+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

Journal of Neuroradiology (2014) xxx, xxx—xxx

Available online at

ScienceDirect www.sciencedirect.com

ORIGINAL ARTICLE

Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners J. Lincot a,∗, F. Veillon a,b, S. Riehm a, N. Babay a, J.-F. Matern a,b, B. Rock a, B. Dallaudière c, N. Meyer b,d a

Department of Radiology I, Strasbourg Hautepierre University Hospital, CHRU Hautepierre, avenue Molière, 67098 Strasbourg cedex, France b Faculty of Medicine, Louis-Pasteur University, 67000 Strasbourg, France c Department of Radiology, Bichat—Claude-Bernard University Hospital, 75018 Paris, France d Laboratory of Biostatistics and Medical Informatics, Louis-Pasteur University, 67000 Strasbourg, France

KEYWORDS Cholesteatoma; Temporal bone; Middle ear masses; Non-echo planar; Diffusion-weighted imaging

Summary Background and purpose: To compare diagnostic performances for cholesteatoma diagnosis of incremental MRI protocols including non-echo planar diffusion-weighted imaging (DWI) performed on 3T and 1.5T scanners. Materials and methods: Thirty-nine patients with suspected cholesteatoma underwent 3T and 1.5T non-echo planar DWI and additional unenhanced T1-, delayed gadolinium-enhanced T1- and high-resolution T2-weighted standard acquisitions. Patients either underwent surgical tympanoplasty (n = 21) or close clinicoradiological follow-up (n = 18). Four radiologists independently and prospectively interpreted two incremental MRI protocols, differing in the magnetic field strength of the diffusion-weighted acquisition and comprising the three standard sequences. At each step, diagnostic performances were expressed as sensitivity, specificity, positive predictive value, negative predictive value and accuracy. Results: Forty middle ear lesions including 21 cholesteatomas were identified. Univariate and multivariate analysis did not demonstrate significant reader, sequence addition or DWI magnetic field effect on diagnostic performances. Concerning non-echo planar DWI alone, sensitivity, specificity, positive predictive value, negative predictive value and accuracy ranged between 90.5—100%, 68.4—100%, 76.9—100%, 90.0—100% and 82.5—95.0, respectively.

Abbreviations: HR, High-Resolution; DGE, Delayed-Gadolinium Enhanced; DWI, Diffusion-Weighted Imaging; EP, Echo Planar; FS, Fat Saturated; PROPELLER, Periodically Rotated Overlapping Parallel Lines with Enhanced Reconstruction; Se, Sensitivity; Sp, Specificity; PPV, Positive Predictive Value; NPV, Negative Predictive Value; GEE, General Estimating Equation; SD, Standard Deviation. ∗ Corresponding author. Tel.: +33 3 88 12 78 65; fax: +33 3 88 12 78 60. E-mail address: [email protected] (J. Lincot). http://dx.doi.org/10.1016/j.neurad.2014.02.003 0150-9861/© 2014 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

2

J. Lincot et al. Conclusion: Non-echo planar DWI for cholesteatoma diagnosis can be performed on 1.5T or 3T scanners indifferently. High sensitivity and negative predictive value and relatively lower specificity and positive predictive value are achieved by a single non-echo planar DWI protocol. © 2014 Elsevier Masson SAS. All rights reserved.

Introduction Cholesteatoma is a locally aggressive middle ear lesion consisting of a desquamative keratinizing epithelium pouch which typically expands over time within the tympanic and/or mastoid cavity, with the potential to cause ossicular destruction, labyrinthine fistulae and destruction of middle cranial fossa. It usually complicates chronic otitis but can occasionally be congenital in nature, occurring behind a normal tympanic membrane [1]. Historically, the diagnosis was suspected on the basis of otoscopic findings. High resolution (HR) CT was the technique of choice for assessing complications and extent [2], especially in the pre- and postoperative periods when tympanoplasty is needed [3]. However, HRCT is unable to directly characterize soft tissue densities within the temporal bone. In the past decade, MRI has shown to be valuable in the initial diagnosis of cholesteatoma [4,5] and in the postoperative follow-up, prior to second-look surgery or for evaluation of recurrence [6—8]. This was first achieved using delayed gadolinium-enhanced (DGE) T1-weighted sequences [8,9], and then using DWI. Echo planar (EP) DWI was the first technique employed [4,7,10,11] but missed small lesions unlike more recent non-EP techniques [6,12]. Recently, a study [13] assessing individual diagnostic accuracies of DGE, EPI and non-EPI-DW MRI sequences in the detection of middle ear cholesteatoma confirmed that, although being the more efficient technique, non-EPI DWI still lacks sensitivity for small lesions. DWI has proven to be superior to DGE T1-weighted imaging but few studies were carried out on their association, with contrasting results [14,15]. Nowadays, cholesteatoma screening tends to be achieved using non-EP DWI exclusively [16] but the influence of standard MRI complementary sequences over diagnostic performances still needs clarification. In another field, it has been demonstrated that the association of DWI and post-contrast T1-weighted imaging is more efficient for parotid gland exploration than each technique alone [17]. Moreover, few studies have looked at the role of 3T non-EP DWI [18,19] and its potential improvement in regard to diagnostic reliability. Consequently, the aim of our work was to prospectively evaluate the possible gain in diagnostic performances offered by 3T non-EP DWI and by unenhanced T1-, DGE T1and HRT2-weighted standard sequences through an incremental MRI reading protocol.

Materials and methods Patients We conducted a prospective study on a cohort of 39 patients (mean age, 47.7 years; range, 15.3—83.7 years; M/F ratio = 23:16) who underwent MRI at our University Hospital. There was no significant difference in age distribution

between the sexes (P = .742). Following otoscopic examination and temporal bone HRCT, 21 patients (mean age, 50.8 years; range, 15.3—83.7 years; M/F ratio = 11:10) were suspected of having an initial presentation of acquired cholesteatoma. Eighteen patients (mean age, 43.7 years; range, 23.5—58.3 years; M/F ratio = 12:6) were clinically and radiologically followed-up in search of residual/recurrent cholesteatoma following surgical tympanoplasty. All patients were similar in demographic variables, signs and symptoms.

MRI technique Written informed consent was obtained from each patient prior to MRI. The MRI protocol included axial unenhanced and delayed post-gadolinium chelate injection fat saturated (FS) T1-weighted sequences, 3D HRT2-weighted acquisition and non-EP DWI, the latter being performed on 1.5T and 3T scanners for each patient on the same day. Non-DW sequences were performed only once for each patient, on either one of the two scanners at random. 1.5T MRI was performed on an Avanto MR unit (Siemens, Erlangen, Germany) with a standard 12-channel head coil coupled with two monochannel surface coils to increase signal-to-noise ratio. 3T MRI was performed on a Signa HDx MR unit (GE Healthcare, Milwaukee, Wis, USA) with a standard 12-channel head coil. The imaging parameters are given in Table 1.

MRI interpretation MRI images were independently interpreted by four blinded radiologists, including two seniors (FV and SR, 25- and 15-year-experience in middle ear imaging, respectively) and two fellows (NB and JFM, one year experience each). Differential diagnoses for cholesteatoma were cholesterol granulomas and hyperplastic granulation tissue. Diagnostic criteria according to signal intensity are given in Table 2. For every patient, each observer independently evaluated two four-step incremental data sets. For the first set, 3T non-EP DWI was initially analyzed as a stand-alone sequence and each abnormal middle ear lesion was classified as either cholesteatomatous or non-cholesteatomatous. Further sequences were then presented in the following predefined study order: T1-weighted FS, DGE T1-weighted FS and 3D HRT2. The same classification (cholesteatomatous vs. non-cholesteatomatous) was made on the basis of the combined results following each successive sequence addition. For the second incremental data set 1.5T non-EP DWI was then evaluated with a similar consecutive presentation of the same non-DW sequences. Minimal and maximal delays between the evaluations of those two data sets were seven and twelve days, respectively.

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

Middle ear cholesteatoma: Compared diagnostic performances of two MRI protocols Table 1

3

MRI sequences parameters according to magnetic field strength.

Sequence

Magnetic field strength (T)

Imaging techniques

Acquisition matrix

Voxel size (mm)

TR (ms)

Axial T1-weighted

1.5 3 1.5 3 1.5 3

SE, FS SE, FS CISS FIESTA HASTE, FSE PROPELLER, FSE

384 × 384 320 × 320 320 × 384 484 × 484 128 × 64 128 × 128

0.6 × 0.6 × 1.5 0.6 × 0.6 × 1.5 0.5 × 0.5 × 0.4 0.4 × 0.4 × 0.3 2.5 × 1.2 × 3 1.5 × 1.5 × 3

719 540 11.6 6.7 4000 5500

3D HR T2-weighted Axial non-EP diffusion-weighted

TE (ms)

13 13 5.8 2.8 107 75

b value (s.mm-2 )

Number of averages

NA NA NA NA 1000 0, 1000

2 2 1 1 5 1.5

SE: spin-echo; FS: fat saturated; CISS: constructive interference in steady state; FIESTA: fast imaging employing steady state acquisition; HASTE: half-fourier acquisition single-shot turbo spin-echo; FSE: fast spin-echo; PROPELLER: periodically rotated overlapping; parallel lines with enhanced reconstruction.

Table 2 MRI diagnostic criteria for each of the three lesions (cholesteatoma, granuloma cholesteatoma and hyperplastic granulation tissue) according to diffusion-, unenhanced T1-, DGE T1- and T2-weighted signals. Lesion

Cholesteatoma Cholesterol granuloma Hyperplastic granulation tissue

Signal intensity b1000 Non-EP DWI

T1-weighted FS

DGE T1-weighted FS

HR T2-weighted

High High Low

Low High Low

Low High High

Mild to high Mild to high High

Post-MRI patient management Radiological consensus was reached among the four observers after incremental MRI reviewing. Patients with diagnosed cholesteatoma, either primarily acquired or residual/recurrent, underwent surgery performed by a senior ENT physician. Each lesion was then classified as either confirmed cholesteatoma or unconfirmed cholesteatoma according to the surgical report. Patients in whom imaging did not suggest the presence of cholesteatoma, were treated conservatively with strict otoscopic survey and radiological follow-up. If no suspicious change was noted in the clinicoradiological presentation, the patient was eventually considered as cholesteatomafree. If subsequent otologic or MRI findings revealed a pejorative evolution, the patient benefited from surgical revision. In all cases, the final decision whether to perform surgery or surveillance was made by ENT physicians on the basis of otoscopic, audiologic and imaging findings.

Data and statistical analysis Diagnostic performance was expressed as Se, Sp, PPV, NPV and accuracy for each observer and at each step of these two MRI incremental data sets. For overall diagnostic performances, averaged values between the four observers were calculated for these five parameters with corresponding 95% CIs. For each surgically confirmed cholesteatoma measurement of the greatest transverse diameter was performed on the non-EP DWI acquisition. We used three univariate general estimating equations (GEE) models to assess the effects of the reader, the MRI

sequence addition and the magnetic field strength of the non-EP DWI on the ability to correctly diagnose true negative and true positive cholesteatomas. A fourth multivariate GEE model was used to assess these three effects altogether. A P value of less than .05 was considered to be as statistically significant. Analyses were computed using the PROC GENMOD procedure (SAS software© , ve.8.02).

Results Cholesteatoma detection Among the study population, twenty patients were diagnosed by MRI consensus with 21 cholesteatomatous lesions, either primarily acquired or residual/recurrent. They all underwent surgical tympanoplasty, which confirmed 21 cholesteatomas (mean size, 11.2 mm; median size, 13 mm; range, 3—19 mm). Nineteen patients were treated conservatively for 19 non-cholesteatomatous lesions. One patient needing secondary myringoplasty had surgical revision, which confirmed the absence of residual/recurrent cholesteatoma. Considering the 18 other patients, analysis of clinical and radiological findings during medical followup led to no suspicious change in the presentation of the lesions. No other patient was secondarily operated and all were finally considered as cholesteatoma-free. Mean and standard deviation (SD) of clinicoradiological follow-up duration were 634 and 89 days, respectively. In total, the study population of 39 patients counted 40 middle ear lesions, among which 21 (52.5%) were acquired cholesteatomas and 19 (47.5%) were non-cholesteatomatous lesions. Individual diagnostic performances for these 40

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

Non-EP DWI

+ T1-weighted FS

+ DGE T1-weighted FS

+ HR T2-weighted

Non-EP DWI

+ T1-weighted FS

+ DGE T1-weighted FS

+ HR T2-weighted

0.952 0.895 0.909 0.944 0.925

0.905 1 1 0.905 0.95O

0.905 1 1 0.905 0.950

0.905 1 1 0.905 0.950

0.952 0.842 0.870 0.941 0.900

0.857 0.947 0.947 0.857 0.900

0.905 1 1 0.905 0.950

0.905 1 1 0.905 0.950

1 0.737 0.808 1 0.875

0.952 0.842 0.870 0.941 0.900

1 0.895 0.913 1 0.950

1 0.947 0,954 1 0.975

0.952 0.895 0.909 0.944 0.925

0,905 0.895 0.905 0.895 0.900

0,952 0.895 0.909 0.944 0.925

0.952 0.947 0.952 0.947 0.950

0.952 0.842 0.870 0.941 0.900

0.952 0.842 0.870 0.941 0.900

0.857 0.895 0.9 0.85 0.875

0.905 0.895 0.905 0.895 0.900

0.905 0.947 0.950 0.900 0.925

0.952 0.842 0.870 0.941 0.900

0.857 0.895 0.900 0.85 0.875

0.905 0.895 0.905 0.895 0.900

0.952 0.684 0.769 0.929 0.825

0.905 0.842 0.864 0.889 0.875

0.905 0.895 0.905 0.895 0.900

0.905 0.947 0.950 0.900 0.925

0.905 1 1 0.905 0.950

0.905 1 1 0.905 0.950

0.905 0.947 0.950 0.900 0.925

0.905 0.947 0.950 0.900 0.925

ARTICLE IN PRESS

FV Se Sp PPV NPV Acc SR Se Sp PPV NPV Acc NB Se Sp PPV NPV Acc JFM Se Sp PPV NPV Acc

1.5T MRI

+Model

3T MRI

NEURAD-468; No. of Pages 9

4

Notes: Se: sensitivity; Sp: specificity; PPV: positive predictive value; NPV: negative predictive value; Acc: accuracy; FV: F. Veillon; SR: S. Riehm; NB: N. Babay; JFM: JF. Matern.

J. Lincot et al.

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

Table 3 Se, Sp, PPV, NPV and accuracies obtained by each reader (FV, SR, NB or JFM) according to magnetic field strength of the interpreted non-EP DW sequence and to incremental MRI reading.

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

Middle ear cholesteatoma: Compared diagnostic performances of two MRI protocols

5

lesions, according to the magnetic field strength of the nonEP DWI acquisition and to the step of the incremental MRI reading, are given in Table 3. Averaged 1.5T and 3T diagnostic performances among the four readers are shown in Fig. 1.

Reader, MRI sequence addition and non-EP DWI magnetic field effects Univariate and multivariate statistical analysis showed no significant reader effect, no sequence addition effect and no magnetic field strength effect in the ability to correctly diagnose true negative and true positive cholesteatomas (P > .05). These results indicate that diagnostic performances did not significantly differ among the four radiologists, between each step of the incremental sequence reading or between the two MRI scanners on which non-EP DWI was performed.

Discussion This study’s incremental MRI reading order was arbitrarily assigned on the basis of the investigators’ experience in middle ear imaging, and designs of previously published studies in order to mimic a real-time MRI exam interpretation. Previous works demonstrated that DGE T1-weighted images were useful for cholesteatoma diagnosis [8,9], and that this sequence is complementary to DWI [15]. We chose to assess non-EP DWI, in light of previous study findings [6,14,18], and to correlate it with unenhanced T1-weighted, DGE T1-weighted and HRT2-weighted images within a multivariate study. This incremental reading order, whilst opened to criticism, represents the investigators’ usual practice in interpreting middle ear MRI. Those standard non-DW sequences were performed only once, on either one of the two scanners randomly, and their results were assumed to be independent of magnetic field strength. Moreover, we chose to assess both 3T and 1.5T non-EP DWI in the same population. To our knowledge, the present work is the first to implement such a comparative analysis. Our results demonstrate that diagnostic performances achieved using a single non-EP DWI sequence for cholesteatoma assessment were not significantly improved by the addition of successive standard MRI sequences in this study’s incremental reading order. Furthermore, diagnostic performances achieved using this self-sufficient non-EP DWI MR protocol were not affected either by magnetic field strength or by reader experience. These results accord with recent publications [6,16,18,19]. As mentioned in Table 3, Se (range, 90.5—100%), Sp (range, 68.4—100%), PPV (range, 76.9—100%), NPV (range, 90.0—100%) and accuracy (range, 82.5—95.0) achieved using this self-sufficient non-EP DWI MR protocol are in the range of previously reported values [14,15,19]. Figs 2 and 3 show typical non-EP DWI aspect at 3T and 1.5T for a characteristic tympano-addito-antral cholesteatoma, appearing as a strong signal hyperintensity at b = 1000s.mm−2 (Fig. 2), and for a negative case showing hyperplastic granulation tissue (Fig. 3). Use of MRI for a screening purpose for cholesteatoma, either primarily acquired or on an operated ear,

Figure 1 Average diagnostic values with 95% CI given for each step of the incremental MRI reading and for the two non-EP DWI magnetic field strengths (light grey, 3T DWI; dark grey, 1.5T DWI).

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

6

J. Lincot et al.

Figure 2 Typical non-EP DWI positive results: 43-year-old male showing a primary acquired cholesteatoma. A and B, 3T and 1.5T non-EP DWI (b = 1000s.mm−2 ), respectively, show a strong tympano-addito-antral hyperintensity (arrow).

necessitates the highest possible Se and NPV in order to avoid missing a potentially evolving lesion. Although not reaching 100%, averaged values of Se and NPV of non-EP DWI alone were 97.6% (95% CI: 95.3%, 99.9%) and 96.7% (95% CI: 93.5%, 99.9%) respectively for 3T MRI; 92.9% (95% CI: 90.5%, 95.2%) and 92.3% (95% CI: 90.3%, 94.2%) for 1.5T MRI, respectively. One particular false-negative case of recurrent cholesteatoma was misdiagnosed as a benign lesion by the four readers on 1.5T non-EP DWI and illustrates the potential discordance between 3T and 1.5T non-EP DWI results. This 52-year-old female had received surgical treatment several years previously for a left ear cholesteatoma and was followed-up by MRI. 3T non-EP DW images showed a nodular mesotympanic mild hyperintensity measuring 3 mm (Fig. 4), which was interpreted as cholesteatoma by all readers. On the contrary, such a suspicious image was not seeable on 1.5T non-EP DWI, which was systematically interpreted as benign. Two of the radiologists corrected their diagnosis upon DGE T1- and HRT2-weighted images, showing no significant enhancement and a heterogeneous mild hyperintensity, respectively. This patient was

eventually operated and tympanoplasty confirmed a small mesotympanic cholesteatoma. It has been demonstrated that non-EP DWI sequences have better diagnostic accuracy than EP DWI for cholesteatoma detection [12], due to reduced slice thickness, higher spatial resolution and annihilation of magnetic susceptibility artifacts caused by air-bone interfaces within the tympanic cavity. But as the present study demonstrates, in accord with previous articles, even non-EP DWI may fail to reach a 100% NPV, especially for smaller lesions < 4 mm [20]. Furthermore, the published size limit for a cholesteatoma to be detectable by non-EP DWI is 2 mm, for both 1.5T [21] or 3T [18] MRI. It should be noted, however, that potentially missed smaller lesions are usually considered less immediately harmful and could benefit from close otologic and radiologic follow-up [22,23], either in a pre- or postoperative chronic otitis imaging context. We conversely observed relatively low values for average Sp and PPV in comparison to those recently reported [14,19,23]. These were 78.9% (95% CI: 70.8%, 87.1%) and 84.0% (95% CI: 78.6%, 89.4%) respectively for 3T MRI; 92.1%

Figure 3 Typical non-EP DWI negative results: 58-year-old male with previous canal wall-up tympanoplasty. A and B, 3T and 1.5T non-EP DWI (b = 1000s.mm−2 ), respectively, show hyperplastic granulation tissue of null hyperintensity within surgical cavity (arrow).

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

Middle ear cholesteatoma: Compared diagnostic performances of two MRI protocols

7

Figure 4 1.5T non-EP DWI false negative case: 52-year-old female treated several years previously for a left ear cholesteatoma; small mesotympanic cholesteatoma misdiagnosed as granulation tissue. A and B, 3T and 1.5T non-EP DWI (b = 1000s.mm−2 ) show a small nodular mild hyperintensity and an ill-defined image of low signal intensity, respectively (arrow). Distance d, 3 mm. C and D, DGE T1- and HRT2-weighted images show no significant enhancement and a heterogeneous mild hyperintensity, respectively.

(95% CI: 86.3%, 97.9%) and 93.2% (95% CI: 88.5%, 98.0%) for 1.5T MRI. This suggests that a single non-EP DWI sequence protocol may carry an increased risk of false-positive cases, misdiagnosing benign lesions (e.g. cholesterol granuloma) as cholesteatomas. Nevertheless, the trend to higher Sp and PPV with additional sequences, as illustrated in Fig. 1, was not significantly demonstrated by our statistical analysis. In addition, the results were quite homogenous among the four observers. One peculiar case was misdiagnosed by the four radiologists as a cholesteatoma on non-EP DWI. This 79-year-old female clinically suffering from chronic otitis never operated and showing a diffuse middle ear filling at HRCT. Non-EP DW images displayed at 3T and 1.5T a small epitympanic signal hyperintensity suggesting a cholesteatomatous nature (Fig. 5). Interestingly, subsequent DGE T1-weighted images helped two of the four readers to correct their diagnosis since this formation showed diffuse enhancement, which clearly was in favor of hyperplastic granulation tissue. HRT2-weighted images also accorded with this final diagnosis showing elevated signal intensity. Although this study did not demonstrate any improvement of non-EP DWI diagnostic performances through the subsequent addition of standard sequences, this finding should be treated with some caution. From our experience, additional sequences may help to correct the diagnosis in difficult cases, such as shown above. Furthermore, the

secondary purposes of middle ear MRI should not be forgotten. Those are highlighting complications, such as labyrinthine fistula or tegmen tympani erosion with subsequent intracranial invasion, and acting as a roadmap prior to either first-stage or second-look surgery. This can be achieved using anatomical sequences such as HRT2weighted as proposed by previous articles [14,24]. One limit of our study is the relatively small number of subjects, which precluded the performance of subgroup analyses between cases of suspected primarily acquired cholesteatoma and cases of suspected residual/recurrent cholesteatoma. Another limitation may arise from the cholesteatoma-free criterion being based solely upon strictly stable otoscopical surveillance and subsequent middle ear imaging instead of systematic surgical findings. This can be justified by our local institution’s usual practice of deferring either first-stage or second-look systematic surgery, on the basis of MRI findings. Moreover, the mean duration of medical follow-up was comparable to recent literature [25] suggesting that non-EPI DW MRI can efficiently select patients for second-look surgery. If associated with clinicoradiological follow-up, it allows avoiding unnecessary surgical revisions when result is judged satisfactory and no secondary reconstruction is needed [26,27]. Only one subject of the studied population benefited from secondary myringoplasty, which corroborates those findings.

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

8

J. Lincot et al.

Figure 5 Non-EP DWI false positive case: 79-year-old female clinically suffering from chronic otitis and never operated; epitympanic hyperplastic granulation tissue misdiagnosed as a cholesteatoma. A and B, 3T and 1.5T non-EP DWI (b = 1000s.mm−2 ), respectively, show a small epitympanic signal hyperintensity (arrow). C and D, DGE T1- and HRT2-weighted images show diffuse enhancement and hyperintensity, respectively.

Conclusions Non-EP DWI for the detection of either newly diagnosed or residual/recurrent acquired cholesteatoma can be performed on 1.5T or 3T hardware indifferently. High Se and NPV values can be achieved by the use of a single nonEP diffusion-weighted protocol. Conversely, such a protocol provides relatively lower values of Sp and PPV. Standard sequences, although not significantly modifying diagnostic performances achieved in our study may keep a role in difficult cases, in diagnosing complications and in planning surgical procedures.

Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.

Acknowledgement None.

References [1] Barath K, Huber AM, Stampfli P, Varga Z, Kollias S. Neuroradiology of cholesteatomas. AJNR Am J Neuroradiol 2011;32(2):221—9.

[2] Mafee MF, Levin BC, Applebaum EL, Campos M, James CF. Cholesteatoma of the middle ear and mastoid. A comparison of CT scan and operative findings. Otolaryngol Clin North Am 1988;21(2):265—93. [3] Alicandri-Ciufelli M, Marchioni D, Grammatica A, et al. Tympanoplasty: an up-to-date pictorial review. J Neuroradiol 2012;39(3):149—57. [4] Vercruysse J-P, De Foer B, Pouillon M, Somers T, Casselman J, Offeciers E. The value of diffusion-weighted MR imaging in the diagnosis of primary acquired and residual cholesteatoma: a surgical verified study of 100 patients. Eur Radiol 2006;16(7):1461—7. [5] Thiriat S, Riehm S, Kremer S, Martin E, Veillon F. Apparent diffusion coefficient values of middle ear cholesteatoma differ from abscess and cholesteatoma admixed infection. AJNR Am J Neuroradiol 2009;30(6):1123—6. [6] Dhepnorrarat RC, Wood B, Rajan GP. Postoperative nonecho-planar diffusion-weighted magnetic resonance imaging changes after cholesteatoma surgery: implications for cholesteatoma screening. Otol Neurotol 2009;30(1): 54—8. [7] Dubrulle F, Souillard R, Chechin D, Vaneecloo FM, Desaulty A, Vincent C. Diffusion-weighted MR imaging sequence in the detection of postoperative recurrent cholesteatoma. Radiology 2006;238(2):604—10. [8] Williams M, Ayache D, Alberti C, et al. Detection of postoperative residual cholesteatoma with delayed contrast-enhanced MR imaging: initial findings. Eur Radiol 2003;13(1):169—74. [9] Ayache D, Williams MT, Lejeune D, Corré A. Usefulness of delayed postcontrast magnetic resonance imaging in the

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

+Model NEURAD-468; No. of Pages 9

ARTICLE IN PRESS

Middle ear cholesteatoma: Compared diagnostic performances of two MRI protocols

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

detection of residual cholesteatoma after canal wall-up tympanoplasty. Laryngoscope 2005;115(4):607—10. Stasolla A, Magliulo G, Parrotto D, Luppi G, Marini M. Detection of postoperative relapsing/residual cholesteatomas with diffusion-weighted echo-planar magnetic resonance imaging. Otol Neurotol 2004;25(6):879—84. Aikele P, Kittner T, Offergeld C, Kaftan H, Hüttenbrink K-B, Laniado M. Diffusion-weighted MR imaging of cholesteatoma in pediatric and adult patients who have undergone middle ear surgery. AJR Am J Roentgenol 2003;181(1):261—5. De Foer B, Vercruysse J-P, Pilet B, et al. Single-shot, turbo spin-echo, diffusion-weighted imaging versus spin-echo planar, diffusion-weighted imaging in the detection of acquired middle ear cholesteatoma. AJNR Am J Neuroradiol 2006;27(7):1480—2. Sharifian H, Taheri E, Borghei P, et al. Diagnostic accuracy of non-echo planar diffusion-weighted MRI versus other MRI sequences in cholesteatoma. Journal of medical imaging and radiation oncology 2012;56(4):398—408. De Foer B, Vercruysse J-P, Bernaerts A, et al. Middle ear cholesteatoma: non-echo planar diffusion-weighted MR imaging versus delayed gadolinium-enhanced T1-weighted MR imaging — value in detection. Radiology 2010;255(3):866—72. Venail F, Bonafe A, Poirrier V, Mondain M, Uziel A. Comparison of echo-planar diffusion-weighted imaging and delayed postcontrast T1-weighted MR imaging for the detection of residual cholesteatoma. AJNR Am J Neuroradiol 2008;29(7):1363—8. De Foer B, Vercruysse J-P, Spaepen M, et al. Diffusion-weighted magnetic resonance imaging of the temporal bone. Neuroradiology 2010;52(9):785—807. Lechner Goyault J, Riehm S, Neuville A, Gentine A, Veillon F. Interest of diffusion-weighted and gadolinium-enhanced dynamic MR sequences for the diagnosis of parotid gland tumors. J Neuroradiol 2010. Pizzini FB, Barbieri F, Beltramello A, Alessandrini F, Fiorino F. HASTE diffusion-weighted 3-Tesla magnetic resonance imaging

[19]

[20]

[21]

[22] [23]

[24] [25]

[26]

[27]

9

in the diagnosis of primary and relapsing cholesteatoma. Otol Neurotol 2010;31(4):596—602. Lehmann P, Saliou G, Brochart C, et al. 3 T MR imaging of postoperative recurrent middle ear cholesteatomas: value of periodically rotated overlapping parallel lines with enhanced reconstruction diffusion-weighted MR imaging. AJNR Am J Neuroradiol 2009;30(2):423—7. Kasbekar AV, Scoffings DJ, Kenway B, et al. Non echo planar, diffusion-weighted magnetic resonance imaging (periodically rotated overlapping parallel lines with enhanced reconstruction sequence) compared with echo planar imaging for the detection of middle-ear cholesteatoma. J Laryngol Otol 2010:1—5. De Foer B, Vercruysse J-P, Bernaerts A, et al. The value of single-shot turbo spin-echo diffusion-weighted MR imaging in the detection of middle ear cholesteatoma. Neuroradiology 2007;49(10):841—8. Khemani S, Singh A, Lingam RK, Kalan A. Imaging of postoperative middle ear cholesteatoma. Clin Radiol 2011;66(8):760—7. Huins CT, Singh A, Lingam RK, Kalan A. Detecting cholesteatoma with non-echo planar (HASTE) diffusionweighted magnetic resonance imaging. Otolaryngol Head Neck Surg 2010;143(1):141—6. Barath K, Huber AM, Stampfli P, Varga Z, Kollias S. Neuroradiology of cholesteatomas. AJNR Am J Neuroradiol 2010:1—9. De Foer B, Vercruysse JP, Bernaerts A, et al. Detection of postoperative residual cholesteatoma with non-echo planar diffusion-weighted magnetic resonance imaging. Otol Neurotol 2008;29(4):513—7. Schwartz KM, Lane JI, Bolster Jr BD, Neff BA. The utility of diffusion-weighted imaging for cholesteatoma evaluation. AJNR Am J Neuroradiol 2011;32(3):430—6. Corrales CE, Blevins NH. Imaging for evaluation of cholesteatoma: current concepts and future directions. Curr Opin Otolaryngol Head Neck Surg 2013;21(5):461—7.

Please cite this article in press as: Lincot J, et al. Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners. J Neuroradiol (2014), http://dx.doi.org/10.1016/j.neurad.2014.02.003

Middle ear cholesteatoma: Compared diagnostic performances of two incremental MRI protocols including non-echo planar diffusion-weighted imaging acquired on 3T and 1.5T scanners.

To compare diagnostic performances for cholesteatoma diagnosis of incremental MRI protocols including non-echo planar diffusion-weighted imaging (DWI)...
1MB Sizes 2 Downloads 4 Views