Re Lest We Abandon Digital Rectal Examination as a Screening Test by ert634


									                                                                                                                        Downloaded from by guest on May 6, 2011
                                                                            Re: Lest We Abandon Digital
                                                                            Rectal Examination as a
                                                                            Screening Test for Prostate

                                                                                The editorial by Basler and Thomp-
                                                                            son (1) on the report by Schroder et al.
                                                                            (2) necessitates comment by me. We are
                                                                            happy that Basler and Thompson agree
                                                                            with the final conclusion of the report
                                                                            that “more sensitive methods will have
                                                                            to be found to replace DRE (digital rec-
                                                                            tal examination) in men with low PSA
                                                                            (prostate-specific antigen) values.” Fur-

Journal of the National Cancer Institute, Vol. 91, No. 15, August 4, 1999                  CORRESPONDENCE 1331
 thermore, we acknowledge that the edi-         report, in spite of rigorous editing, con-    REFERENCES
 torial points out a number of limitations      tradictory information regarding the per-
 of our material. Most of these points are      centage of advanced tumors found for          (1) Basler JW, Thompson IM. Lest we abandon
 correct, such as the possible bias by se-      PSA levels of less than 4 ng/mL was               digital rectal examination as a screening test
                                                                                                  for prostate cancer. J Natl Cancer Inst 1998;
 lecting case subjects for radical prosta-      retained. On page 1821, left column,
 tectomy, the change of the protocol on         line 3, the sentence should read as fol-      (2) Schroder FH, van der Maas P, Beemsterboer
 one occasion during the study period,          lows: “ . . . whereas six tumors (16%)            P, Boeken Kruger A, Hoedemaeker R, Riet-
 that palpable lesions were not directly        showed characteristics of advanced can-           bergen J, Kranse R. Evaluation of the digital
 subjected to biopsy examination, but           cer.” This statement is consistent with           rectal examination as a screening tool for pros-
 were entered into sextant biopsy               information several lines later in the            tate cancer. J Natl Cancer Inst 1998;90:
 schemes, and the fact that only 90% of         same column, where we said that the
                                                                                              (3) Nijs HG, Tordoir DM, Schuurman JH, Kirkels
 eligible men were subjected to biopsy          proportions of men with PSA values of             WJ, Schroder FH. Randomized trials of pros-
 examination; most of these are acknowl-        less than 4.0 ng/mL who had minimal,              tate cancer screening in The Netherlands: as-
 edged in our report.                           moderate, and advanced disease were               sessment of acceptance and motives for atten-
    Several issues raised in the editorial,     42%, 42%, and 16%, respectively. Our              dance. J Med Screen 1997;4:102–6.
 however, require at least short com-           report indeed does not mention the pro-       (4) Gustafsson O, Norming U, Almgard LE, Fred-
                                                                                                  riksson A, Gustavsson G, Harvig B, et al. Di-
 ments; other issues require more com-          portion of cases with metastatic disease,         agnostic methods in the detection of prostate
 ment.                                          which was 0.6%, 14 times lower than in            cancer: a study of a randomly selected popu-
    Screening before randomization              the control arm. PIN has been subject to          lation of 2,400 men. J Urol 1992;148:
 (contamination): Contamination not             a separate report (7).                            1827–31.
 only before, but also after, subjects were         Basler and Thompson state that the        (5) Mettlin C, Lee F, Drago J, Murphy GP. The
 randomly assigned to the screening and         assumption that DRE-detected tumors               American Cancer Society National Prostate
                                                                                                  Cancer Detection Project. Findings on the de-
 control arms is subject to extensive           with PSA levels of less than 4 ng/mL              tection of early prostate cancer in 2,425 men.
 study. The data will be presented else-        might still be curable after 4 years is “an       Cancer 1991;67:2949–58.
 where shortly. In the present context,         unsubstantiated leap of faith.” We fully      (6) Hoedemaeker RF, Rietbergen JB, Kranse R,

                                                                                                                                                     Downloaded from by guest on May 6, 2011
 only the contamination rate before sub-        agree with this statement. Our report             van der Kwast TH, Schroder FH. Comparison
 jects were randomly assigned is of rel-        does not make this assumption. On the             of pathologic characteristics of T1c and non-
 evance. This rate can be evaluated from                                                          T1c cancers detected in a population-based
                                                contrary, we anticipated a discussion of
                                                                                                  screening study, the European Randomized
 the intake form and varies between 9%          this issue. We state that the study group         Study of Screening for Prostate Cancer. World
 and 11%. Characteristics of the refusers       has decided to subject all men with a             J Urol 1997;15:339–45.
 are the subject of a separate study (3).       PSA level of greater than 3 ng/mL to a        (7) Hoedemaeker RF, Kranse R, Rietbergen JB,
 Bias in the direction of selectively ex-       biopsy examination on the basis of the            Schroder FH, van der Kwast TH. Evaluation
 cluding men who might have a positive          experience with the first 10 523 men              of prostate needle biopsies in a population-
 DRE is highly unlikely.                                                                          based screening study: the impact of border-
                                                screened. The question is whether cases
                                                                                                  line lesions. Cancer. In press 1999.
    Low biopsy rate with low PSA:               of prostate cancer detected in men with
 Basler and Thompson suggest that the           a PSA level of less than 3 ng/mL might
 number of cancers in the low PSA                                                             NOTE
                                                still be amenable to curative measures in
 ranges may be substantially underesti-         the second round of screening. In this                                           ¨
                                                                                                 Correspondence to: Fritz H. Schroder, M.D.,
 mated because the DRE-driven biopsy            context, it is of interest that preliminary   Ph.D., Department of Urology, Erasmus Univer-
 rate is lower in the lower ranges of PSA.      data indicate an identical detection rate     sity and Academic Hospital Rotterdam, P.O. Box
 In considering this issue, it is important     with the regimen described in this paper,     1738, 3000 DR Rotterdam, The Netherlands (e-
 to realize that the examiners did not          and the new regimen in which men with         mail:
 know the PSA values before performing          a PSA level of less than 3 ng/mL get an
 the DRE. Furthermore, other authors            all-clear message, while men with a
 (4,5) have encountered the same phe-           PSA level of 3 ng/mL or higher are sub-
 nomenon. The fact that the positive pre-       jected to a biopsy examination (Schro-   ¨
 dictive value (PPV) of DRE increases           der FH: oral communication). This sug-
 with rising PSA in a blind fashion shows       gests that a large proportion of cases of
 that the value of DRE as a test is, in fact,   prostate cancer are missed by DRE in
 PSA dependent.                                 men with PSA levels of 3–3.9 ng/mL.
    Minimal prostate cancer, metastatic             Our report does not state that DRE
 disease, and prostatic intraepithelial         should be abandoned, as suggested by
 neoplasia (PIN): Basler and Thompson           the title of the editorial. The conclusion
 state that the definition of minimal pros-     drawn is that a positive DRE and low
 tate cancer is somewhat arbitrary. The         PSA levels have a less than desirable
 definitions used are documented by             PPV. DRE should be replaced by a bet-
 Hoedemaeker et al. (6). All definitions        ter regimen that detects a larger propor-
 of minimal, moderate, and advanced             tion of cancers with fewer biopsy ex-
 disease are arbitrary because they have        aminations.
 not been clinically tested. Unfortu-
 nately, on pages 1820 and 1821 of our                                            ¨
                                                                     FRITZ H. SCHRODER

1332 CORRESPONDENCE                                                Journal of the National Cancer Institute, Vol. 91, No. 15, August 4, 1999

To top