Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2441060456
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Graaf, P.
Right arrow Articles by Castelijns, J. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Graaf, P.
Right arrow Articles by Castelijns, J. A.
(Radiology 2007;244:273-280.)
© RSNA, 2007

Pediatric Imaging

Eye Size in Retinoblastoma: MR Imaging Measurements in Normal and Affected Eyes1

Pim de Graaf, MD, Dirk L. Knol, PhD, Annette C. Moll, MD, PhD, Saskia M. Imhof, MD, PhD, Antoinette Y. N. Schouten-van Meeteren, MD, PhD, and Jonas A. Castelijns, MD, PhD, BSc

1 From the Departments of Radiology (P.d.G., J.A.C.), Clinical Epidemiology and Biostatistics (D.L.K.), Ophthalmology (A.C.M., S.M.I.), and Pediatric Oncology (A.Y.N.S.), VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands. Received March 13, 2006; revision requested May 17; revision received June 26; accepted July 19; final version accepted October 4. P.d.G. supported in part by grants from ZonMw (Netherlands Organization for Health Research and Development), 's-Gravenhage, the Netherlands; the ODAS Foundation, Delft, the Netherlands; the National Foundation for the Blind and Visually Impaired, Utrecht, the Netherlands; the Blindenhulp Foundation, 's-Gravenhage, the Netherlands; and the Dutch Eye Fund, grant 2004-23, Utrecht, the Netherlands. Address correspondence to P.d.G. (e-mail: p.degraaf{at}vumc.nl).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To evaluate eye size retrospectively by using magnetic resonance (MR) imaging to measure axial length (AL), equatorial diameter (ED), and eye volume (EV) in patients with retinoblastoma and to evaluate the possible effect of retinoblastoma on eye size.

Materials and Methods: Local ethics committee approval was obtained with waiver of informed consent. MR images of 100 patients with retinoblastoma (50 girls, 50 boys; mean age, 19 months; age range, 9 days to 68 months) were scored by one observer (AL, ED, EV, and tumor volume measurements), with a second observer reviewing all measurements. Normal eyes of patients with unilateral retinoblastoma served as controls. Interobserver measurement agreement was evaluated in a random subset of 50 eyes with use of intraclass correlation coefficients. Linear mixed model analysis was used with adjustments for age, laterality, and sex.

Results: Interobserver agreement was good (intraclass correlation coefficients ≥ 0.89). Eyes with retinoblastoma had significantly shorter ALs (95% confidence interval [CI]: –0.57 mm, –0.16 mm; P = .001) and EDs (95% CI: –1.01 mm, –0.66 mm; P < .001) and significantly smaller EVs (95% CI: –336 mm3, –151 mm3; P < .001) than normal eyes. Within patients, a significant negative relationship was found between tumor volume and EV (P < .001).

Conclusion: MR imaging measurements of AL, ED, and EV were significantly smaller in eyes with retinoblastoma than in normal eyes. In addition, in patients with retinoblastoma, the larger the tumor volume, the smaller the eye.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Retinoblastoma is a rare cancer of the eye, and it predominantly affects children younger than 5 years. Tumors can occur in both hereditary (usually bilateral) and nonhereditary (always unilateral) forms. Retinoblastoma is diagnosed on clinical grounds by means of extensive indirect ophthalmoscopy. Complementary investigations in characterizing and staging the extent of disease include patient history, physical examination, ultrasonography (US), and magnetic resonance (MR) imaging of the eyes and brain (1). A number of benign conditions can clinically resemble retinoblastoma, sometimes creating considerable diagnostic difficulty (2). The three conditions that most commonly resemble retinoblastoma are persistent hyperplastic primary vitreous, Coats disease, and ocular toxocariasis (3). Radiologic examinations can provide useful additional information to exclude similar lesions from the differential diagnosis.

Disease in children's eyes may be indicated by variations in ocular biometric parameters. Eyes with malformations, such as persistent hyperplastic primary vitreous or retinopathy of prematurity, are usually small, and small size is used as an important diagnostic criterion. In unilateral Coats disease (4), the affected globe is significantly smaller than the normal eye. Characteristically, eyes with retinoblastoma are assumed to be of normal size (4–10), and it is generally accepted that a small eye is sufficient evidence to rule out retinoblastoma (5,6). If this statement is valid, eye size could be used as an additional diagnostic tool to differentiate between retinoblastoma and other (benign) abnormalities included in the differential diagnosis.

However, there is almost no objective information available in the literature about the size of eyes with retinoblastoma. Galluzzi et al (4) performed a preliminary study in which the axial length (AL) and equatorial diameter (ED) in 18 patients with unilateral retinoblastoma were measured with computed tomography (CT) (all patients) or MR imaging (four patients), and they found that the mean calculated eye volume (EV) (from the measured AL and ED) of affected eyes was not significantly different from that in normal eyes. In contrast to this finding, incidental case reports describe unilateral or bilateral retinoblastoma in microphthalmic eyes (11–16). In the presence of retinoblastoma, even an increase in eye size has been reported (5,6).

Several imaging modalities are available for ocular biometric measurements; of these modalities, US is used most commonly. Existing age-related standard values and growth curves of ocular components are based on US results (17–19), but the frequent presence of extensive tumor calcifications hampers accurate US biometric measurements in eyes with retinoblastoma. Attempts were made to establish AL, ED, and EV standards with CT and MR imaging (4,20–22). However, these data must be interpreted prudently because of unclear methods and variations in the radiologic definitions used for AL and ED measurements and the generally accepted ophthalmologic US definitions. To our knowledge, quantitative MR imaging assessment of AL, ED, and EV, according to ophthalmologic US definitions, has not been reported for patients with retinoblastoma. Thus, the purpose of our study was to evaluate eye size retrospectively by using MR imaging to measure AL, ED, and EV in patients with retinoblastoma, and to evaluate the possible effect of retinoblastoma on eye size.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Patient Selection
Local ethics committee approval was obtained; informed consent was waived. We retrospectively identified all patients with retinoblastoma who were evaluated at our institution between January 1992 and December 2004. Patients eligible for this study were those who underwent MR imaging with and without contrast material enhancement before chemotherapy or radiation therapy and for whom digital MR images were available that enabled adequate diagnostic interpretation. We identified 150 patients with retinoblastoma seen during this period. In 39 patients, digital MR images were unavailable; five patients did not receive intravenous injection of contrast material; and in six patients, the quality of MR images was insufficient. The diagnosis of retinoblastoma was confirmed with extensive fundoscopy, US, and MR imaging. US findings were not collected, because in patients with retinoblastoma US was performed to measure tumor size and to detect tumor calcifications. Accurate measurement of affected eyes with US was hampered by tumor calcifications. One author (P.d.G.) reviewed the patients' clinical records. The final study population consisted of 100 patients (50 boys, 50 girls) with a mean age of 19 months at MR imaging (age range, 9 days to 68 months) (Table 1). Retinoblastoma was familial in 12 patients (12%) and sporadic in 88 (88%). The mean follow-up time after MR imaging was 74 months (range, 9–154 months).


View this table:
[in this window]
[in a new window]

  		Table 1. Demographic Data in 100 Patients with Retinoblastoma

 
Normal and Affected Eyes
Retinoblastoma was unilateral in 61 patients (61%) and bilateral in the remaining 39 patients (39%) (Table 1). In the patients with unilateral retinoblastoma proved during follow-up, normal eyes were selected as controls. Within this group, we found two patients with a proved 13q deletion syndrome, in which microphthalmia is a common finding. Both normal eyes were excluded. Two other patients had already undergone enucleation of the affected eye before MR imaging. Within the patients with bilateral retinoblastoma, two children had undergone enucleation of one affected eye before MR imaging. Eyes of patients with unilateral and bilateral retinoblastoma were divided into three groups: 59 eyes of patients with unilateral retinoblastoma that were normal at follow-up (control group), 59 eyes affected by unilateral retinoblastoma, and 76 eyes affected by bilateral retinoblastoma.

MR Imaging
MR imaging was performed at 1.0 T (Magnetom Impact Expert; Siemens, Erlangen, Germany) and 1.5 T (Magnetom Vision, Siemens; Magnetom Sonata, Siemens) with use of a head coil. Unenhanced transverse and sagittal T1-weighted spin-echo MR images were obtained with 350–500/14–15 (repetition time msec/echo time msec) and four signals acquired. Transverse spin-echo MR images were obtained with 2000–2700/19–60 for intermediate-weighted MR images, 2000–2700/80–120 for T2-weighted MR images, and one signal acquired. After intravenous injection of 0.1 mmol of gadopentetate dimeglumine per kilogram of body weight (Magnevist; Schering, Berlin, Germany), sagittal, transverse, and coronal fat-suppressed T1-weighted spin-echo MR images (400–575/13–15, three signals acquired) were obtained. The section thickness for all images was 3 mm, with an intersection gap of 0.3 mm.

Image Analysis and Definitions
Two observers (P.d.G., J.A.C.; 4 and 12 years of experience with orbital MR imaging, respectively) reviewed the MR images. Both observers were blinded to the patients' identity and clinical records. The first observer measured AL, ED, and EV in normal and affected eyes with a computerized image analysis tool that is available as part of our hospital's picture archiving and communication system (Centricity Radiology RA 600, version 6.1; GE Medical Systems, Milwaukee, Wis). To assess interobserver agreement for the measurements, the second observer independently obtained the same measurements in a random subset of 50 eyes (from the total of 194 eyes).

After finalizing the measurements, the second observer reviewed all measurements obtained by the first observer. New measurements were obtained by means of consensus agreement only if there was discordance between the two observers regarding the initial measurements; otherwise, the original measurements of the first observer were maintained for statistical analysis.

AL and ED measurements were obtained on transverse postcontrast T1-weighted MR images with use of a manually controlled distance-measuring function. T1-weighted MR images are preferable to T2-weighted MR images because of the clear delineation of the choroid. Maximal dimensions of AL and ED were measured on transverse sections, which usually showed both insertions of the lateral and medial rectus muscle, the optic nerve, and the thickest lens section. AL was measured in the optic axis, which is a line that passes through the center of the lens, perpendicular to its anterior and posterior surfaces. AL was defined as the distance between the anterior corneal surface and the anterior surface of the enhancing choroid (Figure). This distance was chosen according to standardized measurements obtained with US in ophthalmology (ocular biometry), which are the basis of existing normal AL-to-age growth curves (17–19).


Figure 1
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
  		Transverse contrast-enhanced T1-weighted (450/14) fat-suppressed spin-echo MR image shows unilateral retinoblastoma of the left eye. AL (vertical arrows) is defined as the distance between the anterior corneal surface and the anterior surface of the enhancing choroid. AL is drawn perpendicular to the anterior and posterior surface of the lens. ED (horizontal arrows) is defined as the maximum distance between inner surfaces of the enhancing choroid, perpendicular to the AL.

 
ED was defined as the maximum distance between inner surfaces of the enhancing choroid, perpendicular to the AL (Figure). EV was measured by manually drawing a region of interest on every section showing the eyeball. Regions of interest were placed on the anterior corneal edge, through the inner surface of the enhancing uvea (corpus ciliare and choroid), and back to the starting point. The area inside the outline was automatically calculated, and areas were multiplied by the MR section thickness plus the intersection gap to calculate a per-section EV. The total EV was obtained by summing the volume calculations for all sections. The total tumor volume of affected eyes was measured to evaluate the relationship between tumor size and AL, ED, and EV.

Statistical Analysis
All data were analyzed with the SPSS statistical software package (SPSS for Windows, version 12.0; SPSS, Chicago, Ill). Interobserver agreement was assessed by using intraclass correlation coefficients (ICCs) calculated with the two-way (eyes x observers) random-effects model. ICC was defined as the ratio of the between-eyes variance to the total variance. The 95% confidence intervals (CIs) were calculated with the delta method after Fisher z transformation. Patient characteristics were compared by means of an independent t test for continuous variables and a binomial test for dichotomous variables. Related continuous variables were compared with a paired t test. Quantitative MR imaging data (AL, ED, and EV) for individual eyes of patients with unilateral or bilateral retinoblastoma were analyzed with linear mixed models to adjust for data clustering within patients (23,24). Unstructured variance-covariance matrices were used.

To enable the effect of retinoblastoma on each of the dependent variables to be studied, while at the same time controlling for confounding effects of age, sex, and laterality, the following parameters were entered into the statistical model (one model for each dependent variable): categorical factors of retinoblastoma (eyes with retinoblastoma vs normal eyes), sex (male vs female), laterality (unilateral vs bilateral), and the quantitative covariate (natural logarithm of age). Use of the natural logarithm of age to obtain a linear relationship with the dependent variables is reported to be advantageous (19). Parameter estimates of the intercept and slope coefficients (explanatory variables) of the fitted relationship are presented with their standard errors. The same model, without random intercept, was used to analyze eyes that showed (metachronous) tumor development in the second eye during follow-up (bilateralization).

In addition, mixed-model analysis was used to look at the relationship between tumor volume and EV. This statistical method provides a powerful tool to disentangle between- and within-patient effects (23,24). Between-patient effect describes the effect of mean tumor volume on EV in each patient. Within-patient effect describes the effect of tumor volume on differences in EV between two eyes of the same patient. To facilitate analysis of this effect, within-patient centering (24) of tumor volume in one eye around the mean tumor volume in each patient was performed. Both mean tumor volume (between) and centered tumor volume (within) are entered into the model, together with the factors of laterality and sex and the covariate (natural logarithm) of age. Testing for statistical interaction between explanatory variables was included in all regression models. Nonsignificant interaction terms were removed stepwise, and for significant interactions the effects were averaged with respect to the appropriate variables. All CIs are 95% CIs, and P values less than .05 were considered to indicate statistical significance for all statistical tests.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Except for age at MR imaging (t = 6.76, P < .001), there were no significant differences between patients with unilateral retinoblastoma and those with bilateral retinoblastoma, including difference in sex or the side of the affected eye.

Interobserver Agreement Assessment
The ICC between the two observers was 0.97 (95% CI: 0.94, 0.98) for measurements of AL and 0.89 (95% CI: 0.82, 0.93) for measurements of ED, which indicated good agreement for both. Interobserver agreement for the volume measurement was especially good (ICC = 0.98; 95% CI: 0.95, 0.99).

AL and ED Measurements
The range in measurements of AL and ED showed a substantial overlap between normal eyes and eyes with retinoblastoma (Table 2). However, mean values for AL and ED were less for eyes with retinoblastoma than for normal eyes. The mean AL values were compared with the mean ED values for the three eye groups. The mean AL was significantly longer than the mean ED within all eye groups for normal eyes (t = 3.28, P = .002), eyes with unilateral retinoblastoma (t = 6.36, P < .001), and eyes with bilateral retinoblastoma (t = 3.07, P = .003). Eyes in female patients had a shorter mean AL and a shorter mean ED than eyes in male patients; this difference was not significant in most eye groups, except for the mean difference in AL for patients with unilateral retinoblastoma (t = 2.47; P = .016).


View this table:
[in this window]
[in a new window]

  		Table 2. Measurements of AL, ED, and EV in 100 Patients with Retinoblastoma

 
Linear mixed model analyses (Table 3) showed a significant relationship between age and AL (P < .001) and between sex and AL (P = .02) after adjustment for other variables. When possible differences in age, sex, and laterality were taken into account, the relationship between AL and the presence of an intraocular retinoblastoma was significant (P = .001); the mean AL in eyes with retinoblastoma was 0.36 mm (95% CI: –0.57 mm, –0.16 mm) smaller than that in normal eyes.


View this table:
[in this window]
[in a new window]

  		Table 3. Linear Mixed Model Effect of Retinoblastoma on AL, ED, and EV

 
Age (P < .001) was independently associated with ED. In contrast to AL, ED showed no significant sex difference (95% CI: –0.007 mm, 0.73 mm; P = .05). For laterality, the EDs of affected eyes in patients with unilateral retinoblastoma were smaller than those in patients with bilateral retinoblastoma, but this difference was also not significant (95% CI: –0.90 mm, 0.003 mm; P = .05). After adjustment for these variables, the relationship between ED and retinoblastoma was highly significant (P < .001); the mean ED in eyes with retinoblastoma was 0.87 mm (95% CI: –1.01 mm, –0.66 mm) smaller than that in normal eyes.

EV Measurements
There was a substantial overlap in EV between normal and affected eyes (Table 2). The volume of eyes was smaller in female patients than in male patients in all groups, but this difference was not significant. With linear mixed model analysis (Table 3), the difference in mean EV between female and male patients remained nonsignificant (P = .11). The relationship between EV and age showed a highly significant relationship (P < .001). Laterality-, sex-, and age-adjusted mean EV appeared to be 244 mm3 smaller for eyes with retinoblastoma than for normal eyes (95% CI: –336 mm3, –151 mm3; P < .001).

Tumor Volume
Tumor volume could be measured in all eyes (n = 59) in patients with unilateral retinoblastoma. Twenty-four of the patients with bilateral retinoblastoma had tumors large enough to obtain measurements in both eyes (n = 48), and nine patients had detectable tumors in one eye (n = 9). In the remaining 19 eyes, the tumor was not detected at MR imaging (n = 12) or MR imaging was performed before the second eye was affected (before bilateralization) (n = 7). Analysis of the latter small subgroup revealed a smaller mean EV of 406 mm3 before clinical presentation of retinoblastoma during follow-up. However, this relationship was not significant (95% CI: –1021 mm3, 209 mm3; P = .19). The mean tumor volume was 1030 mm3 ± 720 (± standard deviation; range, 14–2882 mm3) for the whole group, 1320 mm3 ± 626 (range, 26–2882 mm3) for patients with unilateral retinoblastoma, and 731 mm3 ± 691 (range, 14–2372 mm3) for patients with bilateral retinoblastoma. The linear mixed model for the effect of tumor volume on EV (Table 4) showed significant interactions between age and laterality (P = .003) and between centered tumor volume and laterality (P = .03).


View this table:
[in this window]
[in a new window]

  		Table 4. Linear Mixed Model Effect of Retinoblastoma Tumor Volume on EV

 
To determine the effects of these interactions between laterality and age and between laterality and centered tumor, we analyzed the effects within each laterality group. Again, age showed a significant relationship with EV, but this effect was stronger for eyes of patients with unilateral retinoblastoma (95% CI: 697 mm3, 1016 mm3; P < .001) than for eyes of patients with bilateral retinoblastoma (95% CI: 416 mm3, 675 mm3; P < .001). Within patients, a significant negative relationship was found between tumor volume and EV for eyes of patients with bilateral retinoblastoma (95% CI: –0.39 mm3, –0.16 mm3; P < .001) and patients with unilateral retinoblastoma (95% CI: –0.18, –0.06; P < .001). Between patients, no relationship was present (P = .30).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
In our study, eyes with retinoblastoma and those without retinoblastoma showed significant differences in important biometric variables and EV. Eyes with retinoblastoma had significantly shorter ALs and EDs and significantly smaller EVs than normal eyes, and in the patients with retinoblastoma, a negative relationship between tumor volume and EV was found. We eliminated possible confounding factors, such as sex, age, and laterality, by using a multilevel regression model. Age contributed significantly to the increase in AL, ED, and EV.

There were differences in mean age between the control group and the retinoblastoma group in our study, caused mainly by the inclusion of patients with bilateral retinoblastoma in the retinoblastoma group. Patients with bilateral retinoblastoma generally present at an earlier age than those with unilateral retinoblastoma. Given that patients with bilateral retinoblastoma tend to be younger, the mean AL, ED, and EV in patients with bilateral retinoblastoma are smaller than those in patients with unilateral retinoblastoma. When an adjustment for age was included in the statistical models for AL, ED, and EV, no significant differences were found between affected eyes for bilateral and unilateral retinoblastoma.

The range in AL, ED, and EV measurements showed substantial overlap between normal and affected eyes, and the differences in size between normal and affected eyes were small. We found mean differences of 0.36 mm for AL, 0.87 mm for ED, and 244 mm3 for EV. Although our study revealed that eyes affected by retinoblastoma were smaller than normal eyes in most patients, it must be pointed out that in some patients with unilateral retinoblastoma the affected eye may be as large as the normal eye or even slightly larger. However, a smaller affected eye should no longer be used to rule out retinoblastoma or to favor a diagnosis of persistent hyperplastic primary vitreous, Coats disease, or retinopathy of prematurity. Significantly higher values for mean AL in comparison with mean ED could be explained by the fact that the eye is not a perfect sphere and that protrusion of the cornea and anterior chamber increase the AL.

The negative relationship between retinoblastoma and AL, ED, and EV suggests that growth is inhibited in affected eyes. Given the location of disease within the posterior part of the eye, it may be postulated that the area of the globe usually undergoing maximal growth is mechanically restricted by the biologic stress of the tumor. Supportive evidence comes from the finding of a negative relationship between tumor size and EV in patients with retinoblastoma: The tumor volume is greatest in the eyes with the smallest volume. Although in our patients there was no significant relationship between an increase in tumor volume and an increase in EV, in a group of patients with more advanced (extraocular) disease, this relationship might be significant.

Besides mechanical restriction by the tumor, other factors may influence growth of the affected eye. Although the difference was not significant, the smaller mean EV in the subgroup of eyes in which a metachronous tumor developed during clinical follow-up suggests that (locally acting) growth factors may also influence this process. Since the sample size of this subgroup is small, future research must be performed to evaluate whether a smaller EV can be used as a risk factor for metachronous tumor development in the other eye.

Comparison of our biometric results with those of previous studies is limited in scope because our study was not designed to correlate MR imaging measurements with US biometric results. US biometry is not possible in eyes with retinoblastoma because of artifacts secondary to intratumoral calcifications. To our knowledge, previous validation studies for MR imaging measurement of AL and ED have not been reported in the literature, but measurements of small volumes in the eye with the same volumetric method proved to be accurate (25). Although existing studies on ocular biometry in children have mainly involved AL measurements obtained with US, current MR imaging data in the control group resemble US measurements in the literature (17–19,21,26–28). However, AL measurements may be slightly underestimated with the MR imaging technique because the measured axis can be tilted from the true axis, MR imaging can skip over the plane that includes the true axis, or both (29).

Our study had some limitations. Because patients with retinoblastoma are rare, we used a retrospective design to create a large study group. Patients were recruited during a 12-year period, and three different MR imagers combined with a head coil were used. The MR images were not as good as those that can currently be acquired with high-spatial-resolution surface coils and high field strength. The use of normal eyes in patients with unilateral retinoblastoma as controls may represent a selection bias compared with a random selection from the general population. However, to deal with this data structure, we used a statistical method designed for clustered data. Since it was not a goal of our study to compare results of normal eyes with reported normal values, the lack of MR imaging–based reference data is of minor importance. Furthermore, the variables of age, laterality, and sex were included in all models as explanatory variables to allow us to study the (independent) effect of retinoblastoma on eye size. Considering these points together, we believe that valid comparisons may be performed between eyes with retinoblastoma and our control group. Further investigations in patients with retinoblastoma and age-matched normal control groups are necessary to confirm these differences in ocular biometry and evaluate the diagnostic yield.

In conclusion, MR imaging measurements of AL, ED, and EV were significantly smaller in eyes with retinoblastoma than in normal eyes. In addition, in patients with retinoblastoma, a larger tumor volume corresponded to a smaller eye. These outcomes suggest that the use of eye size as an additional parameter to differentiate between retinoblastoma and (benign) simulating lesions should be considered carefully, and a decreased eye size should not be used to exclude retinoblastoma or to favor a diagnosis of simulating lesions.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   ACKNOWLEDGMENTS
 
The authors acknowledge the support of Karlijn Verduin, MSc, from their Department of Radiology in the preparation of the manuscript.


   FOOTNOTES
 

Abbreviations: AL = axial length • CI = confidence interval • ED = equatorial diameter • EV = eye volume • ICC = intraclass correlation coefficient

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, A.C.M., J.A.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, P.d.G.; clinical studies, A.C.M., S.M.I., A.Y.N.S., J.A.C.; statistical analysis, D.L.K.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. de Graaf P, Barkhof F, Moll AC, et al. Retinoblastoma: MR imaging parameters in detection of tumor extent. Radiology 2005;235:197–207.[Abstract/Free Full Text]
  2. Shields JA, Shields CL, Parsons HM. Differential diagnosis of retinoblastoma. Retina 1991;11:232–243.[CrossRef][Medline]
  3. Shields JA, Parsons HM, Shields CL, Shah P. Lesions simulating retinoblastoma. J Pediatr Ophthalmol Strabismus 1991;28:338–340.[Medline]
  4. Galluzzi P, Venturi C, Cerase A, et al. Coats disease: smaller volume of the affected globe. Radiology 2001;221:64–69.[Abstract/Free Full Text]
  5. Smirniotopoulos JG, Bargallo N, Mafee MF. Differential diagnosis of leukokoria: radiologic-pathologic correlation. RadioGraphics 1994;14:1059–1079.[Abstract]
  6. Smith M, Castillo M. Imaging and differential diagnosis of the large eye. RadioGraphics 1994;14:721–728.[Abstract]
  7. Provenzale JM, Weber AL, Klintworth GK, McLendon RE. Radiologic-pathologic correlation: bilateral retinoblastoma with coexistent pinealoblastoma (trilateral retinoblastoma). AJNR Am J Neuroradiol 1995;16:157–165.[Medline]
  8. Edward DP, Mafee MF, Garcia-Valenzuela E, Weiss RA. Coats' disease and persistent hyperplastic primary vitreous: role of MR imaging and CT. Radiol Clin North Am 1998;36:1119–1131.[CrossRef][Medline]
  9. Brisse HJ, Lumbroso L, Freneaux PC, et al. Sonographic CT, and MR imaging findings in diffuse infiltrative retinoblastoma: report of two cases with histologic comparison. AJNR Am J Neuroradiol 2001;22:499–504.[Abstract/Free Full Text]
  10. O'Brien JM. Retinoblastoma: clinical presentation and the role of neuroimaging. AJNR Am J Neuroradiol 2001;22:426–428.[Medline]
  11. Axelsen I. Retinoblastoma in a microphthalmic eye. Ophthalmologica 1978;176:27–33.[Medline]
  12. Mauger TF, Makley TA, Davidorf FH, Rogers GL. Retinoblastoma, microphthalmia, coloboma, and neuroepithelioma of the pineal body. Ann Ophthalmol 1992;24:290–294.[Medline]
  13. Lee JS, Shin YK, Geun J, Oum BS. Retinoblastoma which developed in microphthalmia. Acta Ophthalmol Scand 1997;75:730–731.[Medline]
  14. Moll AC, Imhof SM, van der Valk P, Schouten-van Meeteren A, Valk J. Case report: a microphthalmic eye in a retinoblastoma patient. Int J Neuroradiol 1999;5:212–216.
  15. Schocket LS, Beaverson KL, Rollins IS, Abramson DH. Bilateral retinoblastoma, microphthalmia, and colobomas in the 13q deletion syndrome. Arch Ophthalmol 2003;121:916–917.[Free Full Text]
  16. Wilson GA, Devaux A, Aroichane M. Retinoblastoma, microphthalmia and the chromosome 13q deletion syndrome. Clin Experiment Ophthalmol 2004;32:101–103.[CrossRef][Medline]
  17. Sampaolesi R, Caruso R. Ocular echometry in the diagnosis of congenital glaucoma. Arch Ophthalmol 1982;100:574–577.[Abstract]
  18. Weiss AH, Kousseff BG, Ross EA, Longbottom J. Simple microphthalmos. Arch Ophthalmol 1989;107:1625–1630.[Abstract]
  19. Jones LA, Mitchell GL, Mutti DO, Hayes JR, Moeschberger ML, Zadnik K. Comparison of ocular component growth curves among refractive error groups in children. Invest Ophthalmol Vis Sci 2005;46:2317–2327.[Abstract/Free Full Text]
  20. Hahn FJ, Chu WK. Ocular volume measured by CT scans. Neuroradiology 1984;26:419–420.[CrossRef][Medline]
  21. Weiss AH, Kousseff BG, Ross EA, Longbottom J. Complex microphthalmos. Arch Ophthalmol 1989;107:1619–1624.[Abstract]
  22. Cheng HM, Singh OS, Kwong KK, Xiong J, Woods BT, Brady TJ. Shape of the myopic eye as seen with high-resolution magnetic resonance imaging. Optom Vis Sci 1992;69:698–701.[CrossRef][Medline]
  23. Fitzmaurice GM, Laird NM, Ware JH. Applied longitudinal analysis. New York, NY: Wiley, 2004.
  24. Snijders TAB, Bosker RJ. Multilevel analysis. London, England: Sage, 1999.
  25. Lemke AJ, Kazi I, Hosten N, Bechrakis NE, Foerster MH, Felix R. MR-based volumetric analysis of small tumor volumes: accuracy of phantom examinations of simulated eye tumors [in German]. Rofo 2003;175:958–962.[Medline]
  26. Fledelius HC, Christensen AC. Reappraisal of the human ocular growth curve in fetal life, infancy, and early childhood. Br J Ophthalmol 1996;80:918–921.[Abstract/Free Full Text]
  27. Gordon RA, Donzis PB. Refractive development of the human eye. Arch Ophthalmol 1985;103:785–789.[Abstract]
  28. Larsen JS. The sagittal growth of the eye. IV. Ultrasonic measurement of the axial length of the eye from birth to puberty. Acta Ophthalmol (Copenh) 1971;49:873–886.
  29. Takei K, Sekine Y, Okamoto F, Hommura S. Measurement of axial length of eyes with incomplete filling of silicone oil in the vitreous cavity using x ray computed tomography. Br J Ophthalmol 2002;86:47–50.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Graaf, P.
Right arrow Articles by Castelijns, J. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Graaf, P.
Right arrow Articles by Castelijns, J. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE