second malignant neoplasms among long-term survivors of … · second malignant neoplasms after...

(CANCER RESEARCH 56. 1564-1570. April 1. I996|

Second Malignant Neoplasms among Long-Term Survivors of Ovarian Cancer

Lois B. Travis,1 Rochelle E. Curtis, John D. Boice, Jr., Charles E. Platz, Benjamin F. Hankey, and

Joseph F. Fraumeni, Jr.Division of Cancer Epidemu>l(}g\ and Genetics [L. B. T.. R. E. C., J. D. B., J. F. F.Â¡,und Cancer Statistics Branch, Sun-eillancc Program, Division of Cancer Prevention

and Control Â¡B.F. H.I, National Cancer Institute, National Institutes of Health. Bethesda. Maryland 20892: and Department of Pathology, University of Iowa, Iowa City, Iowa1C. E. P.I

ABSTRACT

Second malignant neoplasms were evaluated among 32,251 women withovarian cancer, including 4,402 10-year survivors, within the nine population-based registries of the Surveillance, Epidemiology, and End ResultsProgram of the National Cancer Institute (1973-1992) and the Connecticut Tumor Registry (1935-1972). Overall, 1,296 second cancers occurred

against 1,014 expected (observed/expected (O/E), 1.28; 95% confidenceinterval (Cl), 1.21-1.35]. Sites contributing 25 or more excess cancersincluded leukemia (O/E, 4.17; O, 111; 95% CI, 3.43-5.03) and malignancies of colon (O/E, 1.33; O, 188; 95% CI, 1.15-1.54), rectum (O/E, 1.43; O,76; 95% CI, 1.13-1.79), breast (O/E, 1.18; 0,404; 95%, CI 1.07-1.30), andbladder (O/E, 2.07; O, 65; 95% CI, 1.59-2.63). Ocular melanoma (O/E,4.45; O, 8; 95% CI, 1.92-8.77) was also significantly increased. Secondcancer risk was high during all follow-up intervals, and cumulative risk at

20 years was 18.2%, compared with a population expected risk of 11.5%.Statistically significant relationships existed between serous adenocarci-noma of the ovary and breast cancer (O/E, 1.29; 95% CI, 1.06-1.56) and

mucinous ovarian adenocarcinoma and rectal cancer (O/E, 1.95; 95% CI,1.09-3.22). Secondary leukemia appeared linked with antecedent chemo

therapy, whereas radiotherapy was associated with cancers of connectivetissue, bladder, and possibly pancreas. Genetic and reproductive factorspredisposing to ovarian cancer may have contributed to the elevated riskof breast and colorectal neoplasms and possibly ocular melanoma. Thus,excess malignancies following ovarian cancer represent complications ofcurative therapies and/or underlying susceptibility states that have etio-

logical and clinical ramifications.

INTRODUCTION

Cancer of the ovary is the second most common gynecologicalmalignancy in the United States, with an estimated 26,600 casesexpected to develop in 1995 (1). Survival for women with ovariancancer has improved significantly within recent decades (2), such thatthe occurrence of late effects has gained increasing clinical importance (3). Previous investigations have addressed the risk of subsequent neoplasms by type of primary treatment (4, 5) and histology ofovarian cancer (6). Leukemia has been convincingly linked with priorchemotherapy regimens for ovarian cancer (7. 8), but more recenttreatment options have not been evaluated. Although an excess risk ofbreast and colorectal neoplasms seemed associated with histologicaltype of ovarian cancer in one series (6), sparse numbers limitedinterpretation. In the current study, we quantify the risk of secondmalignancies among 32,251 women with ovarian cancer, including4.402 10-year survivors, reported to selected population-based regis

tries within the United States.

MATERIALS AND METHODS

We evaluated all women diagnosed with invasive ovarian cancer as a firstprimary malignancy who were reported to one of nine population-based

registries of the SEER2 program of the National Cancer Institute (197.1-1992)or to the CTR (1935-1972).' In addition to Connecticut. SEER areas include

Hawaii. Iowa. New Mexico, Utah, and the metropolitan regions of Atlanta(1975-1992), Detroit, San Francisco and Oakland, and Seattle and PugetSound (1974-1992). which together constitute -10% of the U.S. population.

Information routinely collected by all SEER registries includes patient demographic data, tumor histology, and vital status. Active tracing of all livingpatients involving patient or physician contact and data linkage with files thatprovide information on vital status are standard follow-up procedures.

SEER and CTR files also typically record from hospital charts the firstcourse of treatment administered to patients, based on one of several broaddesignations. We used this information to identify ovarian cancer patientstreated with chemotherapy and/or radiotherapy. The SEER Program and CTRdo not record therapy given after the initial course. Neither is informationavailable with regard to the specific drugs or dose schedules administered.Treatment of ovarian cancer typically includes surgical resection followed bychemotherapy and/or local radiation, depending on disease stage and site ofinvolvement (9). Surgical management varies from unilateral oophoreclomy orbilateral salpingo-oophorectomy and hysterectomy with omentectomy to pel

vic exenteration (10). In the 1960s, adjuvant chemotherapy frequently consisted of treatment with single alkylating agents, such as chlorambucil. cyclo-

phosphamide, melphalan. or triethylenelhiophosphoramide (thiotepa) (10).with treatment commonly continued until relapse. Melphalan was widely usedthroughout the 1970s, with combination chemotherapy regimens employingcyclophosphamide. Adriamycin. and cisplatin introduced in the latter part ofthe decade. By the early 1980s, platinum-based regimens were considered

optimal chemotherapy. External beam radiotherapy was used for decades aseither an adjuvant to surgery, treatment for nonresectable disease, or palliationof advanced ovarian cancer (10). Postoperative irradiation tor ovarian cancerhas generally consisted of pelvic or whole ahdominopelvic radiotherapy (II).in which fields extend from above the diaphragm to below the pelvic floor.Intraperitoneal instillation of radiocolloids has also been used (12).

SEER and CTR incidence files were searched for second malignant neoplasms that developed at least 2 months after a diagnosis of invasive ovariancancer. Subsequent diagnoses of female genital tract tumors were not includedin any analyses. Second cancer sites were grouped according to the WHOInternational Classification of Diseases for Oncology (13). The risk of secondcancers of the breast, colon, and rectum were stratified by histological categories of ovarian adenocarcinoma specified by Percy et al. (14) and Platz andBenda (15). For secondary colorectal cancer, latter analyses were limited to1-year survivors of ovarian carcinoma, given possible misclassification withmucin-producing adenocarcinomas during the patient's early course.

To estimate risk of second cancer, person-years of observation were com

piled according to age and calendar periods from 2 months after diagnosis ofovarian cancer to date of last follow-up evaluation, date of death, date of

diagnosis of second cancer, or end of study (December 31. 1992). whicheveroccurred first. By close of study. 65% of the women had died. 297r remainedalive, and 6% were lost to follow-up. Cancer incidence rates specific for age(within 5 years), sex, and 5-year calendar year intervals were multiplied by theaccumulated person-years at risk to estimate the number of cancer cases

expected. Statistical tests and 95% CIs were based on the assumption that the

Received 11/27/95; accepted 2/12/96.The co.sls of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicale Ibis fact.

1To whom requests for reprints should be addressed, at National Cancer Institute, EPN

408, Bethesda. MD 20892.

2 The abbreviations used are: SEER, .Surveillance. Epidemiology, und End Results:

CTR. Connecticut Tumor Registry; CI. confidence interval; O/E. observed/expected:ALL. acute lymphoblastic leukemia; ANLL, acute nonlymphocytie leukemia; NOS. nototherwise specified.

1Women diagnosed with ovarian carcinoma within selected SEER registries ( I9KO-

1982) were included in a prior case-control study (see Ref. 36). Patients reported to theCTR were included in previous reports (see Refs. 4 and 3D, with follow-up extended for

the current survey.

1564

Research. on August 15, 2021. © 1996 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

SECOND MALIGNANT NEOPLASMS AFTER OVARIAN CANCER

Table 1 Characteristics of women with ovarian cancer reponed to the CTR 11935-19721 or to the SEER program (1973-19921

All patients were diagnosed with ovarian cancer as a first primary cancer and survived ^2 months. The cohort consisted of 4.742 subjects reported to the CTR ( 1935-1972) and27,509 subjects reported to the SEER program ( 1973-19921. Follow-up of the cohort ended December 31. 1992.

HistologyAll

ovariancancerCarcinoma,all'Adenocarcinoma'Serous'Mucinous*"KndometroiffOther

specifiedtypes'1'NOS"Carcinoma.

other/NOS1Noncarcinoma,

alFNo.

ofpatients32.25130.25427,8877.6023.4242.2799,7774.8052,3671,997Averageage(yr)58.859.759.359.256.158.058.663.864.544.6Person-years

offollow-up133,098120.191113.84429.9402

1.47910,81239,79511,8186,34612,907Average

follow-up(yr)4.14.04.13.96.34.74.12.52.76.5No.of second

primarycancers"1,2961,2181,1603102261293721235878

" Numbers exclude all female genital sites.'' International Classification of Diseases (ICD)-O morphology codes 8010. 8012. 8020-8022. 8031-8033. 8041, 8050. 8052. 8070-8071. 8074. 8120. 8130, 8140-8141. 8230,

8260, 8262. 8310, 8313, 8320, 8323. 8380-8381, 8430. 8440-8442. 8450-8451. 8460-8462. 8470-8473. 8480-8481. 8490. 8504, 8510. 8560, 8570. 8930. 8933. 8940. 8950-8951.8980. 9000. 9013-9015. and 9110.

' ICD-0 morphology codes 8050, 8140-8141. 8260. 8262. 8310. 8313. 8320. 8323, 8380-8381. 8430. 8440-8442. 8450-8451. 8460-8462. 8470-8473. 8480-8481. 8490. 8504.

8560. 8570, 8930. 8933. 8940. 8950-8951, 8980, 9000. 9013-9015. and 91 10.rflCD-0 morphology codes 8441 and 8460-8461.'' ICD-0 morphology codes 8470-8471 and 8480-8481.

' ICD-0 morphology codes 8380-8381 and 8570.K ICD-0 morphology codes 8050. 8141. 8260, 8262, 8310, 8313. 8320, 8323, 8430, 8440. 8442. 8450-8451. 8462. 8472-8473, 8490. 8504. 8560. 8930. 8933. 8940. 8950-8951,

8980. 9000. 9013-9015. and 9110.* ICD-0 morphology code 8140.' ICD-0 morphology codes 8010. 8012. 8020-8022, 8031-8033, 8041, 8052, 8070-8071, 8074. 8120, 8130. 8230. and 8510.' ICD-0 morphology codes 8000-8002. 8004. 8240. 8243. 8245-8246. 8590. 8600. 8620-8622. 8630-8632. 8640, 8650. 8670, 8720, 8800-8801. 8810, 8840. 8890-8891. 8900.

8960, 8981. 8990. 9050. 9052-9054. 9060. 9064. 9070-9071. 9080-9085. 9090-9091. 9100-9101. 9364. 9500. and 9580.

observed numbers of second tumors were distributed as a Poisson variable.Tests for homogeneity and linear trend were conducted using methods described by Breslow et al. (16). All P values are two sided.

RESULTS

Average age at diagnosis for 32,251 women with invasive ovariancancer was 58.8 years, and average follow-up was 4.1 years (Table I ).

Approximately 4,400 women survived 10 or more years. Secondcancers occurred in 1,296 patients (4%). Diagnoses were microscop

ically confirmed in 98% and 97% of initial and subsequent tumors,respectively. Adenocarcinomas constituted 87% of the ovarian cancers, with the growth patterns of 7.602 tumors characterized as serousin morphology.

Altogether, 1296 women developed subsequent cancers, comparedwith 1013.85 expected (OIE, 1.28; 95% CI. 1.21-1.35: Table 2). Therisk of second cancer was slightly higher in the CTR, 1935-1972(OIE. 1.38). than in the SEER Program, 1973-1992 (OIE. 1.25).

Significantly increased risks were observed for all solid tumors

Table 2 Observed ami expected numbers of second malignan! neoplasms among women with ovarian cancer

SiteAllsecondcancers"Allsolidtumors"All

buccalHsophagusStomachSmall

intestineColonRectumLiver,

gallbladderPancreasLarynxLungBreastKidneyBladderMelanomaEye-Brain

und central nervoussystemThyroidBoneConnective

tissueNon-Hodgkin'slymphonuiHodgkin'sdiseaseMultiple

myelomaAllleukemia'Acute

lymphoblasticleukemiaChroniclymphocyticleukemiaAcute

nonlymphocyticleukemiaChronicgranulocyticleukemiaAll

other3Observed1.2961.13333525618876184521234042965258111319383111117784847Expected1,013.85931.7323.127.8826.003.29140.8553.1617.5534.695.66124.30341.3620.8731.4620.891.8013.3111.431.274.6737.153.7914.5526.601.009.329.483.8248.170/El^S*\.22b1.430.630.961.82Ã¬.lÃ¬h1.43*1.031.300.350.991.18''1.39107*1.204.45fc0.831.140.791.931.020.790.764.17''6.97''0.75S.S6h2.100.9895%

Cl1.21-1.351.15-1.290.98-2.000.20-1.480.62-1.420.67-3.971.15-1.541.13-1.790.61-1.620.95-1.740.04-1.280.82-1.181.07-1.300.93-2.001.59-2.630.77-1.771.92-8.770.41-1.480.60-1.940.01

-4.390.88-3.660.72-1.400.16-2.310.38-1.353.43-5.032.79-14.360.30-1.557.07-10.970.90-4.130.72-1.30

" Numbers exclude all female genital sites.h P < 0.05.

Includes Uve leukemias for which cell type was not specified.Includes all solid tumors not itemized within this table, i.e., 47 cancers of unknown (n

1565

= 36) or ill-defined (n â€”¿�II) primary sites.




Table 3 Selected second malignant neoplasms by site, treatment, and time since diagnosis of ovarian cancer

Time since diagnosis (yr)

No. of patients (all) entering intervalRadiotherapy only"Any chemotherapy '

Person-years'

Radiotherapy onlyAny chemotherapyl^t22.248

2.79611.58151.920

7,46722.7095-98,396

1.4393.041

30.7046.0049,6441(1-144.402

I.(XK)1,139

15.8514,1013,37015

+2,147

650316

12,8054.294

714Second

cancerAll

solidtumors'Radiotherapy

onlyAnychemotherapyColon,

allRadiotherapyonlyAny

chemotherapyRectum,allRadiotherapyonlyAny

chemotherapyPancreas,allRadiotherapyonlyAny

chemotherapyBreast,allRadiotherapyonlyAny

chemotherapyKidney,allRadiotherapyonlyAny

chemotherapyBladder,allRadiotherapyonlyAny

chemotherapyConnectivetissue,allRadiotherapy

onlyAnychemotherapyLeukemia,

all'Radiotherapy

onlyAnychemotherapyAcute

lymphohlaslic leukemia,allRadiotherapyonlyAny

chemotherapyAcutenonlymphocytic leukemia,allRadiotherapy

onlyAnychemotherapyObserved375511645292025311133816420789212\2x0(Iii6654441356339O/E'.12'.33.04.07.63).93.35.29.33.07.28.44.32'.35'.30'1.212.450.281.93'1.721.59(E

1.70)(E0.22)(E

0.76)7.or4.67'10.53'11.

00''23.0518.47'16.67''7.96'25.53'Observed2845311248102221251114104224441210234213222520I25218O/E1.34'1.44'1.60'1.53'1.862.32'1.77'0.921.401.410.761.631.34'I.601'1.66'(1.841.231.241.431.711.343.76'10.40'3.035.36'1.9813.66'8.60'(E

0.04)27.01'11.73'5.5526.81'Observed1655833259611405015316II5211372000732110201O/E1

.34'1.90'1

"^1.331.971.601.572.26(E

1.38)1.08(E

1.12)1.031.181.421.081.792.891.603.09'6.86'2.26(E

0.60)(E0.15)(E0.13)2.013.592.807.3229.77(E

0.03)1.61(E

0.30)3.82Observed156708->-|120117113703312161018913303000000000/E-,,â€¢â€¢,/.76'-".21.06.81(E

1.01)1.462.82'2.802.72'-'14.48'(E

0.25)0.810.870.422.311.13(E

0.15)3.95'-'6.0

\'J4.245.33''*16.47'(E

0.03)0.84(E

1.13)(E0.18)(E0.12)(E0.04)(E0.01)(E1.28)(E0.41)(E

0.07)" Includes women who received radiotherapy alone as primary treatment tor ovarian cancer. Patients may have subsequently been given chemotherapy, but these data are not

collected by the SEER program.'' Includes women whose primary treatment for ovarian cancer included chemotherapy alone in - 10,677 1-year survivors) or chemotherapy and radiotherapy (n - 904 1-year

survivors).' Number of person-years within interval.'' One hundred sixty-four solid tumors (O/E. l .04) occurred 2 months-1 year after diagnosis of ovarian cancer.' P < 0.05.' P trend = 0.05." P trend = 0.07.'' P trend = 0.04.' P trend = 0.01.' P trend = 0.03.* P trend = 0.04.' Three leukemias (O/E. 0.72) occurred 2 months-1 year after diagnosis of ovarian cancer.

combined and for cancers of the colon, rectum, breast, bladder, andeye. All ocular tumors, which occurred an average of 6.7 years afterovarian cancer, were unilateral, histologically confirmed malignantmelanomas. Seven- to 9-fold risks were seen for ALL and ANLL.Nonsignificant 2-fold excesses were noted tor chronic granulocytic

leukemia and connective tissue cancer.The risks of all solid tumors and cancers at selected sites are presented

in Table 3 according to primary treatment and time since diagnosis ofovarian cancer. Significantly elevated risks of solid tumors developed 1year after diagnosis of ovarian cancer and persisted throughout thefollow-up period (P trend over time = 0.05). Fifteen-year survivors

experienced significant excesses of cancers of the pancreas, bladder, andconnective tissue. The cumulative risk of all second malignancies twodecades after diagnosis of ovarian cancer was 18.2%, compared with apopulation expected risk of 11.5% (Fig. 1).

Temporal patterns of solid tumor risk differed according to

primary treatment for ovarian cancer. Following radiotherapyalone, excesses of solid tumors increased with time to reach almost2-fold among long-term survivors (P trend = 0.07); an additional88 solid tumors (O/E, 1.42; 95% CI, 1.14-1.75) occurredafter a20 years of follow-up. Although risk of solid tumors afterany chemotherapy was elevated within the 5-9-year interval

after ovarian cancer, nonsignificant excesses occurred in laterintervals.

More than 400 second cancers occurred in the breast, with thedistribution of histolÃ³gica! types similar to general patterns within theSEER and CTR data bases. Significantly elevated risks were notedwithin 10 years after diagnosis of ovarian cancer but decreased tobelow expectation after 15 years of follow-up.

Of 188 colon cancers, 98% were microscopically confirmed. Overall, a 30% excess of second primary colon tumors was observed, withincreased risks 5-14 years after ovarian cancer. Four tumors occurred

1566




20

C-

5 Â«Iâ€¢¿�K

I3

I5 -

All subsequent cancer (18.2%)â€”¿� Population expected risk (11.5%)

Syr

[8,396]

10 yr[4,402]

15 yr

[2,147]

Years From Diagnosis of Ovarian Cancer [number of patients]

Fig. I. Cumulative risk of second malignant neoplasms among 32.251 2-monlh

survivors of ovarian cancer. Percentages in parentheses, actuarial risk at 20 years: bars.95% CIs for point estimates.

within the appendix. Increased risks of rectal cancer, all microscopically verified, were evident within most follow-up periods.

A 2-fold risk of bladder cancer was evident 1-4 years followingdiagnosis of ovarian tumors and increased to 4-fold among 15-yearsurvivors (P trend = 0.01). Significant excesses of bladder cancer wereassociated with radiotherapy for ovarian cancer, with 6-7-fold risksevident ^10 years after primary treatment (P trend = 0.03). Risk among

patients initially given any chemotherapy was not significantly elevated.The distribution of histological types of bladder cancer (54 transitionalcell carcinoma, 5 squamous cell cancer, and 6 other or not specified) wassimilar to general patterns within SEER and CTR files.

Most (86%) renal cancers were diagnosed during life. The 29cancers consisted of 19 renal cell carcinomas, 3 carcinomas NOS, and7 transitional cell carcinomas of the renal pelvis (n = 3) or ureter(n = 4), a pattern similar to de novo cancers in the general population.

Of 45 pancreas cancers, 40 (89%) were microscopically confirmed.Risk was significantly increased among 15-year survivors of ovariancancer (P trend = 0.04). Although late excesses of pancreatic cancer

seemed confined to ovarian carcinoma patients initially treated withradiotherapy, expected numbers within the chemotherapy group werequite small (0.25).

Risk of connective tissue cancer was almost 5-fold among 15-yearsurvivors of ovarian carcinoma (P trend = 0.04). Six sarcomas

(fibrosarcoma, clear cell sarcoma, malignant fibrous histiocytoma.hemangiosarcoma, leiomyosarcoma, and rhabdomyosarcoma) developed al years after any treatment which included radiotherapy; threeof these cancers occurred within soft tissues likely to be in thetreatment field, i.e., abdomen, lower limb, or hip.

Of the 111 leukemias. 84 were ANLL. Significant excesses ofANLL were evident in the 1-4- and 5-9-year intervals after ovariancancer. Significant 60-fold risks for ANLL in the 1-4- and 5-9-year

periods followed combined modality therapy (data not shown).Among subjects given chemotherapy alone, risk of secondary ANLL1-9 years after treatment was 39.58 (O, 19: 95% CI, 23.82-61.82) and17.22 (O, 26; 95% CI, 11.25-25.23), respectively, for those diagnosedwith ovarian cancer in the 1970s or 1980-1992. No secondary ANLL

was reported after treatment with chemotherapy alone among ovariancancer patients diagnosed in the 1960s. Median survival after ANLLwas 2 (range, 0-19) months.

Significantly increased risks for secondary ALL, all of which werehistologically confirmed, were apparent in the 1-4- and 5-9-year

intervals after diagnosis of ovarian cancer. Significant excesses ofALL were restricted to patients initially given any chemotherapy.Survival after ALL was generally poor (median, 4 months; range,0-72 months).

Risks of second cancers of breast, colon, and rectum are stratified byhistological type of ovarian adenocarcinoma in Table 4. A significantlyincreased risk of breast cancer occurred among women with serousadenocarcinoma of the ovary (O/E. l .29), with nonsignificant excessesfollowing endometrioid ovarian tumors (O/E. 1.32). Elevated risks ofcolon cancer occurred after all specified types of ovarian adenocarcinoma. Significantly increased 2-fold risks of rectal cancer developed

among women with mucinous ovarian adenocarcinoma.The large number of second breast cancers permitted further eval

uation of risk by histological type of ovarian adenocarcinoma andpatient age (Table 5). Significantly increased risks of breast cancersoccurred among women diagnosed with ovarian adenocarcinoma before age 50 years (O/E, 1.37) or between 50 and 60 years (O/E, 1.28)but not at age 60 years or older (O/E. 1.04). Patients diagnosed withserous adenocarcinoma of the ovary prior to age 50 years demonstrated a significant 80% excess of breast cancer, with much smallerrisks noted among older women. Less striking age patterns of breastcancer excesses were associated with endometrioid ovarian adenocarcinoma, with nonsignificant 1.3- to 1.5-fold risks present within all

categories.

DISCUSSION

Using registry-based information, 32,251 women with ovarian cancer, including large numbers of long-term survivors, were evaluated

for second cancer risk by primary therapy. An important finding is thesignificantly elevated risk of solid tumors among 15-year survivors of

ovarian cancer, particularly those initially given radiotherapy, with noevidence of a diminution in risk with time. Based on our results, aboutone in five women with ovarian cancer would be expected to developa new malignancy within two decades, including a variety of solid

Table 4 Risk of second cancers of breast, colon, and rectum by histology of first primary ovarian adenocarcinoma"

Second primary cancer

Histology of ovarianadenocarcinoma (no. of patients)'

Breast Colon''

O/E 95% CI O/E 95% CI

Rectum

O/E 95% CI

Serous in =7602)Mucinous(n =3424)Endometroid

(n =2279)Otherspecified types (n =9777)NOS

(n = 4805)1056041Uh38.29'.10.32.13.111.06-1.560.84-1.410.95-1.790.93-1.360.79-1.533130165091.231.441.641.330.760.83-1.740.97-2.060.93-2.660.99-1.750.35-1.4413155255.389ff.36.74'.130.73-2.361.09-3.220.44-3.171.12-2.560.36-2.64

11Includes 27.887 women for whom tumor histology was denoted as adenocarcinoma.* Refer to Table 1 for ICD-0 codes.' No. of 2-month survivors of ovarian adenocarcinoma.

Risk presented for 1-year survivors of ovarian adenocarcinoma; see text.' P < 0.05.

1567




Table 5 Risk of second breast cancer by patient age and histology of first primaryovarian adenocarcinoma"

Histology of ovarianadenocarcinorna'All

histologietypesAge<50yrAge50-59yrAge260yrSerousAge

<50yrAge50-59yrAge^60yrEndometrioidAge

<50yrAge50-59yrAge^60 yrNo.

ofpatients7,1596,65114,0771,8561,8913.855669614996098110152272850HI1219O/E.37''.28''.04.83C.20.15.51.27.2795%CI1.11-1.671.05-1.550.88-1.221.21-2.670.80-1.740.86-1.520.72-2.780.65-2.210.76-1.98

" Includes all patients wilh ovarian adenocarcinoma (n = 27.8871. Risk is also

presented for histologie types of adenocarcinoma (serous and endometrioid) for whichexcess second primary breast cancers were noted (Table 4).

* Refer to Table 1 for ICD-0 codes.r P < 0.05.

tumors and ANLL. as reported previously. However, some new findings were seen, including significant excesses of ocular melanoma andALL after ovarian cancer and possible relationships between histo-

logical type of ovarian adenocarcinoma and subsequent breast andcolorectal neoplasms.

Leukemia. A high risk of ANLL and preleukemia has been documented following ovarian cancer and linked to therapy with melpha-

lan (7, 8), cyclophosphamide (7, 8, 17), or chlorambucil (8, 18) and tocombination chemotherapy with doxorubicin and cisplatin (8). Themarkedly elevated 40-fold risk of ANLL observed among ovarian

cancer patients given chemotherapy in the 1970s likely reflects thecommon use of melphalan. a well-established leukemogen (7, 8. 19).

Increased, but lower, risks for ANLL persisted among patients givenchemotherapy during the later years of our survey, when cisplatin-

containing regimens became more widely used. Because details ofcancer therapy are not available within SEER and CTR files, ANLLrisk cannot be conclusively linked with individual chemotherapeuticagents. Others have remarked on the need to further evaluate theleukemogenicity of cisplatin-based chemotherapy regimens (20, 21).

The increased risk of leukemia in our survey extended to ALL,although data on this cell type are sparse in prior analytic series ofovarian cancer (8, 17). However, ALL is increasingly noted as apossible complication of cancer therapy (22, 23) and may constitute5-10% of secondary acute leukemia (22). The cytogenetic transloca

tion (4; 11)(q21 ;q23), which involves the region targeted by drugs thatinteract with DNA-topoisomerase II (24), has been reported in sec

ondary ALL (22, 23, 25, 26). Although leukemic blasts of theseunusual secondary leukemias resemble lymphocytes morphologicallyand cytochemically, cell differentiation may also be either bipheno-

typic (27) or consistent with an early progenitor myeloid cell (28).Breast Cancer. Because treatment for ovarian cancer frequently

results in ablation of gonadal function by oophorectomy, a subsequentdecreased risk of breast cancer might be expected (29), typically afteran interval of s 10 years (30). Although nonsignificant deficits inbreast neoplasms were evident among long-term survivors in our

study, overall risk was significantly elevated, consistent with severalprior estimates (3, 6, 31 ) which approached 4-fold in one survey (32).

Breast cancer excesses in our series seemed to vary by histologicaltype of ovarian adenocarcinoma, although CIs overlapped considerably. Based on 203 second breast cancers, we found significantlyincreased risks after serous ovarian adenocarcinoma, with excesses ofa similar magnitude following endometrioid tumors. Looking in theother direction, a significantly elevated risk of serous ovarian adenocarcinoma (O, 197; O/E, 1.49) is seen among breast cancer patients

within the SEER and CTR data bases,4 a reciprocity suggestive of

shared etiological influences. Serous variants of ovarian adenocarcinoma also seem overly represented within hereditary ovarian cancersyndromes (33, 34), and risk factors for ovarian carcinoma may differby histological type (35-37). Shah et al. (6) reported nonsignificant

elevated risks of breast cancer after serous ovarian adenocarcinoma,but not mucinous or endometrioid tumors; however, small numbers(/; = 17) limited interpretation.

Although the excess of breast cancer after endometrioid ovarianadenocarcinoma was not significant, the increased risk (O/E, 1.32)may be noteworthy, since women with breast cancer reported to SEERand CTR experienced small excesses of endometrioid ovarian adenocarcinoma (O, 54; O/E, 1.23; P > 0.05).4 In a case-control study of

endometrioid ovarian adenocarcinoma (87 cases), a significant 2-fold

risk was associated with family history of breast cancer (36).In our cohort, elevated risks of breast cancer were confined to

women diagnosed with ovarian adenocarcinoma before age 60 years,which may reflect the greater contribution of genetic factors to tumorsof early onset (38). Young women were especially prone to breastcancer after serous ovarian adenocarcinoma, similar to findings reported by Shah et al. (6). In our series, there was no clear age effecton breast cancer risk among women with endometrioid ovarianadenocarcinoma.

Colorectal Cancer. High-dose radiation for cancer and other dis

eases has induced colon cancer in some (39), but not all (40), patientpopulations. Although excesses of colon cancer were linked with priorradiotherapy in one survey of ovarian cancer patients (31 ). the patternof elevated risks in our cohort seemed more consistent with theinfluence of shared etiological factors. In support of this hypothesis isthe reciprocal finding that significant excesses of ovarian cancer (O,250; O/E, 1.49) follow colon cancer in the SEER and CTR files.4

Some investigators have suggested that nutritional and hormonalinteractions may contribute to the occurrence of multiple primarycancers, including colon, ovary, breast, and uterine corpus (41, 42).Ovarian tumors are also associated with hereditary nonpolyposiscolorectal cancer, suggesting that genetic determinants may influencethe constellation of ovarian and colon cancers (43). Whether defectsin mismatch repair genes described in hereditary nonpolyposis colorectal cancer (44) contribute to elevated risks of colon cancer afterradiotherapy for ovarian cancer (31) should be explored. Excess risksof colon cancer were seen after all cell types of ovarian cancer exceptthe serous type. In several clinical surveys, a relation has beendescribed between mucinous ovarian and appendiceal tumors (45-

48), typically among patients with pseudomyxoma peritonei, a diagnostic entity not included in our study base. However, Shah et al. (6)found significantly elevated risks of colorectal carcinoma (n = 10)

after serous ovarian tumors but not mucinous or endometrioid types.Although we did not find a persuasive relationship between radio

therapy and the risk of rectal cancer, a significant association betweenradiation dose and rectal cancer has been reported following cervicalcancer (40). In our study, the risk of rectal cancer seemed greateramong women with mucinous ovarian adenocarcinoma, but CIs werewide, and misclassification may have affected the pattern seen.

Pancreatic Cancer. The pancreas is considered relatively insensitive to the carcinogenic effects of ionizing radiation (39); however,very high doses may induce cancers at this site (49). The excess seenamong long-term survivors in our survey suggested a late effect of

treatment, but the number of cases was small. The pancreas mayreceive radiation exposure during whole abdominopelvic radiotherapy

4 L. B. Travis, R. E. Curtis, and B. F. Hankey, unpublished data.

1568




for ovarian carcinoma, but SEER and CTR data do not permit correlation of second cancer risk with a specific regimen.

Bladder Cancer. The temporal pattern and distribution of excessbladder cancers among treatment groups in our survey suggest aradiogenic effect, particularly since pelvic irradiation for ovariancancer delivers a high dose (â€”20-60 Gy) to the bladder (5). Com

parable doses to this organ during treatment for cervical cancerresulted in an overall 4-fold risk of bladder neoplasia, with a 6-fold

risk 15 years after therapy (40), consistent with our results. In addition, nonsignificant excesses of bladder cancer followed chemotherapy for ovarian cancer in our survey, perhaps due to the common useof cyclophosphamide. an established bladder carcinogen, in treatmentregimens. Cyclophosphamide has been linked with bladder canceramong ovarian cancer patients (50), and a strong dose-response relationship has been reported following treatment for non-Hodgkin's

lymphoma (51).Connective Tissue Cancer. The pattern of excess risks for con

nective tissue cancers also supports a radiogenic effect. Our findingsare consistent with a follow-up of cervical cancer patients, documenting 3-fold excess risks of connective tissue cancer among 10-year

survivors given radiation doses of >10 Gy (40).Ocular Melanoma. Increased risks observed for ocular melanoma

following ovarian cancer have not been seen previously and deservefurther investigation, particularly to clarify the possible role of genetic(52) and hormonal (53) factors that may be shared by these tumors.Although statistically nonsignificant, an excess of ovarian cancer is seenafter ocular melanoma in the SEER and CTR data bases (O/E, 1.34;P > 0.05).4 Uveal melanoma has been described occasionally in families

prone to breast cancer, particularly in association with the susceptibilitygene BRCA2 (54, 55). Unlike BRCAI, however, this gene has not beenclearly linked with ovarian carcinoma (54). There have been case reportssuggesting an association between bilateral uveal melanoma and othercancers, including ovarian cancer (56). but further studies are needed toclarify the risks and mechanisms involved.

Comment. Various strengths and weaknesses of data on multipleprimary tumors reported to cancer registries must be considered ininterpreting our results. The SEER Program and CTR together providelarge numbers of subjects and population-based data which minimize

the bias due to selection or referral patterns in clinical or hospitalseries. The large sample size also permits quantification of secondcancer risk by site. Because a very high percentage of all cancers inour series was histologically confirmed, it is unlikely that metastaticovarian cancer was mistakenly diagnosed as a second malignancy.Because underreporting of second cancers may occur among long-

term survivors migrating from areas included in the SEER Program orCTR, however, our estimates of increased risk may be conservative.In surveys of multiple primary cancers, one should be mindful that thelarge number of comparisons will generate some statistically significant associations by chance alone.

In our study, classification of ovarian tumors preceded the development of second malignancies, so that risk estimates for secondcancer by cell type of ovarian tumor should not be biased. Furthermore, nondifferential misclassification between designated histologi-

cal types of ovarian adenocarcinoma (57), if present, would dampendifferences in risk (58). There was no evidence that associations ofsecond cancer risk with patient age and histological type of ovarianadenocarcinoma were confounded by treatment.

Nevertheless, our data provide a reasonable estimate of the overall riskof second cancers among women with invasive ovarian cancer. Significant excesses of subsequent malignancies were evident in all time intervals after diagnosis of ovarian cancer and persisted among long-term

survivors, suggesting the need for lifelong medical surveillance. Furtherepidemiological and laboratory studies are needed to clarify the carcino

genic risks associated with modern therapies for ovarian cancer and withshared susceptibility mechanisms, including genetic and reproductivefactors. The multiple tumor complexes associated with ovarian cancerneed further histological and molecular studies in efforts to elucidateetiological mechanisms and to develop targeted screening programs.Meanwhile, in proposing recommendations for the follow-up and man

agement of women with ovarian cancer (59), it is important to recognizetheir long-term predisposition to an array of second cancers.

ACKNOWLEDGMENTS

We are indebted to George Giese and Dennis Buckman for computer support, to Denise Nguy for typing assistance, and to RebeccaAlbert and Debra Reames for clerical services.

REFERENCES

1. Wingo, P. A.. Tong. T.. and Bolcien. S. Cancer statistics. 1995. CA Cancer J. Clin..45: 8-30. 1995.

2. Miller. B. A.. Ries. L. A. G.. Hankey, B. F., Kosary, C. L.. HaÃ¯ras,A.. Devesa, S. S.,and Edwards, B. K. (eds). SEHR, Cancer Statistics Review: 1973-1990. NIH Puh.93-2789. Bethesda, MD: National Cancer Institute, 1993.

3. Kaldor. J. M.. Day. N. E., Band, P., Choi. N. W.. Clarke. E. A.. Coleman. M. P.,Hakama, M.. Koch. M.. Langmark, F., Neal. F. E.. Pettersson. F.. Pompe-Kirn. V.,Prior. P.. and Storni. H. H. Second malignancies following testicular cancer, ovariancancer and Hodgkin's disease: an international collaborative study among cancer

registries. Int. J. Cancer, 39: 571-585, 1987.4. Curtis. R. E.. Hoover. R. N.. Kleinerman. R. A., and Harvey. E. B. Second cancer

following cancer of the female genital system in Connecticut. 1935-1982. NCIMonogr., 68: 113-137. 1985.

5. Prior. P.. and Pope. D. J. Subsequent primary cancers in relation to treatment ofovarian cancer. Br. J. Cancer. 59: 453-459, 1989.

6. Shah, S., Evans. D. G.. Blair. V., Burnell. L. D.. and Birch. J. M. Assessment ofrelative risk of second primary tumors after ovarian cancer and of the usefulness ofdouble primary cases as a source of material for genetic studies within a cancerregistry. Cancer (Phila.). 72: 819-827. 1993.

7. Greene. M. H.. Harris. E. L., Gershenson. D. M.. Malkasian. G. D.. Jr.. Melton, L. J..III. Dembo. A. J.. Bennett, J. M., Moloney, W. C., and Boice. J. D.. Jr. Melphalanmay be a more potent leukemogen than cyclophosphamide. Ann. Intern. Med., 105:360-367, 1986.

8. Kaldor. J. M.. Day, N. E.. Pettersson, F., Clark. E. A.. Pedersen. D.. Mehnert, W..Bell. J.. Host. H.. Prior, P.. Karjalainen. S.. Neal, F., Koch. M.. Band. P., Choi, W..Pompe Kirn. V., Arslan, A., Zaren. B.. Belch. A. R.. Storm. H.. Kittelmann, B..Fraser, P.. and Stovall, M. Leukemia following chemotherapy for ovarian cancer. N.Engl. J. Med., 322: 1-6. 1990.

9. Cannistra, S. A. Cancer of the ovary. N. Engl. J. Med., 329: 1550-1559. 1993.

10. Bagley. C. M.. Jr.. Young. R. C.. Canellos, G. P.. and DeVita. V. T. Treatment ofovarian carcinoma: possibilities for progress. N. Engl. J. Med.. 287: 856-862, 1972.

11. Horowitz. C., and Brady, L. W. Carcinoma of the ovary. In'. C. A. Pere?, and L. W.

Brady (eds.). Principles and Practice of Radiation Oncology, pp. 988-1015.Philadelphia. PA: J. B. Lippincott Co., 1987.

12. Pezner. R. D.. Stevens. K. R., Jr.. Tong. D.. and Allen. C. V. Limited epithelialcarcinoma of the ovary treated with curative intent by the intraperitoneal installationof radiocolloids. Cancer (Phila.). 42: 2563-2571. 1978.

13. WHO. International Classification of Diseases for Oncology. Ed. 2. Geneva.Switzerland: WHO. 1990.

14. Percy. C., Young, J. L., Jr.. Muir. C.. Ries. L.. Hankey. B. F., Sobin, L. H., and Berg.J. W. Cancer. Introduction. Cancer (Phila.). 75: 140-146. 1995.

15. Platz, C. E., and Benda. J. A. Female genital tract cancer. Cancer (Phila.), 75:270-294, 1995.

16. Breslow, N. E., Lubin. J. H.. Marek. P.. and Langholz. P. Multiplicative models andcohort analysis. J. Am. Stat. Assoc., 78: 1-12, 1983.

17. Haas, J. F.. Kittelmann, B., Mehnert. W. H.. Staneczek, W.. Mohner. M.. Kaldor.J. M., and Day. N. E. Risk of leukaemia in ovarian tumour and breast cancer patientsfollowing treatment by cyclophosphamide. Br. J. Cancer, 55: 213-218. 1987.

18. Greene, M. H., Boice. J. D., Jr., Greer. B. E.. Blessing. J. A., and Dembo. A. J. Acutenonlymphocytic leukemia after therapy with alkylating agents for ovarian cancer: astudy of five randomized clinical trials. N. Engl. J. Med., 307: 1416-1421. 1982.

19. Curtis, R. E.. Boice, J. D., Jr., Stovall. M., Bernstein, L., Greenberg. R. S.. Flannery.J. T.. Schwartz, A. G.. Weyer. P.. Moloney, W. C.. and Hoover, R. N. Risk ofleukemia after chemotherapy and radiation treatment for breast cancer. N. Engl. J.Med., 326: 1745-1751, 1992.

20. Greene. M. H. Is cisplatin a human carcinogen? J. Nail. Cancer Inst.. 84: 306-312.1992.

21. Cheruku. R.. Hussain, M., Tyrkus. M.. and Edelstein, M. Myelodysplastic syndromeafter cisplatin therapy. Cancer (Phila.), 72: 213-218, 1993.

22. Hunger, S. P.. Sklar. J.. and Link. M. P. Acute lymphoblastic leukemia occurring asa second malignant neoplasm in childhood: report of three cases and review of theliterature. J. Clin. Oncol.. 10: 156-163, 1992.

23. Auxenfants, E.. Morel. P., Lai. J. L.. Sartiaux. C.. Detourmignies, L., Baulers, F.. and

1569



SECOND MALIGNANT NEOPLASMS AFTKR OVARIAN CANCER

PÃ©naux.P. Secondary acule lymphohlastic leukemia with 1(4;11): report on !wo casesand review of the literature. Ann. Hematol., 65; 143-146. 1992.

24. Pedersen-Bjergaard. J.. and Rowley. J. D. The balanced and the unbalanced chromosome aberrations of acute inyeloid leukemia may develop in different ways and maycontribute differently to malignant transformation. Blood, 83: 2780-2786. 1994.

25. Archimbaud. E.. Charrin, C-. Guyotat, D.. Magaud. J. P.. Gentilhomme. O-. and Piere,D. Acute leukaemia with 1(4;111 in patients previously exposed to carcinogens. Br. J.Macinatoi., (W 467-470. I9S8.

26. Kapelushnik, J., Dube. I.. Wilson. P.. and Greenberg. M. Acute lymphoblasticleukemia with t(4;l!) translwation after osteogenic sarcoma. Cancer (Phila.l. 67:2887-2889. 1991.

27. Chen. S. J.. Chen. Z., DerrÃ©.J.. Le Coniai. M., Valensi, F.. Sigaux. F.. and Berger. R.Are most secondary acute lymphoblastic leukeniias mixed acute leukeniias? Nouv.Res. Fr. Hematol.. il: 17-22. 1989.

28. Palau. F.. Prieto. F.. Badia. L.. Beneyto. M.. Sempere. A., and Borrego, D. Cytoge-netic evidence of involvement of an early progenitor myeloid cell in 4;1 I transloca-lion-associated acute leukemia. Cancer Genet. Cytogenet.. 5/: 79-84. 1991.

29. Boice, J. D.. Jr.. Blettner. M.. Kleinerman. R. A.. Engholm. G.. Stovall. M.. Lisco. H..Austin. D. P.. Bosch. A.. HarÃan.L.. Krementz. E. T.. Latourette. H. B.. Merrill. J. A..Peler. L. J.. Schulz. M. D.. Wactawski. J.. Storm. H. H.. Bjorkholm. E.. Pettersson. F..Bell. C. M. J.. Coleman. M. P.. Fraser. P.. Neal. F. E.. Prior. P.. Choi. N. W.. Hislop.T. G.. Koch. M.. Kreiger. N.. Robb. D.. Robson. D.. Thomson. D. H.. Lochmuller. H..von Fournier, D.. Frischkorn, R-, Kjorslad, K. E., Rimpela, A.. Pejovic, M.. Kirn.V. P.. Stankusova. H., Pisani. P.. Sigurdsson. K.. Hutchison. G. B., and MacMahon.B. Radiation dose and breast cancer risk in patients treated for cancer of the cervix.Int. J. Cancer. 44: 7-16, 1989.

30. MacMahon. B.. Cole. P.. and Brown. J. Etiology of human breasl cancer: a review.J. Nati. Cancer Insl.. 50: 21-42. 1973.

31. Reimer. R. R.. Hoover, R.. Praumeni, J. P.. Jr., and Young. R. C. Second primaryneoplasms following ovarian cancer. J. Nati. Cancer Inst.. 6/: 1195-1197. 1978.

32. Schottenfeld. I)., and Berg. J. Incidence of multiple primary cancers. IV. Cancers ofthe female breast and genital organs. J. Nail. Cancer Inst.. 4f>: 161-170. 1971.

33. Fraumeni. J. F.. Jr.. Grundy. G. W.. Creagan. E. T.. and Everson. R. B. Six familiesprone to ovarian cancer. Cancer (Phila.l, 36: 364-369. 1975.

34. Bewtra. C.. Watson, P., Conway. T.. Read-Hippee. C.. and Lynch, H. T. Hereditaryovarian cancer: a clinicopathological study. Int. J. Gynecol. Pathol.. //: 180-187.

1992.35. S/.amborski. J., C/erwinski, W.. Gadoinska, H., Kowalski. M., and Wacker-Pujdak.

B. Case control study of high-risk factors in ovarian carcinomas. Gynecol. Oncol.. / /:8-16, 1981.

36. Schildkraut. J. M.. and Thompson. W. D. Relationship of epithelial ovarian cancer toother malignancies within families. Genet. Epidemiol.. 5: 355-367. 1988.

37. Schiffman. M.. Hartge, P., Hoover. R. N.. McGowan, L., Lesher. L.. and Noms, H. J.Letter to the editor. Gynecol. Oncol., 33: 129-132. 1989.

38. Amos, C. I.. Shaw. G. L.. Tucker, M. A., and Hange. P. Age at onset for familialepithelial ovarian cancer. JAMA. 2<Vf: 1896-1899. 1992.

39. Committee on the Biologic Effects of loni/.ing Radiation (BEIR V). National Research Council. Health Effects of Exposure to Low Levels of Ionizing Radiation.Washington. DC: National Academy Press. 1990.

40. Boice, J. D., Jr.. Engholm. G., Kleinerman. R. A.. Blettner. M.. Stovall. M.. Lisco. H..Moloncy, W. C.. Austin. D. P., Bosch, A.. Cooktair. D. L., Krementz, E. T..Latourette. H. B.. Merrill. J. A.. Peters. L. J.. Schulz. M. D.. Storm, H. H., Bjorkholm.E.. Pettersson. F.. Bell, C. M. J., Coleman, M. P., Fraser, P.. Neal, F. E., Prior, P..Choi. N. W., Hislop, T. G., Koch. M.. Kreiger. N.. Robb, D., Robson. D.. Thomson.D. H.. LochmÃ¼ller,H., von Fournier. D.. Frischkorn. R.. KjÃ¶rstad.K. Eâ€žRimpeta. A..Pejovic. M. H.. Kirn. V. P., Stankusova. H.. Benino. F.. Sigurdsson. K.. Hutchison.G. B., and MacMahon. B. Radiation dose and second cancer risk in patients treatedfor cancer of the cervix. Radial. Res.. 116: 3-55, 1988.

41. McMichael. A. J.. and Potter J. D. Reproduction, endogenous and exogeneous sexhormones, and colon cancer: a review and hypothesis. J Nati. Cancer Inst.. 65;1201-1207. 1980.

42. Willett. W. C.. and MacMahon. B. Diet and cancerâ€”an overview. N. Engl. J. Med..310: 697-703. 1984.

43. Lynch, H. T.. Lanspa. S.. Smyrk. T.. Boman. B.. Watson. P.. and Lynch. J. Hereditarynonpolyposis colorectal cancer (Lynch syndromes I & II). Genetics, pathology,natural history, and cancer control, pan I. Cancer Genet. Cytogenet.. 53: 143-160.1991.

44. Fishel. R.. Lescoe, M. K., Rao. M. R.. Copeland, N. G.. Jenkins. N. A.. Garber. J.,Kane. M.. and Kolodner. R. The human mutator gene homolog MSH2 and itsassociation with hereditary nonpolyposis colon cancer (published erratum appears inCell. 77: 167. 1994] Cell. 75: 1027-1038. 1993.

45. Young. R. H.. Gilks. C. B.. and Scully. R. P. Mucinous tumors of the appendixassociated with mucinous minors of the ovary and pseudomyxoma peritonei. Aclinicopathological analysis of 22 cases supporting an origin in the appendix. Am. J.Surg. Pathol., 15: 415-429. 1991.

46. Sumithran. E.. and Susil, B. J. Concomitant mucinous tumors of appendix and ovary.Result of a neoplasia- field change? Cancer (Phila.), 70: 2980-2983, 1992.

47. Seidman. J. D.. Elsayed. A. M., Sobin. L. H.. and Tavassoli. P. A. Association ofmucinous tumors of the ovary and appendix. A clinicopalhologic study of 25 cases.Am. J. Surg. Pathol.. 17: 22-34. 1993.

48. Prayson, R. A., Hart, W. R., and Petras. R. E. Pseudomyxoma peritonei. A clinico-

pathologic study of 19 cases with emphasis on site of origin and nature of associatedovarian tumors. Am. J. Surg. Pathol.. IK: 591-603, 1994.

49. Griem. M. L.. Kleinerman. R. A.. Boice. J. D.. Jr.. Stovall, M.. Shefner. D.. andLubin. J. H. Cancer following radiotherapy for peptic ulcer. J. Nail. Cancer Inst.. Xf>:842-849. 1994.

50. Kaldor. J. M.. Day. N. E.. Kittelmann, B.. Petlersson. F.. Langmark, F.. Pedersen, D..Prior. P.. Neal, F.. Karjalainen S.. Bell. J.. Choi. W.. Koch. M.. Band. P.. Pompe-Kirn.

V.. Garton. C.. Staneczek. W.. Zaren. B., Stovall. M., and Boffella. P. Bladdertumours following chemotherapy and radiotherapy for ovarian cancer: a case-controlsludy. Int. J. Cancer. 63: 1-6. 1995.

51. Travis, L. B., Curtis. R. E., Glimelius. B.. Holowaty. E. J.. Van Leeuwen, F. E..Lynch, C. F. Hagenbeek. A.. Slovall. M.. Banks. P. M.. Adami. H-O.. Gospodarow-

icz. M. K.. Wacholder. S.. Inskip. P. D.. Tucker. M. A., and Boice. J. D.. Jr. Bladderand kidney cancer following cyclophosphamide therapy for non-Hodgkin's lym-

phoma. J. Nati. Cancer Insl.. 87: 524-530, 1995.52. Young. L. H.. Egan. K. M.. Walsh. S. M.. and Gragoudas. P. S. Familial uvea!

melanoma. Am. J. Ophthalmol.. 117: 516-520. 1994.53. Hartge. P., Tucker. M. A.. Shields. J. A.. Augsburger. J.. Hoover. R. N.. and

Fraumeni. J. P.. Jr. Case-control study of female hormones and eye melanoma. CancerRes.. 49: 4622-4625. 1989.

54. Wooster. R.. Neuhausen. S. L.. Mangion. J.. Quirk. Y.. Ford. D.. Collins. N.. Nguyen.K.. Seal, S.. Tran, T.. Averill. D., Fields, P., Marshall, G.. Narixl, S., Lenoir, G. M.,Lynch. H.. Feurteun, J., Devilee, P., Cornelisse. C. J.. Menko. F. H.. Daly. P. A..Ormislon. W.. McManus, R.. Pye. C.. Lewis. C. M., Cannon-Albright. L. A.. Peto. J.,

Ponder, B. A. J., Skolnick. M. H.. Easton. D. F.. Goldgar. D. E., and Stratum, M. R.Localization of a breast cancer susceptibility gene. BRCA2. to chromosome I3ql2-13. Science (Washington DC). 265: 2088-2090. 1994.

55. Goldgar, D. E.. Neuhausen. S. L., Steele. L.. Field, P.. Ward. J. H.. Tran. T.. Nguyen.K.. Stratton, M. R.. and Easton. D. F. A 45-year follow-up of kindred 107 and thesearch for BRCA2. NCI Monogr.. 17: 15-19, 1995.

56. Mullaney. J.. Mooney. D.. O'Connor. M.. and McDonald. Ci. S. Bilateral ovarian

carcinoma with bilateral uveal melanoma. Br. J. Ophthalmol.. 6<V:261-267. 1984.

57. Cramer. S. F., Rolh, L. M., Ulbright, T. M., Mazur. M. T.. Nunez. C. A.. Gersell.D. J., Mills. S. E.. and Kraus. F. T. Evaluation of the reproducihility of Ine WorldHealth Organization classification of common ovarian cancers. With emphasis onmethodology. Arch. Pathol. Lab. Med.. ///: 819-829. 1987.

58. Monson R. R. The interpretation of epidemiologie data. In: Occupational Epidemiology, pp. 93-103. Boca Raton. FL: CRC Press, Inc.. 1980.

59. NIH Consensus Development Panel on Ovarian Cancer. Ovarian cancer: screening,treatment, and follow-up. JAMA, 27.J: 491-497, 1995.

1570



1996;56:1564-1570. Cancer Res Lois B. Travis, Rochelle E. Curtis, John D. Boice, Jr., et al. Ovarian CancerSecond Malignant Neoplasms among Long-Term Survivors of

Updated version

http://cancerres.aacrjournals.org/content/56/7/1564

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/56/7/1564To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



second malignant neoplasms among long-term survivors of … · second malignant neoplasms after...

Documents