local environmental influences on uveal melanoma : vitreous humor promotes uveal melanoma invasion,...

Local Environmental Influences on Uveal MelanomaVitreous Humor Promotes Uveal Melanoma Invasion, Whereas the AqueousCan Be Inhibitory

David Canovas, MScIan G. Rennie, MBChBCarmel E. Nichols, BScKaren Sisley, PhD

Academic Unit of Ophthalmology and Orthoptics,Section of Oncology, School of Medicine and Bio-medical Sciences, University of Sheffield, RoyalHallamshire Hospital, Sheffield, United Kingdom.

BACKGROUND. Uveal melanomas of the choroid and ciliary body are aggressive

tumors causing the death of approximately 50% of patients. In contrast, iris mel-

anomas only infrequently metastasize; why these differences exist is not known.

The local environment can regulate cancer growth and development, and it is

probable the aqueous and vitreous humors have an important role in regulating

uveal melanoma behavior.

METHODS. To explore this possibility cultures of uveal melanoma were exposed

to aqueous and vitreous and the effects investigated using invasion and prolifera-

tion assays. ChemiArrays (Chemicon International, Temecula, Calif) were per-

formed to determine which regulatory factors might influence the process.

RESULTS. The vitreous universally promoted uveal melanoma invasion, whereas

the aqueous mainly had no effect or was inhibitory. Tumor location, and the

baseline invasion of the melanoma, affected the ability of aqueous and vitreous

from different patients to regulate invasive behavior. Proliferation was not signifi-

cantly altered as a result of exposure to the aqueous or vitreous. The ability of

the humors to regulate uveal melanomas may involve TIMP-2, TIMP-3, and TGF-

b2, as high expression was found by ChemiArray analysis and there were differ-

ences in the levels of the regulators in the aqueous compared with the vitreous.

CONCLUSIONS. The findings suggest that in situ uveal melanoma development

reflects an interaction between the tumor and the environment of the eye. Exposure

to the aqueous would therefore contribute to the benign nature of iris melanomas,

whereas potential interaction with the vitreous appears to promote the aggressive

behavior of posterior uveal melanomas. Cancer 2008;112:1787–94. � 2008 American

Cancer Society.

KEYWORDS: uveal melanoma, invasion, metastasis, aqueous, vitreous.

T he interaction of tumor cells with the surrounding microenvir-

onment has been shown to be influential to tumor progression

and is mediated by soluble factors, cell-cell, or cell-matrix contacts.1

Uveal melanomas, the commonest primary eye cancer in adults, are

highly aggressive, characteristically metastasizing to the liver and

causing the death of approximately 50% of patients within 10

years.2,3 There are, however, well-recognized differences in the met-

astatic propensities of uveal melanomas, dependent on tumor loca-

tion. Anterior uveal melanomas (iris), representing 5% of all cases of

uveal melanoma,4 are relatively benign and only metastasize in 5%

of patients within 10 years of follow-up.5 In contrast, posterior uveal

melanomas (choroid and in particular ciliary body) are highly

aggressive and produce metastasis in 61% of the cases, with 90% of

deaths occurring within 15 years, thus accounting for the high rates

Address for reprints: Karen Sisley, PhD, AcademicUnit of Ophthalmology and Orthoptics, Universityof Sheffield, K-Floor, School of Medicine & Bio-medical Sciences, Beech Hill Road, Sheffield,S10 2RX, UK; Fax: (011) 0044 114 2766381;E-mail: [email protected]

Supported by Yorkshire Eye Research (grant num-bers 008 and 013) and Yorkshire CancerResearch pump priming.

Received May 31, 2007; revision received August31, 2007; accepted October 30, 2007.

ª 2008 American Cancer SocietyDOI 10.1002/cncr.23358Published online 25 February 2008 in Wiley InterScience (www.interscience.wiley.com).

1787

of metastatic death from this disease.3 Why there is

this behavioral variation is not known, but it is possi-

ble that environmental influences may play a part.

The aqueous and vitreous humors of the eye are

known to contain numerous growth factors and cyto-

kines that have been linked to other ocular diseases.6

Specifically, concentrations of hepatocyte growth fac-

tor/scatter factor (HGF/SF) increase in the vitreous

of patients with diabetic retinopathy,7 and TGF-b in

the aqueous humor has been associated with both

uveitis and glaucoma.8,9 Both HGF and TGF-b have

been implicated in uveal melanoma, with HGF hav-

ing been shown to increase uveal melanoma invasion

and potentially assisting in targeting metastatic

spread to the liver.10,11 In contrast, TGF-b inhibits

growth of uveal melanocytes, and also the prolifera-

tion and invasion of uveal melanomas.11–13 Other

regulators are also likely to exert an effect, including

vascular endothelial growth factor (VEGF), which is

raised in the humors of patients with uveal mela-

noma.14 It is therefore possible that the relatively be-

nign iris melanomas arising in the anterior chamber

could be directly exposed to a number of regulators

within the aqueous that are inhibitory to their

growth. In contrast, posterior uveal melanomas may

be positively influenced by components of the vitre-

ous, which is in close proximity, either through a lack

of inhibition or active promotion of their aggressive

behavior. To investigate this possibility we studied the

effect of exposure to aqueous and vitreous on uveal

melanoma invasion and proliferation. Using ChemiAr-

ray (Chemicon International, Temecula, Calif) we also

looked for differences in the levels of a panel of 70

growth factors and cytokines in the aqueous and vit-

reous from patients with uveal melanoma.

MATERIALS AND METHODSUveal Melanoma Cell Lines and Short-Term CulturesEstablished Sheffield ocular melanoma (SOM) cell

lines SOM 177 (invasive) and SOM 157d (low-inva-

sive) were used as positive and negative controls,

respectively.11,15 Short-term cultures were established

from a total of 12 primary uveal melanomas as

detailed previously.16 All samples were removed after

informed patient consent and protocols followed the

principles of the Declaration of Helsinki. The clini-

cal-pathologic details of the patients are presented in

Table 1. Cells were maintained in RPMI-1640, sup-

plemented with penicillin (100 U/mL), streptomycin

(100 lg/mL), glucose (0.2%), epidermal growth factor

(0.1%), and fetal calf serum (20%), at 378C in an in-

cubator with 5% CO2. All cultures were used experi-

mentally before passage 5. The aqueous and the

vitreous were collected from patients with primary

uveal melanoma at the time of enucleation, centri-

fuged to eliminate tissue particles, and stored at

2208C until required. Samples contaminated with

blood were not used experimentally. The clinicopath-

ologic details of the patients whose aqueous and vit-

reous was used singly in the ChemiArray and

chemoinvasion assay are presented in Table 1.

Patients whose samples were used in proliferation

assays or as part of the pooled aqueous and vitreous

standards have not been detailed.

Chemotaxis and Chemoinvasion AssayStandardization of the protocolThe chemoinvasion assay was performed in tripli-

cate, as detailed previously.11 The maximal effects of

the aqueous and vitreous were determined using the

2 established cell lines (SOM 157d and SOM 177)

and 2 short-term cultures (SOM 433 and 436) (data

not shown). Samples of aqueous were rarely greater

than 150 lL; therefore, both aqueous and vitreous

were diluted in the assay to a final concentration of

10%. The effect of freshly isolated aqueous and vitre-

ous was assessed when included in the upper cham-

ber with the cell line, in the lower chamber as a

chemoattractant, or when added to both. The aque-

ous produced a variable response, but was consis-

tently inhibitory when included directly in the upper

chamber with the cells. The vitreous was a potent

stimulator whether included in the upper or lower

chamber, or both, but maximal effect was seen when

added to the lower chamber (P < .05).

It was necessary to store the aqueous and vitre-

ous before use; recovery from 2208C was therefore

tested to determine whether the ability to regulate

invasion would be affected. Assays were performed

on the 2 established cell lines using freshly acquired

aqueous and vitreous, and subsequently the same

aqueous and vitreous samples were recovered from

2208C and retested. No significant differences were

seen in the ability of frozen samples to elicit a

response when compared with freshly acquired sam-

ples (data not shown). The study was subsequently

designed to address 2 issues: first, if different tumors

responded comparably to the same sample of aque-

ous and vitreous, and second to see if the aqueous

and vitreous from different patients had varying

effects. To test a series of 10 primary uveal melano-

mas for variations in their response after exposure to

the same aqueous and vitreous standard, it was

therefore necessary to use pooled aqueous and vitre-

ous samples previously collected from 10 patients

and stored at 2208C. To investigate if the aqueous

and vitreous from uveal melanoma patients differed,

1788 CANCER April 15, 2008 / Volume 112 / Number 8

the same tumor cell lines needed to be tested against

the aqueous and vitreous from different patients.

Chemoinvasion of uveal melanomasTen freshly derived cultures of primary uveal mela-

nomas were tested against a pooled sample of aque-

ous and vitreous (frozen as aliquots and recovered as

required). The assay was performed as detailed

previously,11 and was modified accordingly. Where

required vitreous was added to the lower chamber at

a final concentration of 10%, and aqueous humor

(final concentration 10%) was added to the top

chambers and the assay was left for 16 hours at 378Cin 5% CO2. As a contrast, additional assays were

undertaken to establish whether the aqueous and vit-

reous from different patients with uveal melanoma

was capable of inducing varying responses. In these

assays the established cell lines (SOM 157d and 177)

were used as the standard against which was tested

the aqueous and vitreous from 2 ciliary body and 4

choroid melanomas. Finally, the invasion of 2 pri-

mary uveal melanomas tested against their respective

aqueous and vitreous samples was measured in

assays.

The invasive cell line (SOM 177) and noninvasive

cell line (SOM 157d) were included in every assay as

positive and negative assay controls, respectively.

The effect of the aqueous or vitreous on the invasion

of uveal melanomas in the chemoinvasion assay was

compared with the appropriate control invasion, and

analyzed using a Student t-test. In all cases P < .05

was taken as significant (Figs. 1, 2).

Proliferation of Uveal Melanomas (MTT Assay)The proliferation of 7 cultures of uveal melanoma

was investigated, and pooled aqueous and vitreous at

TABLE 1Clinicopathologic Details of Patients

SOM Sex Age at diagnosis Tumor location Cell type M T D (mm)* Statusy

330 M 27 Choroid S 14.4 Alive 62 mo

349 F 66 Choroid SB 17.3 Alive 55 mo

368 F 53 Ciliary body/Choroid M 11.25 Alive 41 mo

377 M 63 Ciliary body M — Died liver metastases 15 mo

380 F 63 Ciliary body M 11.5 Alive 44 mo

389 F 68 Choroid M 10.3 Alive 42 mo

450 M 73 Ciliary body M 19.75 Alive 15 mo

453 M 70 Choroid SB 8.5 Alive 21 mo

455 M — Choroid M 18.95 Alive 20 mo


458 M 60 Choroid M 16.95 Alive 13 mo

459 F 75 Choroid E 12.75 Alive 20 mo

462 F 73 Metastatic cancer to choroid — 11.9 Alive 19 mo

464 M 45 Choroid S 14.5 Alive 17 mo

467 F 56 Choroid SB 9.8 Alive 17 mo

474 M 49 Ciliary body/choroid M 14.7 Alive 15 mo

477 M 64 Ciliary body/Choroid M 14.4 Alive 14 mo

486 M 59 Choroid S 11 Alive 12 mo


The clinicopathologic details for the patients investigated in the invasion assays and ChemiArray. Also included are those patients for which samples of the aqueous and vitreous were tested individually (SOM

453, 455, 457, 459, 474, and 477). The clinical details of patients used in proliferation assays and whose samples contributed to the pool standards are not presented. All patients were treated by enucleation.

* MTD indicates mean tumor diameter.y Mean follow-up time of 25 months.

FIGURE 1. The invasion of different short-term cultures of uveal melano-mas in response to stimulation by pooled samples of aqueous and vitreous.

Invasion is measured in comparison to baseline invasion of the control

(unstimulated) for each respective cell line. *P < .05 and yP < .005. SOM

177 and SOM 157d were used as positive and negative controls for each

assay performed.

Aqueous and Vitreous Regulate Melanoma/Canovas et al. 1789

a final concentration of 10% was used. Proliferation

was assessed over 4 days, with triplicates for each

day. To a 96-well plate, 2000 cells in 150 lL serum-

free RPMI-1640 with 0.1% BSA, with aqueous or vit-

reous as required was added per well. On Days 1, 2,

3, and 4, 50 lL of media was removed from the rele-

vant triplicates and substituted with 100 lL of 4, 5-

dimethyl-2-thiazolyl)-2,-5-diphenyl-2H-tetrazolium

bromide (MTT) (1 mg/mL) and incubated at 378C for

4 hours. The media was removed and 200 lL of

DMSO to dissolve the formazan crystal was added.

Absorbance was measured at 560 nm with an MRXII

plate reader (Dynex UK). The means were calculated

for each triplicate and the overall mean and standard

deviation of the mean (SEM) were calculated for

each day.

Cytokine Quantification (ChemiArray)The aqueous and vitreous from a ciliary body, 2

choroid melanomas, and as a comparison a meta-

static bowel cancer to the choroid were investigated

using ChemiArray analysis for over 70 cytokines.

ChemiArrays (ChemiArray Human Cytokine Antibody

Array V Map, Chemicon International, Temecula,

Calif) were prepared following the manufacturer’s

instructions. Briefly, ChemiArray membranes were

incubated with 2 mL ChemiArray blocking buffer for

30 minutes at room temperature. One mL of aqueous

or vitreous (1:10 dilution in ChemiArray blocking

buffer) was placed onto membranes for 2 hours.

After incubation membranes were washed 33 with

2 mL of washing buffer I (5 minutes each wash), and

32 with washing buffer II (5 minutes each wash).

One mL of biotin-conjugated anti-cytokines primary

antibody was added to each membrane and left for

2 hours at room temperature. Membranes were

washed 33 in washing buffer I and 32 in buffer II (5

minutes). Horseradish peroxidase-conjugated strepta-

vidin (1 mL) was added for 2 hours at room tempera-

ture. Finally, membranes were washed, as detailed

previously, and 250 lL of detection buffer I and II

were mixed together, added to the membranes, and

left for 2 minutes. Membranes were allowed to dry

and exposed to x-ray film and the expression of cyto-

kines was measured by densitometry and compared

as a percentage of the positive control. To confirm

that the aqueous and vitreous studied by ChemiArray

were capable of regulating uveal melanomas, che-

moinvasion assays with SOM 177 and 157d were

undertaken using the same aqueous and vitreous

samples.

RESULTSProliferation of Uveal MelanomasThe uveal melanomas tested had wide variation in

their levels of proliferation, but there were no major

differences in the ability of the aqueous and the vit-

reous to regulate proliferation, and the responses

were nonsignificant, reflecting the proliferative ability

of the tumor (data not shown).

Chemotaxis and Chemoinvasion AssayThe regulation of different primary uveal melanomas

by pooled samples of aqueous and vitreous is pre-

sented in Figure 1. The control invasion of the pri-

mary uveal melanomas naturally varied. It was

apparent that the vitreous universally stimulated

invasion, compared with baseline control invasion,

whereas the effects of the aqueous were more mixed.

Stimulation by the vitreous ranged in magnitude

from minimal effect to an 11-fold increase, and it

was noted that low-invasive melanomas (SOM 157d,

389, 458) were proportionally more highly stimulated

than the naturally more invasive melanomas. Results

for the aqueous varied, with tumors inhibited (SOM

330, 377, and 389) unaffected (SOM 458 and 464), or

stimulated at a reduced level compared with the vit-

reous (SOM 349, 368, 380, 459, and 467). The find-

ings did not correlate with the baseline invasion of

the melanoma and both low and highly invasive mel-

anomas could be inhibited by the aqueous. In the

assays where 2 melanomas were tested against their

own aqueous and vitreous (486 and 494), the aque-

FIGURE 2. The regulation of invasion of SOM 177 (invasive) and SOM

157d (low-invasive) by the aqueous and vitreous from different patients with

primary uveal melanomas. Invasion is measured in comparison to baseline

invasion of the control (unstimulated) for each respective cell line. P < .05.


ous inhibited, whereas the vitreous stimulated (data

not shown).

Regulation of invasion by the vitreous and aque-

ous from different patients with uveal melanoma is

presented in Figure 2, and includes samples from

SOM 453 and 457 which were also studied in the

ChemiArray assay. The vitreous from all patients

studied was always stimulatory, as were most of the

aqueous samples tested on SOM 177. In contrast,

the aqueous samples always inhibited to some extent

the low-invasive SOM 157d cell line, and the invasive

SOM 177 was also inhibited by 1 aqueous sample

(474).

ChemiArray Analysis for Cytokine QuantificationAs an initial screen to identify potential regulators,

only 4 samples of aqueous and vitreous were ana-

lyzed, and it was not possible to undertake statistical

analysis. All cytokines were detectable in the aque-

ous and vitreous samples from all cases. There were

variations in the amount detected, and in general

vitreous samples had the same or higher levels for

the majority of cytokines, compared with the aque-

ous samples. Only a few cytokines were found in

greater levels in the aqueous, such as MIP-1b (cases

450 and 462). The most highly expressed cytokines

were consistently TIMP-2, TIMP-3, TGF-b2, PIGF,

NAP-2, HGF, GDNF, CKB8-1, OSM, ANG, TNF-a,TNF-b, and RANTES. Others were highly expressed

in some samples but not all, including GRO, GRO-a,

IL-8, IL-10, MCP-1, MCSF, MIP-1b, and MIP-1 DEL.

There was also variation in the content of the aque-

ous and vitreous from different patients. Lower

levels for most regulators were found for the meta-

static cancer patient (462) and from case 450, and it

is of interest that these were the only cases to have

higher levels of some regulators in the aqueous com-

pared with the vitreous (MIP-1b). Insufficient aque-

ous humor remained from SOM 450 and 462 after

the ChemiArray analysis had been performed; there-

fore, only SOM 453 and 457 could be tested by both

ChemiArray analysis and invasion assay. The

response in the invasion assay confirmed that the

vitreous was stimulatory and the aqueous could be

inhibitory.

DISCUSSIONAfter standardization of the chemoinvasion assay

the vitreous was included in the lower chamber,

whereas the aqueous was most inhibitory when

added directly to the upper chamber. In its simplest

form this arrangement mimics in situ exposure of

uveal melanomas, as tumor cells in the anterior

segment are directly in contact with the aqueous,

whereas posterior uveal melanomas in early stages

are anatomically separated from direct contact with

the vitreous by the neuro-retina and pigment epi-

thelium. We have found pathologic evidence that

suggests that the aqueous and vitreous can alter

uveal melanoma behavior in situ. Through examina-

tion of tumors extending from the posterior section

to the anterior chamber we observed alterations in

cell morphology, with cells directly in contact with

the aqueous adopting a benign appearance (data

not shown).

Pooled vitreous was found to universally stimu-

late different uveal melanomas (Fig. 1), but in a

smaller study we previously showed that the vitreous

had little or no effect.15 This difference probably

reflects the filtration of the vitreous before use in the

earlier study, a procedure which could remove con-

stituents of the vitreous, whereas in this study cen-

trifugation of the vitreous was used. The effects of

the pooled aqueous on invasion, however, were more

diverse, but in the assays where 2 melanomas were

tested with their own respective aqueous and vitre-

ous (data not shown) the aqueous was found on

both occasions to be inhibitory. These findings sug-

gest that the resultant invasion is a manifestation

of how the tumor itself interacts and responds to

regulation by aqueous. As pooling of the samples

would equalize any regulatory differences the effect

of aqueous and vitreous from individual patients

was also considered (Fig. 2). The vitreous from all

patients was again consistently stimulatory. In con-

trast, the high invasive cell line (SOM 177) was both

inhibited and stimulated by the different aqueous

samples, implying variation in the levels of a regula-

tory factor or factors in the humors. The finding,

however, that all aqueous samples inhibited or had

little effect on the low-invasive SOM 157d again sug-

gests that tumor phenotype affects how uveal mela-

nomas respond to regulation by the aqueous (Fig. 2).

As greater variability was observed in response to the

aqueous and vitreous samples from patients with

choroid melanomas (SOM 453, 455, 457, and 459), it

seems likely that the melanomas themselves contri-

bute to differences in the aqueous and vitreous. It is

possible that as tumors from different locations de-

velop they influence the ability of the aqueous and

vitreous to modulate invasion as a mechanism of

self-regulation. For further confirmation of this point,

iris melanomas should ideally be investigated, but

because of their infrequency and small size none

were available during the course of the study. Like-

wise, to extend these comparisons it is important to

consider that all melanomas studied were removed


from patients with large tumors treated by enuclea-

tion; smaller tumors may, of course, behave differ-

ently. Finally, as all tests in this study were

performed with a dilution of 1:10 aqueous and vitre-

ous, it is likely that in vitro the localized effects could

be amplified.

In this investigation ChemiArray analysis (Fig. 3)

established that the same patterns of cytokines were

detectable in all samples, but that levels varied (Ta-

ble 2). Despite these similarities chemoinvasion anal-

ysis confirmed that the aqueous and vitreous of SOM

453 and 457 were still capable of inhibiting and sti-

mulating invasion, respectively (Fig. 2). Differences

in the amount of the regulators present could contri-

bute toward control of invasion; however, in what

manner is unclear, as the aqueous and vitreous from

the same patient tended to have comparable levels

of any given regulator. The aqueous and vitreous

from SOM 450 and the patient with metastatic bowel

cancer (462) did, however, have lower levels of many

regulators when compared with SOM 453 and 457

(Table 2), but as neither case could be analyzed in

the chemoinvasion assay it was therefore not possi-

ble to establish whether the lower levels translated

into regulatory differences.

The high expression of TIMP-2, TIMP-3, and

TGF-b2 observed by ChemiArray is of interest, as

all have been directly associated with uveal mela-

noma development and/or progression. TIMP-3

gene expression has been identified as down-regu-

lated in Class II aggressive uveal melanomas,17,18

probably through hypermethylation,18 and translat-

ing to differential protein expression19 by tumors

with and without disomy of chromosome 3.

Improved survival has been reported for patients

with TIMP-2 melanomas,20 and its up-regulation

correlates with decreased migratory ability.22 Lastly,

TGF-b2 has been found to decrease uveal mela-

noma invasion, regulate adhesive interactions with

hepatic endothelium,11,22 and abnormalities in its

pathway have been associated with the develop-

ment of uveal melanoma.13 The high levels detecta-

ble for these regulators in all samples of aqueous

and vitreous from patients with uveal melanoma

could imply a direct association, possibly as a

method of self-regulation, as the aqueous and vitre-

ous from the patient with metastatic bowel cancel

contained comparatively half the levels for TIMP-2,

TIMP-3, and TGF-b2 (Table 2).

Other regulators of potential interest include

HGF, which increases uveal melanoma invasion and

assists in preferential targeting of the liver.10,11

Increased uveal melanoma invasion has also been

observed in response to stimulation from MIP-1b,GRO, and RANTES,11 and expression of these regula-

tors in the ChemiArray was higher than average and

showed sample variation (Table 2). In contrast, it

was of interest that VEGF, a potent angiogenic stimu-

lator, was not comparatively more highly expressed

(Table 2), as studies have shown raised VEGF in the

aqueous and vitreous from patients with uveal mela-

noma.14,23 It is expected that the aqueous and vitre-

ous contain a myriad array of other factors not

tested for in this preliminary study, and the relatively

low comparative expression of VEGF in this study

may suggest that the aqueous and vitreous of

patients with uveal melanoma show a dramatic rise

in a large number of other potentially more influen-

tial regulators. Further studies comparing the levels

of regulators found in the aqueous and vitreous from

a larger series of uveal melanomas, with comparison

of the levels seen in patients with other ocular con-

ditions, such as those undergoing vitrectomy and

cataract surgery, will be required to confirm this

hypothesis.

FIGURE 3. ChemiArray analysis for cytokines and growth factors of pairedaqueous and vitreous samples from SOM 453. Positive controls top left and

bottom right.


The microenvironment of the tumor has been

shown to radically alter tumor behavior, and uveal

melanoma has a highly variable response to local

environmental pressures, resulting in suppression of

the invasive genotype and phenotype.24 Invasion of

uveal melanoma as shown here ultimately reflects a

combination of the many regulators present within

both the aqueous and vitreous, which amalgamate to

produce either inhibition or stimulation. Further-

more, we have shown here that the inherent charac-

teristics of individual tumors affect how the

melanoma responds to these stimuli. The response

in vivo would therefore reflect an intricate combina-

tion of factors related to the tumor and its micro-

environment, with the environment having the

potential to fundamentally influence the tumor phe-

notype. In summary, our findings suggest that inter-

action between the melanoma and the local

environment provided by the aqueous and vitreous

plays a major role in dictating future melanoma

behavior.

REFERENCES1. Fidler IJ. Critical determinants of metastasis. Semin Cancer

Biol. 2002;12:89–96.

2. Gragoudas ES, Egan KM. Uveal melanoma: a rare malig-

nancy. Ophthalmology. 2000;107:1441–1442.

TABLE 2ChemiArray Analysis for Cytokines and Growth Factors in theAqueous and Vitreous of Patients With Ocular Tumors

Cytokine

SOM450 SOM453 SOM 457 SOM 462

A V A V A V A V

ENA78 9 10 30 26 20 29 12 13GCSF 9 9 31 25 19 27 11 10

GM-CSF 9 6 26 31 19 29 10 10GRO 18 8 35 47 28 44 20 22GRO-A 11 4 29 40 20 38 13 13L309 15 11 21 13 16 19 12 10IL-1A 18 13 32 25 22 26 14 11IL-1B 15 8 25 25 19 22 12 11IL-2 10 7 22 27 18 22 10 11IL-3 6 7 21 25 17 21 7 9IL-4 6 9 27 32 23 29 9 11IL-5 6 7 20 23 17 22 8 8IL-6 7 6 21 21 18 26 8 7

IL-7 6 5 16 20 14 22 8 6IL-8 18 9 27 39 24 40 23 14IL-10 26 10 35 51 31 49 25 32IL-12 12 13 20 11 14 19 13 11IL-13 8 8 21 17 15 19 8 7IL-15 8 8 21 21 16 20 8 8IFN-A 8 9 23 23 19 23 8 11MCP-1 14 23 37 36 28 46 17 29MCP-2 6 10 23 28 21 28 7 10MCP-3 6 10 23 31 21 29 8 10

MCSF 28 23 42 49 44 47 27 24MDL 23 7 23 33 25 29 15 9MIG 8 5 19 32 20 29 14 8MIP-1B 35 21 44 61 43 53 34 8MIP-1DEL 25 41 35 31 32 33 28 28RANTES 21 22 20 37 31 35 19 18SCF 9 10 39 20 19 21 6 8SDF-1 8 11 23 24 21 27 7 11TARC 7 11 20 25 19 26 6 11TGF-B1 9 11 23 32 22 30 7 16TNF-A 30 30 42 49 41 45 25 32

TNF-B 16 18 32 37 34 40 14 14EGF 9 9 22 31 25 28 11 9IGF-1 13 7 22 60 26 32 13 10ANG 31 21 42 39 43 52 28 41OSM 32 59 42 22 43 46 31 28TPO 1 14 24 41 20 27 8 8VEGF 17 15 31 31 29 31 9 13PDGF-BB 10 14 22 26 24 26 7 9LEPTIN 12 13 23 30 25 28 8 12BDNF 20 19 32 38 26 35 13 24

BLC 15 20 25 30 26 28 10 15CKB8-1 37 31 39 48 45 43 34 23EOTAXIN 13 8 18 25 24 21 11 8EOTAXIN- 28 14 31 39 35 37 16 20EOTAXIN- 46 30 42 52 45 48 16 36FGF-4 10 25 25 18 28 25 13 7FGF-6 8 19 22 20 21 25 7 8FGF-7 14 28 26 29 26 27 8 12FGF-9 13 24 27 32 26 28 9 16FIT-3LIGA 11 18 22 29 22 24 7 13FRACTALI 36 30 37 52 39 40 19 22

GCP-2 17 20 26 35 26 27 10 11

(continued)

TABLE 2(continued)

GDNF 30 30 40 50 40 44 33 17

HGF 38 49 45 60 49 57 43 45

IGFBP-1 29 28 32 45 37 38 46 15

IGFBP-2 28 37 35 49 38 47 10 31

IGFBP-3 23 59 38 32 40 38 23 44

IGFBP-4 12 32 25 22 22 22 8 19

IL-16 15 39 30 34 30 27 9 24

IP-10 25 42 39 47 40 35 17 34

LIF 20 34 28 40 28 27 10 17

LIGHT 18 35 28 41 25 26 7 13

MCP-4 13 24 24 37 24 24 7 9

MIP 15 42 31 42 30 31 11 12

MIP-3A 10 19 22 32 23 24 8 7

NAP-2 39 61 47 58 52 50 29 28

NT-3 21 27 34 48 30 38 8 11

NT-4 23 49 44 36 37 39 21 40

OPIGRN 22 44 48 48 42 42 21 47

PARC 20 40 30 41 31 30 23 47

PIGF 39 51 38 52 51 41 31 56

TGF-B2 56 63 65 60 65 66 35 54

TGF-B3 20 30 35 45 34 35 9 14

TIMP-2 68 80 73 73 66 69 35 58

TIMP-3 69 87 68 70 72 71 46 64

The relative levels are expressed as a percentage of the average membrane positive. The aqueous (A)

and vitreous (V) are presented for each case number. Cases SOM 450 and 457 were from patients

with ciliary body melanomas; SOM 453 was from a patient with choroid melanoma; and case 462

from a patient with metastatic cancer.


3. Kujala E, Makitie T, Kivela T. Very long-term prognosis of

patients with malignant uveal melanoma. Invest Ophthal-

mol Vis Sci. 2003;44:4651–4659.

4. Conway RM, Chua WC, Qureshi C, Billson FA. Primary iris

melanoma: diagnostic features and outcome of conserva-

tive surgical treatment. Br J Ophthalmol. 2001;85:848–854.

5. Shields CL, Shields JA, Materin M, Gershenbaum E, Singh

AD, Smith A. Iris melanoma: risk factors in 169 consecutive

patients. Ophthalmology. 2001;108:172–178.

6. Patel JI, Tombran-Tink J, Hykin PG, Gregor ZJ, Cree IA.

Vitreous and aqueous concentrations of proangiogenic,

antiangiogenic factors and other cytokines in diabetic reti-

nopathy patients with macular edema: implications for

structural differences in macular profiles. Exp Eye Res.

2006;82:798–806.

7. Nishimura M, Ikeda T, Ushiyama M, Nanbu A, Kinoshita S,

Yoshimura M. Increased vitreous concentrations of human

hepatocyte growth factor in proliferative diabetic retinopa-

thy. J Clin Endocrinol Metab. 1999;84:659–662.

8. de Boer JH, Limpens J, Orengo-Nania S, de Jong PT, La

Heij E, Kijlstra A. Low mature TGF-beta 2 levels in aqueous

humor during uveitis. Invest Ophthalmol Vis Sci. 1994;

35:3702–3710.

9. Ozcan AA, Ozdemir N, Canataroglu A. The aqueous levels

of TGF-beta 2 in patients with glaucoma. Int Ophthamol.

2004;25:19–22.

10. Hendrix MJ, Seftor EA, Sefter RE, et al. Regulation of uveal mel-

anoma interconverted phenotype by hepatocyte growth factor/

scatter factor (HGF/SF). Am J Pathol. 1998;152:855–863.

11. Woodward JKL, Elshaw SR, Murray AK, et al. Stimulation

and inhibition of uveal melanoma invasion by HGF, GRO,

IL-1alpha and TGF-beta. Invest Ophthalmol Vis Sci. 2002;

43:3144–3152.

12. Hu DN, McCormick SA, Lin AY, Lin JY. TGF-beta 2 inhibits

growth of uveal melanocytes at physiological concentra-

tions. Exp Eye Res. 1998;67:143–150.

13. Myatt N, Aristodemon P, Neale MH, et al. Abnormalities of

the transforming growth factor-beta pathway in ocular

melanoma. J Pathol. 2000;192:511–518.

14. Boyd SR, Tan D, Bunce C, et al. Vascular endothelial growth

factor is elevated in ocular fluids of eyes harbouring uveal

melanoma: identification of a potential therapeutic window.

Br J Ophthalmol. 2002;86:448–452.

15. Canton I, Eves PC, Szabo M, et al. Tumor necrosis factor

alpha increases and alpha melanocyte-stimulating hor-

mone reduces uveal melanoma invasion through fibronec-

tin. J Invest Dermatol. 2003;121:557–563.

16. Elshaw SR, Sisley K, Cross NA, et al. A comparison of ocu-

lar melanocyte and uveal melanoma cell invasion and the

implication of a1b1, a4 b1, and a6 b1 integrins. Br J

Ophthalmol. 2001;85:732–738.

17. Tschentscher F, Husing J, Holter T, et al. Tumor classifica-

tion based on gene expression profiling shows that uveal

melanomas with and without monosomy 3 represent two

distinct entities. Cancer Res. 2003;63:2578–2584.

18. van der Velden PA, Zuidervaart W, Hurks MH, et al. Expres-

sion profiling reveals that methylation of TIMP3 is involved

in uveal melanoma development. Int J Cancer. 2003;106:

472–479.

19. Nareyeck G, Zeschnigk M, von der Haar D, Schilling H,

Bornfeld N, Anastassiou G. Differential expression of tissue

inhibitor of matrix metalloproteinases 3 in uveal mela-

noma. Ophthalmic Res. 2005;37:23–28.

20. El-Shabrawi Y, Ardjomand N, Radner H, Ardjomand N.

MMP-9 is predominantly expressed in epithelioid and not

spindle cell uveal melanoma. J Pathol. 2001;194:201–206.

21. Klisovic DD, Klisovic MI, Effron D, Liu S, Marcucci G, Katz

SE. Depsipeptide inhibitis migration of primary and meta-

static uveal melanoma cell lines in vitro: a potential stra-

tedy for uveal melanoma. Melanoma Res. 2005;15:147–

153.

22. Woodward JKL, Rennie IG, Elshaw SR, Burn JL, Sisley K.

Invasive and non-invasive uveal melanomas have different

adhesive properties. Eye. 2005;19:342–348.

23. Missotten GS, Notting IC, Schlingemann RO, et al. Vascular

endothelial growth factor a in eyes with uveal melanoma.

Arch Ophthalmol. 2006;124:1428–1434.

24. Folberg R, Arbieva Z, Moses J, et al. Tumor cell plasticity in

uveal melanoma: microenvironment directed dampening

of the invasive and metastatic genotype and phenotype

accompanies the generation of vasculogenic mimicry pat-

terns. Am J Pathol. 2006;169:1376–1389.


local environmental influences on uveal melanoma : vitreous humor promotes uveal melanoma invasion,...

Documents