Risk of testicular cancer with cryptorchidism and with testicular biopsy: cohort studyBMJ 1998; 317 doi: http://dx.doi.org/10.1136/bmj.317.7160.729 (Published 12 September 1998) Cite this as: BMJ 1998;317:729
- Henrik M⊘ller, head ()a,
- Dina Cortes, registrar in pathologyb,
- Gerda Engholm, statisticiana,
- J⊘rgen Thorup, head of paediatric surgeryb
- aCentre for Research in Health and Social Statistics, Danish National Research Foundation, Sejr⊘gade 11, DK-2100 Copenhagen ⊘, Denmark,
- bDepartments of Pathology and Paediatric Surgery, National University Hospital, DK-2100 Copenhagen ⊘
- Correspondence to: Dr M⊘ller
Swerdlow et al studied a cohort of 1075 men who had been treated for cryptorchidism.1 In 120 testes that had been biopsied at the time of orchiopexy the relative risk of subsequent development of testicular cancer was 66.7. The corresponding relative risk in 1285 undescended testes that had also been operated on but not biopsied was 6.7. The estimate for testes that were not biopsied is about as expected,2–4 but the high relative risk of testicular cancer in biopsied testes is a new finding not yet substantiated by other studies. A potential weakness of the study is the unclear criteria for biopsy in 9% of the testes in the series; this leaves open the possibility of selection bias if testes with a high risk of subsequent malignancy were preferably biopsied.
To explore the risk of testicular cancer in an unselected population of biopsied undescended testes, we investigated a cohort of 830 men who had surgical treatment of 1026 undescended testes at this hospital between January 1971 and January 1992.
Subjects, methods, and results
An open surgical biopsy sample was taken from all testes at the time of surgery.5 The current analysis is based on one person less than the previous report5 because a duplicate record was discovered and deleted. The clinical data included information on date of birth, date of surgery for cryptorchidism, and results for the biopsied tissue. Information on cancer occurrence, including date of diagnosis and tumour type, were obtained from the Danish cancer registry. Dates of death of cohort members were obtained from the Danish register of deaths. The incidence of testicular cancer among cohort members was compared with the corresponding expected incidence in the total male population in Denmark. The analysis was carried out, firstly, on the 830 men who had had one or both testes biopsied, and, secondly, on the 1026 biopsied testes.
The table shows the occurrence of testicular neoplasms in the 830 men from birth to 31 December 1994. Of the seven cases, one occurred before a biopsy sample was taken from the contralateral testis (case 1), three were diagnosed by histological examination at the time of surgery for cryptorchidism (cases 2-4), two occurred in previously biopsied testes (cases 5 and 7), and one was a contralateral cancer in a man previously operated on for unilateral cryptorchidism (case 6). The three cases that occurred in men who had a testicular biopsy in one or both testes (cases 5-7) corresponded to a relative risk of 2.0, while the two cases in biopsied testes (cases 5 and 7) corresponded to a relative risk of 2.2 (table).
Our data, based on a large series of men operated on for cryptorchidism who had all had a biopsy done at the time of the operation, do not support the finding of Swerdlow et al of a greatly increased risk of testicular cancer in biopsied testes.1 On the contrary, our data suggest a moderately increased (about twofold) risk of testicular cancer in biopsied testes. Histological examination at the time of surgery for cryptorchidism discovered one case of seminoma and two cases of carcinoma in situ (cases 2-4). If these three cases had become clinically manifest as invasive cancer later the number of cases during follow up would have been six and the relative risk would have been about fourfold, which is the rate expected in a population of men treated for cryptorchidism.2–4
Contributors: DC and JT collected and verified the clinical and pathological data and contributed to study design, interpretation, and reporting. GE conducted the statistical analyses and contributed to interpretation and reporting. HM took the initiative in this study and was responsible for interpreting the results and drafting the paper. HM, DC, GE, and JT are all guarantors.
Funding: No specific funding.
Conflict of interest: None.