- Email: [email protected]
Second Nonocular Tumors in Retinoblastoma Survivors
Second Nonocular Tumors in Retinoblastoma Survivors Are They Radiation-induced? DAVID H. ABRAMSON, MD, ROBERT M. ELLSWORTH, MD, F. DAVID KITCHIN, MD, GEORGE TUNG, MD
Abstract: A review of 693 patients with bilateral retinoblastoma and 18 patients with unilateral/germinal retinoblastoma was carried out to find the incidence, time, course and pattern of second nonocular tumors in retinoblastoma survivors. Of 688 patients who survived therapeutic radiation for retinoblastoma, 89 developed second tumors: 62 in the field of radiation and 27 out of the field. Of 23 patients who received no radiation, five developed second tumors: one in the field and four out of the field. The most common tumor found was a sarcoma both in and out of the field of radiation. The incidence of second tumors increases with time, although the mean latent period is 10.4 years. At 10 years, the incidence of second tumors is 20%, at 20 years, it is 50% and at 30 years, 90%. The incidence of tumors in patients treated without radiation and where tumors developed outside the field was 10% at 10 years, 30% at 20 years and 68% at 32 years. There was no relationship between incidence of tumors and dose of therapeutic radiation when analyzed with life tables. A second course of radiation therapy did not increase the incidence of second nonocular tumors. [Key words: betatron, cancer, germinal mutation, orthovoltage, osteogenic sarcoma, radiation, retinoblastoma, second neoplasms, supervoltage.] · Ophthalmology 91:1351-1355, 1984
Retinoblastoma is a rare, malignant tumor of children with present survival rate from the disease in excess of 90%. Twenty-five percent of all patients with retinoblastoma are seen with bilateral disease, and 75% are seen with the disease limited to one eye. 1 The subsequent development of second nonocular t~mors in patients who survive retinoblastoma has been reported in many case reports and series. 2- 14 In our survey of over 2000 cases of retinoblastoma collected from our cases and cases on file at the Armed Forces Institute of Pathology, we noted that more than 98% of those patients who develop second nonocular tumors had originally had bilateral retinoblastoma despite the fact that bilateral retinoblastoma From the Ophthalmic Oncology Center, New York Hospital-Come!! Medical Center. New York, New York. Reprint requests to David H. Abramson, MD, 70 East 66th Street, New York, New York 10021.
was less common in the entire group of patients. 11 These tumors have frequently been attributed to radiation, since bilateral retinoblastoma patients are usually treated with radiation, while unilateral patients usually receive no radiation.15-17 In fact, Sagerman et al demonstrated a doseresponse curve for incidence and radiation dose. The higher the dose given to the ocular tumor, the greater the subsequent incidence of second tumors. 9 t~owever, recent observations have cast serious doubts on this clean-cut relationship between retinoblastoma and the subsequent development of second tumors. For example, we described patients with bilateral retinoblastoma who developed nonocular tumors at sites far distant from the orbital radiation, eg. osteogenic sarcoma of the femur.11·12 One patient with bilateral disease was found to develop a second nonocular tumor of the humerus after treatment with only a localized cobalt plaque to one eye. 13 Three patients with bilateral disease were found in our series to have developed second tumors despite the fact 1351
OPHTHALMOLOGY
•
NOVEMBER 1984
Table 1. Patients with Retinoblastoma No. Cases Bilateral Unilateral
693
18
No. Second Tumors
No. in the Field
No. Out of Field
89 5*
58 2
31
3
* Includes three patients without family history of retinoblastoma.
that they received no radiation. 12 One of these was in the temporal bones and would have been called radiation induced had the patient received any radiation. Additional analysis of our series revealed that the latent periods for the development of the second ocular tumors were not always long. In fact, second nonocular tumors were seen within one year of the diagnosis and treatment of retinoblastoma. The true incidence of this problem is unknown. We have previously shown that 10 to 15% of patients who survived bilateral retinoblastoma have developed second nonocular tumors. But because of the very variable lengths of follow-up on these patients (less than 1 year to greater than 30 years), and be<.:;ause it appears that the longer the follow-up, the higher the incidence of second tumors truely incidence figures can only be attaiq~d with lif~ table analysis. 18 All prior studies but one 16 have neglected to recognize the fact that the patients with the longest follow-up are also the patients who have received the highest doses of radiation. Radiation doses of over 15,000 rad were common thirty years ago and, therefore, have follow-ups of thirty years. However, radiation dose of 3,500 rad have only been given in more recent time and with shorter follow-ups. Thus, we desired to determine whether it was the length of follow-up or the radiation dose that explains the difference in incidence figures. To answer _this question, we performed life table analysis on our patients to determine the true incidence time course and relationship of second non ocular tumo;s to r~diation therapy.
MATERIALS AND METHODS A retrospective analysis of all cases treated through our group and on file at the Ophthalmic Oncology Center of the New York Hospital-Cornell Medical Center was performed. Any patient with a second tumor who was not originall)' tr~ated by us was deleted from this study. Life table analysis of our cases using the technique recommended by Newell and Schullman was done. In particular, whether the patient was unilateral or bilateral, the year of treatment, the duration of follow-up, the total radiation dose delivered, the location of the second tumor the histopathology of this second tumor and the sex of the patient was recorded. The comparison of life tables was performed on a computer employing the Mantei-Hanzel test. 1·9 For the purpose of this study, all tumors in the field of radiation were so classified if they appeared to be orig1352
•
VOLUME 91
•
NUMBER 11
inating in the lids, orbits, periocular sinuses, temporal bones or skin overlying the temporal bone region. Tumors classified as out of the field including all tumors at distant sites such as femur, humerus, etc., and including those cases originating in the thyroid and pineal gland.
RESULTS Reviewed were 1512 cases of retinoblastoma. Bilateral disease was present in 893 patients, and unilateral in 619. Of patients who survived retinoblastoma, 693 bilateral retinoblastoma cases contained sufficient information to be included in this study. Eighteen patients with unilateral retinoblastoma were found to have a family history of retinoblastoma. . Of the 693 patients with bilateral retinoblastoma, 89 developed second tumors, 58 were in the field of radiation and 31 out of the field of radiation. Of the 18 unilateral patients, five second tumors were found, two in the field of radiation and three out ·of the field of radiation (Table 1). Ninety-four second tumors were found; 50 (53%) were male and 44 (47%) were female. A family history of retinoblastoma was positive in 17 of the 94 patients with bilateral disease. Of the five unilateral patients we found with second tumors only two had a positive family history of retinoblastoma. Three patients were found with unilateral retinoblastoma with no family history but with second tumors. The age of diagnosis of the original retinoblastoma varied from 3 days to 6 years (mean, 13 months). The latent period, calculated from the date of the completion of radiotherapy to the date of second tumor diagnosis, ranged from 10 months to 23 years, 9 months (mean, 10.4 years). Of patients treated with radiation, 62 cases of second tumors occurred within the radiation field. Twenty-seven occurred outside the field. Of patients treated without radiation, four developed them away from the head; one developed a tumor in the temporal bone (Table 2). The most common histology of second tumors was osteosarcoma: of the 38 cases found, 26 were in the radiation field and 12 were out of the field. The latent period for the development of osteosarcoma ranged from 1 year, 1 month to 22 years, 9 months. Of osteosarcomas out of the field above, the age at detection was 5 years to 12 years, 2 months. Table 2. Treatment of Patients No. Cases Radiation No Radiation
688 23
No. Second Tumors
No. In the Field
No. Out of Field
89
62
27* 4
5
1
* Includes one case treated with cobalt plaque who developed tumor in humerus.
ABRAMSON, et al.
•
RETINOBLASTOMA SURVIVORS
The next most common second tumor type was soft tissue sarcoma: of the 23 cases found, 22 were in the field and I out of the field. The latent period ranged from 4 years 5 months to 30 years. The distribution of pathologic types in and out of the field are presented in Tables 3 and 4. Of the 94 cases of second tumors, 89 had been treated with radiation; 5 received no radiation, 37 were treated with orthovoltage only (3,500 rad-26,000 rad). Thirtyfive were treated with betatron alone (3,500 rad to 12,000 rad) and 15 were treated with a combination of orthovoltage, betatron, linear accelerator and localized cobalt plaques. One case was treated with radium rods while one was treated with cobalt plaques only. The core of the data has been presented in life tables. Figure 1 shows the overall life table pattern of second tumors for all bilateral retinoblastoma patients, including those who did and did not receive radiation. Figure 2 shows the life table for patients with the germinal mutation who were treated without radiation, as well as those who developed tumors out of the field with radiation. Figure 3 compares all second tumors that developed in the field of radiation to those that developed out of the field of radiation (x 2 = 7.884; P = 0.005 statistically significant). Figure 4 represents the life tables comparing patients who received one course ofbetatron radiation (3,500 rad4,500 rad) with those who received multiple courses of betatron radiation (7,000 rad-9,000 rad) (x2 = 0.055; P = 0.815 not significant). Figure 5 represents the life tables for patients treated with stated doses of orthovoltage radiation of less than 11,000 rad with those of greater than 11 ,000 rad (x 2 = 0.043; P = 0.836 not significant). Table 3. Distribution of Pathologic Types in the Field Pathologic Type
No. Cases
Osteosarcoma Fibrosarcoma Soft Tissue Sarcoma Squamous Cell Carcinoma Rhabdomyosarcoma Anaplastic Tumor Unclassifiable Insufficient Data Glioblastoma Fibrous Histiocytoma Sarcoma Neuroblastoma Meningioma Angiosarcoma Mixed Parotid Lymphoma Mesenchymoma Papillary Carcinoma Neurofibroma
25*
Total * Includes one case without radiation.
6 5
3 3 3 2 2 .2 2 1 1
1 1 1 1
1 1 1 62
Table 4. Distribution of Pathologic Types out of the Field Pathologic Type
No. Cases
Osteosarcoma Malignant Melanoma Pinealoma Papillary Thyroid Carcinoma Ewings Sarcoma Testicular Carcinoma Liposarcoma Hodgkins Leukemia , Unclassitiable Insufficient data Astrocytoma Wilms Neuroepithelioma Adenocarcinoma (Breast)
12
4
3 2 2 1 1
1 1
1 1
1
1 1 1
Total
33
Includes four unilateral cases with second tumors.
DISCUSSION Prior studies and case reports of second nonocular tumors in retinoblastoma survivors have noted that the patient with bilateral retinoblastoma is the one at risk for the later development of second nonocular tumors. For example, in 1976 we reported that 97.5% ofthe patients who developed second tumors had been treated for bilateral retinoblastoma. 11 In the present series, 89 of 94 patients who developed second tumors had been treated for bilateral retinoblastoma. The retinoblastoma gene is intimately related to the proclivity for the development of second tumors. Fifteen percent of patients with unilateral retinoblastoma also harbor the germinal mutation. They are easily identified when there is a family history of retinoblastoma. Indeed, in this series of 18 patients with unilateral retinoblastoma with a family history, two developed second nonocular tumors. Unilateral retinoblastoma patients with the germinal mutation present at an earlier age of diagnosis and usually
"'0
100
All second tumors
E 80
~
'0
c
.,
60
~
40
c.,
20
0 0
"'
'i
.,~
0..
0
2
4
6
8
10
12
14 16
18 20 22 24 26 28 30 32
Years after diagnosis
Fig 1. The overall life table pattern of second tumors for all bilateral retinoblastoma patients, including those who did and did not receive radiation.
1353
OPHTHALMOLOGY
.
100
~
80
l; E
•
Tumors without radiation and tumors out of the field with radiation
.
~
80
c:
60
.<:::
40
"i
100
•
8 Q)
"'
c:
20
~
Q)
Q)
a..
a.. 24 26 28 30 32
0
12
0
Years after diagnosis
0
E 80
~
0
E
BETATRON:
~
---Single course --Multiple course
c: 0 <>
"'
.<:::
"i
c:
60
Q)
.,
60 40 20
Q)
a..
0
40
c
20
Q)
0
--
/
6
/
/
10
12
14
16
18
Years after diagnosis Fig 4. Life tables comparing patients who received one course of betatron radiation (3500 rad-4500 rad) with those who received multiple courses of betatron radiation (7000 rad-9000 rad).
with multiple tumors. 17 The mean age for unilateral patients with the germinal mutation is 2.4 months, versus 24 months for unilateral patients without the germinal mutation. A review of our five unilateral patients who developed second nonocular tumors is instructive: Patient #47 developed an undifferentiated tumor of the cheek and forehead after radiation treatment for unilateral retinoblastoma. She was the third family member in three generations to have retinoblastoma. Patient #81 was treated with unilateral enucleation without radiation and developed a cutaneous malignant melanoma of the chest wall 22 years later. She had a family history of retinoblastoma. Therefore, unilateral patients who have a family history of retinoblastoma are at risk for the development of second non ocular tumors whether they receive therapeutic radiation or not. Patient #49 developed a sphenoid ridge meningioma 19 years 3 months after 3000 rad for unilateral retina1354
4
14
16 18 20 22 24 26 28 30 32
Fig S. Life tables for patients treated with stated doses of orthovoltage radiation ofless than 11,000 rad, compared with those receiving greater than 11,000 rad .
r-"'' 8
2
Years after diagnosis
I I I
~
.... a..
rods
~
(.) Q)
Q) (.)
-~11,000
Q)
"0
-
80
ORTHOVOLTAGE: - - - < 11,000 rods
"0
::J
..r;
16 18 20 22 24 26 28 30 32
Fig 3. Comparison of second tumors that developed in the field of radiation to those that developed out of the field of radiation.
100
100
14
Years after diagnosis
Fig 2. The life tables for patients with the germinal mutation (all bilateral and unilateral patients with a family history) who did not receive therapeutic radiation, or who developed tumors outside the field after therapeutic radiation.
c
'
----'
60 40
Q)
~
0
,
---All tumors in field - - A l l tumors out of field
.<:::
"i
~ 20
.,....
NUMBER 11
"0
8 Q)
.
l; E
"0
c:
VOLUME 91
NOVEMBER 1984 •
blastoma. Although there was no family history of retinoblastoma, the child presented at 6 months of age. Thus, it is probable that this unilaterally affected child also had the germinal mutation for retinoblastoma. Patient #27 had a radium tray held near the side of her face at 3 years of age and subsequently developed an adenocarcinoma of the thyroid. The sensitivity of the thyroid gland to radiation is well-known; we believe that this tumor was radiation induced. There was no family history of retinoblastoma. Patient #71 presented at the age of 11h with unilateral retinoblastoma, and did not have a family history of retinoblastoma; she was treated with an enucleation alone, and 91f2 years later developed the most common malignancy of childhood, leukemia. This child is the only unilaterally affected child without the germinal mutation in our experience to develop a second tumor. Our registry contains almost 700 unilaterally affected individuals, and we believe that in this case the leukemia probably represents no association with retinoblastoma. Indeed, it is striking that of all patients with retinoblastoma who develop second tumors, leukemia is very unusual, and tumors which are less common in the general population (eg. osteogenic sarcoma) are more common. Having demonstrated the importance of the retinoblastoma gene for the development of the second tumors we were interested in determining the impact of radiation
ABRAMSON, et al.
•
RETINOBLASTOMA SURVIVORS
in the genesis of second tumors in this group of patients. Unfortunately for this analysis, more than 95% of all bilaterally affected patients have received therapeutic radiation. A life table analysis was generated using the second tumor patients with the germinal mutation who had been treated without employing radiation, as well as those who developed tumors out of the field of ocular therapeutic radiation. The incidence of tumors in this group increases with time; at 10 years it is 10%, by 20 years it is almost 30%, and by 32 years it is 68%. Thus, without therapeutic radiation or chemotherapy, seven often children with the germinal mutation for retinoblastoma develop second tumors. It is of interest that of the patients who received no radiation, one of five developed a tumor in the temporal bone which would have been classified as radiation induced, had the patient received radiation. For the purpose of this study, we assumed that all tumors that appeared in the field of radiation were radiation induced, and that all tumors that appeared out of the field were not radiation induced. The comparison of those life tables for tumors induced by radiation to those unrelated to radiation (comprised of those tumors out of the field and all tumors in non-irradiated germinal patients) was statistically significant toP= 0.001. Thus, the effect of radiation is to increase the total incidence of second tumors above the already high incidence in these patients, had they received no radiation. Sagerman has published a dose response curve for bilateral retinoblastoma patients suggesting that higher doses of therapeutic radiation cause more second tumors to occur. He demonstrated that at a tumor dose of under 6000 rad, 2.5% of the patients developed second tumors; at doses of 6000 to 10,999 rad, 5.5% developed second tumors; at doses greater than 11 ,000 rad, 32% developed a second tumor. Unfortunately, the follow-up periods for these groups differed. The patients who had been treated with more than 11 ,000 rad were the original group of retinoblastoma patients who had the longest follow-up, while those under 6000 rad had the shorter follow-up. Life table analysis performed by us demonstrates that there is no significant difference in the group treated with less than 11,000 rad when compared to those treated with more than 11,000 rad. Thus, the curves generated by Sagerman emphasize that the groups with longer followup have the highest incidence of second nonocular tumors, not that the higher radiation dose is associated with the second nonocular tumors (Fig 4). Both curves demonstrate the alarming increasing evidence of second tumors as shown with longer follow-up. After enucleating one eye and radiating the second eye with 3500 rad ofbetatron radiation, we have used a second course of 3500 rad months or years later, in a desperate
attempt to salvage vision in the only eye. A comparison of patients treated with the betatron with one course of radiation to those with multiple courses of radiation reveals no statistically increased incidence of second tumors.
REFERENCES 1. Abramson DH. Retinoblastoma: diagnosis and management CA 1982; 32:130-40. 2. Jensen RD, Miller RW. Retinoblastoma: epidemiologic characteristics. N Engl J Med 1971; 285:307-11. 3. Reese AB, Merriam GR Jr, Martin HE. Treatment of bilateral retinoblastoma by irradiation and surgery. Report on 15-year results. Am J Ophthalmol 1949; 32:175-90. 4. Zimmerman LE, Ingalls R. Clinical pathologic conference. Am J Ophthalmol1957; 43:417-26. 5. Forrest AW. Tumors following radiation about the eye. Trans Am Acad Ophthalmol Otolaryngol1961; 65:694-717. 6. Tebbe! RD, Vickery RD. Osteogenic sarcoma following irradiation for retinoblastoma, with report of a case. Am J Ophthalmol1952; 35:811-8. 7. Raivio I, Tarkkanen A. Sarcoma following radiation for retinoblastoma. Acta Ophthalmol 1965; 43:428-9. 8. Soloway HB. Radiation-induced neoplasms following curative therapy for retinoblastoma. Cancer 1966; 19: 1984-8. 9. Sagerman RH, Cassady JR, Tretter P, Ellsworth RM. Radiation induced neoplasia following external beam therapy for children with retinoblastoma. Am J Roentgenol Radium Ther Nucl Med 1969; 105:52935. 10. Shah IC, Arlen M, Miller T. Osteogenic sarcoma developing after radiotherapy for retinoblastoma. Arn Surg 1974; 40:485-90. 11. Abramson DH, Ellsworth RM, Zimmerman LE. Nonocular cancer in retinoblastoma survivors. Trans Am Acad Ophthalmol Otolaryngol 1976; 81:454-7. 12. Abramson DH, Renner HJ, Ellsworth RM. Second tumors in nonirradiated bilateral retinoblastoma. Am J Ophthalmol 1979; 87:624-7. 13. Abramson DH, Ellsworth RM, Kitchin FD. Osteogenic sarcoma of the humerus after cobalt plaque treatment for retinoblastoma. Am J Ophthalmol1980; 90:374-6. 14. Abramson DH, Ellsworth RM, Tretter P, et al. The treatment of bilateral groups I through Ill retinoblastoma with bilateral radiation. Arch Ophthalmol1981; 99:1761-2. 15. Abramson DH, Ellsworth RM, Tretter P, et al. Simultaneous bilateral radiation for advanced bilateral retinoblastoma. Arch Ophthalmol1981; 99:1763-6. 16. Abramson DH, Ellsworth RM, Rosenblatt M, et al. Retreatment of retinoblastoma with external beam irradiation. Arch Ophthalmol1982; 100:1257-60. 17. Abramson DH, Marks RF, Ellsworth RM, et al. The management of unilateral retinoblastoma without primary enucleation. Arch Ophthalmol 1982; 100:1249-52. 18. Merrell M, Shulman LE. Determination of prognosis in chronic disease, illustrated by systemic lupus erythematosis. J Chron Dis 1955; 1:1232. 19. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50:16370.
1355
Abstract: A review of 693 patients with bilateral retinoblastoma and 18 patients with unilateral/germinal retinoblastoma was carried out to find the incidence, time, course and pattern of second nonocular tumors in retinoblastoma survivors. Of 688 patients who survived therapeutic radiation for retinoblastoma, 89 developed second tumors: 62 in the field of radiation and 27 out of the field. Of 23 patients who received no radiation, five developed second tumors: one in the field and four out of the field. The most common tumor found was a sarcoma both in and out of the field of radiation. The incidence of second tumors increases with time, although the mean latent period is 10.4 years. At 10 years, the incidence of second tumors is 20%, at 20 years, it is 50% and at 30 years, 90%. The incidence of tumors in patients treated without radiation and where tumors developed outside the field was 10% at 10 years, 30% at 20 years and 68% at 32 years. There was no relationship between incidence of tumors and dose of therapeutic radiation when analyzed with life tables. A second course of radiation therapy did not increase the incidence of second nonocular tumors. [Key words: betatron, cancer, germinal mutation, orthovoltage, osteogenic sarcoma, radiation, retinoblastoma, second neoplasms, supervoltage.] · Ophthalmology 91:1351-1355, 1984
Retinoblastoma is a rare, malignant tumor of children with present survival rate from the disease in excess of 90%. Twenty-five percent of all patients with retinoblastoma are seen with bilateral disease, and 75% are seen with the disease limited to one eye. 1 The subsequent development of second nonocular t~mors in patients who survive retinoblastoma has been reported in many case reports and series. 2- 14 In our survey of over 2000 cases of retinoblastoma collected from our cases and cases on file at the Armed Forces Institute of Pathology, we noted that more than 98% of those patients who develop second nonocular tumors had originally had bilateral retinoblastoma despite the fact that bilateral retinoblastoma From the Ophthalmic Oncology Center, New York Hospital-Come!! Medical Center. New York, New York. Reprint requests to David H. Abramson, MD, 70 East 66th Street, New York, New York 10021.
was less common in the entire group of patients. 11 These tumors have frequently been attributed to radiation, since bilateral retinoblastoma patients are usually treated with radiation, while unilateral patients usually receive no radiation.15-17 In fact, Sagerman et al demonstrated a doseresponse curve for incidence and radiation dose. The higher the dose given to the ocular tumor, the greater the subsequent incidence of second tumors. 9 t~owever, recent observations have cast serious doubts on this clean-cut relationship between retinoblastoma and the subsequent development of second tumors. For example, we described patients with bilateral retinoblastoma who developed nonocular tumors at sites far distant from the orbital radiation, eg. osteogenic sarcoma of the femur.11·12 One patient with bilateral disease was found to develop a second nonocular tumor of the humerus after treatment with only a localized cobalt plaque to one eye. 13 Three patients with bilateral disease were found in our series to have developed second tumors despite the fact 1351
OPHTHALMOLOGY
•
NOVEMBER 1984
Table 1. Patients with Retinoblastoma No. Cases Bilateral Unilateral
693
18
No. Second Tumors
No. in the Field
No. Out of Field
89 5*
58 2
31
3
* Includes three patients without family history of retinoblastoma.
that they received no radiation. 12 One of these was in the temporal bones and would have been called radiation induced had the patient received any radiation. Additional analysis of our series revealed that the latent periods for the development of the second ocular tumors were not always long. In fact, second nonocular tumors were seen within one year of the diagnosis and treatment of retinoblastoma. The true incidence of this problem is unknown. We have previously shown that 10 to 15% of patients who survived bilateral retinoblastoma have developed second nonocular tumors. But because of the very variable lengths of follow-up on these patients (less than 1 year to greater than 30 years), and be<.:;ause it appears that the longer the follow-up, the higher the incidence of second tumors truely incidence figures can only be attaiq~d with lif~ table analysis. 18 All prior studies but one 16 have neglected to recognize the fact that the patients with the longest follow-up are also the patients who have received the highest doses of radiation. Radiation doses of over 15,000 rad were common thirty years ago and, therefore, have follow-ups of thirty years. However, radiation dose of 3,500 rad have only been given in more recent time and with shorter follow-ups. Thus, we desired to determine whether it was the length of follow-up or the radiation dose that explains the difference in incidence figures. To answer _this question, we performed life table analysis on our patients to determine the true incidence time course and relationship of second non ocular tumo;s to r~diation therapy.
MATERIALS AND METHODS A retrospective analysis of all cases treated through our group and on file at the Ophthalmic Oncology Center of the New York Hospital-Cornell Medical Center was performed. Any patient with a second tumor who was not originall)' tr~ated by us was deleted from this study. Life table analysis of our cases using the technique recommended by Newell and Schullman was done. In particular, whether the patient was unilateral or bilateral, the year of treatment, the duration of follow-up, the total radiation dose delivered, the location of the second tumor the histopathology of this second tumor and the sex of the patient was recorded. The comparison of life tables was performed on a computer employing the Mantei-Hanzel test. 1·9 For the purpose of this study, all tumors in the field of radiation were so classified if they appeared to be orig1352
•
VOLUME 91
•
NUMBER 11
inating in the lids, orbits, periocular sinuses, temporal bones or skin overlying the temporal bone region. Tumors classified as out of the field including all tumors at distant sites such as femur, humerus, etc., and including those cases originating in the thyroid and pineal gland.
RESULTS Reviewed were 1512 cases of retinoblastoma. Bilateral disease was present in 893 patients, and unilateral in 619. Of patients who survived retinoblastoma, 693 bilateral retinoblastoma cases contained sufficient information to be included in this study. Eighteen patients with unilateral retinoblastoma were found to have a family history of retinoblastoma. . Of the 693 patients with bilateral retinoblastoma, 89 developed second tumors, 58 were in the field of radiation and 31 out of the field of radiation. Of the 18 unilateral patients, five second tumors were found, two in the field of radiation and three out ·of the field of radiation (Table 1). Ninety-four second tumors were found; 50 (53%) were male and 44 (47%) were female. A family history of retinoblastoma was positive in 17 of the 94 patients with bilateral disease. Of the five unilateral patients we found with second tumors only two had a positive family history of retinoblastoma. Three patients were found with unilateral retinoblastoma with no family history but with second tumors. The age of diagnosis of the original retinoblastoma varied from 3 days to 6 years (mean, 13 months). The latent period, calculated from the date of the completion of radiotherapy to the date of second tumor diagnosis, ranged from 10 months to 23 years, 9 months (mean, 10.4 years). Of patients treated with radiation, 62 cases of second tumors occurred within the radiation field. Twenty-seven occurred outside the field. Of patients treated without radiation, four developed them away from the head; one developed a tumor in the temporal bone (Table 2). The most common histology of second tumors was osteosarcoma: of the 38 cases found, 26 were in the radiation field and 12 were out of the field. The latent period for the development of osteosarcoma ranged from 1 year, 1 month to 22 years, 9 months. Of osteosarcomas out of the field above, the age at detection was 5 years to 12 years, 2 months. Table 2. Treatment of Patients No. Cases Radiation No Radiation
688 23
No. Second Tumors
No. In the Field
No. Out of Field
89
62
27* 4
5
1
* Includes one case treated with cobalt plaque who developed tumor in humerus.
ABRAMSON, et al.
•
RETINOBLASTOMA SURVIVORS
The next most common second tumor type was soft tissue sarcoma: of the 23 cases found, 22 were in the field and I out of the field. The latent period ranged from 4 years 5 months to 30 years. The distribution of pathologic types in and out of the field are presented in Tables 3 and 4. Of the 94 cases of second tumors, 89 had been treated with radiation; 5 received no radiation, 37 were treated with orthovoltage only (3,500 rad-26,000 rad). Thirtyfive were treated with betatron alone (3,500 rad to 12,000 rad) and 15 were treated with a combination of orthovoltage, betatron, linear accelerator and localized cobalt plaques. One case was treated with radium rods while one was treated with cobalt plaques only. The core of the data has been presented in life tables. Figure 1 shows the overall life table pattern of second tumors for all bilateral retinoblastoma patients, including those who did and did not receive radiation. Figure 2 shows the life table for patients with the germinal mutation who were treated without radiation, as well as those who developed tumors out of the field with radiation. Figure 3 compares all second tumors that developed in the field of radiation to those that developed out of the field of radiation (x 2 = 7.884; P = 0.005 statistically significant). Figure 4 represents the life tables comparing patients who received one course ofbetatron radiation (3,500 rad4,500 rad) with those who received multiple courses of betatron radiation (7,000 rad-9,000 rad) (x2 = 0.055; P = 0.815 not significant). Figure 5 represents the life tables for patients treated with stated doses of orthovoltage radiation of less than 11,000 rad with those of greater than 11 ,000 rad (x 2 = 0.043; P = 0.836 not significant). Table 3. Distribution of Pathologic Types in the Field Pathologic Type
No. Cases
Osteosarcoma Fibrosarcoma Soft Tissue Sarcoma Squamous Cell Carcinoma Rhabdomyosarcoma Anaplastic Tumor Unclassifiable Insufficient Data Glioblastoma Fibrous Histiocytoma Sarcoma Neuroblastoma Meningioma Angiosarcoma Mixed Parotid Lymphoma Mesenchymoma Papillary Carcinoma Neurofibroma
25*
Total * Includes one case without radiation.
6 5
3 3 3 2 2 .2 2 1 1
1 1 1 1
1 1 1 62
Table 4. Distribution of Pathologic Types out of the Field Pathologic Type
No. Cases
Osteosarcoma Malignant Melanoma Pinealoma Papillary Thyroid Carcinoma Ewings Sarcoma Testicular Carcinoma Liposarcoma Hodgkins Leukemia , Unclassitiable Insufficient data Astrocytoma Wilms Neuroepithelioma Adenocarcinoma (Breast)
12
4
3 2 2 1 1
1 1
1 1
1
1 1 1
Total
33
Includes four unilateral cases with second tumors.
DISCUSSION Prior studies and case reports of second nonocular tumors in retinoblastoma survivors have noted that the patient with bilateral retinoblastoma is the one at risk for the later development of second nonocular tumors. For example, in 1976 we reported that 97.5% ofthe patients who developed second tumors had been treated for bilateral retinoblastoma. 11 In the present series, 89 of 94 patients who developed second tumors had been treated for bilateral retinoblastoma. The retinoblastoma gene is intimately related to the proclivity for the development of second tumors. Fifteen percent of patients with unilateral retinoblastoma also harbor the germinal mutation. They are easily identified when there is a family history of retinoblastoma. Indeed, in this series of 18 patients with unilateral retinoblastoma with a family history, two developed second nonocular tumors. Unilateral retinoblastoma patients with the germinal mutation present at an earlier age of diagnosis and usually
"'0
100
All second tumors
E 80
~
'0
c
.,
60
~
40
c.,
20
0 0
"'
'i
.,~
0..
0
2
4
6
8
10
12
14 16
18 20 22 24 26 28 30 32
Years after diagnosis
Fig 1. The overall life table pattern of second tumors for all bilateral retinoblastoma patients, including those who did and did not receive radiation.
1353
OPHTHALMOLOGY
.
100
~
80
l; E
•
Tumors without radiation and tumors out of the field with radiation
.
~
80
c:
60
.<:::
40
"i
100
•
8 Q)
"'
c:
20
~
Q)
Q)
a..
a.. 24 26 28 30 32
0
12
0
Years after diagnosis
0
E 80
~
0
E
BETATRON:
~
---Single course --Multiple course
c: 0 <>
"'
.<:::
"i
c:
60
Q)
.,
60 40 20
Q)
a..
0
40
c
20
Q)
0
--
/
6
/
/
10
12
14
16
18
Years after diagnosis Fig 4. Life tables comparing patients who received one course of betatron radiation (3500 rad-4500 rad) with those who received multiple courses of betatron radiation (7000 rad-9000 rad).
with multiple tumors. 17 The mean age for unilateral patients with the germinal mutation is 2.4 months, versus 24 months for unilateral patients without the germinal mutation. A review of our five unilateral patients who developed second nonocular tumors is instructive: Patient #47 developed an undifferentiated tumor of the cheek and forehead after radiation treatment for unilateral retinoblastoma. She was the third family member in three generations to have retinoblastoma. Patient #81 was treated with unilateral enucleation without radiation and developed a cutaneous malignant melanoma of the chest wall 22 years later. She had a family history of retinoblastoma. Therefore, unilateral patients who have a family history of retinoblastoma are at risk for the development of second non ocular tumors whether they receive therapeutic radiation or not. Patient #49 developed a sphenoid ridge meningioma 19 years 3 months after 3000 rad for unilateral retina1354
4
14
16 18 20 22 24 26 28 30 32
Fig S. Life tables for patients treated with stated doses of orthovoltage radiation ofless than 11,000 rad, compared with those receiving greater than 11,000 rad .
r-"'' 8
2
Years after diagnosis
I I I
~
.... a..
rods
~
(.) Q)
Q) (.)
-~11,000
Q)
"0
-
80
ORTHOVOLTAGE: - - - < 11,000 rods
"0
::J
..r;
16 18 20 22 24 26 28 30 32
Fig 3. Comparison of second tumors that developed in the field of radiation to those that developed out of the field of radiation.
100
100
14
Years after diagnosis
Fig 2. The life tables for patients with the germinal mutation (all bilateral and unilateral patients with a family history) who did not receive therapeutic radiation, or who developed tumors outside the field after therapeutic radiation.
c
'
----'
60 40
Q)
~
0
,
---All tumors in field - - A l l tumors out of field
.<:::
"i
~ 20
.,....
NUMBER 11
"0
8 Q)
.
l; E
"0
c:
VOLUME 91
NOVEMBER 1984 •
blastoma. Although there was no family history of retinoblastoma, the child presented at 6 months of age. Thus, it is probable that this unilaterally affected child also had the germinal mutation for retinoblastoma. Patient #27 had a radium tray held near the side of her face at 3 years of age and subsequently developed an adenocarcinoma of the thyroid. The sensitivity of the thyroid gland to radiation is well-known; we believe that this tumor was radiation induced. There was no family history of retinoblastoma. Patient #71 presented at the age of 11h with unilateral retinoblastoma, and did not have a family history of retinoblastoma; she was treated with an enucleation alone, and 91f2 years later developed the most common malignancy of childhood, leukemia. This child is the only unilaterally affected child without the germinal mutation in our experience to develop a second tumor. Our registry contains almost 700 unilaterally affected individuals, and we believe that in this case the leukemia probably represents no association with retinoblastoma. Indeed, it is striking that of all patients with retinoblastoma who develop second tumors, leukemia is very unusual, and tumors which are less common in the general population (eg. osteogenic sarcoma) are more common. Having demonstrated the importance of the retinoblastoma gene for the development of the second tumors we were interested in determining the impact of radiation
ABRAMSON, et al.
•
RETINOBLASTOMA SURVIVORS
in the genesis of second tumors in this group of patients. Unfortunately for this analysis, more than 95% of all bilaterally affected patients have received therapeutic radiation. A life table analysis was generated using the second tumor patients with the germinal mutation who had been treated without employing radiation, as well as those who developed tumors out of the field of ocular therapeutic radiation. The incidence of tumors in this group increases with time; at 10 years it is 10%, by 20 years it is almost 30%, and by 32 years it is 68%. Thus, without therapeutic radiation or chemotherapy, seven often children with the germinal mutation for retinoblastoma develop second tumors. It is of interest that of the patients who received no radiation, one of five developed a tumor in the temporal bone which would have been classified as radiation induced, had the patient received radiation. For the purpose of this study, we assumed that all tumors that appeared in the field of radiation were radiation induced, and that all tumors that appeared out of the field were not radiation induced. The comparison of those life tables for tumors induced by radiation to those unrelated to radiation (comprised of those tumors out of the field and all tumors in non-irradiated germinal patients) was statistically significant toP= 0.001. Thus, the effect of radiation is to increase the total incidence of second tumors above the already high incidence in these patients, had they received no radiation. Sagerman has published a dose response curve for bilateral retinoblastoma patients suggesting that higher doses of therapeutic radiation cause more second tumors to occur. He demonstrated that at a tumor dose of under 6000 rad, 2.5% of the patients developed second tumors; at doses of 6000 to 10,999 rad, 5.5% developed second tumors; at doses greater than 11 ,000 rad, 32% developed a second tumor. Unfortunately, the follow-up periods for these groups differed. The patients who had been treated with more than 11 ,000 rad were the original group of retinoblastoma patients who had the longest follow-up, while those under 6000 rad had the shorter follow-up. Life table analysis performed by us demonstrates that there is no significant difference in the group treated with less than 11,000 rad when compared to those treated with more than 11,000 rad. Thus, the curves generated by Sagerman emphasize that the groups with longer followup have the highest incidence of second nonocular tumors, not that the higher radiation dose is associated with the second nonocular tumors (Fig 4). Both curves demonstrate the alarming increasing evidence of second tumors as shown with longer follow-up. After enucleating one eye and radiating the second eye with 3500 rad ofbetatron radiation, we have used a second course of 3500 rad months or years later, in a desperate
attempt to salvage vision in the only eye. A comparison of patients treated with the betatron with one course of radiation to those with multiple courses of radiation reveals no statistically increased incidence of second tumors.
REFERENCES 1. Abramson DH. Retinoblastoma: diagnosis and management CA 1982; 32:130-40. 2. Jensen RD, Miller RW. Retinoblastoma: epidemiologic characteristics. N Engl J Med 1971; 285:307-11. 3. Reese AB, Merriam GR Jr, Martin HE. Treatment of bilateral retinoblastoma by irradiation and surgery. Report on 15-year results. Am J Ophthalmol 1949; 32:175-90. 4. Zimmerman LE, Ingalls R. Clinical pathologic conference. Am J Ophthalmol1957; 43:417-26. 5. Forrest AW. Tumors following radiation about the eye. Trans Am Acad Ophthalmol Otolaryngol1961; 65:694-717. 6. Tebbe! RD, Vickery RD. Osteogenic sarcoma following irradiation for retinoblastoma, with report of a case. Am J Ophthalmol1952; 35:811-8. 7. Raivio I, Tarkkanen A. Sarcoma following radiation for retinoblastoma. Acta Ophthalmol 1965; 43:428-9. 8. Soloway HB. Radiation-induced neoplasms following curative therapy for retinoblastoma. Cancer 1966; 19: 1984-8. 9. Sagerman RH, Cassady JR, Tretter P, Ellsworth RM. Radiation induced neoplasia following external beam therapy for children with retinoblastoma. Am J Roentgenol Radium Ther Nucl Med 1969; 105:52935. 10. Shah IC, Arlen M, Miller T. Osteogenic sarcoma developing after radiotherapy for retinoblastoma. Arn Surg 1974; 40:485-90. 11. Abramson DH, Ellsworth RM, Zimmerman LE. Nonocular cancer in retinoblastoma survivors. Trans Am Acad Ophthalmol Otolaryngol 1976; 81:454-7. 12. Abramson DH, Renner HJ, Ellsworth RM. Second tumors in nonirradiated bilateral retinoblastoma. Am J Ophthalmol 1979; 87:624-7. 13. Abramson DH, Ellsworth RM, Kitchin FD. Osteogenic sarcoma of the humerus after cobalt plaque treatment for retinoblastoma. Am J Ophthalmol1980; 90:374-6. 14. Abramson DH, Ellsworth RM, Tretter P, et al. The treatment of bilateral groups I through Ill retinoblastoma with bilateral radiation. Arch Ophthalmol1981; 99:1761-2. 15. Abramson DH, Ellsworth RM, Tretter P, et al. Simultaneous bilateral radiation for advanced bilateral retinoblastoma. Arch Ophthalmol1981; 99:1763-6. 16. Abramson DH, Ellsworth RM, Rosenblatt M, et al. Retreatment of retinoblastoma with external beam irradiation. Arch Ophthalmol1982; 100:1257-60. 17. Abramson DH, Marks RF, Ellsworth RM, et al. The management of unilateral retinoblastoma without primary enucleation. Arch Ophthalmol 1982; 100:1249-52. 18. Merrell M, Shulman LE. Determination of prognosis in chronic disease, illustrated by systemic lupus erythematosis. J Chron Dis 1955; 1:1232. 19. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50:16370.
1355