HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stanley, R. J. Search for Related Content PubMed PubMed Citation Articles by Stanley, R. J. Social Bookmarking                      What's this?

Inherent Dangers in Radiologic Screening

¹ Department of Radiology, University of Alabama at Birmingham, 619 19th St. S., N342 JT, Birmingham, AL 35249-6830.

Received June 11, 2001; accepted after revision June 21, 2001.

 
Presented at the annual meeting of the American Roentgen Ray Society, Seattle, April 2001.

Address correspondence to R. J. Stanley.

Article

Top Article References

 
Over the past few years I have become increasingly concerned about the use of diagnostic imaging for the purpose of screening adults. In my pursuit of further information on this topic, I have leaned heavily on various studies, William Black and Gilbert Welch, from the Dartmouth-Mary Hitchcock Medical Center [1,2,3,4]. I have also found considerable useful information in the published works of Alan Morrison and Philip Cole, which will be referenced later in my talk.

View larger version (107K): [in this window] [in a new window] [as a PowerPoint slide]

Let me initiate this discussion by asking you a question: How many islands are off the coast of Maine? The answer is that it depends on how closely you look and also on what constitutes an island. A large-scale map of the coast of Maine shows nothing that resembles an island. However, when we look at a regional map of the Bangor area, we discover that there are multiple islands, some of which are named and some of which might be too small to deserve a name. If one can see a giant rock only during low tide and not during high tide, does that rock qualify as an island?

Likewise, we may ask the question: How many renal cell cancers exist in an adult screening population? The answer is that it depends on how closely you look and on what constitutes a renal cell carcinoma. The prevalence of these tumors in patients who had no symptoms suggestive of renal cell carcinoma before death depends on how closely their kidneys are examined after death.

The prevalence of clinically unsuspected renal tumors at autopsy, when the threshold size for the macro inspection is 2 cm, is 1-2% [5]. However, when kidneys are examined at autopsy by serial 2- to 3-mm sectioning, the prevalence of tumors exceeds 22% [6]. The current age-specific prevalence rate of kidney cancer is 0.03-0.06% in the general population and 0.3% in men more than 70 years old [7]. Fewer than 0.5% of all deaths in the United States in 1986 were attributed to kidney cancer (8897 cancers of the kidney and renal pelvis divided by 2,105,361 total deaths) [8]. With a cure rate of approximately 50%, only 1% of the adult population develops potentially lethal renal cancer. However, if the detection threshold is lowered to the serial sectioning level (2-3 mm) and histology remains the gold standard, 22% of the screened population could be said to have pathologically proven renal cell carcinoma.

In the Atlas of Tumor Pathology [9], published by the Armed Forces Institute of Pathology, the following statement is found concerning tumors of the kidney, bladder, and related urinary structures: "With the exception of oncocytoma, it has not been possible to define an unequivocally benign renal cortical neoplasm using histologic, immunohistochemical, and ultrastructural studies."

In a similarly respected and influential source, Urologic Surgical Pathology [10], the following statement appears:

The prevalence of small cortical lesions at autopsy suggests that many proliferations of renal cortical epithelium lack the capacity to develop into clinical cancer. However, the search for a means to distinguish between those lesions that have the potential to progress and those that do not has been unsuccessful.

How is "screening" defined? Screening is the systematic testing of asymptomatic individuals for preclinical disease. The purpose is to prevent or delay the development of advanced disease in the subset of patients with preclinical disease through early detection and treatment [11]. Why initiate screening programs? The answer appears to be the compelling intuitive appeal of the idea that early detection of cancer will be rewarding. It appears to be a straightforward approach to cancer control [12].

What is the objective of screening? Operationally, the objective of a screening program is the application of a relatively simple, inexpensive test to a large number of persons in order to classify them as likely or unlikely to have the cancer that is the object of the screen [12]. In terms of outcome, the objective is to reduce morbidity and mortality from that cancer among the persons screened, achieved at a reasonable cost [12].

The concept of a critical point in a disease needs to be emphasized. The critical point is the time when treatment is more effective before that point than after. Screening must find the disease before that time. For example, if the critical point in the disease is just before the development of metastases, then the screening test may be of value. If the critical point occurs long before the screening test is capable of detecting the pathologic morphology, then there is clearly no point in subjecting a person to the screening test.

In a recent edition of Applied Radiology [13], one of the editors, Robert Harris, cited the opinion of Paul Chang, the director of informatics and PACS (picture archiving and communication system) at the University of Pittsburgh, on the entire-body CT scan performed on paying "customers" to screen for potential disease. He called them "yuppie scans," seemingly designed for the self-absorbed, immortality-seeking types prevalent in that generation. In the December 2000 issue of Diagnostic Imaging, an article by Michael Brant-Zawadzki [14], entitled "Screening on Demand: Portent of a Revolution in Medicine," has an introductory subtitle that asks the following question: "When they are willing to pay themselves, can patients be denied information about their own bodies?" My question is, precisely what sort of information will be conveyed to these "patients"? Presumably, the screens will be done on asymptomatic individuals who would theoretically be in the preclinical phase of whatever disease they might have. They don't become patients, apparently, until after these screening studies.

Brant-Zawadski [14], in a reference to lung cancer, has stated, "When detected early, it is quite curable; treated stage I cancers demonstrate a 50% 5-year survival." Unfortunately, increasing the 5-year survival has little or nothing to do with proving that a disease is curable, as I will explain later.

In March 2000, The Wall Street Journal published an article [15] that discussed the Southern California "pioneers" in the use of whole-body screening CT scans. The article stated that the range of referrals to specialists after asymptomatic adults underwent these screening examinations varied from 4% at a clinic headed by Jerry Dalrymple to 80% at a similar clinic headed by Harvey Eisenberg. Obviously, this range suggests a problem in the outcome of these studies. In reference to this concept of screening the general population, Black [15], from Dartmouth Medical School, states, "If you screen the general population, you can wreak havoc on the general population." Dennis Fryback [15], of the University of Wisconsin Medical School, expresses similar concerns: "What we may be finding in screening exams is indolent disease that would never bother people."

To muddy the water a bit, a report from the United States Department of Health and Human Services quoted by Welch et al. has reported [1]:

The 5-year survival rate for all cancers improved from 51% in the early 1980s to almost 60% in the early 1990s.... Since the 1971 National Cancer Act, much of the research into early cancer detection and treatment has paid off.

Are increased 5-year survival rates evidence of success against cancer? In fact, changes in 5-year survival over time bear little relationship to changes in cancer mortality. Instead, they appear primarily related to changing patterns of diagnosis [1].

Borrowing heavily from the published work of Welch and Black [1,2,3,4], let me now briefly review some of the potential screening biases that lead to the confusion about the relationship between survival rates and cancer mortality. What are some of the potential screening biases? In evaluating the effectiveness of a screening program, when survival rates are used as the endpoint, four potential biases arise: lead time, length time, self-selection, and overdiagnosis. Lead time is the interval between detection at screening and the usual time of clinical manifestation. Lead-time bias pertains to comparisons that do not account for the progression of disease over time and are not adjusted for the timing of diagnosis. Figures 1 and 2 show the relationship between detection of a disease by a test during the detectable preclinical phase and how detection increases the survival time without necessarily doing anything to the actual time of death from the disease or other causes.

View larger version (11K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1. —Natural history of disease. Diagram illustrates that preclinical phase begins at onset and ends when signs or symptoms develop. Clinical phase then starts, ending with death. Detectable preclinical phase (DPCP) begins when disease is detectable by a test. Detection (X) during DPCP advances time of diagnosis by duration of lead time. (Reprinted from [4])

View larger version (11K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2. —Lead-time bias. Diagram shows that, with screening, time of diagnosis is advanced by lead time provided by positive test result. If earlier diagnosis has no effect on time of death from disease, then survival with testing is equal to survival without testing plus lead time. (Reprinted from [4])

Length time is the difference in growth rates of tumors in the same organ. Length-time bias pertains to comparisons that do not account for the variability of disease progress. Figure 3 shows how a test might pick up only two cases in patients with rapidly progressive disease, whose detectable preclinical phase is considerably shorter, versus four cases of a more slowly progressive form of the same disease with a much longer detectable preclinical phase.

View larger version (12K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3. —Length-time bias. Diagram shows how probability of detection is related to rate of disease progression. Length of each arrow represents length of detectable preclinical phase, from initial detectability to clinical diagnosis (Dx). Testing at a single moment detects four slowly progressive cases but only two rapidly progressive cases. Cases not detected by test (thin arrows) are diagnosed clinically either before or after time of testing. Thick arrows indicate detected cases. (Reprinted from [4])

The risk in failing to apprehend the significance of lead-time bias and length-time bias is the erroneous attribution of clinical benefit to a new test that permits earlier detection, or new treatment that accompanies the earlier diagnosis, when the interventions provide no benefit or actually harm the patients [2]. Concerning the other two potential screening biases, voluntarily deciding to participate in a screening program may introduce a self-selection bias, and results from the screening detection of lesions of questionable malignancy, such as in situ cancers, introduce an overdiagnosis bias.

Overdiagnosis may be illustrated by a common tumor in men, prostate cancer. The prevalence of microscopic prostate cancer is approximately 50% in men over 60 years old who do not have a clinical diagnosis of prostate cancer. But only 3% of men without a clinical diagnosis of prostate cancer are expected to eventually die from that disease. Thus, the probability that a 60-year-old man with microscopic prostate cancer will eventually die from it is less than or equal to 6% [8, 16]. A large proportion of the other 44% will require no specific therapy for this less aggressive form of the disease.

Considering the detection thresholds and the percentage of breast cancers attributed to carcinoma in situ, when palpation was the method for detecting breast cancer, only 1-5% of the biopsied lesions were carcinoma in situ [17]. Using mammography, a study from 1981 [18] showed that the percentage had risen to 8%. In 1988, a more recent report on mammography [19] showed that carcinoma in situ accounted for 25-30% of the tumors biopsied; and at the present time, basing the diagnosis on microcalcifications only, 40-50% of the biopsied breast cancers are carcinoma in situ [2]. In his textbook on breast cancer, Fentiman [20] stated:

Trial results represent the results of the enthusiasts, the pioneers in this field who with evangelical style have developed screening, and by obsessional attention to minor abnormalities on mammograms have reduced the mortality from breast cancer.

In citing Fentiman's statement, de Koning [21] concludes that trial results will not necessarily be replicated in widespread screening programs.

Let us direct our attention now to screening for lung cancer. The Mayo Clinic developed a randomized clinical trial to screen for lung cancer that was first described in 1986 [22]. Half of the subjects underwent sputum cytology and chest radiography every 4 months for 6 years; the control group was advised to have these same tests performed annually. In the screened group, the lung cancer mortality per 1000 person-years was 3.2 and in the control group, 3.0. In essence, there was virtually no difference between the two groups. In 1989, in an editorial in the Annals of Internal Medicine, David Eddy [23] stated, in reference to the issue of lung cancer screening, "It is important to emphasize that these conclusions are not based on a lack of evidence of an effect, but are based on good evidence that there is no effect."

The Mayo Lung Project has been criticized by many advocates of screening as a flawed study for a variety of reasons, including the relatively short period of follow-up. However, a more recent reevaluation of this same study was done, entitled "Lung Cancer Mortality in the Mayo Lung Project: Impact of Extended Follow-Up," which appeared in the Journal of the National Cancer Institute [24]. The researchers could identify essentially no change from the earlier conclusions, but their study had more statistical power. Lung cancer mortality was 4.4 deaths per 1000 person-years in the screened, intervention group and 3.9 deaths per 1000 person-years in the usual-care group. The median survival for lung cancer patients with early stage disease diagnosed before July 1983 was better in the intervention arm than in the usual-care arm, but the difference was not statistically significant. The authors stated that the apparent discrepancy between survival and mortality was caused by some combination of lead-time, length- and overdiagnosis biases.

In a study from Yale dealing with so-called lung cancer stage migration, published in the New England Journal of Medicine in 1985 [25], the researchers pointed out that diagnostic methods for staging lung cancer patients in the 1950s and 1960s were not identical to staging methods in the 1970s and early 1980s. Thus, a stage I tumor in the earlier years would possibly be called a stage II or stage III disease by more accurate diagnostic methods. The researchers referred to the so-called improvement in survival of the various stages as the Will Rogers phenomenon. Will Rogers is quoted, without an exact citation, as saying, "When the Okies left Oklahoma and moved to California, they raised the average intelligence level in both states" [25].

In the ACR Bulletin of November 2000 [26], two screening advocates were cited. Denise Aberle, chair of the American College of Radiology Imaging Network lung cancer protocol, is quoted as saying [26]:

A randomized controlled trial will enable us to determine the impact of screening on lung cancer-specific and all-cause mortality rates as well as to better assess differences in the biologic behavior of lung cancers.

Claudia Henschke, principal investigator of the Early Lung Cancer Action Project (ELCAP), is also quoted: "It will be important for us not to repeat the same mistakes that were made in the Mayo Lung Project, which may have set back our screening policy for lung cancer for decades" [26]. She also said, "In regards to lung cancer screening, there is no question that it finds cancers early, and that earlier cancers have improved curability; it is a matter of quantitating this effect" [26]. However, in a consensus statement of the Society of Thoracic Radiology, of which Henschke was a coauthor, the following statement appears in the summary of current recommendations [27]:

The appropriate studies, which address lung cancer mortality and cure rates, need to be performed, the data analyzed and validated before the true utility of this test [lung cancer screening with low dose CT] can be determined. Thus, we do not recommend mass screening for lung cancer at this time, but strongly encourage appropriate subjects to participate in trials so that the true effectiveness of lung cancer screening with low dose helical CT can be determined at the earliest possible time.

An earlier part of this report states, "It is recommended that all subjects being screened with CT for lung cancer are done as part of a prospective study" [27]. Thus, it is not entirely clear whether Henschke is opposed to prospective clinical trials or exclusively to the types of randomized clinical trials that were performed in the past.

To minimize adverse side effects in screening for lung cancer, Black [3] recommends the following approaches: Consider a mandatory observation period for small nodules, perform randomized clinical trials, monitor all causes of mortality to avoid missing a major benefit or harm from the screening process, and provide a balanced presentation of potential benefits and risks to all prospective screenees.

In summary, considering the benefits and risks of screening, each incremental lowering of the detection threshold increases the chances that treatment will benefit patients with clinically significant disease, but at the cost of increasing the proportion of diagnosed patients who have clinically insignificant disease. You, my fellow radiologists, must ensure that you have an understanding of clinical epidemiology when embarking on screening programs and not simply fall into the role of the detectors of possible disease in the chest, abdomen, or pelvis. We must take the responsibility for giving sound advice and for selecting the appropriate people to screen.

Carl Zylak stated in his Radiological Society of North America presidential address in 1991 [28]:

It is apparent that researchers who collect clinical data and radiologists who use it have a great deal in common. Both have a critical stake in the accuracy and validity of the information. Radiologists must understand the basic principles of research in order to interpret what is found.

He goes on to say, "...radiology must attract individuals who are willing to bridge the gap between clinical epidemiology and our specialty."

I am indebted to Dr. Zylak for considerable assistance in gathering materials related to the overall topic of screening.

Thanks very much for your attention. I will now treat you to some photos of my two grandsons and my granddaughter. By the time these young people are adults, I hope we will have successfully and intelligently resolved this question of the appropriate role of screening.

References

Top Article References

 

1. Welch HG, Schwartz LM, Woloshin S. Are increasing 5-year survival rates evidence of success against cancer? JAMA 2000;283:2975 -2978[Abstract/Free Full Text]
2. Black WC, Ling A. Is earlier diagnosis really better? the misleading effects of lead time and length biases. AJR 1990;155:625 -630[Free Full Text]
3. Black WC. Overdiagnosis: an under recognized cause of confusion and harm in cancer screening. J Natl Cancer Inst 2000;92:1280 -1282[Free Full Text]
4. Black WC, Welch HG. Screening for disease. AJR 1997;168:3 -11[Abstract/Free Full Text]
5. Hellsten S, Berge T, Wehlin L. Unrecognized renal cell carcinoma: clinical and diagnostic aspects. Scand J Urol Nephrol 1981;15:269 -272[Medline]
6. Xipell JM. The incidence of benign renal nodules (a clinicopathologic study). J Urol 1971;106:503 -506[Medline]
7. Feldman AR, Kessler L, Myers MH, Naughton MD. The prevalence of cancer: estimates based on the Connecticut tumor registry. N Engl J Med 1986;315:1394 -1397[Abstract]
8. Montie JE, Wood DP Jr, Pontes JE, Boyett JM, Levin HS. Adenocarcinoma of the prostate in cystoprostatectomy specimens removed for bladder cancer. Cancer 1989;63:381 -385[Medline]
9. Murphy WM, Beckwith JB, Farrow GA. Tumors of the kidney, bladder, and related urinary structures. In: Atlas of tumor pathology, 3rd series, vol. 11. Washington, DC: Armed Forces Institute of Pathology, 1994
10. Bostwick DG, Eble JN. Urologic surgical pathology. St. Louis: Mosby-Year Book, 1997
11. Morrison AS. Screening in chronic disease, 2nd ed. New York: Oxford University Press, 1992:3 -20
12. Cole P, Morrison AS. Basic issues in population screening for cancer. J Natl Cancer Inst 1980;64:1263 -1272
13. Harris RD. Yuppie scans. Appl Radiol 2001;29:10
14. Brant-Zawadzki MN. Screening on demand: portent of a revolution in medicine. Diagn Imaging 2000;2212:25 -27
15. Spurgeon D, Burton TM. For the very cautious a physical exam now includes a CAT scan. The Wall Street Journal, March 23, 2000: A1
16. Seidman H, Mushinski MH, Gelb SK, Silverberg E. Probabilities of eventually developing or dying of cancer. CA Cancer J Clin 1985;35:36 -56[Abstract/Free Full Text]
17. Schnitt SJ, Silen W, Sadowsky NL, Connolly JL, Harris JR. Ductal carcinoma in situ (intraductal carcinoma) of the breast. N Engl J Med 1988;318:898 -903[Medline]
18. Tabar L, Gad A. Screening for breast cancer: the Swedish trial. Radiology 1981;138:219 -222[Abstract/Free Full Text]
19. Hall FM, Storella JM, Silverstone DZ, Wyshak G. Nonpalpable breast lesions: recommendations for biopsy based on suspicion of carcinoma at mammography. Radiology 1988;167:353 -358[Abstract/Free Full Text]
20. Fentiman I. Detection and treatment of early breast cancer. Philadelphia: Lippincott 1990:618
21. de Koning HJ. Assessment of nationwide cancer-screening programmes. Lancet 2000;355:80 -81[Medline]
22. Sanderson DR. Screening (lung cancer). Chest 1986;89[suppl 4]:324S -326S[Medline]
23. Eddy DM. Screening for lung cancer. (editorial) Ann Intern Med 1989;354:232 -237
24. Marcus PM, Bergstralh EJ, Fagerstrom RM, et al. Lung cancer mortality in the Mayo Lung Project: impact of extended follow-up. J Natl Cancer Inst 2000;92:1308 -1316[Abstract/Free Full Text]
25. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon: stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med 1985;312:1604 -1608[Abstract]
26. CT screening for lung cancer: has the time come? ACR Bulletin 2000;56:15 -17
27. Aberle DR, Gamsu G, Henschke CI, et al. A consensus statement of the Society of Thoracic Radiology: screening for lung cancer with helical computed tomography. Available at: http://www.thoracicrad.org/str99/consensus_statement.htm. Accessed May 18, 2001
28. Zylak CJ. Clinical epidemiology and radiology: the need to bridge the gap. Breast screening: a case study. Radiology 1992;184:607 -611[Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				I. M. Burger, N. E. Kass, J. H. Sunshine, and S. S. Siegelman The Use of CT for Screening: A National Survey of Radiologists' Activities and Attitudes Radiology, July 1, 2008; 248(1): 160 - 168. [Abstract] [Full Text] [PDF]

				C. T. Kolber, G. Zipp, D. Glendinning, and J. J. Mitchell Patient Expectations of Full-Body CT Screening Am. J. Roentgenol., March 1, 2007; 188(3): W297 - W304. [Abstract] [Full Text] [PDF]

				P. J. Pickhardt and A. J. Taylor Extracolonic Findings Identified in Asymptomatic Adults at Screening CT Colonography. Am. J. Roentgenol., March 1, 2006; 186(3): 718 - 728. [Abstract] [Full Text] [PDF]

				J. K. Leader, T. E. Warfel, C. R. Fuhrman, S. K. Golla, J. L. Weissfeld, R. S. Avila, W. D. Turner, and B. Zheng Pulmonary Nodule Detection with Low-Dose CT of the Lung: Agreement Among Radiologists Am. J. Roentgenol., October 1, 2005; 185(4): 973 - 978. [Abstract] [Full Text] [PDF]

				S. J. Swensen, J. R. Jett, T. E. Hartman, D. E. Midthun, S. J. Mandrekar, S. L. Hillman, A.-M. Sykes, G. L. Aughenbaugh, A. O. Bungum, and K. L. Allen CT Screening for Lung Cancer: Five-year Prospective Experience Radiology, April 1, 2005; 235(1): 259 - 265. [Abstract] [Full Text] [PDF]

				R. J. Stanley Screening Revisited Am. J. Roentgenol., December 1, 2004; 183(6): 1537 - 1537. [Full Text] [PDF]

				D. B. Kopans, B. Monsees, and S. A. Feig Screening for "Cancer": When is it Valid?--Lessons from the Mammography Experience Radiology, November 1, 2003; 229(2): 319 - 327. [Abstract] [Full Text] [PDF]

				S. J. Swensen, J. R. Jett, T. E. Hartman, D. E. Midthun, J. A. Sloan, A.-M. Sykes, G. L. Aughenbaugh, and M. A. Clemens Lung Cancer Screening with CT: Mayo Clinic Experience Radiology, March 1, 2003; 226(3): 756 - 761. [Abstract] [Full Text] [PDF]

				R. L. Morin, T. C. Gerber, and C. H. McCollough Radiation Dose in Computed Tomography of the Heart Circulation, February 18, 2003; 107(6): 917 - 922. [Full Text] [PDF]

				L. Berlin Liability for Failure to Order Screening Examinations Am. J. Roentgenol., December 1, 2002; 179(6): 1401 - 1405. [Full Text] [PDF]

				K. Garg, R. L. Keith, T. Byers, K. Kelly, A. L. Kerzner, D. A. Lynch, and Y. E. Miller Randomized Controlled Trial with Low-Dose Spiral CT for Lung Cancer Screening: Feasibility Study and Preliminary Results Radiology, November 1, 2002; 225(2): 506 - 510. [Abstract] [Full Text] [PDF]

				S. J. Swensen CT Screening for Lung Cancer Am. J. Roentgenol., October 1, 2002; 179(4): 833 - 836. [Full Text] [PDF]

				F. M. Hall, E. F. Patz Jr, W. C. Black, and P. C. Goodman Screening for Lung Cancer: Been There and Done That * Dr Patz and colleagues respond: Radiology, September 1, 2002; 224(3): 928 - 929. [Full Text] [PDF]

				D. B. Kopans, N. A. Obuchowski, and R. J. Stanley Chopped Liver...All Over Again Am. J. Roentgenol., May 1, 2002; 178(5): 1288 - 1289. [Full Text] [PDF]

This Article Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stanley, R. J. Search for Related Content PubMed PubMed Citation Articles by Stanley, R. J. Social Bookmarking                      What's this?