ORIGINAL ARTICLE
Quantitative echography in primary uveal melanoma treated by proton beam therapy Carlo Mosci, MD,* Francesco B. Lanza, MD,* Sofia Mosci, PhD,† Annalisa Barla, PhD† ABSTRACT ● RÉSUMÉ Objective: To describe the dynamics of thickness and internal reflectivity after proton beam therapy (PBT) in uveal melanoma. Participants: One hundred and ninety-eight consecutive patients with choroidal or ciliary body melanoma treated by PBT were retrospectively considered. Methods: The post-PBT follow-up included ophthalmologic examination, retinography, and B and A modes of standardized echography every 6 months. A total of 1393 examinations were performed. We take into account 4 tumour categories according to the seventh TNM classification. Results: Before PBT, tumour thickness ranged from 1.5 to 12.5 mm with a mean of 5.9 mm. Its decrease after radiotherapy was best fitted by the sum of a first-order exponential decay and a constant with a decay half-life of 15 months. Based on the fit, tumour thickness stabilized on a constant value representing, on average, 47% of the initial value. Mean internal reflectivity before PBT was 68%. The dynamics of the reflectivity were best fitted by an exponential and a constant, with rise half-life of 11 months, and stability value of 87%. Conclusions: We found that ultrasonographic dynamics of uveal melanoma treated by PBT resembles a function composed of the sum of a constant and a first-order exponential, as previously noted in studies on brachytherapy. Interestingly, after PBT, because of its shorter half-life, internal reflectivity has a faster dynamic response than thickness in large tumours, suggesting that increase of internal reflectivity is a more sensitive indicator of early response to therapy in larger tumours. Objet : Description de la dynamique de l’épaisseur et de la réflectivité interne après la thérapie du faisceau de proton (TFP) du mélanome uvéal. Nature : Étude de cohorte rétrospective. Participants : Cent quatre-vingt-dix-huit patients consécutifs atteints d’un mélanome choroïdien ou du corps ciliaire et ayant reçu une TFP ont été considérés. Méthodes : Après la TFP, le suivi a compris l’examen ophtalmologique, la rétinographie et les modes A et B d’échographie standardisée aux 6 mois. En tout, 1 393 examens ont été effectués. Nous avons tenu compte de quatre catégories de tumeur selon la septième classification TNM. Résultats : Avant la TFP, l’épaisseur des tumeurs variait entre 1,5 et 12,5 mm, avec une moyenne de 5,9 mm. Sa réduction après la radiothérapie a eu la meilleure concordance avec la somme d’une décroissance exponentielle de premier ordre et d’une constante ayant une demi-vie de décroissance de 15 mois. Selon les critères, l’épaisseur de la tumeur s’est stabilisée sur une valeur constante représentant en moyenne 47 % de la valeur initiale. La moyenne de réflectivité interne avant la TFP était de 68 %. La dynamique de la réflectivité concordait le mieux à une valeur exponentielle et constante, avec une hausse de demi-vie de 11 mois et une stabilité de 87 %. Conclusions : Nous avons trouvé que la dynamique ultrasonographique du mélanome uvéal traité par la TFP ressemble à une fonction composée de la somme d’une constante et d’un exponentiel de premier ordre, tel que mentionné précédemment dans les études sur la brachythérapie. Fait intéressant, après la TFP, étant donné sa demi-vie plus courte, la réflectivité interne a une réponse dynamique plus rapide que l’épaisseur d’une grande tumeur, ce qui suggère que l’accroissement de la réflectivité interne est un indicateur plus sensible de la rapidité de la réaction à la thérapie des plus grandes tumeurs.
Uveal melanoma is the most common primary malignancy of the eye, although it is still a rare cancer with an estimated incidence in the United States of about 6 to 7 cases per 1 million people.1 According to a recent European study, the observed incidence in Italy is lower, about 2 to 4 cases per 1 million people.2 The accuracy of clinical diagnosis, which relies on ophthalmoscopic examination and ultrasound imaging technique, has improved greatly over the last 40 years. In 1990, the Collaborative Ocular Melanoma Study, the largest prospective study of treatment of choroidal melanoma ever
undertaken, reported a misdiagnosis rate of only 0.48% (2/413 eyes with clinical diagnosis of choroidal melanoma).3 The development of ophthalmic ultrasound has been useful for supporting the diagnosis of melanoma in eyes with clear media, suggesting the diagnosis of melanoma in eyes with opaque media, providing reliable measurements of apical tumour height and basal diameters, screening for extraocular tumour extension, ensuring proper plaque placement during brachytherapy, and allowing the monitoring of the response in eye-conserving therapies.4 In particular, B-scan modality is used to localize
From the *Galliera Hospital, Ocular Oncology Center; and †Department of Computer and Information Science, Genoa University, Genoa, Italy.
Can J Ophthalmol 2014;49:60–65 0008-4182/14/$-see front matter & 2014 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcjo.2013.09.007
Originally received Jan. 28, 2013. Final revision Aug. 8, 2013. Accepted Sep. 19, 2013 Correspondence to Carlo Mosci, MD, Ocular Oncology Center, Galliera Hospital, Mura delle Cappuccine 14, IT-16128 Genoa, Italy; carlo. [email protected]
60
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. the tumour, to provide evidence of extrascleral extension, and to measure the diameter and the height of the lesion.5,6 In the 1960s, Ossoinig7 introduced a specific subset of diagnostic ultrasound, that is, standardized echography, that allows to assess the topographic, quantitative, and kinetic features of a tumour. Standardized A-scan, with an S-shaped amplification curve to emphasize the quantification of echo amplitudes, allows an accurate measurement of tumour height8–10 and internal reflectivity.11,12 We study these 2 factors in a large set of patients before and after treatment by proton beam therapy (PBT).
Table 1—Characteristics of tumours before radiotherapy
T1 tumours T2 tumours T3 þ T4 tumours
n
Tumour Thickness, mean ⫾ SD (mm)
Tumour Internal Reflectivity, mean ⫾ SD (%)
58 82 54
3.6 ⫾ 1.0 5.4 ⫾ 1.4 8.9 ⫾ 1.7
81 ⫾ 12 70 ⫾ 13 53 ⫾ 14
The postradiotherapy dynamics of height and internal reflectivity were compared among the 3 groups using the t test and analysis of variance statistics.
RESULTS METHODS We retrospectively included in the study all patients treated with PBT at the Centre Lacassagne in Nice (France) between March 2000 and December 2007. For each patient, a total dose of 60 Gy was delivered in 4 sessions over 4 consecutive days, as previously described in previous reports.13,14 The patients who suffered local recurrence were excluded. We performed clinical and echographic examinations using B-scans and standardized A-scans (CineScan S; Quantel Medical, Z.I. Le Brézet, France) before and after radiotherapy. Follow-up examinations were scheduled every 6 months, but to increase the precision in the temporal dynamics analysis, the measurements were gathered in 3-month groups between 6 and 42 months and in 6-month groups after 45 months, and the mean and SD of each group were calculated. The echographic examination included measurements of base diameter (B-scan), tumour thickness (B- and A-scan), and internal reflectivity (A-scan performed after calibration to “tissue sensitivity” as defined by standardized echography, in which the retina’s reflectivity is 100% and the amplification curve is S-shaped). We paid close attention to an accurate perpendicular orientation of the A-scan sound beam with respect to the point of maximal tumour elevation and the inner sclera. The measurements were obtained by placing calipers on the peak of the surface spike of the tumour and the inner scleral spike. For the evaluation of tissue reflectivity, we used the Quantitative Echography-I software, which calculates a percentage of the average height of the inner tissue spikes. The same experienced examiner performed all scans. The postradiotherapy thickness and internal reflectivity data were fitted with polynomial, hyperbolic, logarithmic, and first- and second-order exponential equations using the least-square fitting method (MATLAB, Version R2008a; The MathWorks, Natick, Mass.). The best fitting equation was chosen using the goodness-of-fit analysis where the error was weighted on the number of samples for each group. Melanomas were classified into 4 classes according to the seventh TNM classification15 T1, T2, T3, or T4 tumours. Because of the small number of T4 tumours (8 cases), we grouped T3 and T4 tumours in the same group.
One hundred and ninety-eight patients with a mean age of 63 ⫾ 13 years (range 23–90 years) were included in the study, with 99 male (50%) and 99 female (50%) patients. We excluded 17 patients who suffered local recurrence (7.9%) after a mean time of 27 ⫾ 24.6 months; 5 of them had T2 tumours, and 12 had T3þT4 tumours. The recurrence pattern was marginal in 7 patients (41%) and diffuse in 10 patients (59%). Each patient was examined, on average, 7 times (range 3–15 times) for a total of 1393 ultrasonography examinations. Mean follow-up time was 60.0 ⫾ 26 months, and median follow-up was 60 months. Before radiotherapy, the mean tumour thickness was 5.9 ⫾ 2.5 mm (range 1.5–12.5 mm), whereas the mean internal reflectivity was 68% ⫾ 17% (range 20%–98%). Characteristics of tumours before radiotherapy are summarized in Table 1. Afterward we fitted the dynamics of the relative thickness [h(t)/h0*100] and internal reflectivity [R(t)] both for the entire set of patients and separately for T1, T2, and T3þT4 tumours. We used various mathematical functions, but each time the function that fitted best was the sum of a constant with first-order exponential decay for the relative thickness [h/h0*100 ¼ A þ B*exp(–Ct)], and with a first-order exponential rise for reflectivity [R ¼ A – B*exp(–Ct)]. In particular, parameter A tells us the stability value, that is, the average value at which tumours stabilize; B is related to a fit for the initial value (for relative thickness, A þ B 100, that is, the initial relative tumour thickness, which simply is 100%; for reflectivity, A – B mean[R(0)], that is, the average initial reflectivity). From C, we can estimate the half-life of the decay for tumour thickness and the halflife of the rise for internal reflectivity (Table 2). The plots in Figure 1 present the comparison of the distribution of tumour thickness at different time points. Until 24 months postradiotherapy, tumour thickness Table 2—Half-lives of thickness and reflectivity for tumours
All tumours T1 tumours T2 tumours T3þT4 tumours
Tumour Thickness Half-life (mo)
Tumour Internal Reflectivity Half-life (mo)
14.96 9.94 11.33 14.67
11.12 9.36 14.68 7.86
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
61
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al.
Fig. 1 — Comparison of the distribution of tumour thickness at different time points. PBT, proton beam therapy.
appears in evolution, because tumours at 6 months after radiotherapy are significantly smaller than tumours at radiotherapy (p o 0.001), and after 12, 18, and 24 months postradiotherapy are still, respectively, smaller than after 6, 12, and 18 months after the treatment (p o 0.001, p o 0.001, and p ¼ 0.016). Conversely, between 24 and 30 months, the tumour thickness stabilizes and the paired t test does not detect any significant change (p ¼ 0.31). Figure 2 shows the decrease in tumour thickness postradiotherapy for the tumours. Before PBT, tumour
thickness ranged from 1.5 to 12.5 mm with a mean of 5.9 mm. Its decrease after radiotherapy was best fitted by the sum of a first-order exponential decay and a constant with a decay half-life of 14.96 months. Based on the fit, tumour thickness stabilized on a constant value representing, on average, 47% of the initial value. T3þT4 tumours stabilized on a lower percentage of their initial thickness, but the time required to stabilize is higher with respect to T1 and T2 tumours. We also performed the comparison of the distribution of tumour thickness at different time points for each
Fig. 2 — Decrease in tumour thickness postradiotherapy. PBT, proton beam therapy.
62
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al.
Fig. 3 — Comparison of the distribution of the tumour internal reflectivity at different time points. PBT, proton beam therapy.
patient subgroup. Although for T1 tumours the first significant p value is found when comparing thickness at 6 and 12 months postradiotherapy (p ¼ 0.07), for T2 tumours it is found when comparing thickness at 18 and 24 months postradiotherapy (p ¼ 0.06), and for T3þT4 tumours it is found when comparing thickness at 12 and 18 months postradiotherapy (p ¼ 0.2). Plots in Figure 3 present the comparison of the distribution of the tumour internal reflectivity at different time points. Until 18 months postradiotherapy, the tumours reflectivity appears in evolution, because tumours at 6 months after radiotherapy have a higher reflectivity
than tumours at radiotherapy (p o 0.001), and tumours at 12 and 18 months postradiotherapy have, respectively, a still higher reflectivity than tumours after 6 and 12 months postradiotherapy (p o 0.001 and p ¼ 0.03). Conversely, between 18 and 24 months, the tumour reflectivity stabilizes and the paired t test does not detect any significant change (p ¼ 0.27). Figure 4 shows the increase in tumour internal reflectivity postradiotherapy. Mean internal reflectivity before PBT was 68%. The dynamics of the reflectivity were best fitted by an exponential and a constant, with rise half-life of 11.12 months, and stability value of 87%.
Fig. 4 — Increase in tumour internal reflectivity postradiotherapy. PBT, proton beam therapy.
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
63
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. Larger tumours (T3þT4 group), which initially had a significantly lower internal reflectivity than T2 (t test; p o 0.001) and T1 tumours (t test; p o 0.001), had a higher initial rise in internal reflectivity (2.43% per month) compared with T2 (0.83%) and T1 tumours (0.77% per month). Nevertheless, T1 tumours stabilized on a higher final internal reflectivity (about 91%) than T2 (about 87%) and T3þT4 tumours (about 81%). We also performed the comparison of the distribution of tumour reflectivity at different time points, for T1, T2, and T3þT4 tumours. For T2 and T3þT4 patient groups, the first significant p value is found when comparing reflectivity at 18 and 12 months postradiotherapy (respectively, p ¼ 0.07 and p ¼ 0.5), whereas for the T1 patient group, the first significant p value is found when comparing reflectivity at 12 and 6 months postradiotherapy (p ¼ 0.4).
DISCUSSION The goal of all forms of radiation therapy is to destroy the reproductive integrity of a tumour. Ionizing radiation damages the DNA of the malignant cells, which may be misrepaired and also may disrupt the integrity of the chromosome. The effects of this damage become manifest during mitosis, at which point the cells cannot successfully replicate. The timing would depend on the proliferation kinetics of the cellular constituents and it will be longer in tumours with slow turnover, including ocular melanoma. Other consequences of irradiation include changes in growth factors and signal transduction pathways, apoptosis, and the regulation of the cell cycle.16,17 Histopathologic studies demonstrated that PBT causes degenerative changes in the tumour (necrosis, fibrosis, balloon cells), decreases the ability of tumour cells to reproduce (fewer mitotic figures), and damages blood supply.18 However, it was shown that postirradiation, most tumours shrink but do not disappear.19–21 In 2002, Kaiserman et al.22 described echographic changes of tumour thickness and internal reflectivity in uveal melanoma treated by brachytherapy. In our study, we analyze the same dynamics in uveal melanomas treated by PBT. After brachytherapy, Kaiserman et al.22 reported a decrease in height in 96.6% of tumours treated. In our study, we observed local recurrence rate of 7.9% (17 patients) with a marginal pattern in 7 patients and a diffuse pattern in 10 patients. To minimize the bias caused by local treatment failures, we excluded these patients from our data set. Moreover, because of the small number of the diffuse pattern relapses, a reliable statistical comparison with the responsive tumours was impossible. In our data set, as well as in the Kaiserman et al. study,22 large tumours have a lower initial internal reflectivity compared with small tumours. However, after
64
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
brachytherapy, large tumours are associated with a slower increase in internal reflectivity, because the rise half-life of large tumours is much larger than the rise half-life of small tumours (8.7 vs 5 months). All tumours then stabilize on a similar final value for internal reflectivity (about 70%). Conversely, after PBT, the rise half-life of large and small tumours is more similar (about 9.4 and 7.9 months), although the rate of the increase ([h – h0]/[t – t0]) is higher in large tumours (2.4% vs 0.8% per month for small tumours). Nevertheless, the stability value of the internal reflectivity is higher in small tumours. As observed by Kaiserman et al. in the brachytherapy study,22 the functional form of height dynamics and internal reflectivity is composed of 2 terms: a constant and a monoexponential decay. This suggested the existence of 2 populations in each tumour, one radioresistant component with higher internal reflectivity and the other radiosensitive with lower internal reflectivity. Interestingly, from our analysis it appears that in T3-T4 tumours, internal reflectivity presents faster variations with respect to tumour thickness, because it has a shorter halflife (7.86 vs 14.67 months). This behaviour is confirmed by the analysis at different time points. In fact, whereas for tumour thickness the first significant p value occurs when comparing tumours at 24 and 18 months postradiotherapy, for internal reflectivity this already happens when comparing tumours at 18 and 12 months postradiotherapy. This means that internal reflectivity stabilizes earlier than tumour thickness. We correlate these findings with early response to radiotherapy that determine cellular changes and loss of fluid that precede a secondary process with ingrowth of macrophages and fibrosis.23 These data suggest ophthalmologists should rely on internal reflectivity for a more sensitive indicator of early response to therapy for large tumours, whereas in small and medium tumours internal reflectivity and tumour thickness are almost interchangeable, although tumour thickness can be preferred. To our knowledge, this is the first attempt to calculate the function describing the dynamics of tumour height and internal reflectivity in a large series of uveal melanomas treated by PBT. As Kaiserman et al.’s22 work is for brachytherapy, we consider our study as a reference model for the follow-up of melanomas treated by PBT.
Disclosure: The authors have no proprietary or commercial interest in any materials discussed in this article. REFERENCES 1. Egan KM, Seddon JM, Glynn RJ, Gragoudas ES, Albert DM. Epidemiologic aspects of uveal melanoma. Surv Ophthalmol. 1988;32:239-51. 2. Virgili G, Gatta G, Ciccolallo L, et al. EUROCARE Working Group. Incidence of uveal melanoma in Europe. Ophthalmology. 2007;114:2309-15.
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. 3. Collaborative Ocular Melanoma Study Group. Accuracy of diagnosis of choroidal melanomas in the Collaborative Ocular Melanoma Study. COMS report no. 1. Arch Ophthalmol. 1990;108:1268-73. 4. Boldt HC, Byrne SF, et al. Baseline echographic characteristics of tumors in eyes of patients enrolled in the Collaborative Ocular Melanoma Study: COMS report no. 29. Ophthalmology. 2008; 115:1390-7, 1397.e1-2. 5. Coleman DJ, Abramson DH, Jack RL, Franzen LA. Ultrasonic diagnosis of tumors of the choroid. Arch Ophthalmol. 1974;91: 344-54. 6. Fuller DG, Snyder WB, Hutton WL, Vaiser A. Ultrasonographic features of choroidal malignant melanomas. Arch Ophthalmol. 1979; 97:1465-72. 7. Ossoinig KC. Standardized echography: basic principles, clinical applications, and results. Int Ophthalmol Clin. 1979;19:127-210. 8. Nicholson DH, Frazier-Byrne S, Chiu MT, Schiffman J, Hughes JR, Novinski EK. Echographic and histologic tumor height measurements in uveal melanoma. Am J Ophthalmol. 1985;100:454-7. 9. Haritoglou C, Neubauer AS, Herzum H, Freeman WR, Mueller AJ. Interobserver and intraobserver variability of measurements of uveal melanomas using standardised echography. Br J Ophthalmol. 2002; 86:1390-4. 10. Kook D, Kreutzer TC, Wolf A, Haritoglou C. Variability of standardized echographic ultrasound using 10 mHz and highresolution 20 mHz B scan in measuring intraocular melanoma. Clin Ophthalmol. 2011;5:477-82. 11. Ossoinig KC, Bigar F, Kaefring SL. Malignant melanoma of the choroid and ciliary body: a differential diagnosis in clinical echography. Bibl Ophthalmol. 1975;83:141-54. 12. Hodes BL, Choromokos E. Standardized A-scan echographic diagnosis of choroidal malignant melanomas. Arch Ophthalmol. 1977;95:593-7. 13. Mosci C, Mosci S, Barla A, Squarcia S, Chauvel P, Iborra N. Proton beam radiotherapy of uveal melanoma: Italian patients treated in Nice, France. Eur J Ophthalmol. 2009;19:654-60.
14. Mosci C, Lanza FB, Barla A, et al. Comparison of clinical outcomes for patients with large choroidal melanoma after primary treatment with enucleation or proton beam radiotherapy. Ophthalmologica. 2012;227:190-6. 15. Edge SD, Byrd DR, Carducci MA, Compton CC, Fritz AG, Greene FL. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2009. 16. Gragoudas ES, Lane AM, Munzenrider J, Egan KM, Li W. Evidence-based estimates of outcome in patients irradiated for intraocular melanoma. Arch Ophthalmol. 2002;120:1665-71. 17. Gragoudas ES. Proton beam irradiation of uveal melanomas: the first 30 years. The Weisenfeld Lecture. Invest Ophthalmol Vis Sci. 2006; 47:4666-73. 18. Saornil MA, Egan KM, Gragoudas ES, Seddon JM, Walsh SM, Albert DM. Histopathology of proton beam-irradiated vs enucleated uveal melanomas. Arch Ophthalmol. 1992;110:1112-8. 19. Boudinet M, Berges O, Le Huerou JY, Lumbroso-Le Rouic L, Desjardins L, Laugier P. Quantitative echography in the follow-up of patients treated with proton-beam irradiation for primary choroidal melanomas. Ultrasound Med Biol. 2007;33:1046-56. 20. Ravazzoni L, Mosci C, Polizzi A, Schenome M, Soldati MR, Buono C. Ultrasonographic follow-up of patients with choroidal melanoma following conservative treatment. Ophthalmologica. 1998;212(Suppl 1):77-8. 21. Romani A, Baldeschi L, Genovesi-Ebert F, et al. Ultrasonographic follow-up of primary choroidal malignant melanoma after proton beam irradiation therapy. Ophthalmologica. 1998;212(Suppl 1): 50-2. 22. Kaiserman I, Anteby I, Chowers I, Blumenthal EZ, Kliers I, Pe’er J. Changes in ultrasound findings in posterior uveal melanoma after Ruthenium 106 brachytherapy. Ophthalmology. 2002;109: 1137-41. 23. Seddon JM, Gragoudas ES, Albert DM. Ciliary body and choroidal melanomas treated by proton beam irradiation. Histopathologic study of eyes. Arch Ophthalmol. 1983;101:1402-8.
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
65
Quantitative echography in primary uveal melanoma treated by proton beam therapy Carlo Mosci, MD,* Francesco B. Lanza, MD,* Sofia Mosci, PhD,† Annalisa Barla, PhD† ABSTRACT ● RÉSUMÉ Objective: To describe the dynamics of thickness and internal reflectivity after proton beam therapy (PBT) in uveal melanoma. Participants: One hundred and ninety-eight consecutive patients with choroidal or ciliary body melanoma treated by PBT were retrospectively considered. Methods: The post-PBT follow-up included ophthalmologic examination, retinography, and B and A modes of standardized echography every 6 months. A total of 1393 examinations were performed. We take into account 4 tumour categories according to the seventh TNM classification. Results: Before PBT, tumour thickness ranged from 1.5 to 12.5 mm with a mean of 5.9 mm. Its decrease after radiotherapy was best fitted by the sum of a first-order exponential decay and a constant with a decay half-life of 15 months. Based on the fit, tumour thickness stabilized on a constant value representing, on average, 47% of the initial value. Mean internal reflectivity before PBT was 68%. The dynamics of the reflectivity were best fitted by an exponential and a constant, with rise half-life of 11 months, and stability value of 87%. Conclusions: We found that ultrasonographic dynamics of uveal melanoma treated by PBT resembles a function composed of the sum of a constant and a first-order exponential, as previously noted in studies on brachytherapy. Interestingly, after PBT, because of its shorter half-life, internal reflectivity has a faster dynamic response than thickness in large tumours, suggesting that increase of internal reflectivity is a more sensitive indicator of early response to therapy in larger tumours. Objet : Description de la dynamique de l’épaisseur et de la réflectivité interne après la thérapie du faisceau de proton (TFP) du mélanome uvéal. Nature : Étude de cohorte rétrospective. Participants : Cent quatre-vingt-dix-huit patients consécutifs atteints d’un mélanome choroïdien ou du corps ciliaire et ayant reçu une TFP ont été considérés. Méthodes : Après la TFP, le suivi a compris l’examen ophtalmologique, la rétinographie et les modes A et B d’échographie standardisée aux 6 mois. En tout, 1 393 examens ont été effectués. Nous avons tenu compte de quatre catégories de tumeur selon la septième classification TNM. Résultats : Avant la TFP, l’épaisseur des tumeurs variait entre 1,5 et 12,5 mm, avec une moyenne de 5,9 mm. Sa réduction après la radiothérapie a eu la meilleure concordance avec la somme d’une décroissance exponentielle de premier ordre et d’une constante ayant une demi-vie de décroissance de 15 mois. Selon les critères, l’épaisseur de la tumeur s’est stabilisée sur une valeur constante représentant en moyenne 47 % de la valeur initiale. La moyenne de réflectivité interne avant la TFP était de 68 %. La dynamique de la réflectivité concordait le mieux à une valeur exponentielle et constante, avec une hausse de demi-vie de 11 mois et une stabilité de 87 %. Conclusions : Nous avons trouvé que la dynamique ultrasonographique du mélanome uvéal traité par la TFP ressemble à une fonction composée de la somme d’une constante et d’un exponentiel de premier ordre, tel que mentionné précédemment dans les études sur la brachythérapie. Fait intéressant, après la TFP, étant donné sa demi-vie plus courte, la réflectivité interne a une réponse dynamique plus rapide que l’épaisseur d’une grande tumeur, ce qui suggère que l’accroissement de la réflectivité interne est un indicateur plus sensible de la rapidité de la réaction à la thérapie des plus grandes tumeurs.
Uveal melanoma is the most common primary malignancy of the eye, although it is still a rare cancer with an estimated incidence in the United States of about 6 to 7 cases per 1 million people.1 According to a recent European study, the observed incidence in Italy is lower, about 2 to 4 cases per 1 million people.2 The accuracy of clinical diagnosis, which relies on ophthalmoscopic examination and ultrasound imaging technique, has improved greatly over the last 40 years. In 1990, the Collaborative Ocular Melanoma Study, the largest prospective study of treatment of choroidal melanoma ever
undertaken, reported a misdiagnosis rate of only 0.48% (2/413 eyes with clinical diagnosis of choroidal melanoma).3 The development of ophthalmic ultrasound has been useful for supporting the diagnosis of melanoma in eyes with clear media, suggesting the diagnosis of melanoma in eyes with opaque media, providing reliable measurements of apical tumour height and basal diameters, screening for extraocular tumour extension, ensuring proper plaque placement during brachytherapy, and allowing the monitoring of the response in eye-conserving therapies.4 In particular, B-scan modality is used to localize
From the *Galliera Hospital, Ocular Oncology Center; and †Department of Computer and Information Science, Genoa University, Genoa, Italy.
Can J Ophthalmol 2014;49:60–65 0008-4182/14/$-see front matter & 2014 Canadian Ophthalmological Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcjo.2013.09.007
Originally received Jan. 28, 2013. Final revision Aug. 8, 2013. Accepted Sep. 19, 2013 Correspondence to Carlo Mosci, MD, Ocular Oncology Center, Galliera Hospital, Mura delle Cappuccine 14, IT-16128 Genoa, Italy; carlo. [email protected]
60
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. the tumour, to provide evidence of extrascleral extension, and to measure the diameter and the height of the lesion.5,6 In the 1960s, Ossoinig7 introduced a specific subset of diagnostic ultrasound, that is, standardized echography, that allows to assess the topographic, quantitative, and kinetic features of a tumour. Standardized A-scan, with an S-shaped amplification curve to emphasize the quantification of echo amplitudes, allows an accurate measurement of tumour height8–10 and internal reflectivity.11,12 We study these 2 factors in a large set of patients before and after treatment by proton beam therapy (PBT).
Table 1—Characteristics of tumours before radiotherapy
T1 tumours T2 tumours T3 þ T4 tumours
n
Tumour Thickness, mean ⫾ SD (mm)
Tumour Internal Reflectivity, mean ⫾ SD (%)
58 82 54
3.6 ⫾ 1.0 5.4 ⫾ 1.4 8.9 ⫾ 1.7
81 ⫾ 12 70 ⫾ 13 53 ⫾ 14
The postradiotherapy dynamics of height and internal reflectivity were compared among the 3 groups using the t test and analysis of variance statistics.
RESULTS METHODS We retrospectively included in the study all patients treated with PBT at the Centre Lacassagne in Nice (France) between March 2000 and December 2007. For each patient, a total dose of 60 Gy was delivered in 4 sessions over 4 consecutive days, as previously described in previous reports.13,14 The patients who suffered local recurrence were excluded. We performed clinical and echographic examinations using B-scans and standardized A-scans (CineScan S; Quantel Medical, Z.I. Le Brézet, France) before and after radiotherapy. Follow-up examinations were scheduled every 6 months, but to increase the precision in the temporal dynamics analysis, the measurements were gathered in 3-month groups between 6 and 42 months and in 6-month groups after 45 months, and the mean and SD of each group were calculated. The echographic examination included measurements of base diameter (B-scan), tumour thickness (B- and A-scan), and internal reflectivity (A-scan performed after calibration to “tissue sensitivity” as defined by standardized echography, in which the retina’s reflectivity is 100% and the amplification curve is S-shaped). We paid close attention to an accurate perpendicular orientation of the A-scan sound beam with respect to the point of maximal tumour elevation and the inner sclera. The measurements were obtained by placing calipers on the peak of the surface spike of the tumour and the inner scleral spike. For the evaluation of tissue reflectivity, we used the Quantitative Echography-I software, which calculates a percentage of the average height of the inner tissue spikes. The same experienced examiner performed all scans. The postradiotherapy thickness and internal reflectivity data were fitted with polynomial, hyperbolic, logarithmic, and first- and second-order exponential equations using the least-square fitting method (MATLAB, Version R2008a; The MathWorks, Natick, Mass.). The best fitting equation was chosen using the goodness-of-fit analysis where the error was weighted on the number of samples for each group. Melanomas were classified into 4 classes according to the seventh TNM classification15 T1, T2, T3, or T4 tumours. Because of the small number of T4 tumours (8 cases), we grouped T3 and T4 tumours in the same group.
One hundred and ninety-eight patients with a mean age of 63 ⫾ 13 years (range 23–90 years) were included in the study, with 99 male (50%) and 99 female (50%) patients. We excluded 17 patients who suffered local recurrence (7.9%) after a mean time of 27 ⫾ 24.6 months; 5 of them had T2 tumours, and 12 had T3þT4 tumours. The recurrence pattern was marginal in 7 patients (41%) and diffuse in 10 patients (59%). Each patient was examined, on average, 7 times (range 3–15 times) for a total of 1393 ultrasonography examinations. Mean follow-up time was 60.0 ⫾ 26 months, and median follow-up was 60 months. Before radiotherapy, the mean tumour thickness was 5.9 ⫾ 2.5 mm (range 1.5–12.5 mm), whereas the mean internal reflectivity was 68% ⫾ 17% (range 20%–98%). Characteristics of tumours before radiotherapy are summarized in Table 1. Afterward we fitted the dynamics of the relative thickness [h(t)/h0*100] and internal reflectivity [R(t)] both for the entire set of patients and separately for T1, T2, and T3þT4 tumours. We used various mathematical functions, but each time the function that fitted best was the sum of a constant with first-order exponential decay for the relative thickness [h/h0*100 ¼ A þ B*exp(–Ct)], and with a first-order exponential rise for reflectivity [R ¼ A – B*exp(–Ct)]. In particular, parameter A tells us the stability value, that is, the average value at which tumours stabilize; B is related to a fit for the initial value (for relative thickness, A þ B 100, that is, the initial relative tumour thickness, which simply is 100%; for reflectivity, A – B mean[R(0)], that is, the average initial reflectivity). From C, we can estimate the half-life of the decay for tumour thickness and the halflife of the rise for internal reflectivity (Table 2). The plots in Figure 1 present the comparison of the distribution of tumour thickness at different time points. Until 24 months postradiotherapy, tumour thickness Table 2—Half-lives of thickness and reflectivity for tumours
All tumours T1 tumours T2 tumours T3þT4 tumours
Tumour Thickness Half-life (mo)
Tumour Internal Reflectivity Half-life (mo)
14.96 9.94 11.33 14.67
11.12 9.36 14.68 7.86
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
61
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al.
Fig. 1 — Comparison of the distribution of tumour thickness at different time points. PBT, proton beam therapy.
appears in evolution, because tumours at 6 months after radiotherapy are significantly smaller than tumours at radiotherapy (p o 0.001), and after 12, 18, and 24 months postradiotherapy are still, respectively, smaller than after 6, 12, and 18 months after the treatment (p o 0.001, p o 0.001, and p ¼ 0.016). Conversely, between 24 and 30 months, the tumour thickness stabilizes and the paired t test does not detect any significant change (p ¼ 0.31). Figure 2 shows the decrease in tumour thickness postradiotherapy for the tumours. Before PBT, tumour
thickness ranged from 1.5 to 12.5 mm with a mean of 5.9 mm. Its decrease after radiotherapy was best fitted by the sum of a first-order exponential decay and a constant with a decay half-life of 14.96 months. Based on the fit, tumour thickness stabilized on a constant value representing, on average, 47% of the initial value. T3þT4 tumours stabilized on a lower percentage of their initial thickness, but the time required to stabilize is higher with respect to T1 and T2 tumours. We also performed the comparison of the distribution of tumour thickness at different time points for each
Fig. 2 — Decrease in tumour thickness postradiotherapy. PBT, proton beam therapy.
62
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al.
Fig. 3 — Comparison of the distribution of the tumour internal reflectivity at different time points. PBT, proton beam therapy.
patient subgroup. Although for T1 tumours the first significant p value is found when comparing thickness at 6 and 12 months postradiotherapy (p ¼ 0.07), for T2 tumours it is found when comparing thickness at 18 and 24 months postradiotherapy (p ¼ 0.06), and for T3þT4 tumours it is found when comparing thickness at 12 and 18 months postradiotherapy (p ¼ 0.2). Plots in Figure 3 present the comparison of the distribution of the tumour internal reflectivity at different time points. Until 18 months postradiotherapy, the tumours reflectivity appears in evolution, because tumours at 6 months after radiotherapy have a higher reflectivity
than tumours at radiotherapy (p o 0.001), and tumours at 12 and 18 months postradiotherapy have, respectively, a still higher reflectivity than tumours after 6 and 12 months postradiotherapy (p o 0.001 and p ¼ 0.03). Conversely, between 18 and 24 months, the tumour reflectivity stabilizes and the paired t test does not detect any significant change (p ¼ 0.27). Figure 4 shows the increase in tumour internal reflectivity postradiotherapy. Mean internal reflectivity before PBT was 68%. The dynamics of the reflectivity were best fitted by an exponential and a constant, with rise half-life of 11.12 months, and stability value of 87%.
Fig. 4 — Increase in tumour internal reflectivity postradiotherapy. PBT, proton beam therapy.
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
63
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. Larger tumours (T3þT4 group), which initially had a significantly lower internal reflectivity than T2 (t test; p o 0.001) and T1 tumours (t test; p o 0.001), had a higher initial rise in internal reflectivity (2.43% per month) compared with T2 (0.83%) and T1 tumours (0.77% per month). Nevertheless, T1 tumours stabilized on a higher final internal reflectivity (about 91%) than T2 (about 87%) and T3þT4 tumours (about 81%). We also performed the comparison of the distribution of tumour reflectivity at different time points, for T1, T2, and T3þT4 tumours. For T2 and T3þT4 patient groups, the first significant p value is found when comparing reflectivity at 18 and 12 months postradiotherapy (respectively, p ¼ 0.07 and p ¼ 0.5), whereas for the T1 patient group, the first significant p value is found when comparing reflectivity at 12 and 6 months postradiotherapy (p ¼ 0.4).
DISCUSSION The goal of all forms of radiation therapy is to destroy the reproductive integrity of a tumour. Ionizing radiation damages the DNA of the malignant cells, which may be misrepaired and also may disrupt the integrity of the chromosome. The effects of this damage become manifest during mitosis, at which point the cells cannot successfully replicate. The timing would depend on the proliferation kinetics of the cellular constituents and it will be longer in tumours with slow turnover, including ocular melanoma. Other consequences of irradiation include changes in growth factors and signal transduction pathways, apoptosis, and the regulation of the cell cycle.16,17 Histopathologic studies demonstrated that PBT causes degenerative changes in the tumour (necrosis, fibrosis, balloon cells), decreases the ability of tumour cells to reproduce (fewer mitotic figures), and damages blood supply.18 However, it was shown that postirradiation, most tumours shrink but do not disappear.19–21 In 2002, Kaiserman et al.22 described echographic changes of tumour thickness and internal reflectivity in uveal melanoma treated by brachytherapy. In our study, we analyze the same dynamics in uveal melanomas treated by PBT. After brachytherapy, Kaiserman et al.22 reported a decrease in height in 96.6% of tumours treated. In our study, we observed local recurrence rate of 7.9% (17 patients) with a marginal pattern in 7 patients and a diffuse pattern in 10 patients. To minimize the bias caused by local treatment failures, we excluded these patients from our data set. Moreover, because of the small number of the diffuse pattern relapses, a reliable statistical comparison with the responsive tumours was impossible. In our data set, as well as in the Kaiserman et al. study,22 large tumours have a lower initial internal reflectivity compared with small tumours. However, after
64
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
brachytherapy, large tumours are associated with a slower increase in internal reflectivity, because the rise half-life of large tumours is much larger than the rise half-life of small tumours (8.7 vs 5 months). All tumours then stabilize on a similar final value for internal reflectivity (about 70%). Conversely, after PBT, the rise half-life of large and small tumours is more similar (about 9.4 and 7.9 months), although the rate of the increase ([h – h0]/[t – t0]) is higher in large tumours (2.4% vs 0.8% per month for small tumours). Nevertheless, the stability value of the internal reflectivity is higher in small tumours. As observed by Kaiserman et al. in the brachytherapy study,22 the functional form of height dynamics and internal reflectivity is composed of 2 terms: a constant and a monoexponential decay. This suggested the existence of 2 populations in each tumour, one radioresistant component with higher internal reflectivity and the other radiosensitive with lower internal reflectivity. Interestingly, from our analysis it appears that in T3-T4 tumours, internal reflectivity presents faster variations with respect to tumour thickness, because it has a shorter halflife (7.86 vs 14.67 months). This behaviour is confirmed by the analysis at different time points. In fact, whereas for tumour thickness the first significant p value occurs when comparing tumours at 24 and 18 months postradiotherapy, for internal reflectivity this already happens when comparing tumours at 18 and 12 months postradiotherapy. This means that internal reflectivity stabilizes earlier than tumour thickness. We correlate these findings with early response to radiotherapy that determine cellular changes and loss of fluid that precede a secondary process with ingrowth of macrophages and fibrosis.23 These data suggest ophthalmologists should rely on internal reflectivity for a more sensitive indicator of early response to therapy for large tumours, whereas in small and medium tumours internal reflectivity and tumour thickness are almost interchangeable, although tumour thickness can be preferred. To our knowledge, this is the first attempt to calculate the function describing the dynamics of tumour height and internal reflectivity in a large series of uveal melanomas treated by PBT. As Kaiserman et al.’s22 work is for brachytherapy, we consider our study as a reference model for the follow-up of melanomas treated by PBT.
Disclosure: The authors have no proprietary or commercial interest in any materials discussed in this article. REFERENCES 1. Egan KM, Seddon JM, Glynn RJ, Gragoudas ES, Albert DM. Epidemiologic aspects of uveal melanoma. Surv Ophthalmol. 1988;32:239-51. 2. Virgili G, Gatta G, Ciccolallo L, et al. EUROCARE Working Group. Incidence of uveal melanoma in Europe. Ophthalmology. 2007;114:2309-15.
Tumour thickness and internal reflectivity after proton beam therapy—Mosci et al. 3. Collaborative Ocular Melanoma Study Group. Accuracy of diagnosis of choroidal melanomas in the Collaborative Ocular Melanoma Study. COMS report no. 1. Arch Ophthalmol. 1990;108:1268-73. 4. Boldt HC, Byrne SF, et al. Baseline echographic characteristics of tumors in eyes of patients enrolled in the Collaborative Ocular Melanoma Study: COMS report no. 29. Ophthalmology. 2008; 115:1390-7, 1397.e1-2. 5. Coleman DJ, Abramson DH, Jack RL, Franzen LA. Ultrasonic diagnosis of tumors of the choroid. Arch Ophthalmol. 1974;91: 344-54. 6. Fuller DG, Snyder WB, Hutton WL, Vaiser A. Ultrasonographic features of choroidal malignant melanomas. Arch Ophthalmol. 1979; 97:1465-72. 7. Ossoinig KC. Standardized echography: basic principles, clinical applications, and results. Int Ophthalmol Clin. 1979;19:127-210. 8. Nicholson DH, Frazier-Byrne S, Chiu MT, Schiffman J, Hughes JR, Novinski EK. Echographic and histologic tumor height measurements in uveal melanoma. Am J Ophthalmol. 1985;100:454-7. 9. Haritoglou C, Neubauer AS, Herzum H, Freeman WR, Mueller AJ. Interobserver and intraobserver variability of measurements of uveal melanomas using standardised echography. Br J Ophthalmol. 2002; 86:1390-4. 10. Kook D, Kreutzer TC, Wolf A, Haritoglou C. Variability of standardized echographic ultrasound using 10 mHz and highresolution 20 mHz B scan in measuring intraocular melanoma. Clin Ophthalmol. 2011;5:477-82. 11. Ossoinig KC, Bigar F, Kaefring SL. Malignant melanoma of the choroid and ciliary body: a differential diagnosis in clinical echography. Bibl Ophthalmol. 1975;83:141-54. 12. Hodes BL, Choromokos E. Standardized A-scan echographic diagnosis of choroidal malignant melanomas. Arch Ophthalmol. 1977;95:593-7. 13. Mosci C, Mosci S, Barla A, Squarcia S, Chauvel P, Iborra N. Proton beam radiotherapy of uveal melanoma: Italian patients treated in Nice, France. Eur J Ophthalmol. 2009;19:654-60.
14. Mosci C, Lanza FB, Barla A, et al. Comparison of clinical outcomes for patients with large choroidal melanoma after primary treatment with enucleation or proton beam radiotherapy. Ophthalmologica. 2012;227:190-6. 15. Edge SD, Byrd DR, Carducci MA, Compton CC, Fritz AG, Greene FL. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2009. 16. Gragoudas ES, Lane AM, Munzenrider J, Egan KM, Li W. Evidence-based estimates of outcome in patients irradiated for intraocular melanoma. Arch Ophthalmol. 2002;120:1665-71. 17. Gragoudas ES. Proton beam irradiation of uveal melanomas: the first 30 years. The Weisenfeld Lecture. Invest Ophthalmol Vis Sci. 2006; 47:4666-73. 18. Saornil MA, Egan KM, Gragoudas ES, Seddon JM, Walsh SM, Albert DM. Histopathology of proton beam-irradiated vs enucleated uveal melanomas. Arch Ophthalmol. 1992;110:1112-8. 19. Boudinet M, Berges O, Le Huerou JY, Lumbroso-Le Rouic L, Desjardins L, Laugier P. Quantitative echography in the follow-up of patients treated with proton-beam irradiation for primary choroidal melanomas. Ultrasound Med Biol. 2007;33:1046-56. 20. Ravazzoni L, Mosci C, Polizzi A, Schenome M, Soldati MR, Buono C. Ultrasonographic follow-up of patients with choroidal melanoma following conservative treatment. Ophthalmologica. 1998;212(Suppl 1):77-8. 21. Romani A, Baldeschi L, Genovesi-Ebert F, et al. Ultrasonographic follow-up of primary choroidal malignant melanoma after proton beam irradiation therapy. Ophthalmologica. 1998;212(Suppl 1): 50-2. 22. Kaiserman I, Anteby I, Chowers I, Blumenthal EZ, Kliers I, Pe’er J. Changes in ultrasound findings in posterior uveal melanoma after Ruthenium 106 brachytherapy. Ophthalmology. 2002;109: 1137-41. 23. Seddon JM, Gragoudas ES, Albert DM. Ciliary body and choroidal melanomas treated by proton beam irradiation. Histopathologic study of eyes. Arch Ophthalmol. 1983;101:1402-8.
CAN J OPHTHALMOL — VOL. 49, NO. 1, FEBRUARY 2014
65
