal and enhancement after contrast quantification of diffusion without perfusion contamination injection. Lymph node assessment is most hindered by [30, 32, 33 ]. reactive lymph node that can be slightly enlarged similar to Acquiring multiple b values yields techniques such as nodal metastasis. Furthermore, normal sized nodes can still intravoxel incoherent motion (IVIM) and diffusion kurtosis contain tumor. See Table 1 for a detailed description of the imaging (DKI). IVIM imaging can distinguish between signal intensities on anatomical MRI post-treatment. pure molecular diffusion and motion of water molecules in the capillary network through a single DWI acquisition 123 Curr Radiol Rep (2018) 6:2 Page 3 of 15 2 Table 1 Use of conventional anatomical MRI for treatment evaluation Anatomical MRI Primary tumor Lymph nodes sequence T1 without fat Anatomical details Anatomical localization of node levels suppression Tumor: ; compared to fat Metastatic lymph nodes: Size :/:: (suggested cut-off [ 7–10 mm for level II and [ 5–7 mm for all other levels). Round shape Fat infiltration by tumor or inflammation: similar ;/;; Reactive lymph nodes: Size =/: (can be false-positive using above cut-off); Oval with fatty hilum Necrosis: ;; round, oval, well circumscribed Location of lymph node and level in relation to location primary tumor Fibrosis: Linear commonly ;;, but can be ;/= as well T2 with and without Fat suppression useful for the detection Fat suppression needed to identify abnormal nodes fat suppression of abnormalities T2 without fat suppression for anatomical details T2 without fat suppression for Metastatic lymph nodes: : slightly heterogeneous; more commonly an anatomical details irregular border; possible extra-nodal extension Edema, fat infiltration by tumor or Reactive lymph nodes: = inflammation: similar :/:: Necrosis: :: round, oval, well circumscribed Perineural spread: : Fibrosis: Linear commonly ;;, but can be ;/= as well. T1 post-contrast with Fat infiltration by tumor or Fat suppression needed to identify abnormal lymph nodes fat suppression inflammation: similar :/:: Metastatic lymph nodes: :/::, thick, irregular rim enhancement in case of Edema or necrosis: no enhancement necrosis Fibrosis: no enhancement after Reactive lymph nodes: =/: 6-12 months. Most commonly :/:: Perineural spread: :: High signal intensity is indicated as :, low signal intensity is indicated as ; and intermediate signal as = technique if both low b values (\ 200 s/mm ) and high contrast agents. DCE perfusion has been reported as a b values ([ 200 s/mm ) are used. The relationship between technique which is able to characterize perfusion and •• •• • signal intensities and multiple b values can be assessed. vascularization of tissues [24 , 25 , 30–33 ]. However, •• Real diffusion of water molecules (D) can be distinguished this has not always been histologically confirmed [25 , from the contribution of perfusion to the signal decay (D*) 38, 39]. Ktrans is the most commonly derived quantitative and the contribution of perfusion to the diffusion signal (f). parameters representing capillary permeability and seems •• •• Another, multiple b value method, DKI, represents the to be to most consistent parameter [24 , 25 , 40]. extent to which the diffusion pattern of the water molecules DSC perfusion exploits the susceptibility-induced signal deviates from a perfect Gaussian curve that is assumed loss after administration of contrast on T2-weighted calculating standard ADC values. Table 2 includes the sequences, most commonly a quick T2* gradient echo most commonly used parameters for the different diffusion sequence. It is based on inhomogeneity of the magnetic techniques. field during the passage of a short bolus of contrast through a capillary bed [27]. As result on the T2* sequence, blood MR Perfusion products, calcifications and aerated structures result in artificial signal loss. Mean transit time, blood flow and Perfusion is defined as the steady-state delivery of blood to blood volume can be calculated. However, in the head and tissue. Several perfusion techniques are available; dynamic neck area a multitude of artifacts are present (e.g. volun- contrast-enhanced (DCE) perfusion, dynamic susceptibility tary/involuntary motion, breathing, air-to-tissue surface •• • • contrast (DSC) perfusion and arterial spin labeling (ASL) artifacts) [10 , 11, 12, 13 , 14 ], affecting the reliability of all yielding different parameters (see Table 2). the results acquired with DSC. DCE perfusion is most commonly used for the head and ASL is a perfusion technique without injection of con- neck area. DCE is based on the T1 relaxivity effects of trast. Arterial blood is magnetized below the volume of 123 2 Page 4 of 15 Curr Radiol Rep (2018) 6:2 Table 2 Use of functional MRI for treatment evaluation Functional Most used parameters During treatment primary After treatment After treatment lymph MRI tumor and lymph nodes primary tumor nodes sequence Diffusion DWI: ADC, ADC-ratio (= ADC /ADC Locoregional control: %ADC Tumor: Metastatic lymph nodes: 2000 1000 : tumor and lymph nodes ADC;; and b800- ADC ;; and b800–1000 9 100%) Locoregional failure: %ADC ; 1000 :: ::. Suggested ADC -3 IVIM: D, D*, f tumor and lymph nodes. Cut- cut-off 1.1 9 10 Peritumoral off range 14–24% [32, 55, 65] mm /s DKI: skewness of distribution inflammation: Reactive lymph nodes: ADC ;/= and b800–1000 =/: ADC ;/= and b800–1000 =/: Necrosis/ apoptosis: IVIM-derived D and f contradicting literature ADC :/:: and [35, 38] b800–1000 :/:: DKI: ? Edema: ADC =/: and b800–1000 =/: Fibrosis: ADC = and b800–1000 = IVIM/DKI: ? Perfusion DCE: AUC, Ktrans, rate constant, Local control: Tumor: Metastatic lymph nodes: extravascular volume and plasma space Ktrans =/: [42]. Ktrans :, blood blood flow :, blood volume or flow volume :, blood volume :, Ktrans ? AUC =/: [42]. DSC: blood volume, blood flow, mean transit flow :, wash out Reactive lymph nodes: Plasma flow =/: [66]. time, wash out : blood flow =/:, blood Local failure: ASL: blood flow Peritumoral volume =/:, Ktrans ? Ktrans ;/= [42] inflammation: =/ AUC ;/= [42] Necrosis/ Plasma flow ;/= [66] apoptosis: all ; Regional control (lymph Edema: ;/=/: ? nodes): ? Fibrosis: all ; Spectroscopy Concentration of lactate (1.3 ppm), N-acetyl- Increased choline, decreased creatine and increase choline/creatine ratio in aspartate (2.0 ppm), creatine (3.0 ppm) and primary tumor recurrence and nodal metastasis is suggested, although choline (3.2 ppm). Ratios can be calculated insufficient data available to reliably provide insight [62–64] See technique section of the paper for explanation of the most commonly used parameters. Suggested cut-off values are given if available. High values are indicated as :, low values are indicated as ; and intermediate values are indicated as = . References are given if relevant with numbers corresponding to the reference listed in the text ADC apparent diffusion coefficient, ASL arterial spin labeling, AUC area under the curve, IVIM intravoxel incoherent motion, D diffusion of water molecules, D* perfusion contribution to the signal decay, DCE dynamic contrast enhanced, DKI diffusion kurtosis imaging, DSC dynamic susceptibility enhanced, f contribution of perfusion to the diffusion signal, Ktrans capillary permeability, ppm parts per million interest. After a certain period, the magnetized blood flows head and neck cancer using an Locker–Locker sequence into the volume of interest and its derived signal is mea- [41] or a pseudo-continuous sequence [42]. sured. Blood flow can be calculated, which could reflect neovascularity and angiogenic activity of malignancy [32]. MR Spectroscopy ASL also uses T1 relaxation, but is challenging as timing of the signal read-out should be precise. Acquiring the MRS is a technique that detects the presence of specific volume of interest too late, and the magnetized arterial metabolites. Different metabolites have small differences blood has already passed. However, ASL is feasible in in their intrinsic vibration frequency and thereby result in small differences in signal of H protons. Spectroscopy is 123 Curr Radiol Rep (2018) 6:2 Page 5 of 15 2 thus well-suited to detect changes in the components of Also plasma flow has shown to react in patients under- tissue due to tumor after suppression of the abundant water going induction chemotherapy for the regional tumor [47]. signal [43]. Single voxel and multivoxel techniques are The median baseline tumor plasma flow was 53 ml/100 ml/ able to characterize tissue including the measurements of min in 25 responders and 24 ml/100 ml/min in 12 non- lactate, N-acetylaspartate, creatine and choline. Spec- responders. In lymph nodes, differences were not signifi- troscopy should be regarded as complimentary to the cantly different between non-responders and responders already acknowledged functional MRI techniques in [47]. After appropriate validation, this method may be assessing HNSCC. potentially used to guide treatment modification in patients. MR Spectroscopy Response Evaluation During Therapy To the best of our knowledge, only one in vitro study of Diffusion Weighted Imaging tumor specimens by has shown significantly elevated pre- treatment choline-to-creatine ratios in a poor response A rise in ADC is seen after the treatment in HNSCC group, but these findings could not be confirmed in an (Fig. 1) and can be seen already in the first few weeks in vivo human study using choline/creatine ratios as well •• • [24 , 29 ]. This percentage increase in ADC has been choline/water ratios [48]. •• shown to be a predictor of treatment response [24 ]. A smaller mean ADC in the first 3 weeks after treatment start was shown in patients with disease failure compared to Imaging Primary Tumor Site Post-therapy • • those with disease control [29 , 33 , 44]. Three other studies found thresholds of\ 14–24% to be predictive for Diffusion Weighted Imaging regional failure in using clinical outcome data with at least 2 years follow-up [26, 45, 46]. However, it must be noted Anatomical MRI is mandatory for an accurate delineation that imaging is generally not performed within the first of anatomical details (see Table 1). However, anatomical couple of weeks in standard clinical practice. MRI is hindered by interpretation difficulties in the •• It is of great importance to interpret ADC analysis in detection of local primary tumor recurrence [10 , • • conjunction with anatomical imaging. Areas of necrosis 11, 12, 13 , 14 ]. A diffusion-derived b 800 or b1000 map may take longer to resolve than solid areas. In the interim, provides high lesion-to-background contrast, outperform- the necrosis may become organized and show a fall in ing conventional T2-weighted sequences in this aspect. The •• ADC value [24 ]. Therefore, it is critical to identify sites accompanying ADC indicates whether the high signal on of necrosis that need to be excluded from ADC analysis the b value map is indeed due to tumor recurrence if low •• [24 ]. Furthermore, the development of mature scar tissue signal is seen on the ADC map. If the high signal on the may also decrease the ADC value [27]. The same holds for b value map is accompanied by high signal on the ADC compact fibrosis which can demonstrate lowered ADC map it is not due to tumor and represents T2-shine-through, values and low to intermediate T2 signal. or increased diffusivity (see also Table 2 for interpretation of functional MRI). Fibrosis also lacks diffusion restriction MR Perfusion (Fig. 2). A large meta-analysis showed a higher diagnostic accuracy for ADC compared to anatomical MRI. Vascular HNSCCs are thought to have better treatment Anatomical MRI yielded a pooled sensitivity and speci- response compared to less vascular HNSCCs because of ficity of 84 and 82%, respectively. ADC showed a pooled better delivery of chemotherapeutic agents and greater sensitivity and specificity of 89 and 86%, respectively •• •• radiosensitivity [24 ]. On the other hand, vascular tumors [22 ]. More recent studies demonstrate a similar diag- may have a poorer outcome because they are thought they nostic accuracy for ADC values [46]. Even higher b values •• have greater metastatic potential [24 ]. Reports suggest up to b2000 do not increase the diagnostic accuracy that a fall in blood volume is associated with poor overall [44, 49]. Using both a b1000 and b2000 and ADC ratio survival. On the other hand, an increased area under the (= ADC /ADC 9 100%) can be calculated. The 2000 1000 curve is associated with local control [39]. The early rise in ADC might increase the diagnostic accuracy although ratio volume transfer (Ktrans) is speculated to result from results are variable with a sensitivity and specificity of 63 damaged blood vessels causing them to temporarily and 84%, respectively, for one study [44]. This is a small become leakier, which potentially could increase the study with 32 patients, thus should be further studied in a delivery of chemotherapeutic agents into the tumor. large population. 123 2 Page 6 of 15 Curr Radiol Rep (2018) 6:2 Fig. 1 Tumor response confirmed on diffusion. A 54-year-old patient anatomical MRI with some residual high T2 signal and enhancement. with a tumor at the retromolar trigonum showing high T2 signal, Diffusion restriction aided in the differentiation between residual enhancement and diffusion restriction before treatment. Follow-up tumor and post-therapy inflammation. Lack of diffusion restriction in 6 months after radiation therapy showed at least partial response on this patient was in keeping with post-therapy changes Diffusion restriction results from high cellularity as in studies to differentiate treatment changes from tumor tumor, but can be also induced due to inflammation and recurrence or residual with DCE or DSC perfusion are abscesses. Moreover, restricted diffusivity can be seen in lacking. Although, visual assessment is possible (see also normal structures (e.g. Waldeyer’s ring or normal lymph Table 2 for interpretation of functional MRI), further nodes) because these structures have an inherent high quantification is currently hindered by standardization of •• • • cellularity [10 , 11, 12, 13 , 14 ]. Apoptosis and tumor scan parameters and thresholds. In our experience, the area necrosis can lead to decreased cellularity resulting in an under the curve (AUC) summing the enhancement in a •• •• • increased diffusivity [24 , 25 , 29 ]. This should be kept certain voxel, delineates abnormalities most easily with in mind when interpreting DWI. high values for tumor. Relative enhancement provides more insight in the magnitude of enhancement compared MR Perfusion with the pre-contrast values. Region of interest analyses could demonstrate relative enhancements curves with the A cross-sectional study demonstrated significant differ- internal carotid artery as reference. A rapid wash-in com- ences between DCE perfusion parameters comparing the parable with the carotid artery followed by a wash out or blood volume of scar tissue and tumor recurrence in plateau phase is indicative of tumor (Fig. 3), while slowly HNSCC [50]. Its potential use in treatment follow-up was progressive enhancements indicate benign treatment also shown in a small retrospective study [51]. Although changes (Fig. 4). DSC is not the most used perfusion method in the head and neck area, a higher wash-in on DSC has been related with tumor recurrence instead of treatment changes in a prospective study [33 ]. However, diagnostic accuracy 123 Curr Radiol Rep (2018) 6:2 Page 7 of 15 2 Fig. 2 Fibrosis on follow-up MRI confirmed with diffusion. A 67-year-old patient with a T3 vallecula tumor showed fibrosis after radiation therapy with low signal on T1 and T2, no enhancement and no diffusion restriction MR Spectroscopy [54–56]. This is also demonstrated in lymph nodes between 5 and 10 mm [54–56]. However, mean ADC values for -3 2 MRS is not routinely used for the treatment evaluation of benign lymph nodes range from to 1.1 to 1.6 9 10 mm / HNSCC. However, the presence of choline as indication of s, while HNSCC metastatic nodes range between to 0.78 -3 2 •• •• proliferation and cell membrane turnover yield high and 1.1 9 10 mm /s [24 , 25 ]. A threshold of 1.1, specificity of 100%, although false-negative are frequently therefore, seems most appropriate to use, although overlap present, resulting in a very low sensitivity of 44% [52]. could result in false-positive and false-negative results. The diagnostic accuracy for post-treatment lymph nodes using the IVIM or DKI methods might be better using multiple Imaging Lymph Nodes Post-therapy b values. This remains speculative currently as diagnostic accuracy studies are lacking post-therapy. The values of the Diffusion Weighted Imaging known decrease of kurtosis of lymph nodes during treat- ment [57, 58] should be further established. The IVIM- Treatment evaluation of regional lymph node is less stud- derived D values represent pure diffusion without perfusion ied than the primary tumor site. A higher diagnostic components. Significantly higher D values are demon- •• accuracy for ADC over anatomical MRI is suggested [22 , strated in patients with regional failure in line with the • • 53–55]. Anatomical MRI sensitivity and specificity ranged ADC results [29 , 33 ]. However, another study showed no •• between 67–90 and 33–97%, respectively [22 ]. For ADC, significant rise in D values but a higher initial f value this was 78 and 88% in one study and 73 and 100% in (perfusion fraction) in locoregional failure compared to another study [45, 53]. However, the difference was sta- locoregional control [53]. tistically not significant. Benign lymph nodes demonstrate higher ADC values compared to malignant lymph nodes 123 2 Page 8 of 15 Curr Radiol Rep (2018) 6:2 Fig. 3 Tumor recurrence differentiated using diffusion and perfu- findings could be due to both tumor recurrence as well as inflamma- sion. A 57-year-old patient with a total resection of a pT2N0Mx tion. Functional MRI demonstrated findings in keeping with tumor lateral tongue carcinoma. Because of small free resection margins, a recurrence. Diffusion restriction was shown with high b1000 and low second resection was performed 1 month later with a submandibulec- ADC values. Perfusion demonstrated increased AUC. Relative tomy and free radial forearm flap reconstruction. Anatomical MRI enhancement of the tumor (blue) showed a wash-in comparable to showed changes during follow-up 6 months after resection with high the carotid artery (purple) with plateau phase indicative for tumor. signal on T2 with and without fat suppression. There is enhancement Tumor recurrence was pathologically confirmed (Color figure online) post gadolinium. Anatomical MRI was difficult to interpret as these Perfusion-Weighted Imaging DSC MRI perfusion. The capillary permeability (Ktrans) correlates with the hypoxia-induced transcription factor in A few recent studies have demonstrated differences in the tissue, which is known to stimulate angiogenesis [59]. perfusion parameters between benign lymph nodes and However, interpretation of MR perfusion in post-therapy •• • malignant lymph nodes [24 , 33 , 41, 42]. Perfusion of lymph nodes is difficult and it remains to be elucidated nodal metastasis might be increased (Fig. 5). Metastatic whether differentiation of malignant and benign lymph lymph nodes demonstrate higher blood flow and blood nodes can be done reliable (Fig. 6). volume compared to benign lymph nodes on CT perfusion [41, 42], which thus would be expected to be similar for 123 Curr Radiol Rep (2018) 6:2 Page 9 of 15 2 Fig. 4 Benign perfusion profile post-therapy. A 45-year-old patient suppression. A benign perfusion profile is seen with slowly progres- with a T1 tongue carcinoma after resection. The primary site showed sive relative enhancement (Color figure online) some enhancement after gadolinium injection on the T1 with fat MR Spectroscopy Second, it must be stressed that functional MRI remains technically challenging to perform due to artifacts (i.e. Acquiring MRS in lymph nodes is currently not clinically breathing, swallowing, involuntary motion and air-tissue •• • • •• •• • • applicable as the region of interest should be placed sep- interfaces) [10 , 11, 12, 13 , 14 , 24 , 25 , 29 , 32, 33 ]. arately on each suspicious lymph node by a radiologist on Moreover, acquisition parameters have yet to be stan- site. If the technically challenges are overcome, the dardized. Examples of protocols for the functional MRI increased choline, decreased creatine and subsequently sequences of the head and neck are described and could be •• increased choline/creatine ratio of metastatic nodes need to used as a guide when implementing these sequences [24 ]. be confirmed in larger studies [60–62]. Diffusion-derived interpretation is mainly done using mean ADC values. Diffusion showed good reproducibility for baseline scans for the ADC value of the primary tumor Limitations and nodal metastasis [63]. The reproducibility of the ADC during treatment is also suggested to be good [64]. Mean The limitations and potential pitfalls of the functional MRI values of the tumor or metastatic lymph node are not sequences should be kept in mind during the interpretation. representative when they consist of both highly and poorly First, the lack of anatomical information at high b values in cellular (necrotic) portions. Mean ADC values should be DWI is a drawback because of suppressed signal in most of measured in the areas with high cellularity only to over- •• •• the normal tissues. Therefore, DWI should not be inter- come this limitation [24 , 25 ]. Even then, ADC inter- preted alone, but in correlation with anatomical sequences. pretation remains challenging. A recent study suggested a This is also true for perfusion and spectroscopy which reduced field of view (FOV) might increase accuracy [65]. means that all functional MRI sequences can never be used Moreover, it has been suggested that multiple b values are without the use of anatomical sequences. Moreover, all more accurate as this method is able to distinguish the functional sequences are currently hindered by high vari- perfusion component resulting in a pure diffusion value. ability of cut-offs and parameters used. This perfusion might influence the ADC value, although 123 2 Page 10 of 15 Curr Radiol Rep (2018) 6:2 Fig. 5 Nodal metastasis with positive diffusion and perfusion. Same (arrow head). Peripheral enhancement corresponded with high AUC patient as in Fig. 4 showing a lymph node metastasis with necrotic (arrows) (Color figure online) center with high T2 signal and no enhancement or increased perfusion some consider the influence of perfusion below clinical artifacts and currently best suited to perform in patients relevance [65]. As the clinical implication of multiple with HNSCC. Post-processing of perfusion is more com- b values is not yet firmly established, the acquisition of plex due to the nonspecific nature of vessel leakage multiple b values in clinical setting can be questioned. resulting in possible false-negatives and false-positive However, multiple b values are clearly preferred in a results. Perfusion post-processing also has a greater range research setting. of methods and functional parameters for analysis that are •• •• With respect to DCE perfusion, an increased scan available if compared to DWI [24 , 25 , 41]. This adds to duration with approximately 7–10 min is most hindering the complexity of perfusion imaging and its clinical clinical applicability next to the potential artifact as dis- implementation. •• cussed above [24 ]. DCE perfusion is least influenced by 123 Curr Radiol Rep (2018) 6:2 Page 11 of 15 2 Fig. 6 Normal lymph node and nodal metastasis with diffusion and enlarged and demonstrated slightly restricted diffusion as is also seen perfusion. A 66-year-old patient with a right sided pT1N1Mx floor of in normal lymphoid tissue. Perfusion showed a high AUC and relative the mouth SCC demonstrated recurrent lymph nodes after postoper- enhancement with a rapid wash-in with plateau phase for both lymph ative radiation therapy. An enlarged metastasis lymph node was seen nodes, although most pronounced in the metastatic lymph node. on the right side with diffusion restriction and increased relative Interpretation of the perfusion of lymph nodes remains difficult and enhancement and AUC (arrow). A contralateral lymph node was not should be further investigated (Color figure online) Studies with regards to MRS suggest a higher choline- should be placed by a radiologist to ensure correct place- to-creatine ratio in patients with poor prognosis, which ment in the anatomically difficult head and neck area. corresponds with expected high rates of proliferation and Furthermore, motion artifact from the carotid artery, long membrane biosynthesis in aggressive tumors (increased scan durations and complex post-processing hinders clini- rate of metabolism) [48]. However, MRS is not commonly cal applicability [52, 66]. used due to its technical challenges. The region of interest 123 2 Page 12 of 15 Curr Radiol Rep (2018) 6:2 Differentiation between malignant and benign post- Future Developments and Challenges treatment effects in HNSCC is of importance to guide Differentiation between malignancy and benign post-treat- clinical decisions. As anatomical MRI is not able to reli- ably differentiate post-therapy effect from tumor, func- ment effects such as fibrosis in HNSCC is of importance to guide clinical decisions. The head and neck is an area sen- tional techniques have been investigated and shown to be promising. This review showed that DWI can increase the sitive for artifacts and functional MR imaging requires advanced MRI post-processing software to evaluate diagnostic accuracy significantly for the primary tumor site and might also increase the diagnostic accuracy for the HNSCC. Combined functional sequences are required to fully appreciate HNSCC post-therapy, in addition to the region lymph nodes after therapy. Diffusion is most easy to implement and is recommended to perform routinely in a necessary anatomical sequences. This would result in long clinical setting in HNSCC follow-up. Its use during treat- scan durations, but new developments could overcome time ment to predict outcome is interesting, but evidence is too issues. A possible role of hybrid integrated PET/MR imag- low to implement. ing might be demonstrated offering the potential to acquired anatomical and function data using different modalities. Although perfusion parameters might be increased in tumor residual or recurrence and nodal metastasis, its However, future research is needed to evaluate PET/MRI and its appropriate applications compared to existing tech- diagnostic accuracy has yet to be established and is not routinely used clinically. DCE is least hindered by artifact niques [67] and whether PET/MRI is of greater clinical value than PET/CT and retrospective image fusion tech- and might be performed clinically if local experience is present. niques [68]. HNSCC is common and local residual and/or Spectroscopy research is promising, but evidence is too recurrence and nodal metastasis are seen in many patients. sparse for clinical implementation in the near future. The Diffusion is already frequently used. However, diffusion role of hybrid PET/MR imaging is to be established. with multiple b values and perfusion required further con- firmation of their added value in the post-therapy setting Acknowledgements This study was funded by a Mandema sti- before wide-spread implementation. This is even more the pendium from the University of Groningen (AH). case for spectroscopy. Future studies should focus on the added value of the different functional MRI sequences Compliance with Ethical Guidelines preferable by large prospective longitudinal multicenter Conflict of interest Roland Nooij, Jan Hof, Peter Jan van Laar, and studies comparing all sequences in the same population. Anouk van der Hoorn each declare that they have no conflicts of These studies are needed to assess the diagnostic accuracy interest. of the functional MRI sequences separately and in combina- tion. Another important aspect of these studies should be to Human and Animal Rights and Informed Consent This article does not contain any studies with human or animal subjects per- define the optimum time for assessment of metabolic and formed by the authors. physiological MRI parameters using functional techniques. The functional parameters should be tested in relation to the Open Access This article is distributed under the terms of the histopathological changes in HNSCC, treatment effects and Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted patient outcomes. These new trials must result in standardized use, distribution, and reproduction in any medium, provided you give cut-off values and ratios for the anatomical and functional appropriate credit to the original author(s) and the source, provide a MRI sequences to precisely define post-therapy changes from link to the Creative Commons license, and indicate if changes were tumor progression. The use of standardized cut-off values made. might remain arbitrary because of the use of different MRI systems. Nevertheless, it would be a valuable guideline for References the clinician in daily practice. Despite these possible limita- tions, implications into clinical practice would be an impor- Paper of particular interest, published recently, have been tant step in making an accurate treatment decisions for highlighted as: HNSCC patients. Of importance •• Of major importance Conclusions 1. Fitzmaurice C, Allen C, Barber RM, Barregard L, Bhutta ZA, Brenner H, et al. Global, regional, and national cancer incidence, In summary, this review analyzed the role of specific mortality, years of life lost, years lived with disability, and dis- functional MRI modalities in differentiating benign post- ability-adjusted life-years for 32 cancer groups, 1990 to 2015: a treatment effects from recurrence and/or residual malig- systematic analysis for the global burden of disease study. JAMA nancy and metastases in HNSCC. 123 Curr Radiol Rep (2018) 6:2 Page 13 of 15 2 Oncol. 2017;3:524–48. https://doi.org/10.1001/jamaoncol.2016. imaging. PET/CT and anatomical MRI are discussed. Functional 5688. MRI is not discussed. 2. Gatta G, Botta L, Sanchez MJ, Anderson LA, Pierannunzio D, 15. Omura G, Saito Y, Ando M, Kobayashi K, Ebihara Y, Yamasoba Licitra L. Prognoses and improvement for head and neck cancers T, et al. Salvage surgery for local residual or recurrent pharyngeal diagnosed in Europe in early 2000s: the EUROCARE-5 popula- cancer after radiotherapy or chemoradiotherapy. Laryngoscope. tion-based study. Eur J Cancer. 2015;51:2130–43. https://doi.org/ 2014;124:2075–80. https://doi.org/10.1002/lary.24695. 10.1016/j.ejca.2015.07.043. 16. Camisasca DR, Silami MANC, Honorato J, Dias FL, de Faria 3. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA PAS, Lourenc¸o SdQC. Oral squamous cell carcinoma: clinico- Cancer J Clin. 2017;67:7–30. https://doi.org/10.3322/caac.21387. pathological features in patients with and without recurrence. 4. Jou A, Hess J. Epidemiology and molecular biology of head and ORL J Otorhinolaryngol Relat Spec. 2011;73:170–6. https://doi. neck cancer. Oncol Res Treat. 2017;40:328–32. https://doi.org/ org/10.1159/000328340. 10.1159/000477127. 17. Guo T, Qualliotine JR, Ha PK, Califano JA, Kim Y, Saunders JR, 5. Argiris A, Karamouzis MV, Raben D, Ferris RL. Head and neck et al. Surgical salvage improves overall survival for patients with cancer. Lancet. 2008;371:1695–709. https://doi.org/10.1016/ HPV-positive and HPV-negative recurrent locoregional and dis- S0140-6736(08)60728-X. tant metastatic oropharyngeal cancer. Cancer. 2015;121:1977–84. 6. Seiwert TY, Salama JK, Vokes EE. The chemoradiation para- https://doi.org/10.1002/cncr.29323. digm in head and neck cancer. Nat Clin Pract Oncol. 18. Kim S-A, Roh J-L, Kim JS, Lee JH, Lee SH, Choi S-H, et al. 2007;4:156–71. https://doi.org/10.1038/ncponc0750. (18)F-FDG PET/CT surveillance for the detection of recurrence 7. Cooper JS, Pajak TF, Forastiere A, Jacobs J, Fu KK, Ang KK, in patients with head and neck cancer. Eur J Cancer. et al. Precisely defining high-risk operable head and neck tumors 2017;72:62–70. https://doi.org/10.1016/j.ejca.2016.11.009. based on RTOG#85-03 and #88-24: targets for postoperative 19. Castaldi P, Leccisotti L, Bussu F, Micciche` F, Rufini V. Role radiochemotherapy. Head Neck. 1998;20:588–94. https://doi.org/ of 18F-FDG PET-CT in head and neck squamous cell carcinoma. 10.1002/(SICI)1097-0347(199810)20:7\588:AID-HED2[3.0. Acta Otorhinolaryngol Ital. 2013;33:1–8. CO;2-F. 20. Helsen N, Roothans D, van den Heuvel B, van den Wyngaert T, 8. Jacob LA, Chaudhuri T, Lakshmaiah KC, Babu KG, Dasappa L, van den Weyngaert D, Carp L, et al. 18F-FDG-PET/CT for the Babu MCS, et al. Current status of systemic therapy for recurrent detection of disease in patients with head and neck cancer treated and/or metastatic squamous cell carcinoma of the head and neck. with radiotherapy. PLoS ONE. 2017;12:e0182350. https://doi. Indian J Cancer. 2016;53:471–7. https://doi.org/10.4103/0019- org/10.1371/journal.pone.0182350. 509X.204786. 21. Mak D, Corry J, Lau E, Rischin D, Hicks RJ. Role of FDG-PET/ 9. Overgaard J, Hansen HS, Specht L, Overgaard M, Grau C, CT in staging and follow-up of head and neck squamous cell Andersen E, et al. Five compared with six fractions per week of carcinoma. Q J Nucl Med Mol Imaging. 2011;55:487–99. •• conventional radiotherapy of squamous-cell carcinoma of head 22. van der Hoorn A, van Laar PJ, Holtman GA, Westerlaan HE. and neck: DAHANCA 6 and 7 randomised controlled trial. Diagnostic accuracy of magnetic resonance imaging techniques Lancet. 2003;362:933–40. https://doi.org/10.1016/S0140- for treatment response evaluation in patients with head and neck 6736(03)14361-9. tumors, a systematic review and meta-analysis. PLoS One •• 10. Saito N, Nadgir RN, Nakahira M, Takahashi M, Uchino A, 2017;12:e0177986. https://doi.org/10.1371/journal.pone. Kimura F, et al. Posttreatment CT and MR imaging in head and 0177986. A meta-analysis summarizing all available diagnostic neck cancer: What the radiologist needs to know. Radiographics accuracy of DWI studies for the treatment evaluation of SCC 2012;32:1261–82. https://doi.org/10.1148/rg.325115160. Normal head and neck tumors for the primary tumor site and lymph postoperative imaging findings and complications in patients with nodes. HNSCC on CT and anatomical MRI are demonstrated. Some 23. de Bondt RBJ, Nelemans PJ, Bakers F, Casselman JW, Peutz- examples of the applicability of DWI are also discussed from a Kootstra C, Kremer B, et al. Morphological MRI criteria improve radiologist perspective. the detection of lymph node metastases in head and neck squa- 11. Maroldi R, Ravanelli M, Farina D. Magnetic resonance for mous cell carcinoma: multivariate logistic regression analysis of laryngeal cancer. Curr Opin Otolaryngol Head Neck Surg. MRI features of cervical lymph nodes. Eur Radiol. 2014;22:131–9. https://doi.org/10.1097/MOO.0000000000000036. 2009;19:626–33. https://doi.org/10.1007/s00330-008-1187-3. •• 12. Bhatnagar P, Subesinghe M, Patel C, Prestwich R, Scarsbrook 24. Jansen JFA, Parra C, Lu Y, Shukla-Dave A. Evaluation of head AF. Functional imaging for radiation treatment planning, and neck tumors with functional MR imaging. Magn Reson response assessment, and adaptive therapy in head and neck Imaging Clin N Am 2016;24:123-33. https://doi.org/10.1016/j. cancer. Radiographics. 2013;33:1909–29. https://doi.org/10. mric.2015.08.011. This review describes the use of functional 1148/rg.337125163. MRI sequences in head and neck cancer including detailed 13. Al-Shwaiheen FA, Wang SJ, Uzelac A, Yom SS, Ryan WR. description of protocols. •• The advantages and drawbacks of routine magnetic resonance 25. King AD, Thoeny HC. Functional MRI for the prediction of imaging for long-term posttreatment locoregional surveillance of treatment response in head and neck squamous cell carcinoma: oral cavity squamous cell carcinoma. Am J Otolaryngol potential and limitations. Cancer Imaging 2016;16:23. https://doi. 2015;36:415–23. https://doi.org/10.1016/j.amjoto.2015.01.024. org/10.1186/s40644-016-0080-6. Review of role and potential of Analysis of routinely performed MRI with regards to treated functional MRI during and after treatment, focussing on DWI squamous cell carcinoma of the oral cavity. Advantages and (also grants insight in % rise ADC early in treatment in unfa- drawbacks with regards to false positives and false negatives on vourable treatment outcome) and Ktrans from DCE-MRI. Also MRI in patients without suspicious clinical or exam findings are discusses possibility of MRS. discussed. 26. King AD, Chow K-K, Yu K-H, Mo FKF, Yeung DKW, Yuan J, 14. Gage KL, Thomas K, Jeong D, Stallworth DG, Arrington JA. et al. Head and neck squamous cell carcinoma: diagnostic per- Multimodal imaging of head and neck squamous cell carcinoma. formance of diffusion-weighted MR imaging for the prediction of Cancer Control 2017;24:172–9. Evaluation of advanced HNSCC treatment response. Radiology. 2013;266:531–8. https://doi.org/ for treatment planning and follow up through multimodal 10.1148/radiol.12120167. 123 2 Page 14 of 15 Curr Radiol Rep (2018) 6:2 27. Marzi S, Piludu F, Sanguineti G, Marucci L, Farneti A, Terrenato therapy for oral cancer. J Magn Reson Imaging. 2012;36:589–97. I, et al. The prediction of the treatment response of cervical nodes https://doi.org/10.1002/jmri.23704. using intravoxel incoherent motion diffusion-weighted imaging. 39. Cao Y, Popovtzer A, Li D, Chepeha DB, Moyer JS, Prince ME, Eur J Radiol. 2017;92:93–102. https://doi.org/10.1016/j.ejrad. et al. Early prediction of outcome in advanced head-and-neck 2017.05.002. cancer based on tumor blood volume alterations during therapy: a 28. Liang L, Luo X, Lian Z, Chen W, Zhang B, Dong Y, et al. Lymph prospective study. Int J Radiat Oncol Biol Phys. node metastasis in head and neck squamous carcinoma: efficacy 2008;72:1287–90. https://doi.org/10.1016/j.ijrobp.2008.08.024. of intravoxel incoherent motion magnetic resonance imaging for 40. Ng S-H, Lin C-Y, Chan S-C, Yen T-C, Liao C-T, Chang JT-C, the differential diagnosis. Eur J Radiol. 2017;90:159–65. https:// et al. Dynamic contrast-enhanced MR imaging predicts local doi.org/10.1016/j.ejrad.2017.02.039. control in oropharyngeal or hypopharyngeal squamous cell car- 29. Paudyal R, Oh JH, Riaz N, Venigalla P, Li J, Hatzoglou V, et al. cinoma treated with chemoradiotherapy. PLoS ONE. Intravoxel incoherent motion diffusion-weighted MRI during 2013;8:e72230. https://doi.org/10.1371/journal.pone.0072230. chemoradiation therapy to characterize and monitor treatment 41. Fujima N, Kudo K, Yoshida D, Homma A, Sakashita T, Tsuka- response in human papillomavirus head and neck squamous cell hara A, et al. Arterial spin labeling to determine tumor viability in carcinoma. J Magn Reson Imaging 2017;45:1013–23. https://doi. head and neck cancer before and after treatment. J Magn Reson org/10.1002/jmri.255. This study showed the utility of IVIM DWI Imaging. 2014;40:920–8. https://doi.org/10.1002/jmri.24421. for early assessment of treatment response of chemoradiation 42. Fujima N, Kudo K, Tsukahara A, Yoshida D, Sakashita T, therapy in HNSCC patients. D and f values as indicators are also Hommaet A, et al. Measurement of tumor blood flow in head and discussed. neck squamous cell carcinoma by pseudo-continuous arterial spin 30. Sasaki M, Sumi M, Eida S, Katayama I, Hotokezaka Y, Naka- labeling: comparison with dynamic contrast-enhanced MRI. mura T. Simple and reliable determination of intravoxel inco- J Magn Reson Imaging. 2015;41:983–91. https://doi.org/10.1002/ herent motion parameters for the differential diagnosis of head jmri.24885. and neck tumors. PLoS ONE. 2014;9:e112866. https://doi.org/10. 43. Devpura S, Barton KN, Brown SL, Palyvoda O, Kalkanis S, Naik 1371/journal.pone.0112866. VM, et al. Vision 20/20: the role of Raman spectroscopy in early 31. Zheng D, Chen Y, Liu X, Chen Y, Xu L, Ren W, et al. Early stage cancer detection and feasibility for application in radiation response to chemoradiotherapy for nasopharyngeal carcinoma therapy response assessment. Med Phys. 2014;41:050901. https:// treatment: value of dynamic contrast-enhanced 3.0 T MRI. doi.org/10.1118/1.4870981. J Magn Reson Imaging. 2015;41:1528–40. https://doi.org/10. 44. Hwang I, Choi SH, Kim Y-J, Kim KG, Lee AL, Yun TJ, et al. 1002/jmri.24723. Differentiation of recurrent tumor and posttreatment changes in 32. Sumi M, Nakamura T. Head and neck tumors: assessment of head and neck squamous cell carcinoma: application of high perfusion-related parameters and diffusion coefficients based on b-value diffusion-weighted imaging. AJNR Am J Neuroradiol. the intravoxel incoherent motion model. AJNR Am J Neurora- 2013;34:2343–8. https://doi.org/10.3174/ajnr.A3603. diol. 2013;34:410–6. https://doi.org/10.3174/ajnr.A3227. 45. Vandecaveye V, Dirix P, de Keyzer F, de Beeck K, vander 33. Abdel Razek AAK, Gaballa G, Ashamalla G, Alashry MS, Poorten V, Roebben I, et al. Predictive value of diffusion- Nada N. Dynamic susceptibility contrast perfusion-weighted weighted magnetic resonance imaging during chemoradiotherapy magnetic resonance imaging and diffusion-weighted magnetic for head and neck squamous cell carcinoma. Eur Radiol. resonance imaging in differentiating recurrent head and neck 2010;20:1703–14. https://doi.org/10.1007/s00330-010-1734-6. cancer from postradiation changes. J Comput Assist Tomogr 46. Matoba M, Tuji H, Shimode Y, Toyoda I, Kuginuki Y, Miwa K, 2015;39:849–54. https://doi.org/10.1097/rct.0000000000000311. et al. Fractional change in apparent diffusion coefficient as an This study shows the possibilities of ADC-values and DSC per- imaging biomarker for predicting treatment response in head and fusion-weighted MR imaging, in the differentiation of recurrent neck cancer treated with chemoradiotherapy. AJNR Am J Neu- head and neck cancer from postradiation changes. roradiol. 2014;35:379–85. https://doi.org/10.3174/ajnr.A3706. 34. Verduijn GM, Bartels LW, Raaijmakers CPJ, Terhaard CHJ, 47. Bernstein JM, Kershaw LE, Withey SB, Lowe NM, Homer JJ, Pameijer FA, van den Berg CAT. Magnetic resonance imaging Slevin NJ, et al. Tumor plasma flow determined by dynamic protocol optimization for delineation of gross tumor volume in contrast-enhanced MRI predicts response to induction hypopharyngeal and laryngeal tumors. Int J Radiat Oncol Biol chemotherapy in head and neck cancer. Oral Oncol. Phys. 2009;74:630–6. https://doi.org/10.1016/j.ijrobp.2009.01. 2015;51:508–13. https://doi.org/10.1016/j.oraloncology.2015.01. 014. 013. 35. Bertrand M, Tollard E, Francois A, Bouchetetemble P, Marie PJ, 48. Bezabeh T, Odlum O, Nason R, Kerr P, Sutherland D, Patel R, Dehesdin D, et al. CT scan, MR imaging and anatomopathologic et al. Prediction of treatment response in head and neck cancer by correlation in the glottic carcinoma T1-T2. Rev Laryngol Otol magnetic resonance spectroscopy. AJNR Am J Neuroradiol. Rhinol (Bord). 2010;131:51–7. 2005;26:2108–13. 36. Ravanelli M, Farina D, Rizzardi P, Botturi E, Prandolini P, 49. Acampora A, Manzo G, Fenza G, Busto G, Serino A, Manto A. Mangili S, et al. MR with surface coils in the follow-up after High b-Value Diffusion MRI to differentiate recurrent tumors endoscopic laser resection for glottic squamous cell carcinoma: from posttreatment changes in head and neck squamous cell feasibility and diagnostic accuracy. Neuroradiology. carcinoma: a single center prospective study. Biomed Res Int. 2013;55:225–32. https://doi.org/10.1007/s00234-012-1128-3. 2016;2016:2865169. https://doi.org/10.1155/2016/2865169. 37. Zbaren P, Becker M, Lang H. Pretherapeutic staging of laryngeal 50. Choi SH, Lee JH, Choi YJ, Park JE, Sung YS, Kim N, et al. carcinoma. Clinical findings, computed tomography, and mag- Detection of local tumor recurrence after definitive treatment of netic resonance imaging compared with histopathology. Cancer. head and neck squamous cell carcinoma: histogram analysis of 1996;77:1263–73. https://doi.org/10.1001/archotol.1997. dynamic contrast-enhanced T1-weighted perfusion MRI. AJR 01900090016003. Am J Roentgenol. 2017;208:42–7. https://doi.org/10.2214/AJR. 38. Chikui T, Kitamoto E, Kawano S, Sugiura T, Obara M, Simonetti 16.16127. AW, et al. Pharmacokinetic analysis based on dynamic contrast- 51. Choi YJ, Lee JH, Sung YS, Yoon RG, Park JE, Nam SY, et al. enhanced MRI for evaluating tumor response to preoperative Value of dynamic contrast-enhanced MRI to detect local tumor 123 Curr Radiol Rep (2018) 6:2 Page 15 of 15 2 recurrence in primary head and neck cancer patients. Medicine. 60. King AD, Yeung DKW, Ahuja AT, Leung SF, Tse GMK, van 2016;95:e3698. https://doi.org/10.1097/MD.0000000000003698. Hasselt AC. In vivo proton MR spectroscopy of primary and 52. King AD,Yeung DKW, YuK-H,MoFKF,HuC-W,Bhatia KS, nodal nasopharyngeal carcinoma. AJNR Am J Neuroradiol. et al. Monitoring of treatment response after chemoradiotherapy for 2004;25:484–90. head and neck cancer using in vivo 1H MR spectroscopy. Eur 61. Shah GV, Gandhi D, Mukherji SK. Magnetic resonance spec- Radiol. 2010;20:165–72. https://doi.org/10.1007/s00330-009-1531-2. troscopy of head and neck neoplasms. Top Magn Reson Imaging. 53. Berrak S, Chawla S, Kim S, Quon H, Sherman E, Loevner LA, 2004;15:87–94. et al. Diffusion weighted imaging in predicting progression free 62. Chawla S, Kim S, Loevner LA, Quon H, Wang S, Mutale F, et al. survival in patients with squamous cell carcinomas of the head Proton and phosphorous MR spectroscopy in squamous cell and neck treated with induction chemotherapy. Acad Radiol. carcinomas of the head and neck. Acad Radiol. 2009;16:1366–72. 2011;18:1225–32. https://doi.org/10.1016/j.acra.2011.06.009. https://doi.org/10.1016/j.acra.2009.06.001. 54. Jin GQ, Yang J, Liu LD, Su DK, Wang DP, Zhao SF, et al. The 63. Hoang JK, Choudhury KR, Chang J, Craciunescu OI, Yoo DS, diagnostic value of 1.5-T diffusion-weighted MR imaging in Brizel DM. Diffusion-weighted imaging for head and neck detecting 5 to 10 mm metastatic cervical lymph nodes of squamous cell carcinoma: quantifying repeatability to understand nasopharyngeal carcinoma. Medicine (Baltimore). 2016;95: early treatment-induced change. AJR Am J Roentgenol. e4286. https://doi.org/10.1097/MD.0000000000004286. 2014;203:1104–8. https://doi.org/10.2214/AJR.14.12838. 55. Holzapfel K, Duetsch S, Fauser C, Eiber M, Rummeny EJ, Gaa J. 64. Vaid S, Chandorkar A, Atre A, Shah D, Vaid N. Differentiating Value of diffusion-weighted MR imaging in the differentiation recurrent tumours from posttreatment changes in head and neck between benign and malignant cervical lymph nodes. Eur J cancers: does diffusion-weighted MRI solve the eternal dilemma? Radiol. 2009;72:381–7. https://doi.org/10.1016/j.ejrad.2008.09. Clin Radiol. 2017;72:74–83. https://doi.org/10.1016/j.crad.2016. 034. 09.019. 56. Abdel Razek AAK, Soliman NY, Elkhamary S, Alsharaway MK, 65. Vidiri A, Minosse S, Piludu F, Curione D, Pichi B, Spriano G, Tawfik A. Role of diffusion-weighted MR imaging in cervical et al. Feasibility study of reduced field of view diffusion- lymphadenopathy. Eur Radiol. 2006;16:1468–77. https://doi.org/ weighted magnetic resonance imaging in head and neck tumors. 10.1007/s00330-005-0133-x. Acta Radiol. 2017;58:292–300. https://doi.org/10.1177/ 57. Tyagi N, Riaz N, Hunt M, Wengler K, Hatzoglou V, Young R, 0284185116652014. et al. Weekly response assessment of involved lymph nodes to 66. Abdel Razek AAK, Poptani H. MR spectroscopy of head and radiotherapy using diffusion-weighted MRI in oropharynx squa- neck cancer. Eur J Radiol. 2013;82:982–9. https://doi.org/10. mous cell carcinoma. Med Phys. 2016;43:137. https://doi.org/10. 1016/j.ejrad.2013.01.025. 1118/1.4937791. 67. Torigian DA, Zaidi H, Kwee TC, Saboury B, Udupa JK, Cho ZH, 58. Hauser T, Essig M, Jensen A, Laun FB, Mu¨nter M, Maier-Hein et al. PET/MR imaging: technical aspects and potential clinical KH, et al. Prediction of treatment response in head and neck applications. Radiology. 2013;267:26–44. https://doi.org/10. carcinomas using IVIM-DWI: evaluation of lymph node metas- 1148/radiol.13121038. tasis. Eur J Radiol. 2014;83:783–7. https://doi.org/10.1016/j. 68. Loeffelbein DJ, Souvatzoglou M, Wankerl V, Martinez-Mo¨ller ejrad.2014.02.013. A, Dinges J, Schwaiger M, et al. PET-MRI fusion in head-and- 59. Jansen JFA, Carlson DL, Lu Y, Stambuk HE, Moreira AL, Singh neck oncology: current status and implications for hybrid PET/ B, et al. Correlation of a priori DCE-MRI and (1)H-MRS data MRI. J Oral Maxillofac Surg. 2012;70:473–83. https://doi.org/10. with molecular markers in neck nodal metastases: initial analysis. 1016/j.joms.2011.02.120. Oral Oncol. 2012;48:717–22. https://doi.org/10.1016/j.oraloncology. 2012.02.001. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Current Radiology Reports Springer Journals

Functional MRI for Treatment Evaluation in Patients with Head and Neck Squamous Cell Carcinoma: A Review of the Literature from a Radiologist Perspective

Download PDF
15 pages

Loading next page...
 
/lp/springer-journals/functional-mri-for-treatment-evaluation-in-patients-with-head-and-neck-UGHom0RZyr

References (77)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Imaging / Radiology
eISSN
2167-4825
DOI
10.1007/s40134-018-0262-z
Publisher site
See Article on Publisher Site

Abstract

Curr Radiol Rep (2018) 6:2 https://doi.org/10.1007/s40134-018-0262-z ENT IMAGING (A A JACOBI-POSTMA, SECTION EDITOR) Functional MRI for Treatment Evaluation in Patients with Head and Neck Squamous Cell Carcinoma: A Review of the Literature from a Radiologist Perspective 1 1 2,3 2,3 • • • Roland P. Nooij Jan J. Hof Peter Jan van Laar Anouk van der Hoorn Published online: 22 January 2018 The Author(s) 2018. This article is an open access publication Abstract Keywords MRI  Treatment evaluation  Primary tumor Purpose of review To show the role of functional MRI in Lymph nodes  Head/neck squamous cell carcinoma patients treated for head and neck squamous cell Review carcinoma. Recent findings MRI is commonly used for treatment Abbreviations evaluation in patients with head and neck tumors. How- ADC Apparent diffusion coefficient ever, anatomical MRI has its limits in differentiating ASL Arterial spin labeling between post-treatment effects and tumor recurrence. DCE Dynamic contrast enhanced Recent studies showed promising results of functional MRI DKI Diffusion kurtosis imaging for response evaluation. DSE Dynamic susceptibility enhanced DWI Diffusion weighted imaging Summary This review analyzes possibilities and limita- tions of functional MRI sequences separately to obtain HNSCC Head and neck squamous cell carcinoma insight in the post-therapy setting. Diffusion, perfusion and IVIM Intravoxel incoherent motion spectroscopy show promise, especially when utilized complimentary to each other. These functional MRI sequences aid in the early detection which might improve survival by increasing effectiveness of salvage therapy. Introduction Future multicenter longitudinal prospective studies are needed to provide standardized guidelines for the use of Head and neck cancer affects 550,000 new cases and functional MRI in daily clinical practice. 380,000 deaths worldwide annually [1–3]. Head and neck squamous cell carcinomas comprise over 90% of the head and neck carcinomas [4]. Patients frequently present with a locally advanced stage for which the current therapy is This article is part of the Topical collection on ENT Imaging. multimodal including surgery, radiation therapy and/or chemotherapy [5–8]. Many patients demonstrate unfavor- & Anouk van der Hoorn able treatment response, with locoregional recurrence seen a.van.der.hoorn@umcg.nl in about 30–60% [7]. This is in about 2/3 due to primary tumor recurrence, 1/3 due to regional nodal metastasis and Department of Radiology, Medical Spectrum Twente, Enschede, The Netherlands in 1/3 due to both primary tumor recurrence as well as regional nodal metastasis [9]. Department of Radiology, University Medical Center Groningen, University of Groningen, Hanzeplein 1, Conventional anatomical MRI techniques are commonly P. O. Box 30.001, 9700 RB Groningen, The Netherlands used for treatment evaluation, but are often not able to have Medical Imaging Center, University Medical Center reliable assess treatment response. Surgery as well as Groningen, University of Groningen, Groningen, The chemoradiotherapy induce false positives on imaging as a Netherlands 123 2 Page 2 of 15 Curr Radiol Rep (2018) 6:2 result of inducing benign changes involving architectural Higher diagnostic accuracy than 84% sensitivity and •• • • distortion, fibrosis and/or necrosis [10 , 11, 12, 13 , 14 ]. 82% specificity post-therapeutically is needed to differen- These benign treatment-induced changes should be differ- tiate treatment effects from true malignancy for the local entiated from residual and/or recurrent tumor on imaging to tumor site and the regional lymph nodes to reliable either prevent unjustified alteration in treatment plan, e.g. salvage initiate new therapy, adjust the current therapy or discon- therapy or (dis)continuation of therapy. Early detection of tinue unjustified therapy. local recurrence could lead to timely salvage therapy which can lead to an increase in overall survival [15–17]. Post-treatment surveillance can consist of ultrasound, Technical Background of Functional MRI •• • • PET-CT [18–21] and MRI [10 , 11, 12, 13 , 14 ]. Several Techniques recent studies have shown the potential usefulness of functional MRI techniques for treatment evaluation in Diffusion Weighted Imaging •• • • patients with head and neck tumors [10 , 11, 12, 13 , 14 , •• • • 22 , 23–28, 29 , 30–33 ]. Diffusion-weighted imaging is DWI measures cellular density and cytoarchitecture using used to image changes in cytoarchitecture and measure the measurement of water diffusivity. Random diffusion cellular density. Perfusion-weighted MRI techniques can results from the Brownian motion of water molecules. identify tumor-induced neovascularization. Changes in Motion of water molecules is hindered, restricted, by concentrations of metabolites are shown with magnetic interactions with other molecules and cellular barriers such resonance spectroscopy (MRS). as fibers, cell membranes and macromolecules. Diffusion This review will analyze the functional MR imaging abnormalities of water molecules thus reflect changes of sequences with regards to their possibilities and limitations tissue organization at a cellular level affecting the MR in head and neck squamous cell carcinoma. Clinical signal of a DWI sequence as can be seen in a number of •• • •• implications, applicability and possibilities of these processes including malignancy [10 , 11, 12, 13 , 14, 22 , •• •• • • sequences for treatment evaluation will be addressed. 24 , 25 , 26–28, 29 , 30, 32, 33 ]. DWI sequences are based on a T2-weighted sequence. At least two b values are needed to analyze motion of Role of Conventional Anatomical MRI in Head water. DWI is done at different b values (in s/mm ), which and Neck SCC represent the duration between the gradient pulses used. Simplified, it is the time that water is allowed to diffuse Conventional anatomical MRI techniques are used for before the distance is measured. Most commonly, a b0 and treatment evaluation. MRI is superior to CT yielding higher b800 or b1000 value are used for head and neck imaging. • • anatomical detail [11, 12, 13 , 14 , 34–37]. Anatomical Diffusion is quantified using ADC in mm /s. Having MRI to assess HNSCC should include a T1 without fat measured at least two different b values (e.g. b0 and b800), suppression, T2 with and without fat suppression and T1 the logarithm of relative signal intensity of a tissue is post-contrast with fat suppression. These sequences are plotted on the y axis against the b values on the x axis. The used to analyze certain characteristics of the primary tumor slope of the line fitted through the plots describes the ADC. • • and possible nodal involvement [11, 12, 13 , 14 , 34–37]. This mono-exponential fitting represents a rough approxi- However, anatomical MRI techniques are often unable mation of ADC and is most often used in clinical routine. to accurately identify treatment response showing a pooled This parameter is independent of the magnetic field sensitivity and specificity for local treatment response strength. Lower values indicate more restricted diffusion. •• evaluation in HNSCC of 84 and 82%, respectively [22 ]. However, mono-exponential fitted ADC values cannot This is due to benign treatment effects such as inflamma- separate the pure molecular diffusion from the motion of tion, fibrosis and necrosis as a result of surgery and water molecules in the capillary network [32]. Low b val- chemoradiotherapy. These post-therapy changes show ues are most influenced by the capillary component which overlapping signal characteristics with tumor. Most prob- influences the ADC values. Multi-exponential models lematic for the primary site is that inflammation and tumor using several b values are more suitable for accurate both show high T2 signal and enhancement after contrast quantification of diffusion without perfusion contamination injection. Lymph node assessment is most hindered by [30, 32, 33 ]. reactive lymph node that can be slightly enlarged similar to Acquiring multiple b values yields techniques such as nodal metastasis. Furthermore, normal sized nodes can still intravoxel incoherent motion (IVIM) and diffusion kurtosis contain tumor. See Table 1 for a detailed description of the imaging (DKI). IVIM imaging can distinguish between signal intensities on anatomical MRI post-treatment. pure molecular diffusion and motion of water molecules in the capillary network through a single DWI acquisition 123 Curr Radiol Rep (2018) 6:2 Page 3 of 15 2 Table 1 Use of conventional anatomical MRI for treatment evaluation Anatomical MRI Primary tumor Lymph nodes sequence T1 without fat Anatomical details Anatomical localization of node levels suppression Tumor: ; compared to fat Metastatic lymph nodes: Size :/:: (suggested cut-off [ 7–10 mm for level II and [ 5–7 mm for all other levels). Round shape Fat infiltration by tumor or inflammation: similar ;/;; Reactive lymph nodes: Size =/: (can be false-positive using above cut-off); Oval with fatty hilum Necrosis: ;; round, oval, well circumscribed Location of lymph node and level in relation to location primary tumor Fibrosis: Linear commonly ;;, but can be ;/= as well T2 with and without Fat suppression useful for the detection Fat suppression needed to identify abnormal nodes fat suppression of abnormalities T2 without fat suppression for anatomical details T2 without fat suppression for Metastatic lymph nodes: : slightly heterogeneous; more commonly an anatomical details irregular border; possible extra-nodal extension Edema, fat infiltration by tumor or Reactive lymph nodes: = inflammation: similar :/:: Necrosis: :: round, oval, well circumscribed Perineural spread: : Fibrosis: Linear commonly ;;, but can be ;/= as well. T1 post-contrast with Fat infiltration by tumor or Fat suppression needed to identify abnormal lymph nodes fat suppression inflammation: similar :/:: Metastatic lymph nodes: :/::, thick, irregular rim enhancement in case of Edema or necrosis: no enhancement necrosis Fibrosis: no enhancement after Reactive lymph nodes: =/: 6-12 months. Most commonly :/:: Perineural spread: :: High signal intensity is indicated as :, low signal intensity is indicated as ; and intermediate signal as = technique if both low b values (\ 200 s/mm ) and high contrast agents. DCE perfusion has been reported as a b values ([ 200 s/mm ) are used. The relationship between technique which is able to characterize perfusion and •• •• • signal intensities and multiple b values can be assessed. vascularization of tissues [24 , 25 , 30–33 ]. However, •• Real diffusion of water molecules (D) can be distinguished this has not always been histologically confirmed [25 , from the contribution of perfusion to the signal decay (D*) 38, 39]. Ktrans is the most commonly derived quantitative and the contribution of perfusion to the diffusion signal (f). parameters representing capillary permeability and seems •• •• Another, multiple b value method, DKI, represents the to be to most consistent parameter [24 , 25 , 40]. extent to which the diffusion pattern of the water molecules DSC perfusion exploits the susceptibility-induced signal deviates from a perfect Gaussian curve that is assumed loss after administration of contrast on T2-weighted calculating standard ADC values. Table 2 includes the sequences, most commonly a quick T2* gradient echo most commonly used parameters for the different diffusion sequence. It is based on inhomogeneity of the magnetic techniques. field during the passage of a short bolus of contrast through a capillary bed [27]. As result on the T2* sequence, blood MR Perfusion products, calcifications and aerated structures result in artificial signal loss. Mean transit time, blood flow and Perfusion is defined as the steady-state delivery of blood to blood volume can be calculated. However, in the head and tissue. Several perfusion techniques are available; dynamic neck area a multitude of artifacts are present (e.g. volun- contrast-enhanced (DCE) perfusion, dynamic susceptibility tary/involuntary motion, breathing, air-to-tissue surface •• • • contrast (DSC) perfusion and arterial spin labeling (ASL) artifacts) [10 , 11, 12, 13 , 14 ], affecting the reliability of all yielding different parameters (see Table 2). the results acquired with DSC. DCE perfusion is most commonly used for the head and ASL is a perfusion technique without injection of con- neck area. DCE is based on the T1 relaxivity effects of trast. Arterial blood is magnetized below the volume of 123 2 Page 4 of 15 Curr Radiol Rep (2018) 6:2 Table 2 Use of functional MRI for treatment evaluation Functional Most used parameters During treatment primary After treatment After treatment lymph MRI tumor and lymph nodes primary tumor nodes sequence Diffusion DWI: ADC, ADC-ratio (= ADC /ADC Locoregional control: %ADC Tumor: Metastatic lymph nodes: 2000 1000 : tumor and lymph nodes ADC;; and b800- ADC ;; and b800–1000 9 100%) Locoregional failure: %ADC ; 1000 :: ::. Suggested ADC -3 IVIM: D, D*, f tumor and lymph nodes. Cut- cut-off 1.1 9 10 Peritumoral off range 14–24% [32, 55, 65] mm /s DKI: skewness of distribution inflammation: Reactive lymph nodes: ADC ;/= and b800–1000 =/: ADC ;/= and b800–1000 =/: Necrosis/ apoptosis: IVIM-derived D and f contradicting literature ADC :/:: and [35, 38] b800–1000 :/:: DKI: ? Edema: ADC =/: and b800–1000 =/: Fibrosis: ADC = and b800–1000 = IVIM/DKI: ? Perfusion DCE: AUC, Ktrans, rate constant, Local control: Tumor: Metastatic lymph nodes: extravascular volume and plasma space Ktrans =/: [42]. Ktrans :, blood blood flow :, blood volume or flow volume :, blood volume :, Ktrans ? AUC =/: [42]. DSC: blood volume, blood flow, mean transit flow :, wash out Reactive lymph nodes: Plasma flow =/: [66]. time, wash out : blood flow =/:, blood Local failure: ASL: blood flow Peritumoral volume =/:, Ktrans ? Ktrans ;/= [42] inflammation: =/ AUC ;/= [42] Necrosis/ Plasma flow ;/= [66] apoptosis: all ; Regional control (lymph Edema: ;/=/: ? nodes): ? Fibrosis: all ; Spectroscopy Concentration of lactate (1.3 ppm), N-acetyl- Increased choline, decreased creatine and increase choline/creatine ratio in aspartate (2.0 ppm), creatine (3.0 ppm) and primary tumor recurrence and nodal metastasis is suggested, although choline (3.2 ppm). Ratios can be calculated insufficient data available to reliably provide insight [62–64] See technique section of the paper for explanation of the most commonly used parameters. Suggested cut-off values are given if available. High values are indicated as :, low values are indicated as ; and intermediate values are indicated as = . References are given if relevant with numbers corresponding to the reference listed in the text ADC apparent diffusion coefficient, ASL arterial spin labeling, AUC area under the curve, IVIM intravoxel incoherent motion, D diffusion of water molecules, D* perfusion contribution to the signal decay, DCE dynamic contrast enhanced, DKI diffusion kurtosis imaging, DSC dynamic susceptibility enhanced, f contribution of perfusion to the diffusion signal, Ktrans capillary permeability, ppm parts per million interest. After a certain period, the magnetized blood flows head and neck cancer using an Locker–Locker sequence into the volume of interest and its derived signal is mea- [41] or a pseudo-continuous sequence [42]. sured. Blood flow can be calculated, which could reflect neovascularity and angiogenic activity of malignancy [32]. MR Spectroscopy ASL also uses T1 relaxation, but is challenging as timing of the signal read-out should be precise. Acquiring the MRS is a technique that detects the presence of specific volume of interest too late, and the magnetized arterial metabolites. Different metabolites have small differences blood has already passed. However, ASL is feasible in in their intrinsic vibration frequency and thereby result in small differences in signal of H protons. Spectroscopy is 123 Curr Radiol Rep (2018) 6:2 Page 5 of 15 2 thus well-suited to detect changes in the components of Also plasma flow has shown to react in patients under- tissue due to tumor after suppression of the abundant water going induction chemotherapy for the regional tumor [47]. signal [43]. Single voxel and multivoxel techniques are The median baseline tumor plasma flow was 53 ml/100 ml/ able to characterize tissue including the measurements of min in 25 responders and 24 ml/100 ml/min in 12 non- lactate, N-acetylaspartate, creatine and choline. Spec- responders. In lymph nodes, differences were not signifi- troscopy should be regarded as complimentary to the cantly different between non-responders and responders already acknowledged functional MRI techniques in [47]. After appropriate validation, this method may be assessing HNSCC. potentially used to guide treatment modification in patients. MR Spectroscopy Response Evaluation During Therapy To the best of our knowledge, only one in vitro study of Diffusion Weighted Imaging tumor specimens by has shown significantly elevated pre- treatment choline-to-creatine ratios in a poor response A rise in ADC is seen after the treatment in HNSCC group, but these findings could not be confirmed in an (Fig. 1) and can be seen already in the first few weeks in vivo human study using choline/creatine ratios as well •• • [24 , 29 ]. This percentage increase in ADC has been choline/water ratios [48]. •• shown to be a predictor of treatment response [24 ]. A smaller mean ADC in the first 3 weeks after treatment start was shown in patients with disease failure compared to Imaging Primary Tumor Site Post-therapy • • those with disease control [29 , 33 , 44]. Three other studies found thresholds of\ 14–24% to be predictive for Diffusion Weighted Imaging regional failure in using clinical outcome data with at least 2 years follow-up [26, 45, 46]. However, it must be noted Anatomical MRI is mandatory for an accurate delineation that imaging is generally not performed within the first of anatomical details (see Table 1). However, anatomical couple of weeks in standard clinical practice. MRI is hindered by interpretation difficulties in the •• It is of great importance to interpret ADC analysis in detection of local primary tumor recurrence [10 , • • conjunction with anatomical imaging. Areas of necrosis 11, 12, 13 , 14 ]. A diffusion-derived b 800 or b1000 map may take longer to resolve than solid areas. In the interim, provides high lesion-to-background contrast, outperform- the necrosis may become organized and show a fall in ing conventional T2-weighted sequences in this aspect. The •• ADC value [24 ]. Therefore, it is critical to identify sites accompanying ADC indicates whether the high signal on of necrosis that need to be excluded from ADC analysis the b value map is indeed due to tumor recurrence if low •• [24 ]. Furthermore, the development of mature scar tissue signal is seen on the ADC map. If the high signal on the may also decrease the ADC value [27]. The same holds for b value map is accompanied by high signal on the ADC compact fibrosis which can demonstrate lowered ADC map it is not due to tumor and represents T2-shine-through, values and low to intermediate T2 signal. or increased diffusivity (see also Table 2 for interpretation of functional MRI). Fibrosis also lacks diffusion restriction MR Perfusion (Fig. 2). A large meta-analysis showed a higher diagnostic accuracy for ADC compared to anatomical MRI. Vascular HNSCCs are thought to have better treatment Anatomical MRI yielded a pooled sensitivity and speci- response compared to less vascular HNSCCs because of ficity of 84 and 82%, respectively. ADC showed a pooled better delivery of chemotherapeutic agents and greater sensitivity and specificity of 89 and 86%, respectively •• •• radiosensitivity [24 ]. On the other hand, vascular tumors [22 ]. More recent studies demonstrate a similar diag- may have a poorer outcome because they are thought they nostic accuracy for ADC values [46]. Even higher b values •• have greater metastatic potential [24 ]. Reports suggest up to b2000 do not increase the diagnostic accuracy that a fall in blood volume is associated with poor overall [44, 49]. Using both a b1000 and b2000 and ADC ratio survival. On the other hand, an increased area under the (= ADC /ADC 9 100%) can be calculated. The 2000 1000 curve is associated with local control [39]. The early rise in ADC might increase the diagnostic accuracy although ratio volume transfer (Ktrans) is speculated to result from results are variable with a sensitivity and specificity of 63 damaged blood vessels causing them to temporarily and 84%, respectively, for one study [44]. This is a small become leakier, which potentially could increase the study with 32 patients, thus should be further studied in a delivery of chemotherapeutic agents into the tumor. large population. 123 2 Page 6 of 15 Curr Radiol Rep (2018) 6:2 Fig. 1 Tumor response confirmed on diffusion. A 54-year-old patient anatomical MRI with some residual high T2 signal and enhancement. with a tumor at the retromolar trigonum showing high T2 signal, Diffusion restriction aided in the differentiation between residual enhancement and diffusion restriction before treatment. Follow-up tumor and post-therapy inflammation. Lack of diffusion restriction in 6 months after radiation therapy showed at least partial response on this patient was in keeping with post-therapy changes Diffusion restriction results from high cellularity as in studies to differentiate treatment changes from tumor tumor, but can be also induced due to inflammation and recurrence or residual with DCE or DSC perfusion are abscesses. Moreover, restricted diffusivity can be seen in lacking. Although, visual assessment is possible (see also normal structures (e.g. Waldeyer’s ring or normal lymph Table 2 for interpretation of functional MRI), further nodes) because these structures have an inherent high quantification is currently hindered by standardization of •• • • cellularity [10 , 11, 12, 13 , 14 ]. Apoptosis and tumor scan parameters and thresholds. In our experience, the area necrosis can lead to decreased cellularity resulting in an under the curve (AUC) summing the enhancement in a •• •• • increased diffusivity [24 , 25 , 29 ]. This should be kept certain voxel, delineates abnormalities most easily with in mind when interpreting DWI. high values for tumor. Relative enhancement provides more insight in the magnitude of enhancement compared MR Perfusion with the pre-contrast values. Region of interest analyses could demonstrate relative enhancements curves with the A cross-sectional study demonstrated significant differ- internal carotid artery as reference. A rapid wash-in com- ences between DCE perfusion parameters comparing the parable with the carotid artery followed by a wash out or blood volume of scar tissue and tumor recurrence in plateau phase is indicative of tumor (Fig. 3), while slowly HNSCC [50]. Its potential use in treatment follow-up was progressive enhancements indicate benign treatment also shown in a small retrospective study [51]. Although changes (Fig. 4). DSC is not the most used perfusion method in the head and neck area, a higher wash-in on DSC has been related with tumor recurrence instead of treatment changes in a prospective study [33 ]. However, diagnostic accuracy 123 Curr Radiol Rep (2018) 6:2 Page 7 of 15 2 Fig. 2 Fibrosis on follow-up MRI confirmed with diffusion. A 67-year-old patient with a T3 vallecula tumor showed fibrosis after radiation therapy with low signal on T1 and T2, no enhancement and no diffusion restriction MR Spectroscopy [54–56]. This is also demonstrated in lymph nodes between 5 and 10 mm [54–56]. However, mean ADC values for -3 2 MRS is not routinely used for the treatment evaluation of benign lymph nodes range from to 1.1 to 1.6 9 10 mm / HNSCC. However, the presence of choline as indication of s, while HNSCC metastatic nodes range between to 0.78 -3 2 •• •• proliferation and cell membrane turnover yield high and 1.1 9 10 mm /s [24 , 25 ]. A threshold of 1.1, specificity of 100%, although false-negative are frequently therefore, seems most appropriate to use, although overlap present, resulting in a very low sensitivity of 44% [52]. could result in false-positive and false-negative results. The diagnostic accuracy for post-treatment lymph nodes using the IVIM or DKI methods might be better using multiple Imaging Lymph Nodes Post-therapy b values. This remains speculative currently as diagnostic accuracy studies are lacking post-therapy. The values of the Diffusion Weighted Imaging known decrease of kurtosis of lymph nodes during treat- ment [57, 58] should be further established. The IVIM- Treatment evaluation of regional lymph node is less stud- derived D values represent pure diffusion without perfusion ied than the primary tumor site. A higher diagnostic components. Significantly higher D values are demon- •• accuracy for ADC over anatomical MRI is suggested [22 , strated in patients with regional failure in line with the • • 53–55]. Anatomical MRI sensitivity and specificity ranged ADC results [29 , 33 ]. However, another study showed no •• between 67–90 and 33–97%, respectively [22 ]. For ADC, significant rise in D values but a higher initial f value this was 78 and 88% in one study and 73 and 100% in (perfusion fraction) in locoregional failure compared to another study [45, 53]. However, the difference was sta- locoregional control [53]. tistically not significant. Benign lymph nodes demonstrate higher ADC values compared to malignant lymph nodes 123 2 Page 8 of 15 Curr Radiol Rep (2018) 6:2 Fig. 3 Tumor recurrence differentiated using diffusion and perfu- findings could be due to both tumor recurrence as well as inflamma- sion. A 57-year-old patient with a total resection of a pT2N0Mx tion. Functional MRI demonstrated findings in keeping with tumor lateral tongue carcinoma. Because of small free resection margins, a recurrence. Diffusion restriction was shown with high b1000 and low second resection was performed 1 month later with a submandibulec- ADC values. Perfusion demonstrated increased AUC. Relative tomy and free radial forearm flap reconstruction. Anatomical MRI enhancement of the tumor (blue) showed a wash-in comparable to showed changes during follow-up 6 months after resection with high the carotid artery (purple) with plateau phase indicative for tumor. signal on T2 with and without fat suppression. There is enhancement Tumor recurrence was pathologically confirmed (Color figure online) post gadolinium. Anatomical MRI was difficult to interpret as these Perfusion-Weighted Imaging DSC MRI perfusion. The capillary permeability (Ktrans) correlates with the hypoxia-induced transcription factor in A few recent studies have demonstrated differences in the tissue, which is known to stimulate angiogenesis [59]. perfusion parameters between benign lymph nodes and However, interpretation of MR perfusion in post-therapy •• • malignant lymph nodes [24 , 33 , 41, 42]. Perfusion of lymph nodes is difficult and it remains to be elucidated nodal metastasis might be increased (Fig. 5). Metastatic whether differentiation of malignant and benign lymph lymph nodes demonstrate higher blood flow and blood nodes can be done reliable (Fig. 6). volume compared to benign lymph nodes on CT perfusion [41, 42], which thus would be expected to be similar for 123 Curr Radiol Rep (2018) 6:2 Page 9 of 15 2 Fig. 4 Benign perfusion profile post-therapy. A 45-year-old patient suppression. A benign perfusion profile is seen with slowly progres- with a T1 tongue carcinoma after resection. The primary site showed sive relative enhancement (Color figure online) some enhancement after gadolinium injection on the T1 with fat MR Spectroscopy Second, it must be stressed that functional MRI remains technically challenging to perform due to artifacts (i.e. Acquiring MRS in lymph nodes is currently not clinically breathing, swallowing, involuntary motion and air-tissue •• • • •• •• • • applicable as the region of interest should be placed sep- interfaces) [10 , 11, 12, 13 , 14 , 24 , 25 , 29 , 32, 33 ]. arately on each suspicious lymph node by a radiologist on Moreover, acquisition parameters have yet to be stan- site. If the technically challenges are overcome, the dardized. Examples of protocols for the functional MRI increased choline, decreased creatine and subsequently sequences of the head and neck are described and could be •• increased choline/creatine ratio of metastatic nodes need to used as a guide when implementing these sequences [24 ]. be confirmed in larger studies [60–62]. Diffusion-derived interpretation is mainly done using mean ADC values. Diffusion showed good reproducibility for baseline scans for the ADC value of the primary tumor Limitations and nodal metastasis [63]. The reproducibility of the ADC during treatment is also suggested to be good [64]. Mean The limitations and potential pitfalls of the functional MRI values of the tumor or metastatic lymph node are not sequences should be kept in mind during the interpretation. representative when they consist of both highly and poorly First, the lack of anatomical information at high b values in cellular (necrotic) portions. Mean ADC values should be DWI is a drawback because of suppressed signal in most of measured in the areas with high cellularity only to over- •• •• the normal tissues. Therefore, DWI should not be inter- come this limitation [24 , 25 ]. Even then, ADC inter- preted alone, but in correlation with anatomical sequences. pretation remains challenging. A recent study suggested a This is also true for perfusion and spectroscopy which reduced field of view (FOV) might increase accuracy [65]. means that all functional MRI sequences can never be used Moreover, it has been suggested that multiple b values are without the use of anatomical sequences. Moreover, all more accurate as this method is able to distinguish the functional sequences are currently hindered by high vari- perfusion component resulting in a pure diffusion value. ability of cut-offs and parameters used. This perfusion might influence the ADC value, although 123 2 Page 10 of 15 Curr Radiol Rep (2018) 6:2 Fig. 5 Nodal metastasis with positive diffusion and perfusion. Same (arrow head). Peripheral enhancement corresponded with high AUC patient as in Fig. 4 showing a lymph node metastasis with necrotic (arrows) (Color figure online) center with high T2 signal and no enhancement or increased perfusion some consider the influence of perfusion below clinical artifacts and currently best suited to perform in patients relevance [65]. As the clinical implication of multiple with HNSCC. Post-processing of perfusion is more com- b values is not yet firmly established, the acquisition of plex due to the nonspecific nature of vessel leakage multiple b values in clinical setting can be questioned. resulting in possible false-negatives and false-positive However, multiple b values are clearly preferred in a results. Perfusion post-processing also has a greater range research setting. of methods and functional parameters for analysis that are •• •• With respect to DCE perfusion, an increased scan available if compared to DWI [24 , 25 , 41]. This adds to duration with approximately 7–10 min is most hindering the complexity of perfusion imaging and its clinical clinical applicability next to the potential artifact as dis- implementation. •• cussed above [24 ]. DCE perfusion is least influenced by 123 Curr Radiol Rep (2018) 6:2 Page 11 of 15 2 Fig. 6 Normal lymph node and nodal metastasis with diffusion and enlarged and demonstrated slightly restricted diffusion as is also seen perfusion. A 66-year-old patient with a right sided pT1N1Mx floor of in normal lymphoid tissue. Perfusion showed a high AUC and relative the mouth SCC demonstrated recurrent lymph nodes after postoper- enhancement with a rapid wash-in with plateau phase for both lymph ative radiation therapy. An enlarged metastasis lymph node was seen nodes, although most pronounced in the metastatic lymph node. on the right side with diffusion restriction and increased relative Interpretation of the perfusion of lymph nodes remains difficult and enhancement and AUC (arrow). A contralateral lymph node was not should be further investigated (Color figure online) Studies with regards to MRS suggest a higher choline- should be placed by a radiologist to ensure correct place- to-creatine ratio in patients with poor prognosis, which ment in the anatomically difficult head and neck area. corresponds with expected high rates of proliferation and Furthermore, motion artifact from the carotid artery, long membrane biosynthesis in aggressive tumors (increased scan durations and complex post-processing hinders clini- rate of metabolism) [48]. However, MRS is not commonly cal applicability [52, 66]. used due to its technical challenges. The region of interest 123 2 Page 12 of 15 Curr Radiol Rep (2018) 6:2 Differentiation between malignant and benign post- Future Developments and Challenges treatment effects in HNSCC is of importance to guide Differentiation between malignancy and benign post-treat- clinical decisions. As anatomical MRI is not able to reli- ably differentiate post-therapy effect from tumor, func- ment effects such as fibrosis in HNSCC is of importance to guide clinical decisions. The head and neck is an area sen- tional techniques have been investigated and shown to be promising. This review showed that DWI can increase the sitive for artifacts and functional MR imaging requires advanced MRI post-processing software to evaluate diagnostic accuracy significantly for the primary tumor site and might also increase the diagnostic accuracy for the HNSCC. Combined functional sequences are required to fully appreciate HNSCC post-therapy, in addition to the region lymph nodes after therapy. Diffusion is most easy to implement and is recommended to perform routinely in a necessary anatomical sequences. This would result in long clinical setting in HNSCC follow-up. Its use during treat- scan durations, but new developments could overcome time ment to predict outcome is interesting, but evidence is too issues. A possible role of hybrid integrated PET/MR imag- low to implement. ing might be demonstrated offering the potential to acquired anatomical and function data using different modalities. Although perfusion parameters might be increased in tumor residual or recurrence and nodal metastasis, its However, future research is needed to evaluate PET/MRI and its appropriate applications compared to existing tech- diagnostic accuracy has yet to be established and is not routinely used clinically. DCE is least hindered by artifact niques [67] and whether PET/MRI is of greater clinical value than PET/CT and retrospective image fusion tech- and might be performed clinically if local experience is present. niques [68]. HNSCC is common and local residual and/or Spectroscopy research is promising, but evidence is too recurrence and nodal metastasis are seen in many patients. sparse for clinical implementation in the near future. The Diffusion is already frequently used. However, diffusion role of hybrid PET/MR imaging is to be established. with multiple b values and perfusion required further con- firmation of their added value in the post-therapy setting Acknowledgements This study was funded by a Mandema sti- before wide-spread implementation. This is even more the pendium from the University of Groningen (AH). case for spectroscopy. Future studies should focus on the added value of the different functional MRI sequences Compliance with Ethical Guidelines preferable by large prospective longitudinal multicenter Conflict of interest Roland Nooij, Jan Hof, Peter Jan van Laar, and studies comparing all sequences in the same population. Anouk van der Hoorn each declare that they have no conflicts of These studies are needed to assess the diagnostic accuracy interest. of the functional MRI sequences separately and in combina- tion. Another important aspect of these studies should be to Human and Animal Rights and Informed Consent This article does not contain any studies with human or animal subjects per- define the optimum time for assessment of metabolic and formed by the authors. physiological MRI parameters using functional techniques. The functional parameters should be tested in relation to the Open Access This article is distributed under the terms of the histopathological changes in HNSCC, treatment effects and Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted patient outcomes. These new trials must result in standardized use, distribution, and reproduction in any medium, provided you give cut-off values and ratios for the anatomical and functional appropriate credit to the original author(s) and the source, provide a MRI sequences to precisely define post-therapy changes from link to the Creative Commons license, and indicate if changes were tumor progression. The use of standardized cut-off values made. might remain arbitrary because of the use of different MRI systems. Nevertheless, it would be a valuable guideline for References the clinician in daily practice. Despite these possible limita- tions, implications into clinical practice would be an impor- Paper of particular interest, published recently, have been tant step in making an accurate treatment decisions for highlighted as: HNSCC patients. Of importance •• Of major importance Conclusions 1. Fitzmaurice C, Allen C, Barber RM, Barregard L, Bhutta ZA, Brenner H, et al. Global, regional, and national cancer incidence, In summary, this review analyzed the role of specific mortality, years of life lost, years lived with disability, and dis- functional MRI modalities in differentiating benign post- ability-adjusted life-years for 32 cancer groups, 1990 to 2015: a treatment effects from recurrence and/or residual malig- systematic analysis for the global burden of disease study. JAMA nancy and metastases in HNSCC. 123 Curr Radiol Rep (2018) 6:2 Page 13 of 15 2 Oncol. 2017;3:524–48. https://doi.org/10.1001/jamaoncol.2016. imaging. PET/CT and anatomical MRI are discussed. Functional 5688. MRI is not discussed. 2. Gatta G, Botta L, Sanchez MJ, Anderson LA, Pierannunzio D, 15. Omura G, Saito Y, Ando M, Kobayashi K, Ebihara Y, Yamasoba Licitra L. Prognoses and improvement for head and neck cancers T, et al. Salvage surgery for local residual or recurrent pharyngeal diagnosed in Europe in early 2000s: the EUROCARE-5 popula- cancer after radiotherapy or chemoradiotherapy. Laryngoscope. tion-based study. Eur J Cancer. 2015;51:2130–43. https://doi.org/ 2014;124:2075–80. https://doi.org/10.1002/lary.24695. 10.1016/j.ejca.2015.07.043. 16. Camisasca DR, Silami MANC, Honorato J, Dias FL, de Faria 3. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA PAS, Lourenc¸o SdQC. Oral squamous cell carcinoma: clinico- Cancer J Clin. 2017;67:7–30. https://doi.org/10.3322/caac.21387. pathological features in patients with and without recurrence. 4. Jou A, Hess J. Epidemiology and molecular biology of head and ORL J Otorhinolaryngol Relat Spec. 2011;73:170–6. https://doi. neck cancer. Oncol Res Treat. 2017;40:328–32. https://doi.org/ org/10.1159/000328340. 10.1159/000477127. 17. Guo T, Qualliotine JR, Ha PK, Califano JA, Kim Y, Saunders JR, 5. Argiris A, Karamouzis MV, Raben D, Ferris RL. Head and neck et al. Surgical salvage improves overall survival for patients with cancer. Lancet. 2008;371:1695–709. https://doi.org/10.1016/ HPV-positive and HPV-negative recurrent locoregional and dis- S0140-6736(08)60728-X. tant metastatic oropharyngeal cancer. Cancer. 2015;121:1977–84. 6. Seiwert TY, Salama JK, Vokes EE. The chemoradiation para- https://doi.org/10.1002/cncr.29323. digm in head and neck cancer. Nat Clin Pract Oncol. 18. Kim S-A, Roh J-L, Kim JS, Lee JH, Lee SH, Choi S-H, et al. 2007;4:156–71. https://doi.org/10.1038/ncponc0750. (18)F-FDG PET/CT surveillance for the detection of recurrence 7. Cooper JS, Pajak TF, Forastiere A, Jacobs J, Fu KK, Ang KK, in patients with head and neck cancer. Eur J Cancer. et al. Precisely defining high-risk operable head and neck tumors 2017;72:62–70. https://doi.org/10.1016/j.ejca.2016.11.009. based on RTOG#85-03 and #88-24: targets for postoperative 19. Castaldi P, Leccisotti L, Bussu F, Micciche` F, Rufini V. Role radiochemotherapy. Head Neck. 1998;20:588–94. https://doi.org/ of 18F-FDG PET-CT in head and neck squamous cell carcinoma. 10.1002/(SICI)1097-0347(199810)20:7\588:AID-HED2[3.0. Acta Otorhinolaryngol Ital. 2013;33:1–8. CO;2-F. 20. Helsen N, Roothans D, van den Heuvel B, van den Wyngaert T, 8. Jacob LA, Chaudhuri T, Lakshmaiah KC, Babu KG, Dasappa L, van den Weyngaert D, Carp L, et al. 18F-FDG-PET/CT for the Babu MCS, et al. Current status of systemic therapy for recurrent detection of disease in patients with head and neck cancer treated and/or metastatic squamous cell carcinoma of the head and neck. with radiotherapy. PLoS ONE. 2017;12:e0182350. https://doi. Indian J Cancer. 2016;53:471–7. https://doi.org/10.4103/0019- org/10.1371/journal.pone.0182350. 509X.204786. 21. Mak D, Corry J, Lau E, Rischin D, Hicks RJ. Role of FDG-PET/ 9. Overgaard J, Hansen HS, Specht L, Overgaard M, Grau C, CT in staging and follow-up of head and neck squamous cell Andersen E, et al. Five compared with six fractions per week of carcinoma. Q J Nucl Med Mol Imaging. 2011;55:487–99. •• conventional radiotherapy of squamous-cell carcinoma of head 22. van der Hoorn A, van Laar PJ, Holtman GA, Westerlaan HE. and neck: DAHANCA 6 and 7 randomised controlled trial. Diagnostic accuracy of magnetic resonance imaging techniques Lancet. 2003;362:933–40. https://doi.org/10.1016/S0140- for treatment response evaluation in patients with head and neck 6736(03)14361-9. tumors, a systematic review and meta-analysis. PLoS One •• 10. Saito N, Nadgir RN, Nakahira M, Takahashi M, Uchino A, 2017;12:e0177986. https://doi.org/10.1371/journal.pone. Kimura F, et al. Posttreatment CT and MR imaging in head and 0177986. A meta-analysis summarizing all available diagnostic neck cancer: What the radiologist needs to know. Radiographics accuracy of DWI studies for the treatment evaluation of SCC 2012;32:1261–82. https://doi.org/10.1148/rg.325115160. Normal head and neck tumors for the primary tumor site and lymph postoperative imaging findings and complications in patients with nodes. HNSCC on CT and anatomical MRI are demonstrated. Some 23. de Bondt RBJ, Nelemans PJ, Bakers F, Casselman JW, Peutz- examples of the applicability of DWI are also discussed from a Kootstra C, Kremer B, et al. Morphological MRI criteria improve radiologist perspective. the detection of lymph node metastases in head and neck squa- 11. Maroldi R, Ravanelli M, Farina D. Magnetic resonance for mous cell carcinoma: multivariate logistic regression analysis of laryngeal cancer. Curr Opin Otolaryngol Head Neck Surg. MRI features of cervical lymph nodes. Eur Radiol. 2014;22:131–9. https://doi.org/10.1097/MOO.0000000000000036. 2009;19:626–33. https://doi.org/10.1007/s00330-008-1187-3. •• 12. Bhatnagar P, Subesinghe M, Patel C, Prestwich R, Scarsbrook 24. Jansen JFA, Parra C, Lu Y, Shukla-Dave A. Evaluation of head AF. Functional imaging for radiation treatment planning, and neck tumors with functional MR imaging. Magn Reson response assessment, and adaptive therapy in head and neck Imaging Clin N Am 2016;24:123-33. https://doi.org/10.1016/j. cancer. Radiographics. 2013;33:1909–29. https://doi.org/10. mric.2015.08.011. This review describes the use of functional 1148/rg.337125163. MRI sequences in head and neck cancer including detailed 13. Al-Shwaiheen FA, Wang SJ, Uzelac A, Yom SS, Ryan WR. description of protocols. •• The advantages and drawbacks of routine magnetic resonance 25. King AD, Thoeny HC. Functional MRI for the prediction of imaging for long-term posttreatment locoregional surveillance of treatment response in head and neck squamous cell carcinoma: oral cavity squamous cell carcinoma. Am J Otolaryngol potential and limitations. Cancer Imaging 2016;16:23. https://doi. 2015;36:415–23. https://doi.org/10.1016/j.amjoto.2015.01.024. org/10.1186/s40644-016-0080-6. Review of role and potential of Analysis of routinely performed MRI with regards to treated functional MRI during and after treatment, focussing on DWI squamous cell carcinoma of the oral cavity. Advantages and (also grants insight in % rise ADC early in treatment in unfa- drawbacks with regards to false positives and false negatives on vourable treatment outcome) and Ktrans from DCE-MRI. Also MRI in patients without suspicious clinical or exam findings are discusses possibility of MRS. discussed. 26. King AD, Chow K-K, Yu K-H, Mo FKF, Yeung DKW, Yuan J, 14. Gage KL, Thomas K, Jeong D, Stallworth DG, Arrington JA. et al. Head and neck squamous cell carcinoma: diagnostic per- Multimodal imaging of head and neck squamous cell carcinoma. formance of diffusion-weighted MR imaging for the prediction of Cancer Control 2017;24:172–9. Evaluation of advanced HNSCC treatment response. Radiology. 2013;266:531–8. https://doi.org/ for treatment planning and follow up through multimodal 10.1148/radiol.12120167. 123 2 Page 14 of 15 Curr Radiol Rep (2018) 6:2 27. Marzi S, Piludu F, Sanguineti G, Marucci L, Farneti A, Terrenato therapy for oral cancer. J Magn Reson Imaging. 2012;36:589–97. I, et al. The prediction of the treatment response of cervical nodes https://doi.org/10.1002/jmri.23704. using intravoxel incoherent motion diffusion-weighted imaging. 39. Cao Y, Popovtzer A, Li D, Chepeha DB, Moyer JS, Prince ME, Eur J Radiol. 2017;92:93–102. https://doi.org/10.1016/j.ejrad. et al. Early prediction of outcome in advanced head-and-neck 2017.05.002. cancer based on tumor blood volume alterations during therapy: a 28. Liang L, Luo X, Lian Z, Chen W, Zhang B, Dong Y, et al. Lymph prospective study. Int J Radiat Oncol Biol Phys. node metastasis in head and neck squamous carcinoma: efficacy 2008;72:1287–90. https://doi.org/10.1016/j.ijrobp.2008.08.024. of intravoxel incoherent motion magnetic resonance imaging for 40. Ng S-H, Lin C-Y, Chan S-C, Yen T-C, Liao C-T, Chang JT-C, the differential diagnosis. Eur J Radiol. 2017;90:159–65. https:// et al. Dynamic contrast-enhanced MR imaging predicts local doi.org/10.1016/j.ejrad.2017.02.039. control in oropharyngeal or hypopharyngeal squamous cell car- 29. Paudyal R, Oh JH, Riaz N, Venigalla P, Li J, Hatzoglou V, et al. cinoma treated with chemoradiotherapy. PLoS ONE. Intravoxel incoherent motion diffusion-weighted MRI during 2013;8:e72230. https://doi.org/10.1371/journal.pone.0072230. chemoradiation therapy to characterize and monitor treatment 41. Fujima N, Kudo K, Yoshida D, Homma A, Sakashita T, Tsuka- response in human papillomavirus head and neck squamous cell hara A, et al. Arterial spin labeling to determine tumor viability in carcinoma. J Magn Reson Imaging 2017;45:1013–23. https://doi. head and neck cancer before and after treatment. J Magn Reson org/10.1002/jmri.255. This study showed the utility of IVIM DWI Imaging. 2014;40:920–8. https://doi.org/10.1002/jmri.24421. for early assessment of treatment response of chemoradiation 42. Fujima N, Kudo K, Tsukahara A, Yoshida D, Sakashita T, therapy in HNSCC patients. D and f values as indicators are also Hommaet A, et al. Measurement of tumor blood flow in head and discussed. neck squamous cell carcinoma by pseudo-continuous arterial spin 30. Sasaki M, Sumi M, Eida S, Katayama I, Hotokezaka Y, Naka- labeling: comparison with dynamic contrast-enhanced MRI. mura T. Simple and reliable determination of intravoxel inco- J Magn Reson Imaging. 2015;41:983–91. https://doi.org/10.1002/ herent motion parameters for the differential diagnosis of head jmri.24885. and neck tumors. PLoS ONE. 2014;9:e112866. https://doi.org/10. 43. Devpura S, Barton KN, Brown SL, Palyvoda O, Kalkanis S, Naik 1371/journal.pone.0112866. VM, et al. Vision 20/20: the role of Raman spectroscopy in early 31. Zheng D, Chen Y, Liu X, Chen Y, Xu L, Ren W, et al. Early stage cancer detection and feasibility for application in radiation response to chemoradiotherapy for nasopharyngeal carcinoma therapy response assessment. Med Phys. 2014;41:050901. https:// treatment: value of dynamic contrast-enhanced 3.0 T MRI. doi.org/10.1118/1.4870981. J Magn Reson Imaging. 2015;41:1528–40. https://doi.org/10. 44. Hwang I, Choi SH, Kim Y-J, Kim KG, Lee AL, Yun TJ, et al. 1002/jmri.24723. Differentiation of recurrent tumor and posttreatment changes in 32. Sumi M, Nakamura T. Head and neck tumors: assessment of head and neck squamous cell carcinoma: application of high perfusion-related parameters and diffusion coefficients based on b-value diffusion-weighted imaging. AJNR Am J Neuroradiol. the intravoxel incoherent motion model. AJNR Am J Neurora- 2013;34:2343–8. https://doi.org/10.3174/ajnr.A3603. diol. 2013;34:410–6. https://doi.org/10.3174/ajnr.A3227. 45. Vandecaveye V, Dirix P, de Keyzer F, de Beeck K, vander 33. Abdel Razek AAK, Gaballa G, Ashamalla G, Alashry MS, Poorten V, Roebben I, et al. Predictive value of diffusion- Nada N. Dynamic susceptibility contrast perfusion-weighted weighted magnetic resonance imaging during chemoradiotherapy magnetic resonance imaging and diffusion-weighted magnetic for head and neck squamous cell carcinoma. Eur Radiol. resonance imaging in differentiating recurrent head and neck 2010;20:1703–14. https://doi.org/10.1007/s00330-010-1734-6. cancer from postradiation changes. J Comput Assist Tomogr 46. Matoba M, Tuji H, Shimode Y, Toyoda I, Kuginuki Y, Miwa K, 2015;39:849–54. https://doi.org/10.1097/rct.0000000000000311. et al. Fractional change in apparent diffusion coefficient as an This study shows the possibilities of ADC-values and DSC per- imaging biomarker for predicting treatment response in head and fusion-weighted MR imaging, in the differentiation of recurrent neck cancer treated with chemoradiotherapy. AJNR Am J Neu- head and neck cancer from postradiation changes. roradiol. 2014;35:379–85. https://doi.org/10.3174/ajnr.A3706. 34. Verduijn GM, Bartels LW, Raaijmakers CPJ, Terhaard CHJ, 47. Bernstein JM, Kershaw LE, Withey SB, Lowe NM, Homer JJ, Pameijer FA, van den Berg CAT. Magnetic resonance imaging Slevin NJ, et al. Tumor plasma flow determined by dynamic protocol optimization for delineation of gross tumor volume in contrast-enhanced MRI predicts response to induction hypopharyngeal and laryngeal tumors. Int J Radiat Oncol Biol chemotherapy in head and neck cancer. Oral Oncol. Phys. 2009;74:630–6. https://doi.org/10.1016/j.ijrobp.2009.01. 2015;51:508–13. https://doi.org/10.1016/j.oraloncology.2015.01. 014. 013. 35. Bertrand M, Tollard E, Francois A, Bouchetetemble P, Marie PJ, 48. Bezabeh T, Odlum O, Nason R, Kerr P, Sutherland D, Patel R, Dehesdin D, et al. CT scan, MR imaging and anatomopathologic et al. Prediction of treatment response in head and neck cancer by correlation in the glottic carcinoma T1-T2. Rev Laryngol Otol magnetic resonance spectroscopy. AJNR Am J Neuroradiol. Rhinol (Bord). 2010;131:51–7. 2005;26:2108–13. 36. Ravanelli M, Farina D, Rizzardi P, Botturi E, Prandolini P, 49. Acampora A, Manzo G, Fenza G, Busto G, Serino A, Manto A. Mangili S, et al. MR with surface coils in the follow-up after High b-Value Diffusion MRI to differentiate recurrent tumors endoscopic laser resection for glottic squamous cell carcinoma: from posttreatment changes in head and neck squamous cell feasibility and diagnostic accuracy. Neuroradiology. carcinoma: a single center prospective study. Biomed Res Int. 2013;55:225–32. https://doi.org/10.1007/s00234-012-1128-3. 2016;2016:2865169. https://doi.org/10.1155/2016/2865169. 37. Zbaren P, Becker M, Lang H. Pretherapeutic staging of laryngeal 50. Choi SH, Lee JH, Choi YJ, Park JE, Sung YS, Kim N, et al. carcinoma. Clinical findings, computed tomography, and mag- Detection of local tumor recurrence after definitive treatment of netic resonance imaging compared with histopathology. Cancer. head and neck squamous cell carcinoma: histogram analysis of 1996;77:1263–73. https://doi.org/10.1001/archotol.1997. dynamic contrast-enhanced T1-weighted perfusion MRI. AJR 01900090016003. Am J Roentgenol. 2017;208:42–7. https://doi.org/10.2214/AJR. 38. Chikui T, Kitamoto E, Kawano S, Sugiura T, Obara M, Simonetti 16.16127. AW, et al. Pharmacokinetic analysis based on dynamic contrast- 51. Choi YJ, Lee JH, Sung YS, Yoon RG, Park JE, Nam SY, et al. enhanced MRI for evaluating tumor response to preoperative Value of dynamic contrast-enhanced MRI to detect local tumor 123 Curr Radiol Rep (2018) 6:2 Page 15 of 15 2 recurrence in primary head and neck cancer patients. Medicine. 60. King AD, Yeung DKW, Ahuja AT, Leung SF, Tse GMK, van 2016;95:e3698. https://doi.org/10.1097/MD.0000000000003698. Hasselt AC. In vivo proton MR spectroscopy of primary and 52. King AD,Yeung DKW, YuK-H,MoFKF,HuC-W,Bhatia KS, nodal nasopharyngeal carcinoma. AJNR Am J Neuroradiol. et al. Monitoring of treatment response after chemoradiotherapy for 2004;25:484–90. head and neck cancer using in vivo 1H MR spectroscopy. Eur 61. Shah GV, Gandhi D, Mukherji SK. Magnetic resonance spec- Radiol. 2010;20:165–72. https://doi.org/10.1007/s00330-009-1531-2. troscopy of head and neck neoplasms. Top Magn Reson Imaging. 53. Berrak S, Chawla S, Kim S, Quon H, Sherman E, Loevner LA, 2004;15:87–94. et al. Diffusion weighted imaging in predicting progression free 62. Chawla S, Kim S, Loevner LA, Quon H, Wang S, Mutale F, et al. survival in patients with squamous cell carcinomas of the head Proton and phosphorous MR spectroscopy in squamous cell and neck treated with induction chemotherapy. Acad Radiol. carcinomas of the head and neck. Acad Radiol. 2009;16:1366–72. 2011;18:1225–32. https://doi.org/10.1016/j.acra.2011.06.009. https://doi.org/10.1016/j.acra.2009.06.001. 54. Jin GQ, Yang J, Liu LD, Su DK, Wang DP, Zhao SF, et al. The 63. Hoang JK, Choudhury KR, Chang J, Craciunescu OI, Yoo DS, diagnostic value of 1.5-T diffusion-weighted MR imaging in Brizel DM. Diffusion-weighted imaging for head and neck detecting 5 to 10 mm metastatic cervical lymph nodes of squamous cell carcinoma: quantifying repeatability to understand nasopharyngeal carcinoma. Medicine (Baltimore). 2016;95: early treatment-induced change. AJR Am J Roentgenol. e4286. https://doi.org/10.1097/MD.0000000000004286. 2014;203:1104–8. https://doi.org/10.2214/AJR.14.12838. 55. Holzapfel K, Duetsch S, Fauser C, Eiber M, Rummeny EJ, Gaa J. 64. Vaid S, Chandorkar A, Atre A, Shah D, Vaid N. Differentiating Value of diffusion-weighted MR imaging in the differentiation recurrent tumours from posttreatment changes in head and neck between benign and malignant cervical lymph nodes. Eur J cancers: does diffusion-weighted MRI solve the eternal dilemma? Radiol. 2009;72:381–7. https://doi.org/10.1016/j.ejrad.2008.09. Clin Radiol. 2017;72:74–83. https://doi.org/10.1016/j.crad.2016. 034. 09.019. 56. Abdel Razek AAK, Soliman NY, Elkhamary S, Alsharaway MK, 65. Vidiri A, Minosse S, Piludu F, Curione D, Pichi B, Spriano G, Tawfik A. Role of diffusion-weighted MR imaging in cervical et al. Feasibility study of reduced field of view diffusion- lymphadenopathy. Eur Radiol. 2006;16:1468–77. https://doi.org/ weighted magnetic resonance imaging in head and neck tumors. 10.1007/s00330-005-0133-x. Acta Radiol. 2017;58:292–300. https://doi.org/10.1177/ 57. Tyagi N, Riaz N, Hunt M, Wengler K, Hatzoglou V, Young R, 0284185116652014. et al. Weekly response assessment of involved lymph nodes to 66. Abdel Razek AAK, Poptani H. MR spectroscopy of head and radiotherapy using diffusion-weighted MRI in oropharynx squa- neck cancer. Eur J Radiol. 2013;82:982–9. https://doi.org/10. mous cell carcinoma. Med Phys. 2016;43:137. https://doi.org/10. 1016/j.ejrad.2013.01.025. 1118/1.4937791. 67. Torigian DA, Zaidi H, Kwee TC, Saboury B, Udupa JK, Cho ZH, 58. Hauser T, Essig M, Jensen A, Laun FB, Mu¨nter M, Maier-Hein et al. PET/MR imaging: technical aspects and potential clinical KH, et al. Prediction of treatment response in head and neck applications. Radiology. 2013;267:26–44. https://doi.org/10. carcinomas using IVIM-DWI: evaluation of lymph node metas- 1148/radiol.13121038. tasis. Eur J Radiol. 2014;83:783–7. https://doi.org/10.1016/j. 68. Loeffelbein DJ, Souvatzoglou M, Wankerl V, Martinez-Mo¨ller ejrad.2014.02.013. A, Dinges J, Schwaiger M, et al. PET-MRI fusion in head-and- 59. Jansen JFA, Carlson DL, Lu Y, Stambuk HE, Moreira AL, Singh neck oncology: current status and implications for hybrid PET/ B, et al. Correlation of a priori DCE-MRI and (1)H-MRS data MRI. J Oral Maxillofac Surg. 2012;70:473–83. https://doi.org/10. with molecular markers in neck nodal metastases: initial analysis. 1016/j.joms.2011.02.120. Oral Oncol. 2012;48:717–22. https://doi.org/10.1016/j.oraloncology. 2012.02.001.

Journal

Current Radiology Reports – Springer Journals

Published: Jan 22, 2018

There are no references for this article.