Correspondence

Cancer and Venous Thromboembolism

N Engl J Med 1998; 339:703September 3, 1998

Article

To the Editor:

In their article on cancer and venous thromboembolism, Sørensen et al. (April 23 issue)1 reported an increased incidence of the diagnosis of cancer in the first six months of follow-up after the diagnosis of primary deep venous thrombosis or pulmonary embolism. They also noted that “the estimated risk of all types of cancer during the first year of follow-up decreased with increasing age at first discharge with venous thromboembolism.” However, there is a possible confounding factor that does not appear to have been considered in their article or in the editorial that accompanied it.2

As the authors themselves state in their introduction, it is very well known that there is an association between cancer and venous thromboembolism. It is therefore possible that the physicians caring for a patient presenting with primary venous thromboembolism will look closely for a primary cancer. They are more likely to do so in younger patients. It is generally accepted that intensive screening is unlikely to be rewarding, but abdominal computed tomography or ultrasound examinations may well be regarded as minimal investigation. Such testing could lead to the early (presymptomatic) diagnosis of cancers of the pancreas, liver, and ovary. Similarly, the apparent early excess of prostate cancer may be due to the performance of routine digital rectal examination at the time of admission for the venous thromboembolism. These factors taken together would produce exactly the pattern described.

A record-linkage study such as the one conducted by Sørensen et al. cannot tell us the number of patients whose cancers were found because the presence of a venous thromboembolism led the physician to screen the patient — that is, the early excess is due to a variety of lead-time biases.

Andrew J.W. Hilson, F.R.C.P.
Royal Free Hospital, London NW3 2QG, United Kingdom

2 References

1. 1

   Sorensen HT, Mellemkjaer L, Steffensen FH, Olsen JH, Nielsen GL. The risk of a diagnosis of cancer after primary deep venous thrombosis or pulmonary embolism. N Engl J Med 1998;338:1169-1173
   Full Text | Web of Science | Medline

2. 2

   Buller H, ten Cate JW. Primary venous thromboembolism and cancer screening. N Engl J Med 1998;338:1221-1222
   Full Text | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: We do not find that our use of routine data is substantially affected by surveillance bias. The following factors argue against bias. First, 40 percent of the patients with cancer who received a diagnosis within the first year of follow-up had distant metastases at the time of diagnosis. Second, the risk of cancer one to four years after the episode of deep venous thrombosis and pulmonary embolism was not reduced. And third, a similar Swedish study1 reported almost identical results. The Swedish study included patients in whom venous thromboembolism was diagnosed between 1965 and 1983, before the extensive use of computed tomographic scanning and ultrasonography had been established. These facts speak against the possibility of surveillance bias due to diagnostic drift.

Henrik Toft Sørensen, Dr.Med.Sci.
University of Aarhus, DK-8000 Aarhus C, Denmark

Lene MellemkjÆr, Ph.D.
Jørgen H. Olsen, Dr.Med.Sci.
Danish Cancer Society, 2100 Copenhagen Ø, Denmark

1 References

1. 1

   Baron JA, Gridley G, Weiderpass E, Nyren O, Linet M. Venous thromboembolism and cancer. Lancet 1998;351:1077-1080
   CrossRef | Web of Science | Medline