Back to the index
Choroidal Melanoma

Enucleation or Observation? A New Approach.

Willem A. Manschot, MD, Hennie A. van Peperzeel, MD

Application of the knowledge of growth rate of malignant neoplasms on choroidal melanomas shows that (1) fatalities within two years after enucleation only exceptionally can be due to dissemination of tumor cells during enucleation because the growth rate of uveal melanomas generally is far too low; (2) the apparent relationship between enucleation and death by metastasis actually is a relationship in time between diagnosis of the primary tumor and death by metastasis; (3) simultaneous diagnosis of the primary tumor and its metastases rarely is possible; (4) long intervals between enucleation and death by metastasis are explained by a low growth rate of metastases. Early enucleation, especially of small melanomas, is obligatory, because it prevents (further) dissemination of tumor cells. Preoperative local radiotherapy of two doses of 400 rads each will devitalize about 90% of the primary cells and largely will prevent possible iatrogenic dissemination.
     (Arch Ophthalmol 98:71-77,1980)

   As long ago as 1882, Fuchs1 wrote that all eyes containing a uveal sarcoma were treated by enucleation. Already in 1999, Pawel2 emphasized that every uveal sarcoma not treated by enucleation definitely would prove fatal. This conviction was repeated by Stallard3 in 1964. For a century, enucleation has remained the accepted practice, the only exceptions occurring in cases where the patient has refused enucleation or cases where a melanoma has been diagnosed in the only functioning eye. Two follow-up series of such cases have been published4,5; both show 100% death by metastatic disease.
   Ten-year follow-up data on large series of patients whose enucleated eyes harbored a melanoma6-13 have disclosed a corrected mortality (death by metastasis) varying between 40% and 60%. Raivio4 found a 20-year corrected surrvival rate of 43% in 214 patients. A 50% ten-year survival rate represents a favorable figure among the various human cancers.
   Occasionally, the effectiveness of enucleation was doubted. Von Hippel14,15 concluded that there was not enough evidence that early enucleation would save noticeably more lives than later enucleation. He stressed that entended follow-up periods were necessary to evaluate the results of the operation. Later, Chisholm16 reported five cases of metastasis 19 to 36 years after enucleation.
   Westerveld-Brandon and Zeeman7 interpreted data on 90 patients on the basis of being or not being in accordance with the statistical expectation of life. They concluded that the conventional indication for immediate enucleation could not be maintained for patients aged 60 or more and that its justifiability for younger patients had not been demonstrated with any certainty.
   The development of a melanoma in an only eye and the refusal of patients to have the eye enucleated have led to attempts to develop alternatives to enucleation, whereby the tumor would be eradicated while the eye remained preserved and some useful vision retained. These alternatives are excision of a melanoma in the anterior uvea,17 transscleral diathermy,18 light coagulation,19 radiotherapy,20 local excision from the posterior uvea,21 cryotherapy,22 and, more recently, the policy of nontreatment with careful observation in patients with small and asymptomatic tumors.23
   This nontreatment therapy recently has acquired authoritative advocates in Zimmerman et al24 who summarized that, from certain impressions, they believe that the following conclusions are warranted:

   1. Enucleation may have for many patients an adverse rather than a beneficial effect with respect to the development of metastatic disease.
   2. A long-term follow-up study of untreated patients is indicated.
   3. New techniques for enucleation designed to prevent the dissemination of tumor cells must be developed and tested to enable the ophthalmic surgeon to remove safely the tumor-containing eye that has developed such complications as uncontrollable glaucoma, panophthalmitis or proptosis from extraocular extension.

   With regard to the last part of the final conclusion, the authors indirectly invite ophthalmic surgeons to enucleate only those eyes containing a melanoma in which the disease has reached the ultimate stages of uncontrollable glaucoma, panophthalmitis, or proptosis from extraocular extension. In particular, the last complication will hardly be met anymore in medically advanced countries, unless the demand for long-term follow-up studies of untreated patients gains ground.


   Treatment of (small) melanomas by early enucleation has for a century been based on the axiom that a malignant tumor that has the potential for production of metastatic disease should be treated by complete surgical removal as early as possible. This complete surgical removal is easy to perform in most uveal melanomas. This basic concept in the treatment of melanomas has been substantiated by two well-documented facts:
   1. Prognosis in small melanomas is much better than in large melanomas. This has been proved by long-term follow-up studies25-28 in which the prognosis was related to tumor size. Davidorf and Lang27 found a ten-year follow-up period, a corrected death rate of 0% in 18 p with histopathologically proved malignant melanomas of less than 7-mm diameter and 2-mm elevation, of 23% in 27 patients with melanomas of less than 10-mm diameter and 3-mm elevation, and of 46% in 82 patients with larger melanomas. Warren,26 after a five-year follow-up period, also found 0% death by metastasis in small melanomas as against 57% in large melanomas. Zimmerman and McLean11 recently stated: "There is a striking variation in prognosis when small and large tumours are compared, large tumours carrying a worse prognosis than small tumours."
   One of the reasons for a noticeably worse prognosis in patients with a large melanoma is that the latter contain epithelioid cells in a much higher percentage than is found in small melanomas.11,26
   The lethality in the epithelioid cell group after a 15-year follow-up is 2.5 times that in the spindle B cell group.28 Small melanomas, however, may also contain epithelioid cells. Zimmerman and McLean11 remarked that, in their institute, one third of the melanomas less than 10 mm in diameter had a significant component of epithelioid cells; the mortality in these cases was almost 50%, as against 7% in the smallest spindle cell tumors.
   The problem for clinicians is that it is not possible for them to differentiate clinically between small epithelioid cell and small spindle cell melanomas.
   2. Prognosis in melanomas without extrachoroidal invasion is much better than in melanomas with extrachoroidal invasion and extrascleral extension. Jensen28 noted after a 15-year follow-up period that the corrected survival rate of patients without extrascleral extension was 49%, as against 27% of patients with extrascleral outgrowth. Raivio4 found a corrected ten-year survival rate of 71% in patients without extrachoroidal invasion, of 52% in patients where Bruch's membrane had been ruptured by the melanoma, of 37% where the sclera had been invaded, and of 12% where an extrascleral outgrowth had occurred. Davidorf and Lang27 reported that ten (20%) of their 50 small melanomas of less than 10-mm diameter and 3-mm elevation had extended outside the choroid; four of these already had an extrascleral outgrowth. Small posteriorly located extrascleral extensions rarely can be diagnosed clinically; this is another compelling reason to enucleate every eye containing a small choroidal melanoma.
   From the data available it appears unreasonable, therefore, to allow a small melanoma to become a large(r) one, or to allow an intrachoroidal melanoma to become an extrachoroidal and extrascleral melanoma.


   These arguments are based only on impressions. Zimmerman et al24 recently summarized these impressions as follows:
   1. "Very small tumors have rarely been associated with a fatal outcome." This impression is in contradiction with the already cited statement by the same authors that one third of their smallest melanomas had a significant component of epithelioid cells and that the mortality in these cases was almost 50%. On the other hand, it is certain that a great number of these patients with a small uveal melanoma have been saved because during this century almost all eyes containing a melanoma have been enucleated.
   2. "Our knowledge of the natural history of uveal melanoma is very meager." However, the natural history of uveal melanoma already was known in 19002 and more recently was sharply defined by Stallard3 as "the expectancy of inevitable defeat by laissez-faire." The only two published follow-up series4,5 of untreated patients both show 100% death by metastasis.
   3. "The prognosis of small epithelioid cell melanomas treated by enucleation is almost as bad as that of large melanomas and the management of such tumors may have little effect on the outcome." This impression refutes impression 1. It also denies the well-documented experiences that nontreatment of choroidal melanomas means 100% death by metastasis,4,5 while enucleation means that the lives of more than 50% of these patients will be saved,11 and early enucleation may save even 100%.27
   4. "Enucleation might be more harmful than beneficial." This impression has been worked out in detail and has been summarized by Zimmerman et al as follows:

   A reappraisal of survival data on patients with uveal melanomas has led us to these impressions: (a) that the mortality rate before enucleation is low, estimated at 1% per year; (6) that the mortality rate rises abruptly following enucleation; and (c) that approximately two thirds of the fatalities could be attributed to the dissemination of tumor emboli at the time of enucleation.

   It appears that there is an easier and more logical explanation of these impressions than the supposition by Zimmerman et al24 that most of the fatalities in cases of uveal melanoma should be attributed to iatrogenic dissemination of tumor emboli at the time of enucleation.
   The biological behavior and the prognosis of uveal melanomas are determined by the same factors that determine the behavior and the prognosis of all other human malignant neoplasms.
   Consideration of the established knowledge of the growth rate of malignant neoplasms, its relation to the histological types of tumors, and its relation to the survival of the patients shows that impressions a and b can be interpreted as a feature of the natural history of uveal melanomas and may have no relation whatsoever to the surgical procedure of enucleation. It will also be shown, on theoretical grounds, that impression c basically is incorrect.


figure 1

   Collins et al29 have evolved a method of measurement of growth rate of human tumors for clinical application by measuring the increase in diameter of the shadows of pulmonary metastases in roentgenograms. The growth rate was expressed in terms of "doubling time" (Td), ie, the time necessary for a metastasis to double its volume. Doubling of the diameter of a tumor nodule needs three volume-Tds. The diameter of the metastases can be plotted logarithmically with a linear time scale and the result is a straight line, the slope of which indicates the growth rate (Fig 1). Collins et al have taken some poetic license in using a kilogram instead of a nodule diameter of 10cm. They probably did so to make us visualize the rapid increase of the size. Also a 1-kg nodule is correct only in case the specific gravity of the tissue is 1.0 and the nodule is a cube instead of a sphere, because the content of a sphere is about half that of a cube.
   Doubling times of 11 to 164 days were observed in various types of tumors. Later, other investigators30,31 found a still greater variability of four to 745 days. Different pulmonary metastases in one given patient showed exactly the same growth rate. The survival time - the time between diagnosis of the tumor and death of the patient - increases as the Td of the tumor increases.32
   The growth rate was found to approach exponential growth in 79 of 86 cases.30 It has been recognized for a long time that the growth of most untreated tumors is well described by the Gompertzian or exponential growth curves.33 The relatively few deviations of the growth curve from the exponential type can be explained by the occurrence of quiescent cells, which do not contribute to tumor growth, and by cell death. The growth of a primary tumor runs an identical exponential course,34,35 but it may differ from the growth rate of its metastases when the primary tumor has a more complicated composition than that of its metastases, or when the metastases are located in an environment that promotes the tumor growth rate.
   Davies36 stated that with a fair degree of certainty we can predict that, in the metastases that do occur, the morphologic and functional features of the primary growth will be closely duplicated in the great majority of cases, over 80%.
   Collins and co-workers29 calculated that a 10-mm nodule, a likely size for early diagnosis of metastasis in the lung, would be achieved in 30 doublings. With slowly growing metastases, extrapolation may indicate very long intervals between the time of dissemination and the time of diagnosis of, or death by, metastases. Since the primary tumor in such cases has a similarly low growth rate, the primary tumor must have existed a long time before the metastases originated. After dissemination, 35 to 40 Tds are needed to effect a tumor load that is incompatible with life.29,37

figure 2

   The clinician can only observe the ultimate period of any given tumor process and, especially with slowly growing tumors, he often is unaware that the beginning of the process must be sought far in the past. After 24 doublings, a tumor has reached a diameter of no more than 2.5 mm; but after six further doublings, the diameter is already 10 mm, and after another six doublings it is 40 mm (Fig 2). Thus, the same growth rate appears very slow in the beginning and very fast in a later stage.
   If we apply these oncological data to uveal melanomas, the apparently unsolved problems regarding their natural history appear to be solved.
   So far no Tds of metastases of uveal melanomas have been calculated. However, the Td of metastases of skin melanomas has been determined eight times; the values were < 10, 31, 40, 71, 102, 120, 148, and 266 days.30,38,39 It is generally accepted that uveal melanomas are more slowly growing tumors than skin melanomas. It seems warranted, therefore, to assume for uveal melanomas with a significant component of epithelioid cells that the Td varies between 30 to 100 days, and for spindle cell melanomas that the Td varies between 100 to 350 days.
   Earliest clinical diagnosis of a uveal melanoma is possible when it has reached a diameter of 5 mm; this time corresponds to 27 doublings of the volume of a tumor originating from one single cell. Thirty doublings are needed to achieve a 10-mm diameter, while a diameter of 20 mm is reached after 33 Tds. However, small choroidal melanomas are not spherical but disk-shaped. If we assume their diameter to be eight times their thickness, their volume is one eighth that of a sphere with the same diameter. Clinical diagnosis, therefore, is possible three Tds earlier than mentioned above. On the other hand, clinical measurements of a choroidal melanoma often represent only the tip of the iceberg; histopathological intrachoroidal measurements frequently are much larger than could be expected clinically.
   The survival time after enucleation of a uveal melanoma varies greatly; this is due to the fact that uveal melanomas have various cellular components. Jensen10 found in 126 cases of metastasizing uveal melanomas that most metastases occurred between two to four years after enucleation, while a second peak-incidence was found six to seven years after the globe had been removed. Microscopic examination of the primary tumors showed that the early peak had been caused by melanomas of the epithelioid cell type or of the mixed cell type with predominantly epithelioid cells, while the second peak originated from melanomas of the spindle cell type and of the mixed cell type with predominantly spindle cells.
   The apparent close relationship in time between the time of enucleation and the time of diagnosis of metastases, with its main peak after two to four years, is actually a close relationship in time between the time of diagnosis of the primary tumor and the time of diagnosis of its metastases.
   Assuming that time 10 Td of the primary tumor - when the latter is formed by only 1,024 cells - is the time that dissemination starts, the longest period of dissemination in large melanomas of 20-mm diameters (diagnosis and enucleation at time 33 Td) is 33 - 10 = 23 Td; in small melanomas of 5-mm diameters (diagnosis and enucleation at time 27 Td) this period is 27 - 10 = 17 Td.
   Diagnosis of metastases occurs between times 30 Td and 36 Td after dissemination; at these times the diameter of the metastases is 10 mm and 40 mm, respectively. Assuming that diagnosis of metastases in large uveal melanomas occurs at time 30 Td after dissemination, the interval between the time of diagnosis of the primary growth and the time of diagnosis of its metastases varies between 7 (30 - 23) Td in cases with very early dissemination (eg, time 10 Td) and 30 Td in cases with very late dissemination just before enucleation. In small uveal melanomas this interval varies between 13 (30 - 17) Td in cases with very early dissemination and 30 Td in cases with dissemination just before enucleation.

figure 3figure 4

   Figures 3. and 4 are the graphic representations of these calculations.
   Incorporating in these calculations the values of Td = 30 to 100 days in epithelioid cell melanomas, and of 100 to 350 days in spindle cell melanomas, Table 1 shows the maximum possible variations in the interval between the time of enucleation (time of diagnosis of the primary uveal melanoma) and the time of diagnosis of metastases, in epithelioid cell melanomas and in spindle cell melanomas, respectively.
   Some data in this table had to be chosen arbitrarily. In general, however, extreme values on both sides have been used. All other justifiably chosen data will lie between the values used and will present a still narrower relation between the times of enucleation and the times of diagnosis of metastasis.
   For example, diagnosis and enucleation of the primary melanoma may occur at time 30 Td when its diameter is 10 mm. When dissemination is assumed to have started at time 18 Td and when diagnosis of metastases took place at time 30 Td after dissemination, the period of dissemination will be 12 Td and the interval between enucleation and diagnosis of metastases will be 30 - 12 = 18 Td.
   Incorporating into this calculation the values of Td = 30 to 100 days in epithelioid cell melanomas and 100 to 300 days in spindle cell melanomas. Table 2 presents the maximum possible variation in the interval between the time of enucleation and of diagnosis of metastases in an epithelioid cell melanoma and a spindle cell melanoma, respectively.
   The two peaks of diagnosis of metastasis after enucleation, as noted by Jensen,10 are represented clearly in these tables. So far, the time of enucleation has almost always been immediately after the time of diagnosis of the uveal melanoma. This is the reason that the time of diagnosis of metastases has always been related to the time of enucleation; but actually, the apparent close relation in time between the time of diagnosis of metastases and the enucleation has been a close relation in time between the time of diagnosis of metastases and the diagnosis of the primary uveal melanoma. This relation is independent of surgical intervention.

figure 5

   The fact that metastases of uveal melanomas are rarely seen prior to or simultaneously with the diagnosis of the primary tumor is also easy to explain in this way. Simultaneous diagnosis of a primary uveal melanoma and its metastases when the uveal melanoma has a diameter of 10 mm (Td39) is only possible in the hypothetical case that dissemination has taken place at the first doubling that has constituted the origin of the uvual melanoma, because diagnosis of metastases is generally not possible before time 30 Td after dissemination. Simultaneous diagnoses of a uveal melanoma of 20 mm - the diameter reached at time 33 Td - and its metastases is possible only when dissemination has occurred before time 4 Td of the primary melanoma, which at this time consists of no more than 16 cells (Fig 5).
   Thus, only dissemination of tumor cells in the earliest stage of a uveal melanoma formed by two to 16 cells allows simultaneous diagnosis of a uveal melanoma and its metastases. Whether Td = 30 or 250 days does not alter this rule.
   The foregoing mathematical constructions are valid only in the usual situation where the growth rate of the primary melanoma is equal to that of its metastases. In case of different growth rates, the growth rate of the metastases is always faster than that of the primary tumor. Only if the primary melanoma is of the mixed cell type and its metastases are formed by the epithelioid cell type do the two growth lines not run a parallel course and, thus, may cross each other. Diagnosis of metastases might then be possible simultaneously with or prior to the diagnosis of the primary uveal melanoma.
   Occasionally, extremely long intervals are found between the time of enucleation and the time of death by metastases - the longest reported interval being 36 years. This can be explained easily by a very low growth rate of the metastases. It is not necessary to assume that there are host defense mechanisms or that tumor emboli have remained dormant for decades. The Td of exceptionally slow-growing spindle cell melanomas might be longer than 350 days. But even if Td = 350 days, death by metastases (35 to 40 Td after dissemination of tumor cells) should not occur before 33 to 38 years after dissemination.
   Generally, the interval between diagnosis of metastases and death is short. Jensen10 found an interval of six months or less in 85% of 126 cases; the longest interval in his series was four years. Other reports40,41 mention a mean interval of seven months and of one year, respectively.
   Let us return to the last of the four impressions on which the arguments in favor of observation of small melanomas were based, namely "Enucleation might be more harmful then beneficial." In its turn, this impression had been based on three other impressions,24 as follows:
   1. "The mortality rate before enucleation is low, estimated at 1% per year." Even this low estimation seems too high. It has been shown above that simultaneous diagnosis of the primary uveal melanoma and its metastases is possible only if dissemination occurs when the uveal melanoma consists of no more than two to 16 cells or if a uveal melanoma of the mixed cell type should produce metastases of the epithelioid cell type.
   2. "The mortality rate rises abruptly following enucleation, reaching a peak of 8% during the second year after enucleation." It appears from Table I that this peak is not dependent on surgical intervention, but on, first, the limited number of TDs between dissemination and the time of diagnosis (enucleation) of an epithelioid cell melanoma and, second, the limited number of TDs between dissemination of tumor cells and diagnosis of, or death by, metastases in epithelioid cell melanomas. According to Zimmerman et al,24 the peak in their mathematically constructed model of death by metastases, which starts abruptly about three months after enucleation and continues to rise sharply until 18 months (550 days) after enucleation, is caused by the shower of emboli at the time of enucleation. If this is true, in a large group of uveal melanomas the Td has to be 2.5 to 16 days, because death by metastasis occurs 35 to 40 Tds after dissemination. This means that the growth rate of a large group of uveal melanomas has to be much faster than that of skin melanomas and has to equal that of the most rapidly growing human malignant neoplasms, namely, embryonal carcinomas of the testes. This is highly unlikely.
   3. "Approximately two thirds of the fatalities can be attributed to the dissemination of tumor emboli at the time of enucleation." The same argument used in the preceding paragraph is valid against this hypothetical impression. Death by metastasis in a large group of uveal melanomas cannot occur within the first 18 months after dissemination of tumor cells. It appears from Tables 1 and 2 that the fatalities in subsequent years also are a consequence of the natural history of these melanomas. Only incidentally may there be a relation with surgery, but such relation can never be proved or denied.


   The natural history of untreated melanomas is clear: death by metastasis occurs 35 to 40 Tds after tumor emboli have settled. No clinician will be able to ascertain whether settling of tumor emboli already has occurred at the time of diagnosis of a uveal melanoma. His medical duty is to give the patient the greatest chance of survival. This will be realized by early enucleation of all choroidal melanomas, but especially of small melanomas, in order to prevent the initial or later dissemination of tumor cells. The chance that metastatic spread has not yet occurred is greater in small melanomas.
   Reduction of the period of dissemination is possible only for a very limited number of doubling times. Shortening of the patient's delay in attendance is not possible, but ophthalmic centers often will be able to reduce the physician's delay in treatment by early diagnosis and immediate enucleation. All other treatments, such as photocoagulation, radiation, and excision of the choroidal melanoma, and certainly also the attitude of observation, will be harmful to the patient because dissemination of tumor cells is promoted, either actively or passively.
   Many of the patients with a small melanoma who have not received any treatment during recent years will survive for a long time because their melanoma is of the slowly growing spindle cell type with a Td = 100 to 350 days. Their eyes will be saved for many years, but this will be at the cost of the lives of a number of other patients with small epithelioid cell melanomas, whose lives could have been saved by early enucleation Also, a number of patients with a small spindle cell melanoma will die unnecessarily after many years because of the delay in treatment. Shammas and Blodi13 rightly concluded: "This delay in treatment might shift the prognosis from a relatively good, almost identical to that in the normal population, to a poor one."
   The second recommendation is prevention of possible incidental dissemination of tumor cells at the time of enucleation by a surgical procedure that gets as near to a no-touch technique as possible. Also, preoperative local radiotherapy consisting of two consecutive daily doses of 400 rads (4 Gray), doses that are sufficient to kill 90% of the tumor cells,42,43 can be helpful in preventing artificially induced distant metastases. This radiation does not retard enucleation for more than two days, it does not cause postoperative complications or difficulties with the prosthesis, and it does not jeopardize histopathological examination, although it might influence some histopathological parameters. Uniform histopathological material would be obtained if all patients were to receive this treatment.
   Postoperative radiation is indicated only in cases of extrascleral extension of the melanoma in which removal of the tumor is thought to be incomplete. In such cases, a total dose of 5,000 to 6,000 rads (50 to 60 Gray) should be given within about five weeks postoperatively.
   Prevention of incidental iatrogenic dissemination and destruction of micrometastases present before enucleation can, theoretically, be performed by postoperative adjuvant therapeutic measures such as chemotherapy and immunotherapy. However, effective combinations of cytotoxic agents in melanomas are not known. Such adjuvant therapies often are a heavy burden on older patients and it is questionable whether the toxicity of such therapy does not outweigh possible beneficial effects. Immunotherapy against melanoma is also still in an experimental stage.
   Doubling time of pulmonary metastases should be determined in patients who have refused enucleation and in whom metastases have been diagnosed in a later stage. If autopsy is performed, measures should be taken for postmortem enucleation of the eye. The primary tumor shou ld be compared with the metastases and the characteristics and measurements of their cells, nuclei, and nucleoli should be determined.44 In this connection tissue culture of tumor cells or culture in nude mice may be of help to determine growth characteristics.
   It seems inappropriate to introduce a discussion on immunological factors that might effect a patient's response to a choroidal melanoma. The prematurity of such a discussion appears from the following quotations from a recent authoritative book on secondary spread of cancer45: (1) "There is considerable controversy as to whether neoantigens function to produce potential tumour rejection," and (2) "In human malignant disease the evidence for tumour associated antigens is still largely circumstantial and so cannot firmly be correlated with tumour rejection."
   The assumption that enucleation should enhance metastatic growths because it should be associated with a fall in antibody levels also finds no substantiation in this book: "The capacity of the tumour-bearing host to reject the second tumour challenge waned as the primary tumour in creased in size (eclipse phenomen) but resistance was restored following resection of the primary tumour."
   Ophthalmology has to accept that in many cases little can be accomplished after a choroidal melanoma has been diagnosed. At the time of diagnosis, these melanomas have existed for 27 or more TDs, which means for many years and sometimes even for decades. It clearly appears from Fig 2 that all tumors remain invisible and undemonstrable during the major period of their growth. Dissemination of tumor cells may have occurred long before the time of diagnosis of the primary melanoma. We can prevent (further) dissemination only by immediate enucleation. And it has been proved that we are doing a lot of harm by doing nothing, namely, reducing the corrected survival from 100% to zero for patients with a very small melanoma and from 50% to zero for all patients.

B. R. Vermeer, physicist, designed Fig 3,4, and 5. Fellow members of the Netherlands National Committee on Ocular Melanomas, Jaap G. van Andel, MD; Prof Roy O. van der Heul, PhD; and Prof Jenno A. Oosterhuis, PhD, critically evaluated the manuscript, which reflects the views and recommendations of the committee.

Back to the index

1. Fuchs E: Das Sarcom des Uvealtractus. Vienna, Wilhelm Braummüller, 1882.
2. Pawel E: Beitrag zur Lehre von den Chorioi-dealsarkomen. Albrecht Von Graefes Arch Ophthalmol 49:71-124, 1900.
3. Stallard HB: Malignant melanoma of the choroid treated with radioactive applicators, in Boniuk M (ed): Ocular and Adnexal Tumors. St Louis, CV Mosby Co, 1964, p 322.
4. Raivio I: Uveal melanoma in Finland: An epidemiological; clinical, histological and prognostic study. Acta Ophthalmol, suppl 133, 1977, pp 45-46.
5. Sobanski J, Zeydler L, Goetz J: Uber die Therapie des intraokularen Melanoma malignum. Klin Monastsbl Augenheilkd 146: 70-76, 1965.
6. Wilder HC, Paul EV: Malignant melanoma of the choroid and ciliary body: A study of 2,535 cases. Milit Surg 109:370-375, 1951.
7. Westerveld-Brandon ER, Zeeman WPC: The prognosis of melanomablastomata of the choroid. Ophthalmologica 134:20-29, 1957.
8. Vannas S: Zur Prognose der malignen Geschwtilste der Aderhaut. Klin Monatsbl Augenheilkd 135:678-690, 1959.
9. Paul EV, Parnell BL, Fraker M: Prognosis of malignant melanomas of the choroid and ciliary body. Int Ophthalmol Clin 2:387-402, 1962.
10. Jensen OA: Malignant melanomas of the uvea in Denmark, 1943-1952: A clinical, histopathological and prognostic study. Acta Ophthalmol, suppl 75, 1963, p 133.
11. Zimmerman LE, McLean IW: Changing concepts of the prognosis and management of small malignant melanomas of the choroid. Trans Ophthalmol Soc UK 95:487-494, 1975.
12. McLean 1W, Foster WD, Zimmerman LE: Prognostic factors in small malignant melanomas of choroid and ciliary body. Arch Ophthalmol 95:48-58, 1977.
13. Shammas HF, Blodi FC: Prognostic factors in choroidal and ciliary body melanomas. Arch Ophthalmol 95:63-69, 1977;
14. Von Hippel E: Ist die Frühenukleation bei Aderhautsatkomen uneingeschränkt zu verlangen? Ber Dtsch Ophthalmol Ges 34:147-155, 1922.
15. Von Hippel E: Zur Prognose der Uvealsarkome. Albrecht Von Graefes Arch Ophtalmol 124:206-220, 1930.
16. Chishoim JF: A long-term follow-up of malignant melanomas of the choroid based on the Terry and Johns series. Am J Ophtalmol 36:61-73, 1953.
17. Schubert F: Operation eines "Leukosarkoms" der Choiroidea mit Erhaltung des Auges .Dauerheilung. Wien Klin Wochenschr 24:677-678, 1925.
18. Weve HJM: Ueber operative Behandlung von intraokularen Tumoren mit Erhaltung des Bulbus. Arch Augenheilkd 110:482-491, 1937.
19. Meyer-Schwickerath G: Neue Indikationen der Lichtkoagulation. Ber Dtsch Ophthalmol Ges 60:197-201,1957.
20. Stallard HB: Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol 50:147-155; 1966.
21. Stallard HB: Partial choroidectomy. Br J Ophthalmol 50:660-662, 1966.
22. Lincoff H, McLean J, Long R: The cryosurgical treatment of intraocular tumors. Am J Ophthalmol 63:389-399,1967.
23. Curtin VT, Cavender JC: Natural course of selected malignant melanomas of the choroid and ciliary body. Mod Probl Ophthalmol 12:523-527, 1974.
24. Zimmerman LE, McLean IW, Foster WD: Does enucleation of the eye containing a malignant melanoma prevent or accelerate the dissemination of tumor cells? Br J Opthalmol 62:420-425, 1978.
25. Flocks M, Gerende JH, Zimmerman LE: The size and shape of malignant melanomas of the choroid and ciliary body in relation to the prognosis and histological characteristics: A statistical study of 210 tumors. Trans Am Acad Ophthalmol Otolaryngol 59:740-758, 1955.
26. Warren RM: Prognosis of malignant melanomas of the choroid and ciliary body, in Blodi FC (ed): Current Concepts in Ophthalmology. St Louis, CV Mosby Co, 1974, vol 4.
27. Davidorf FH, Lang JR: The natural history of malignant melanoma of the choroid: small vs. large tumors. Trans Am Acad Ophthalmol Otolaryngol 79:310-320, 1975.
28. Jensen OA: Malignant melanomas of the human uvea: Recent follow-up of cases in Denmark, 1943-1952. Acta Ophthalmol 48:1113-1128, 1970.
29 Collins VP, Loeffler RK, Tivey H: Observations on growth rates of human tumors. AJR 76:988-1000, 1956.
30. Breur K: Growth rate and radiosensitivity of human tumours: I. Growth rate of human tumours. Eur J Cancer 2:157-169, 1966.
31. Van Peperzeel HA: Patterns of Tumor Growth after Radiation, thesis, Grdningen, the Netherlands, Van Denderen, 1970.
32. Malaise EP, Chavaudra N, Charbit A, et al: Relationship between the growth rate of human metastases, survival and pathological type. Eur J Cancer 10:451-459, 1974.
33. Norton L, Simon R: Growth curve of an experimental solid tumor following radiotherapy. J Natl Cancer Inst 58:1735-1741, 1977.
34. Spratt JS, Spjut HJ, Roper CL: The frequency distribution of the growth rates and the estimated duration of primary pulmonary carcinomas. Acta Union International Contre Cancer 19:6-7, 1963.
35. Garland LH, Coulson W, Wollin E: The rate of growth and apparent duration of untreated primary bronchial carcinoma. Cancer 16:694-707, 1963.
36. Davies JNP: Overview of the biology and pathology of metastasis, in Day SB, Myers WPL, Stansly P, et al (ed): Cancer Invasion and Metastasis: Biologic Mechanisms and Therapy: Progress in Cancer Research and Therapy. New York, Raven Press, 1977, vol 5, p 15.
37. Breur K: Growth Rate and Radiosensitivity of Human Tumours, thesis. The Hague, the Netherlands, Mouton & Co, 1965.
38. Rëmke P, Bernheim J, Van der Vorm DH, et al: Effect of B.C.G. stimulation on the growth rate of pulmonary metastases in three patients with melanoma, in Mathé G and Weiner R (eds): Recent Results in Cancer Research. New York, Springer, 1974, vol 47, pp 470-472.
39. Ishihara T, Kikuchi K, Ikeda T, et al: Metastatic pulmonary diseases: Biologic factors and modes of treatment. Chest 63:277-232, 1973.
40. Einhorn LH, Burgess M, Gottlieb JA: Metastatic pattern of choroidal melanoma. Cancer 34:1001-1004, 1974.
41. Char DH: Metastatic choroidal melanoma. Am J Ophthalmol 86;76-80, 1978.
42. Hoye RC, Smith PR: Effectiveness of small amounts of preoperative irradiations in preventing the growth of tumor cells disseminated at surgery. Cancer 14:284-295, 1961.
43. Powers WE, Tolmach U. Preoperative radiation therapy: Biological basis and experimental investigations. Nature 201:272-273, 1964.
44. Gamel JW, McLean IW, Foster WD, et al: Uveal melanomas: Correlation of cytologic features with prognosis. Cancer 41:1897-1901, 1978.
45. Pimm MV, Baldwin RW: Immunology and immunotherapy of experimental and clinical metastases, in Baldwin RW (ed): Secondary Spread of Cancer. New York, Academic Press Inc, 1978, pp 163-209.

Accepted for publication May 23, 1979. From the Institute of Pathology (Department of Ophthalmic Pathology), Erasmus University, Rotterdam, the Netherlands (Dr Manschot), and the Academic Hospital, Department of Radiotherapy, Utrecht, the Netherlands (Prof Dr van Peperzeel). Reprint requests to Institute of Pathology, Erasmus University, P0 Box 1738, Rotterdam, the Netherlands (Dr Manschot).
Reprinted from the Archives of Ophthalmology, January 1980, Volume 98. (c) 1980, American Medical Association.
Back to the index