Carcinoembryonic Antigen Its Role in the Evaluation of Intraocular Joseph
B.
Tumors
Michelson, MD; Norman T. Felberg, PhD; Jerry A. Shields, MD
\s=b\ Plasma
carcinoembryonic antigen
studied in 60 patients with histologically confirmed intraocular neoplasms including 56 malignant melanomas of the uvea and four metastatic tumors to the choroid. While 45% of the patients with primary uveal melanomas, as well as 75% of the patients with metastatic disease demonstrated elevated plasma CEA levels, both patients who exhibited metastatic lesions of entodermal origin demonstrated plasma CEA values that clearly fell into a separate, highly elevated category, consistent with metastatic disease or pancreatic or colorectal carcinoma. Thus, in the patient seen with a nonpigmented choroidal mass that may represent either a choroidal hemangioma, amelanotic melanoma, or metastatic tumor, plasma CEA levels may be useful in the differential diagnosis. If the clinician suspects a metastatic tumor from an occult primary site, highly elevated CEA levels may indicate that the lesion is of entodermal origin.
(CEA) was
Malignant
(Arch Ophthalmol 94:414-416, 1976)
A
report suggests that met¬ astatic carcinoma to the eye is more frequent than previously thought, and that it, in fact, may be the most common malignant tumor of the eye.1 Ferry and Font list as the primary tissue of origin breast (39.7%), lung (29.5%), and kidney (4%) as the leading three sources for ocular metastasis.' They separate the gas¬ trointestinal malignant tumors by their respective organ.1 Many pa¬ tients with metastatic malignant tu¬ mors are not examined by an ophthal¬ mologist unless the ocular symptoms are the initial manifestation of the disease. In patients seen with ocular metastasis from an occult source, a means of directing one's attention to the primary site would be of definite value. A carcinoembryonic antigen (CEA) thought to be specific for adenocarcinoma of the colon was first described by Gold and Freedman in 1965'; later investigations have in¬ cluded other gastrointestinal tract malignant tumors that manifested elrecent
Submitted for publication Dec 17, 1974. From the Department of Molecular Biology of the Research Institute (Drs Michelson and Felberg) and the Oncology Unit of the Retina Service (Dr Shields) of the Wills Eye Hospital and Research Institute, Philadelphia. Reprint requests to Wills Eye Research Institute, 1601 Spring Garden St, Philadelphia, PA 19130 (Dr Michelson).
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
evated
CEA
values.'
Subsequent
studies, using radioimmunoassay' have demonstrated elevated plasma CEA levels in a variety of other tu¬ mors, some nonentodermally derived, "'
as
well
ease
as
in
some
nonmalignant
dis¬
states.6-8
METHODS Plasma CEA assays were performed by the indirect and direct method, employing the CEA-Roche CEA assay kit," which uti¬ lizes the Hansen zirconyl phosphate gel
technique. All patients had their conditions eval¬ uated by the Oncology Unit of the Retina Service, Wills Eye Hospital, Philadelphia, and subsequently had a histologically proven diagnosis of intraocular malignant tumor. Each patient was studied by the usual diagnostic modalities.10 RESULTS
Of the 60 patients with intraocular malignant tumors, 56 had primary malignant melanomas of the uvea,
and four had metastatic tumors to the uvea. The primary uveal melanomas included three iris tumors, and 53 choroidal and ciliary body tumors. The four metastatic tumors were all to the choroid. Two were from pri¬ mary lesions in the gastrointestinal tract, one from the breast, and one was metastatic melanoma from the skin to the choroid. In each case, clini-
cal
diagnosis of the
metastatic ocular
Table 1.—Plasma CEA Level Distribution in Patients With Proven Intraocular Malignant Tumor*
tumor was confirmed histologically. The two eyes with métastases from
gastrointestinal origin
were
hemorrhage.
The second patient had a plasma CEA level greater than 20 ng/ml by the indirect method and 330 ng/ml by direct assay. Once again, medical in¬ vestigations at another institution failed to detect a primary lesion. Fol¬ low-up of this patient six weeks after enucleation revealed a CEA value of greater than 20 ng/ml and 583 ng/ml by indirect and direct assay, respec¬
tively.
CEA Level,
enu¬
cleated because no evidence of a pri¬ mary malignant tumor could be found on the initial medical evaluation. The eye with metastatic melanoma to the choroid was obtained at autopsy. Plasma CEA values for these pa¬ tients are presented in Table 1. Of 56 patients with uveal melanomata, 25 or 44.6% demonstrated elevated plasma CEA values, ie, above 2.5 ng/ml.6"8 The highest value recorded in this series was 9.9 ng/ml (median value was 2.2 ng/ml, and the mean for the group was 2.7 ng/ml). As seen in Table 2, the preoperative and post¬ operative plasma CEA values of sev¬ eral patients studied by serial assay would indicate that with removal of the tumor burden, plasma CEA values decrease. The two patients whose conditions were later diagnosed as metastatic colorectal adenocarcinoma were ini¬ tially admitted with a working diag¬ nosis of "metastatic tumor to the cho¬ roid of uncertain origin." The first patient was seen with a plasma CEA level of 18.7 ng/ml. Thorough medical evaluation failed to reveal a primary lesion. Six weeks later, the plasma CEA values were greater than 20 ng/ml by the indirect method, and 164 ng/ml by the direct assay. Enucleation was performed for therapeu¬ tic as well as diagnostic purposes, and the histologie findings demonstrated cords of mucin secreting adenocarci¬ noma to the choroid. This patient sub¬ sequently succumbed to a colorectal
One patient who was seen with a metastatic lesion to the choroid, had a history of carcinoma of the breast and subsequent mastectomy. The pa-
Histologically
Diagnosis
0.0-2.5
2.6-5.0
5.1-10.0
56
55.4
30.4
14.3
Primary uveal malignant melanoma Colorectal adenocarcinoma metastatic to the choroid Breast carcinoma metastatic to the choroid Primary cutaneous melanoma metastatic to the choroid Nonsmoking volunteers8
Distribution values given
as
ng/ml
Subjects
>10
100 100
892
100 2.8
97.0
0.2
percentage.
Table 2.—Plasma CEA Levels in Ocular Melanoma Patients
CEA, ng/ml Patient 1
5'
Cell Type
Preoperative
Posttherapy
Time, Mo
Spindle A Epithelioid
8.1 5.4
8
Mixed Mixed
1.0 7.1
5.5 4.3 1.3 3.8 4.1
...
5.4
*
Patient did not have enucleation performed; small lesion lation.
tient exhibited a plasma CEA value of 2.5 ng/ml. Subsequent enucleation at another institution revealed meta¬ static breast carcinoma to the cho¬ roid. The last patient with metastatic tumor in this study was admitted with a "solid pigmented fundus le¬ sion," and a plasma CEA value of 2.9 ng/ml. The patient had a primary cutaneous melanoma removed previ¬ ously. Autopsy six weeks later dem¬ onstrated melanomatosis of the CNS, lung, liver, pleural fluid, and choroid. Three patients with presumed hemangiomata of the choroid and one with an iris cyst, none of whom had had enucleations, all demonstrated plasma CEA values that fell within normal limits (less than 2.5 ng/ml). COMMENT
In the 56 patients with primary uveal melanomas, there is no clear re¬ lationship between plasma CEA val¬ ues and cell type, tumor mass, tumor extension, or age of the patient. Such data are consistent with those found for primary cutaneous melanomata (J.P. Vandevoorde, PhD, oral commu¬ nication, March 1974). Clearly, the majority of patients with positive
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
being
treated with
photocoagu¬
melanomas, both iris and choroidal, have CEA levels within 2.5 to 5.0 ng/ml limits, which is within the same range as the levels of an in¬ creased number of smoking volun¬ teers, patients with nonmalignant disease, and patients with entodermal inflammation.08 While no concur¬ rent diseases, such as colitis, cirrhosis, or pancreatitis, were noted in the pa¬ tients' history, such subclinical dis¬ ease could have gone undetected and could, therefore, account for an ele¬ vated plasma CEA level unrelated to the neoplasm in question. The patient with metastatic breast carcinoma demonstrated a value of 2.5 ng/ml when seen. Such a value is consistent with that found by Reynoso et al7 in a study of 25 patients with disseminated breast cancer; it also concurs with the study of Hansen et al,8 in which 73% of patients with breast carcinoma had a value of less than 5.0 ng/ml. The two patients with metastat¬ ic gastrointestinal adenocarcinoma, both demonstrated plasma CEA lev¬ els well above the range of the 58 pa¬ tients mentioned thus far. Although we had no patient with metastatic
to the eye from the lung, one extrapolate from published data. Seventy-two percent or more of such patients would be expected to have an elevated CEA level.8" In one study of 77 patients considered "inoperable" (ie, either locally advanced or dis¬ seminated disease), who might there¬ fore be expected to be candidates for cancer can
ocular métastases, a mean value of 7.8 ng/ml was found in the plasma. Plasma CEA levels would appear to be of value in the differential diag¬ nosis of uveal tumors. Values greater than 20 ng/ml are found predomi¬ nantly in patients with gastrointesti¬ nal adenocarcinoma (pancreatic and colorectal), while values up to 10 ng/ml could be found in a variety of patients with malignant tumors, in¬ flammations, and in an otherwise healthy smoking population.8 The sharply elevated CEA values of the two patients with metastatic colo¬ rectal adenocarcinoma to the choroid are consistent with those reported in patients with pancreatic or colorectal carcinoma or generalized metastatic
disease.8 The ocular tumors from these patients stained positively for CEA by the immunohistologic tech¬ nique of Hsu,'- indicating that the ocular tumors in both of these entodermally derived cases were possibly contributing to the CEA-secreting tu¬ mor burden.1-·" Neither the eye from the patient with metastatic breast carcinoma (with normal plasma CEA level), nor those of any of three pa¬ tients with choroidal melanomas tested (two with elevated plasma CEA) gave positive results for the presence of CEA by the immuno¬ histologic technique. This raises the question of the source of this antigen in the plasma of patients with pri¬ mary uveal melanoma. We have not been able to demonstrate CEA by this method in tumor tissue from retinoblastoma patients with elevated plasma CEA, although the plasma values decrease with removal of the tumor burden." One of the most difficult diagnostic problems occasionally encountered by the ophthalmologist is in differ-
entiating an amelanotic melanoma or choroidal hemangioma from a meta¬
static tumor to the choroid from an occult primary.1"" The use of fluores¬ cein angiography, radioactive phos¬ phorus uptake studies, and other mo¬ dalities have definite diagnostic limitations.'"·'" Plasma CEA determi¬ nation may play a future role as a possible diagnostic adjunct in such cases, especially in cases of gastroin¬ testinal malignant tumors.
This investigation was supported in part by Public Health Service training grant EY-00084 and General Research Support grant 5S01RR05510 of the National Eye Institute, and by the Retina Research and Development Foun¬ dation of Philadelphia, and the Lions Club of Pennsylvania. Mr. Leonard Foster gave technical assistance. H.J. Hansen, PhD, J.P. Vandevoorde, PhD, and J. Primus, PhD, of Hoffmann-La Roche, Nutley, N.J. supplied reagents employed in these studies and performed the CEA assays. H. Robbin, MD, Department of Pathology, Thomas Jefferson Hospital, Philadelphia supplied tissue. The mem¬ bers of the Retina Service of the Wills Eye Hospital, Philadelphia, allowed us to examine their patients.
References 1. Ferry AP, Font FL: Carcinoma metastatic to the eye and orbit. Arch Ophthalmol 92:276-286,
1974. 2. Gold
mor
P, Freedman SD: Demonstration of tuspecific antigens in human colonic carcino-
by immunological tolerance and absorption techniques. J Exp Med 121:439-459, 1965. 3. Gold P, Freedman SD: Specific carcinoembryonic antigens of the human digestive system. J Exp Med 122:467-481, 1965. 4. Thomson DMP, Krupey J, Freedman SD, et al: The radioimmunoassay of circulating carcinoembryonic antigen of the human digestive mata
system. Proc Natl Acad Sci 64:161-167, 1969. 5. Hansen HJ, Lance KP, Krupey J: Demon-
an ion sensitive antigenic site on carcinoembryonic antigen using zirconyl phosphate gel. Clin Res 19:143, 1971. 6. Groover JR, Rogers AI: Immunologic tests for the detection of gastrointestinal cancers: Status report on carcinoembryonic antigen
stration of
(CEA) and alpha-fetoprotein (AFP). So
Med J
66:1218-1221, 1973.
7. Reynoso G, Chu FM, Holyoke TM, et al: Carcinoembryonic antigens in patients with dif-
ferent cancers. JAMA 22:361-365, 1972. 8. Hansen JH, Snyder JJ, Miller E, et al: Carcinoembryonic antigen (CEA) assay: A laboratory adjunct in the diagnosis and management of cancer. Human Pathol 5:139-147, 1974. 9. CEA-Roche, Carcinoembryonic Antigen Assay: An in vitro test to aid in the management and diagnosis of cancer: in A Clinical Monograph, Nutley, NJ, Roche Clinical Laboratories, Inc, 1974. 10. Shields JA, McDonald PR: Improvements in the diagnosis of posterior uveal melanoma. Arch Ophthalmol 91:259-264, 1974. 11. Vincent RG, Chu TM: Carcinoembryonic antigen in patients with carcinoma of the lung. J Thorac Cardiovasc Surg 66:320-328, 1973. 12. Hsu KC, Zimmerman EA, Rodin L, et al:
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
Application of tumor tissues with antibody to tumor associated antigen (TAA), in Embryonic and Fetal Antigens in Cancer: Proceedings of the 2nd Conference, Oak Ridge National Laboratory, Oak Ridge, Tenn, 1972, vol 2. 13. Michelson JB, Felberg NT, Shields JA, et al: Carcinoembryonic antigen-positive metastatic adenocarcinoma of the choroid. Arch Ophthalmol 93:794-796, 1975. 14. Michelson JB, Felberg NT, Shields JA: Fetal antigens in retinoblastoma. Cancer, to be
published.
15. Shields JA, Zimmerman LE: Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 89:466-471, 1973. 16. Shields JA, Hagler WS, Federman JL, et al: The significance of the 32P uptake test in the diagnosis of posterior uveal melanomas. Trans Am Acad Ophthalmol Otolaryngol, to be published.
B.
Tumors
Michelson, MD; Norman T. Felberg, PhD; Jerry A. Shields, MD
\s=b\ Plasma
carcinoembryonic antigen
studied in 60 patients with histologically confirmed intraocular neoplasms including 56 malignant melanomas of the uvea and four metastatic tumors to the choroid. While 45% of the patients with primary uveal melanomas, as well as 75% of the patients with metastatic disease demonstrated elevated plasma CEA levels, both patients who exhibited metastatic lesions of entodermal origin demonstrated plasma CEA values that clearly fell into a separate, highly elevated category, consistent with metastatic disease or pancreatic or colorectal carcinoma. Thus, in the patient seen with a nonpigmented choroidal mass that may represent either a choroidal hemangioma, amelanotic melanoma, or metastatic tumor, plasma CEA levels may be useful in the differential diagnosis. If the clinician suspects a metastatic tumor from an occult primary site, highly elevated CEA levels may indicate that the lesion is of entodermal origin.
(CEA) was
Malignant
(Arch Ophthalmol 94:414-416, 1976)
A
report suggests that met¬ astatic carcinoma to the eye is more frequent than previously thought, and that it, in fact, may be the most common malignant tumor of the eye.1 Ferry and Font list as the primary tissue of origin breast (39.7%), lung (29.5%), and kidney (4%) as the leading three sources for ocular metastasis.' They separate the gas¬ trointestinal malignant tumors by their respective organ.1 Many pa¬ tients with metastatic malignant tu¬ mors are not examined by an ophthal¬ mologist unless the ocular symptoms are the initial manifestation of the disease. In patients seen with ocular metastasis from an occult source, a means of directing one's attention to the primary site would be of definite value. A carcinoembryonic antigen (CEA) thought to be specific for adenocarcinoma of the colon was first described by Gold and Freedman in 1965'; later investigations have in¬ cluded other gastrointestinal tract malignant tumors that manifested elrecent
Submitted for publication Dec 17, 1974. From the Department of Molecular Biology of the Research Institute (Drs Michelson and Felberg) and the Oncology Unit of the Retina Service (Dr Shields) of the Wills Eye Hospital and Research Institute, Philadelphia. Reprint requests to Wills Eye Research Institute, 1601 Spring Garden St, Philadelphia, PA 19130 (Dr Michelson).
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
evated
CEA
values.'
Subsequent
studies, using radioimmunoassay' have demonstrated elevated plasma CEA levels in a variety of other tu¬ mors, some nonentodermally derived, "'
as
well
ease
as
in
some
nonmalignant
dis¬
states.6-8
METHODS Plasma CEA assays were performed by the indirect and direct method, employing the CEA-Roche CEA assay kit," which uti¬ lizes the Hansen zirconyl phosphate gel
technique. All patients had their conditions eval¬ uated by the Oncology Unit of the Retina Service, Wills Eye Hospital, Philadelphia, and subsequently had a histologically proven diagnosis of intraocular malignant tumor. Each patient was studied by the usual diagnostic modalities.10 RESULTS
Of the 60 patients with intraocular malignant tumors, 56 had primary malignant melanomas of the uvea,
and four had metastatic tumors to the uvea. The primary uveal melanomas included three iris tumors, and 53 choroidal and ciliary body tumors. The four metastatic tumors were all to the choroid. Two were from pri¬ mary lesions in the gastrointestinal tract, one from the breast, and one was metastatic melanoma from the skin to the choroid. In each case, clini-
cal
diagnosis of the
metastatic ocular
Table 1.—Plasma CEA Level Distribution in Patients With Proven Intraocular Malignant Tumor*
tumor was confirmed histologically. The two eyes with métastases from
gastrointestinal origin
were
hemorrhage.
The second patient had a plasma CEA level greater than 20 ng/ml by the indirect method and 330 ng/ml by direct assay. Once again, medical in¬ vestigations at another institution failed to detect a primary lesion. Fol¬ low-up of this patient six weeks after enucleation revealed a CEA value of greater than 20 ng/ml and 583 ng/ml by indirect and direct assay, respec¬
tively.
CEA Level,
enu¬
cleated because no evidence of a pri¬ mary malignant tumor could be found on the initial medical evaluation. The eye with metastatic melanoma to the choroid was obtained at autopsy. Plasma CEA values for these pa¬ tients are presented in Table 1. Of 56 patients with uveal melanomata, 25 or 44.6% demonstrated elevated plasma CEA values, ie, above 2.5 ng/ml.6"8 The highest value recorded in this series was 9.9 ng/ml (median value was 2.2 ng/ml, and the mean for the group was 2.7 ng/ml). As seen in Table 2, the preoperative and post¬ operative plasma CEA values of sev¬ eral patients studied by serial assay would indicate that with removal of the tumor burden, plasma CEA values decrease. The two patients whose conditions were later diagnosed as metastatic colorectal adenocarcinoma were ini¬ tially admitted with a working diag¬ nosis of "metastatic tumor to the cho¬ roid of uncertain origin." The first patient was seen with a plasma CEA level of 18.7 ng/ml. Thorough medical evaluation failed to reveal a primary lesion. Six weeks later, the plasma CEA values were greater than 20 ng/ml by the indirect method, and 164 ng/ml by the direct assay. Enucleation was performed for therapeu¬ tic as well as diagnostic purposes, and the histologie findings demonstrated cords of mucin secreting adenocarci¬ noma to the choroid. This patient sub¬ sequently succumbed to a colorectal
One patient who was seen with a metastatic lesion to the choroid, had a history of carcinoma of the breast and subsequent mastectomy. The pa-
Histologically
Diagnosis
0.0-2.5
2.6-5.0
5.1-10.0
56
55.4
30.4
14.3
Primary uveal malignant melanoma Colorectal adenocarcinoma metastatic to the choroid Breast carcinoma metastatic to the choroid Primary cutaneous melanoma metastatic to the choroid Nonsmoking volunteers8
Distribution values given
as
ng/ml
Subjects
>10
100 100
892
100 2.8
97.0
0.2
percentage.
Table 2.—Plasma CEA Levels in Ocular Melanoma Patients
CEA, ng/ml Patient 1
5'
Cell Type
Preoperative
Posttherapy
Time, Mo
Spindle A Epithelioid
8.1 5.4
8
Mixed Mixed
1.0 7.1
5.5 4.3 1.3 3.8 4.1
...
5.4
*
Patient did not have enucleation performed; small lesion lation.
tient exhibited a plasma CEA value of 2.5 ng/ml. Subsequent enucleation at another institution revealed meta¬ static breast carcinoma to the cho¬ roid. The last patient with metastatic tumor in this study was admitted with a "solid pigmented fundus le¬ sion," and a plasma CEA value of 2.9 ng/ml. The patient had a primary cutaneous melanoma removed previ¬ ously. Autopsy six weeks later dem¬ onstrated melanomatosis of the CNS, lung, liver, pleural fluid, and choroid. Three patients with presumed hemangiomata of the choroid and one with an iris cyst, none of whom had had enucleations, all demonstrated plasma CEA values that fell within normal limits (less than 2.5 ng/ml). COMMENT
In the 56 patients with primary uveal melanomas, there is no clear re¬ lationship between plasma CEA val¬ ues and cell type, tumor mass, tumor extension, or age of the patient. Such data are consistent with those found for primary cutaneous melanomata (J.P. Vandevoorde, PhD, oral commu¬ nication, March 1974). Clearly, the majority of patients with positive
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
being
treated with
photocoagu¬
melanomas, both iris and choroidal, have CEA levels within 2.5 to 5.0 ng/ml limits, which is within the same range as the levels of an in¬ creased number of smoking volun¬ teers, patients with nonmalignant disease, and patients with entodermal inflammation.08 While no concur¬ rent diseases, such as colitis, cirrhosis, or pancreatitis, were noted in the pa¬ tients' history, such subclinical dis¬ ease could have gone undetected and could, therefore, account for an ele¬ vated plasma CEA level unrelated to the neoplasm in question. The patient with metastatic breast carcinoma demonstrated a value of 2.5 ng/ml when seen. Such a value is consistent with that found by Reynoso et al7 in a study of 25 patients with disseminated breast cancer; it also concurs with the study of Hansen et al,8 in which 73% of patients with breast carcinoma had a value of less than 5.0 ng/ml. The two patients with metastat¬ ic gastrointestinal adenocarcinoma, both demonstrated plasma CEA lev¬ els well above the range of the 58 pa¬ tients mentioned thus far. Although we had no patient with metastatic
to the eye from the lung, one extrapolate from published data. Seventy-two percent or more of such patients would be expected to have an elevated CEA level.8" In one study of 77 patients considered "inoperable" (ie, either locally advanced or dis¬ seminated disease), who might there¬ fore be expected to be candidates for cancer can
ocular métastases, a mean value of 7.8 ng/ml was found in the plasma. Plasma CEA levels would appear to be of value in the differential diag¬ nosis of uveal tumors. Values greater than 20 ng/ml are found predomi¬ nantly in patients with gastrointesti¬ nal adenocarcinoma (pancreatic and colorectal), while values up to 10 ng/ml could be found in a variety of patients with malignant tumors, in¬ flammations, and in an otherwise healthy smoking population.8 The sharply elevated CEA values of the two patients with metastatic colo¬ rectal adenocarcinoma to the choroid are consistent with those reported in patients with pancreatic or colorectal carcinoma or generalized metastatic
disease.8 The ocular tumors from these patients stained positively for CEA by the immunohistologic tech¬ nique of Hsu,'- indicating that the ocular tumors in both of these entodermally derived cases were possibly contributing to the CEA-secreting tu¬ mor burden.1-·" Neither the eye from the patient with metastatic breast carcinoma (with normal plasma CEA level), nor those of any of three pa¬ tients with choroidal melanomas tested (two with elevated plasma CEA) gave positive results for the presence of CEA by the immuno¬ histologic technique. This raises the question of the source of this antigen in the plasma of patients with pri¬ mary uveal melanoma. We have not been able to demonstrate CEA by this method in tumor tissue from retinoblastoma patients with elevated plasma CEA, although the plasma values decrease with removal of the tumor burden." One of the most difficult diagnostic problems occasionally encountered by the ophthalmologist is in differ-
entiating an amelanotic melanoma or choroidal hemangioma from a meta¬
static tumor to the choroid from an occult primary.1"" The use of fluores¬ cein angiography, radioactive phos¬ phorus uptake studies, and other mo¬ dalities have definite diagnostic limitations.'"·'" Plasma CEA determi¬ nation may play a future role as a possible diagnostic adjunct in such cases, especially in cases of gastroin¬ testinal malignant tumors.
This investigation was supported in part by Public Health Service training grant EY-00084 and General Research Support grant 5S01RR05510 of the National Eye Institute, and by the Retina Research and Development Foun¬ dation of Philadelphia, and the Lions Club of Pennsylvania. Mr. Leonard Foster gave technical assistance. H.J. Hansen, PhD, J.P. Vandevoorde, PhD, and J. Primus, PhD, of Hoffmann-La Roche, Nutley, N.J. supplied reagents employed in these studies and performed the CEA assays. H. Robbin, MD, Department of Pathology, Thomas Jefferson Hospital, Philadelphia supplied tissue. The mem¬ bers of the Retina Service of the Wills Eye Hospital, Philadelphia, allowed us to examine their patients.
References 1. Ferry AP, Font FL: Carcinoma metastatic to the eye and orbit. Arch Ophthalmol 92:276-286,
1974. 2. Gold
mor
P, Freedman SD: Demonstration of tuspecific antigens in human colonic carcino-
by immunological tolerance and absorption techniques. J Exp Med 121:439-459, 1965. 3. Gold P, Freedman SD: Specific carcinoembryonic antigens of the human digestive system. J Exp Med 122:467-481, 1965. 4. Thomson DMP, Krupey J, Freedman SD, et al: The radioimmunoassay of circulating carcinoembryonic antigen of the human digestive mata
system. Proc Natl Acad Sci 64:161-167, 1969. 5. Hansen HJ, Lance KP, Krupey J: Demon-
an ion sensitive antigenic site on carcinoembryonic antigen using zirconyl phosphate gel. Clin Res 19:143, 1971. 6. Groover JR, Rogers AI: Immunologic tests for the detection of gastrointestinal cancers: Status report on carcinoembryonic antigen
stration of
(CEA) and alpha-fetoprotein (AFP). So
Med J
66:1218-1221, 1973.
7. Reynoso G, Chu FM, Holyoke TM, et al: Carcinoembryonic antigens in patients with dif-
ferent cancers. JAMA 22:361-365, 1972. 8. Hansen JH, Snyder JJ, Miller E, et al: Carcinoembryonic antigen (CEA) assay: A laboratory adjunct in the diagnosis and management of cancer. Human Pathol 5:139-147, 1974. 9. CEA-Roche, Carcinoembryonic Antigen Assay: An in vitro test to aid in the management and diagnosis of cancer: in A Clinical Monograph, Nutley, NJ, Roche Clinical Laboratories, Inc, 1974. 10. Shields JA, McDonald PR: Improvements in the diagnosis of posterior uveal melanoma. Arch Ophthalmol 91:259-264, 1974. 11. Vincent RG, Chu TM: Carcinoembryonic antigen in patients with carcinoma of the lung. J Thorac Cardiovasc Surg 66:320-328, 1973. 12. Hsu KC, Zimmerman EA, Rodin L, et al:
Downloaded From: http://archopht.jamanetwork.com/ by a University of Iowa User on 06/17/2015
Application of tumor tissues with antibody to tumor associated antigen (TAA), in Embryonic and Fetal Antigens in Cancer: Proceedings of the 2nd Conference, Oak Ridge National Laboratory, Oak Ridge, Tenn, 1972, vol 2. 13. Michelson JB, Felberg NT, Shields JA, et al: Carcinoembryonic antigen-positive metastatic adenocarcinoma of the choroid. Arch Ophthalmol 93:794-796, 1975. 14. Michelson JB, Felberg NT, Shields JA: Fetal antigens in retinoblastoma. Cancer, to be
published.
15. Shields JA, Zimmerman LE: Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 89:466-471, 1973. 16. Shields JA, Hagler WS, Federman JL, et al: The significance of the 32P uptake test in the diagnosis of posterior uveal melanomas. Trans Am Acad Ophthalmol Otolaryngol, to be published.
