Detecting ultrasonographic hollowness in small choroidal melanocytic tumors using 10 MHz and 20 MHz ultrasonography: a comparative study

Transcription

1 Graefes Arch Clin Exp Ophthalmol (2014) 252: DOI /s ONCOLOGY Detecting ultrasonographic hollowness in small choroidal melanocytic tumors using 10 MHz and 20 MHz ultrasonography: a comparative study Antonio Piñeiro-Ces & María José Rodríguez Alvarez & María Santiago & Manuel Bande & María Pardo & Carmela Capeáns & María José Blanco Received: 3 February 2014 /Revised: 17 July 2014 /Accepted: 22 July 2014 /Published online: 8 August 2014 # Springer-Verlag Berlin Heidelberg 2014 Abstract Background The purpose of this work was to compare the detection of ultrasonographic hollowness (UH), as a risk sign for evolution from small choroidal melanocytic tumors (SCMT) to uveal melanoma (UM), between conventional ultrasonography (standardized 8 MHz ultrasonography and B-mode 10 MHz ultrasonography) and high-resolution 20 MHz ultrasonography. Methods Fifty SCMTs from 50 eyes were included in this work. In all cases, ultrasonographic studies were performed using: 8 MHz standardized A-mode, 10 MHz B-mode, and posterior pole 20 MHz B-mode. Comparison between the presence and the absence of UH were carried out between the ultrasonographic images. Results There were no statistically significant differences between the SCMT dimensions obtained using the 8 10 and 20 MHz techniques. UH was detected in 12 and 20 cases by means of ten and 20 MHz probes respectively. The difference between these proportions was statistically different from zero (McNemar test, p-value=0.008). Cases without UH by 20 MHz have lower height values than cases with UH. However, these differences were not found by 10 MHz A. Piñeiro-Ces (*): M. Santiago : M. Bande : C. Capeáns : M. J. Blanco Surgical Retina and Ocular Oncology Unit, Servizo de Oftalmoloxía, Hospital de Conxo, Complexo Hospitalario de Santiago, c/ramón Baltar s/n, Santiago de Compostela, Spain antonio.pineiro@usc.es M. J. Rodríguez Alvarez Department of Statistics and Operational Research, University of Vigo, Vigo, Spain M. Pardo Grupo Obesidómica, Laboratorio de Endocrinología Molecular y Celular, Complexo Hospitalario Universitario de Santiago, Santiago de Compostela, Spain ultrasonography. By receiver operating characteristic (ROC) study, specificity was better by 20 MHz than 10 MHz ultrasonography when the value of tumor height as marker for UH was studied. Conclusions UH is easier to detect by 20 MHz than by 10 MHz ultrasonography. This ultrasonographic sign appears to be correlated with the height of the tumor. Thus, we believe UH estimation by 20 MHz ultrasonography could be used as a significant predictive factor for SCMT growth. Keywords Uveal melanoma. Small choroidal melanocytic lesions. Ocular ultrasonography. Ultrasonographic hollowness Introduction Uveal melanoma (UM) is an ophthalmic malignant disease, and is the most common primary intraocular cancer in adults. In the last 2 decades, life expectancy after ocular therapy (enucleation and/or different methods of radiotherapy) has not improved. Even with local success from conservative treatments, patients remain at risk for developing metastases for more than 12 years after diagnosis and treatment [1]. In the Collaborative Ocular Melanoma Study (COMS), the Kaplan Meier analysis estimated that the 2-, 5-, and 10-year posttreatment metastasis rates were 10, 25, and 34 % respectively [2]. To identify the early stages of the disease, there have been efforts to detect clinical and ultrasonographic signals within small choroidal melanocytic tumors (SCMTs) that predict a high risk of progression to malignant lesions (UM). These works have resulted in publications about that transformation of SCMTs to a UM [3 9]. One of the more extensive studies was published by Shields et al. (2009) [6], who found several

2 2006 Graefes Arch Clin Exp Ophthalmol (2014) 252: clinical and ultrasonographic risk factors for malignant transformation, including: visual acuity diminution, visual field defects, flashes and floaters, a SCMT with a height greater than 2 mm, the presence of subretinal fluid and orange pigment, a tumor margin within 3 mm of the optic disc, ultrasonographic hollowness (UH), and no halo. Subsequently, the median hazard ratios (HRs) were 3 for patients with one to two risk factors, 5 for patients with three to four risk factors, 9 for patients with five to six risk factors, and 21 for patients with all seven risk factors. The highest HR, 31, was found to have a combination of the following factors: symptoms, orange pigment, margin near disc, UH, and no halo. Contrary to previously reported results [9], this study found that the absence of drusen in the SCMT was not a risk factor. Two echographic signs were identified as risk factors for progression to malignant lesions: UH and a height greater than 2 mm. UH, or sound attenuation, was first recognized as a predictive growth feature of SCMTs that turned into a UM by Shields et al. [6]. This quantitative echographic sign, detected in B-mode ultrasonography, consisted of a progressive decrease of internal reflectivity from the apex to the base of the tumor, not characteristic of choroidal nevi [10]. It is probably equivalent to the kappa angle in A-mode ultrasonography [11]. The lower reflectivity on an A-scan and acoustic hollowing on a B-scan at the base of a large UM may also be due, in part, to the more homogeneous nature of the tumor in this region. Occasionally, a UM produces very strong sound attenuation that is related to a decreased reflectivity of the orbital echoes behind the tumor. Thus, a quantitative ultrasonographic sign was recognized for the first time as a risk factor for the malignant transformation of a SCMT. Previously, another ultrasonographic sign height greater than 2 mm was identified [4] as a risk factor for malignant transformation of a SCMT. Usually, ultrasonography in intraocular tumors has been performed by standardized and B-mode ultrasonography techniques [12 15]. However, the latest technological advances in ultrasound, particularly with the advent of high frequency transducers, have further increased the scope of their use, and have provided additional clinical applications in ophthalmology. In the last decade, ultrasound probes with higher frequency and higher resolution have been commercialized, such as the 20 MHz probe used for posterior pole studies. Frequency is important in ultrasound because it is directly related to resolution and inversely related to penetration of ultrasonographic images. Moreover, the ultrasonic wavelength is inversely related to the frequency. Thus, the 75 μm wavelength of ultrasound at 20 MHz is half the wavelength of a 10 MHz transducer and, therefore, is theoretically capable of providing a 2-fold improvement in resolution. This technique may be better for imaging choroidal lesions, such as a UM and SCMT, because it provides a more accurate visual of the epiescleral space, resulting in more reproducible lesion height measurements [16 18]. The purpose of this work was to consider the presence of UH as a risk sign for the evolution of a SCMT to a UM, comparing conventional ultrasonography (standardized 8 MHz ultrasonography and B-mode 10 MHz ultrasonography) and high-resolution 20 MHz ultrasonography. Material and methods This study included 50 SCMTs from 50 eyes belonging to 47 consecutive patients referred for a first visit between June 2012 and November 2012 to the Retinal Surgery and Ocular Oncology Unit at the Complexo Hospitalario Universitario de Santiago (Santiago de Compostela, Spain). The criterion for inclusion in this study was SCMTs not involving the iris or ciliary body. The work was approved by the local institutional review board, and informed consent was obtained from all patients included in this study. In all cases, patients underwent a complete ophthalmologic and ultrasonographic examination. The ophthalmologic study included visual acuity and refractive analysis, an anterior segment study, intraocular pressure measurement, fundus examination, and retinography of each SCMT. All ultrasonographic examinations were performed by the same physician (Antonio Piñeiro-Ces). In all cases, both ultrasonic studies were conducted using: 1) an 8 MHz probe for standardized A-mode ultrasonography and a 10 MHz probe for B-mode ultrasonography (Eyecubed I3 ABD System, Innovative Imaging Ellex, Adelaide, Australia); and 2) a posterior pole probe of 20 MHz for the B-mode ultrasonography (Aviso, Quantel Medical, Clermont-Ferrand, France). In B-mode ultrasonographies, the examination was started with the maximum gain (90 db for the 10 MHz probe and 110 db for the 20 MHz probe). Then, the gain was adjusted for each probe to produce optimal images of the SCMTs. Ultrasonographic scans were performed with the patients placed supine on a reclining chair. Viscotears 0.2 % (polyacrylic acid from Novartis Farmaceutica SA, Barcelona, Spain) was used as a coupling gel, and all scans were performed with the probe placed directly on the sclera after the instillation of local anesthetic drops. Ultrasonography was performed in each case by placing the probe on the conjunctiva peripheral to the cornea. The gaze of patients during measurement of the ultrasonographic mass was directly at the meridian location of the choroidal tumor. Three measures were made during each ultrasonographic examination: longitudinal base, transverse base (Bmode of 10 and 20 MHz), and height (A-mode of 8 MHz standardized and 20 MHz). To standardized echographic technique, specific criteria described by Ossoinig were used to evaluate intraocular height [19, 20]. The tumor was centered in the echogram, and the probe was angled back and forth, thus sweeping the acoustic section across the lesion. When

3 Graefes Arch Clin Exp Ophthalmol (2014) 252: Table 1 Mean and standard deviation from SCMT ultrasonographic measurements. P-value by means of paired Student's t-test 8 10 Mhz 20 Mhz Pairwise difference P-value LB (2.434) (2.731) (1.155) TB (1.908) (1.855) (1.369) H (0.487) (0.547) (0.301) LB longitudinal base; TB transverse base; H height this technique was performed, the echogram was monitored to direct the sound beam perpendicular to both the apex of the tumor and the inner sclera at the tumor base. The inner sclera was identified as the first line at the base of the tumor that was continuous with the surrounding fundus. If the retina remained attached to the apex of the tumor, the spike included both the retina and the tumor surface. If the B-scan examination showed the presence of apical subretinal exudative fluid, measurements were taken from the tumor surface spike and not from the retina. Comparisons between the presence and the absence of UH were performed between echograms at the same gain level. UH was considered positive when a clear difference in the internal reflectivity was detected between the gain from the apical and base areas of the tumor. For the statistical analysis, quantitative data are expressed as the mean (standard deviation, SD) and categorical data as percentages. Differences between SCMT dimensions (longitudinal and transverse base and height) obtained using the 8 10 and 20 MHz techniques were statistically analyzed using the Student s t-test for paired data. A Bland Altman analysis [21] was also performed to compute the limits of agreement between these measurements. The proportion of cases determined as UH-positive by means of both ultrasonography techniques was compared using a McNemar test. Finally, receiver operating characteristic (ROC) analysis was performed to evaluate the value of tumor height as a marker for the presence of UH for both 10 MHz and 20 MHz techniques separately. In all cases, a p-value<0.05 was considered statistically significant. All the analyses were performed using the R software, version [22]. Results An observational study was performed in 50 SCMTs from 50 eyes (right eye: n=22; left eye: n=28) belonging to 47 consecutive patients (31 women and 15 men; mean age: years; three patients with an SCMT in each eye). As determined by the ophthalmoscopic and ultrasonographic studies, the SCMT was localized in the posterior pole in 29 cases, in the equator in 14 cases, and in the periphery in seven cases. The 8 10 and 20 MHz scans were both completed in 28 cases. Twenty-two could not be studied by ultrasonography because they were too flat to be correctly measured and quantitatively analyzed (less than 500 μm inheight). There were no significant differences between the SCMT dimensions obtained using the 8 10 and 20 MHz techniques (paired Student's t-test, Table 1). Bland Altman plots for these dimensions are presented in Fig. 1, which shows no systematic trend in any of the studied dimensions, as well as small mean differences and narrow limits of agreement. However, UH was detected in 12 cases with the 10 MHz probe, while the 20 MHz probe found this ultrasonographic sign in 20 cases (Fig. 2). These 12 UH-positive cases detected by the 10 MHz probe presented the same sign as the 20 MHz probe (Table 2). Following this study, all cases of SCMT detected as UH-positive by means of 10 MHz were also Fig. 1 Bland Altman plots to analyze concordance between 8 10 and 20 MHz ultrasonographies to measure longitudinal and transverse bases (LB and TB) and height (H). The middle line indicates the pairwise mean difference. The outer lines indicate 95 % limits of agreement (mean±1.96 standard deviation)

4 2008 Graefes Arch Clin Exp Ophthalmol (2014) 252: Fig. 2 Examples of ultrasonography in SCMT. Case 18 (a and b), case 25 (c and d) and case 31 (e and f). a, c and e Longitudinal ultrasonography by means of 10 MHz ultrasonography. b, d and f Longitudinal ultrasonography by means of 20 MHz ultrasonography. Significant graphical differences by internal reflectivity and associated A-mode were found UH-positive by means of 20 MHz ultrasonography. Sixteen cases were found to be UH-negative by means 10 MHz, and Table 2 Contingency table between UH estimated by 10 and 20 MHz ultrasonography 10 MHz 20 MHz Total UH UH + Total UH UH + 8 (100 %) 8 (40 %) 16 (57.1 %) 0 (0 %) 12 (60 %) 12 (42.9 %) 8 (100 %) 20 (100 %) 28 (100 %) eight of them were found to be UH-positive by means of 20 MHz ultrasonography. In proportional terms, by means of 10 MHz, UH was positive in 42.8 % of cases and, by means of 20 MHz, UH was positive in 71.4 % of cases. The difference between these proportions is statistically different from zero (using McNemar test p=0.008). This result indicates that 20 MHz ultrasonography detects more cases of UH positivity than 10 MHz ultrasonography. We also compared the height of each SCMT with the presence of UH. Figure 3 depicts the distribution of the tumor height for each considered group (positive and negative UH by means of 10 and 20 MHz). It is worth remembering that the groups with positive and negative UH determinations are different for 10 and 20 MHz. As can be observed in Fig. 3, cases with UH-negative status by 20 MHz have lower height values than cases with a UH-positive status (Mann Whitney test, p = 0.002). Using 10 MHz, height differences between both groups were not statistically significant (Mann Whitney test, p=0.179). We also studied the height of SCMTs, considering height as a marker for the positive presence of UH, by ROC study, in terms of sensitivity and specificity (Fig. 4). For different values of height, sensitivity and specificity were calculated to determine the presence or absence of UH for 10 and 20 MHz separately. In both cases sensitivity was similar (the black and grey solid lines were very similar). However, as shown in the anterior graph of Fig. 3 the specificity is better in the case of 20 MHz compared with 10 MHz (the dotted grey line is increasing faster than the black line). For instance, when using 20 MHz and for a cutoff value of height of 1 mm, the sensitivity and specificity were 0.9 and respectively. For the same height value, in the case of 10 MHz, sensitivity was 0.92 (similar to 20 MHz), but specificity was only Discussion UM has well-known clinical signs: however, there are also several well-known ultrasonographic characteristics. These ultrasonographic signs should be considered in UM diagnosis. The cardinal ultrasonographic features include: solid consistency, collar button shape, low to medium reflectivity, regular internal structure, and internal blood flow. Additional factors, including the presence of sound attenuation or acoustic hollowness, the extent of associated exudative retinal detachment, and the status of adjacent scleral/orbital interface, are also important considerations [23]. Measurements of each UM or SCMT by ultrasonography are very important to classify each case, with follow-up

5 Graefes Arch Clin Exp Ophthalmol (2014) 252: Fig. 3 Superposed and separate presentations of heights and UH+ or UH responses with 10 and 20 MHz Superposed Separately observations after conservative treatments. Classically, these measurements were carried out by means of 10 MHz (longitudinal and transverse bases) and standardized 8 MHz ultrasonography (height). In this work, we did not find differences in these measurements (obtained by the classical method and by 20 MHz B-mode ultrasonography). However, a different result was obtained by other authors [24]. They compared 10 and 20 MHz ultrasonographic measurements of conservatively treated or non-treated UMs, and found several differences. When these authors divided groups of UMs by height, they found no differences in the group of UMs with a height less than 3.0 mm. This result would be similar to our findings in SCMTs. Sound attenuation, or acoustic hollowness, is represented by a decrease in echo strength occurring from left to right in the echogram, and results of lower reflectivity at the base of the tumor. On a B-scan, this decrease in reflectivity at the tumor base has been described by some authors as acoustic hollowing [25]. The lower reflectivity on an A- scan and acoustic hollowing on a B-scan at the base of a large UM may also be due, in part, to the more homogeneous nature of the tumor in this region. As suggested

6 2010 Graefes Arch Clin Exp Ophthalmol (2014) 252: Fig. 4 Presence of positive UH by ROC study, in terms of sensitivity and specificity, for different values of height by Shields et al. [6], this sign should be considered as a risk factor for SCMT. As our study showed, UH is more easily detected by 20 MHz than by 10 MHz ultrasonography, and this ultrasonographic sign appears to be correlated with the height of the tumor. It is likely that this finding occurs more frequently with 20 MHz because the echoes decrease faster than 10 MHz echoes when entering an SCMT with initial histological architecture compatible with a UM. The present study is the first to compare the 20 MHz B- scan probe with the conventional 10 MHz B-scan probe to detect the presence of a UH in SCMT, a risk factor for malignant transformation. The authors suggest that a more extensive and prospective study should be performed on SCMT, mainly to confirm if UH detected by 20 MHz is really a risk factor sign, and its relation to the other known signs of risk. This study could be improved by following each case over a series of time points. Conflict of interest All authors certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript. References 1. Collaborative Ocular Melanoma Study Group (2006) Twelve-year mortality rates and prognostic factors. Randomized trial of iodine 125 brachytherapy for choroidal melanoma. COMS report no. 28. Arch Ophthalmol 124: Diener-West M, Reynolds SM, Agugliaro DJ et al (2005) Development of metastatic disease after enrollment in the COMS trials for treatment of choroidal melanoma: Collaborative Ocular Melanoma Study Group, report no. 26. Arch Ophthalmol 123: Augsburger JJ, Schroeder RP, Territo C et al (1989) Clinical parameters predictive of enlargement of melanocytic lesions. Br J Ophthalmol 73: Shields CL, Shields JA, Kiratli H et al (1995) Risk factors for growth and metastasis of small choroidal melanocytic lesions. Ophthalmology 102: Shields CL, Cater J, Shields JA et al (2000) Combination of clinical factors predictive of growth of small choroidal melanocytic tumors. Arch Ophthalmol 118: Shields CL, Furuta M, Berman EL et al (2009) Choroidal nevus transformation into melanoma: analysis of 2514 consecutive cases. Arch Ophthalmol 127: Collaborative Ocular Melanoma Study Group (1997) Factors predictive of growth and treatment of small choroidal melanoma. COMS report no. 5. Arch Ophthalmol 115: Singh AD, Kalyani P, Topham A (2005) Estimating the risk of malignant transformation of a choroidal nevus. Ophthalmology 112: Singh AD, Mokashi AA, Bena JF, Jacques R, Rundle PA, Rennie IG (2006) Small choroidal melanocytic lesions: features predictive of growth. Ophthalmology 113: Ossoinig KC, Lohmeyer M (1990) Choroidal nevi: diagnosis with standardized echography. In: Sampaolesi R (ed) Ultrasonography in ophthalmology, 12th edn. Kluwer, Dordrecht, p Minning CA Jr, Davidrof FH (1982) Ossoinig s angle of ultrasonic absorption and its role in the diagnosis of malignant melanoma. Ann Ophthalmol 14: Collaborative Ocular Melanoma Study Group (2003) Comparison of clinical, echographic, and histopathological measurements from eyes with medium-sized choroidal melanoma in the Collaborative Ocular Melanoma Study. COMS report no. 21. Arch Ophthalmol 121: Byrne SF, Marsh MJ, Boldt HC, Green RL, Johnson RN, Wilson DJ (2002) Consistency of observations from echograms made centrally in the Collaborative Ocular Melanoma Study. Ophthalmic Epidemiol 9: Collaborative Ocular Melanoma Study (1999) Echography (ultrasound) procedures for the Collaborative Ocular Melanoma Study (COMS), report no. 12, part II. J Ophthalmic Nurs Technol 18: Collaborative Ocular Melanoma Study (1999) Echography (ultrasound) procedures for the Collaborative Ocular Melanoma Study (COMS), report no. 12, part I. J Ophthalmic Nurs Technol 18: Berson M, Grégoire JM, Gens F et al (1999) High frequency (20 MHz) ultrasonic devices: advantages and applications. Eur J Ultrasound 10: Hewick SA, Fairhead AC, Culy JC, Atta HR (2004) A comparison of 10 MHz and 20 MHz ultrasound probes in imaging the eye and orbit. Br J Ophthalmol 88: Coleman DJ, Silverman RH, Chabi A et al (2004) High-resolution ultrasonic imaging of the posterior segment. Ophthalmology 111: Ossoinig KC (1979) Standardized echography: basic principles, clinical applications, and results. Int Ophthalmol Clin 19: Ossoinig KC, Bigar F, Kaefring SL (1975) Malignant melanoma of the choroid and ciliary body: a differential diagnosis in clinical echography. Bibl Ophthalmol 83: Bland JM, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:

7 Graefes Arch Clin Exp Ophthalmol (2014) 252: R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. URL: Accessed 15 Jan Frazier Byrne S, Green RL (2002) Intraocular tumors. In: Frazier Byrne S, Green RL (eds) Ultrasound of the eye and orbit, 2nd edn. Mosby, St. Louis, pp Kook D, Kreutzer TC, Wolf A, Haritoglou C (2011) Variability of standardized echographic ultrasound using 10 MHz and highresolution 20 MHz B scan in measuring intraocular melanoma. Clin Ophthalmol 5: Goldberg MF, Hodes BL (1977) Ultrasonographic diagnosis of choroidal malignant melanoma. Surv Ophthalmol 22:29 40