IMAGING IN PRACTICE

DANIEL LOCKWOOD, MD DAVID EINSTEIN, MD WILLIAM DAVROS, PhD CME Department of Diagnostic , Department of Diagnostic Radiology, Head, Section of Medical Physics, Department of CREDIT Cleveland Clinic Cleveland Clinic Diagnostic Radiology, Department of Biomedical Engineering, Cleveland Clinic

Diagnostic imaging: Radiation dose and patients’ concerns

■ ABSTRACT 45-YEAR-OLD WOMAN WHO has never A had a mammogram comes to see her Exposure to ionizing radiation during diagnostic radio- general practitioner for a general medical logic procedures carries small but real risks, and children, examination. The physician recommends that young adults, and pregnant women are especially vulner- she undergo screening every able. Exposure of patients to diagnostic energy levels of year as part of a program of health mainte- ionizing radiation should be kept to the minimum neces- nance, but the patient expresses concern about sary to provide useful clinical information and allay this. Her older sister developed breast cancer at patients’ concerns about radiation-related risks. age 50, after 5 years of regular mammographic testing, and she fears that radiation from the ■ screening mammography tests may have con- KEY POINTS tributed to the development of cancer. Use CT with discretion: it accounts for two thirds of the How should the clinician counsel this cumulative patient dose from diagnostic radiologic patient? procedures, and the cumulative dose from CT is rising as ■ THE GOOD AND THE BAD technological advances increase the number of indications for and the capabilities of CT. The potential harm from ionizing radiation is an issue that faces every physician and patient Carcinogenesis and teratogenesis are the main concerns considering diagnostic imaging. with ionizing radiation. The risk increases as the radiation With the exception of ultrasonography dose increases. There is no minimum threshold, and the and magnetic resonance imaging (MRI), diag- risk is cumulative: a dose of 1 mSv (0.1 rem) once a year nostic radiologic tests involve ionizing radia- for 10 years is equivalent to a single dose of 10 mSv (1 tion—photons with enough energy to ionize rem). (ie, strip electrons from the nuclei of) atoms with which they interact. It can consist of x- Whenever practical, choose an imaging test that uses less rays, such as in plain film and radiation or no radiation, and lengthen the periods computed (CT), or of gamma rays from radiopharmaceuticals used in between follow-up imaging tests. . Exposure to ionizing radiation during diag- Some patients may avoid screening mammography nostic radiologic procedures carries small but because of fear of radiation-induced cancer, yet this test real risks. Ionizing radiation can damage living uses a very small radiation dose (0.6 mSv, much less than cells by causing undesired chemical reactions the annual dose from background radiation, 3.6 mSv), that alter the structure of macromolecules and technologic improvements are lowering the dose within the cell. Children, young adults, and required. pregnant women are especially vulnerable. On the other hand, the images produced can con- tain critical diagnostic information that may

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 73 • NUMBER 6 JUNE 2006 583

Downloaded from www.ccjm.org on September 29, 2021. For personal use only. All other uses require permission. RADIATION DOSE LOCKWOOD AND COLLEAGUES

TABLE 1 undertake extensive calculations to accurately estimate the dose of radiation received by a Estimated effective radiation dose specific patient in a specific study. of common diagnostic imaging tests* The concept of effective dose, measured in millisieverts (mSv) or “roentgen equivalents STUDY EFFECTIVE DOSE IN MILLISIEVERTS† man” (rem; 10 mSv = 1 rem) allows many of these factors to be compared and controlled for. Chest radiography, posteroanterior and lateral 0.06 But remember that everyone is constantly Screening mammography 0.6 exposed to naturally occurring ionizing radia- Gastric emptying study 1.4 tion, commonly called background radiation. -ureter-bladder radiography 1.7 Some comes from radioactive elements pre- CT of the head 1.8 sent in the earth since its formation (primor- Lumbar spine radiography 2.1 dial radionuclides), such as uranium and the Background radiation, annual dose 3.6 natural products of its decay, radium and the Radionuclide bone scan 4.4 gas radon. Other background radiation is in Ventilation-perfusion (V/Q) scan 6.8 the form of cosmic rays, high-energy particles CT of the pelvis 7.1 CT of the abdomen 7.6 that constantly bombard the atmosphere and CT of the chest 7.8 create radioisotopes of carbon and nitrogen. Barium enema radiography 8.7 The average annual effective dose from back- CT of coronary arteries 10 ground radiation is estimated at 3.6 mSv (0.36 Positron emission tomography, whole body 14 rem). Small bowel series (barium swallow x-ray study) 15 Some diagnostic procedures involve an Intravenous pyelography 10.0–20.0 effective dose of radiation that is a tiny frac- Whole-body screening CT 22.5 tion of that from background radiation, Three-phase hepatic CT scan 29.9 whereas many impart several times that Dual-isotope myocardial rest 32.5 amount (TABLE 1). and stress perfusion CT study CT urographic study 44.1 ■ RADIATION RISKS OF IMAGING *All values are for procedures performed at Cleveland Clinic †10 mSv (millisieverts) = 1 rem Deterministic vs stochastic effects The damaging effects of ionizing radiation are catagorized as deterministic or stochastic. Deterministic effects occur only when the greatly benefit the patient. Therefore, the risks dose has reached a threshold, beyond which and benefits must be considered before pro- the effects increase in severity as the dose ceeding with any diagnostic test involving ion- increases. is the imaging proce- izing radiation. Exposure to ionizing radiation dure for which deterministic effects are a main should be kept as low as reasonably achievable concern: it can damage the skin, leading to (the “ALARA” principle) while still answer- inflammation, epilation, and necrosis. ing the clinical question at hand. More worrisome are the stochastic effects In this article, we review the risks and carcinogenesis and teratogenesis, which benefits of diagnostic imaging and then offer increase in likelihood but not in severity as practical ways to maximize its benefits while the radiation dose increases. Stochastic effects minimizing its risks. have no minimum threshold, and the risk is cumulative. For example, a dose of 1 mSv (0.1 ■ QUANTIFYING THE RADIATION DOSE rem) once a year for 10 years is equivalent to a single dose of 10 mSv (1 rem). Quantifying the radiation dose is not a simple The risk of stochastic effects is often dis- matter. The energy and quantity of the pho- cussed in terms of the “linear no-threshold” tons, the size of the patient, and the vulnera- model, which states that risk varies linearly bility of irradiated tissues must be factored into with dose and assumes that no minimum or any estimate. Medical physicists often must threshold dose is needed to increase risk.

584 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 73 • NUMBER 6 JUNE 2006

Downloaded from www.ccjm.org on September 29, 2021. For personal use only. All other uses require permission. The BEIR VII estimates cancer risk are likely to live long enough for a developing A widely accepted estimate of the risk of radi- tumor to become symptomatic. Conversely, ation-induced carcinogenesis in diagnostic older patients are less vulnerable because imaging comes from the National Research more of their cells are differentiated, and they Council Committee on the Biological Effects are more likely to die of unrelated causes dur- of Ionizing Radiation (BEIR VII) of the ing the latent period of tumor development. National Academy of Sciences. The BEIR VII No one yet has directly studied the effects states that an effective dose of 10 mSv (1 rem) of diagnostic radiation on humans. Theories to a working-age adult results in a 1 in 1,000 about the damaging effects of diagnostic radi- lifetime risk of developing radiation-induced ation are based on studies of populations such cancer. Or, if 10,000 adults receive this dose, as atomic bomb survivors, patients with anky- around 10 of them will develop radiation- losing spondylitis and mastitis treated with induced cancer during their lifetime. The rel- radiation in the early 20th century, and radi- ative risk is small, however, since 4,200 people um watch-dial painters. Doses were calculated out of 10,000 are expected to develop cancer retrospectively, and most people in these for other reasons. cohorts received effective doses that were much larger than the doses from today’s diag- Pregnancy nostic radiologic procedures. Ionizing radiation can be both carcinogenic and Despite uncertainty about the true risks of teratogenic to the fetus. The National Council exposure to levels of radiation used in diag- on Radiation Protection and the American nostic imaging, the linear no-threshold model College of Obstetricians and Gynecologists is nearly universally accepted. This concept maintain that a cumulative effective dose to the and overwhelming evidence that larger radia- fetus of less than 50 mSv (5 rem) is not associ- tion doses are carcinogenic have led radiolo- ated with any increased risks—and none of the gists to follow the principle of using the low- studies listed in TABLE 1 exceeds this. est possible radiation dose necessary to pro- Nevertheless, the use of diagnostic imag- vide the diagnostic information that answers ing in pregnant patients requires careful con- the clinical question. A cumulative sideration. The fetus is most sensitive to the effective dose teratogenic effects of ionizing radiation during ■ WAYS TO MINIMIZE PATIENT EXPOSURE organogenesis, ie, from the second to the to the fetus eighth week of development. But exposure at When ordering a diagnostic radiologic proce- of < 50 mSv is even up to 20 weeks of development increases dure, consider the following principles: the risk of microcephaly, mental retardation, Use CT with discretion. CT accounts for considered safe and growth retardation, and radiation expo- two thirds of the cumulative patient dose from sure at all gestational ages increases the risk of diagnostic radiologic procedures. The cumula- childhood leukemia. tive dose from CT is rising as technological advances increase the number of indications ■ THE LIMITATIONS OF RISK ASSESSMENT for and the capabilities of CT. For example, the newer machines with multiple detectors The BEIR VII risk estimate and the concept are faster than the older machines, allowing of effective dose have significant limitations. imaging in multiple phases after contrast Most importantly, they do not consider age, a administration. CT urography consists of very important factor for several reasons. First, three consecutive CT examinations of the solid tumors have an asymptomatic latent abdomen and pelvis, and it exposes the phase, usually 10 to 40 years. Second, rapidly patient to the highest radiation dose of any dividing and undifferentiated cells are more commonly used diagnostic imaging studies sensitive to radiation than are fully differenti- (TABLE 1). ated cells, and therefore younger patients are Minimize imaging of pregnant women. much more vulnerable to the carcinogenic For example, consider renal ultrasonography effects of radiation than are older patients, as rather than CT of the abdomen and pelvis to more of their cells are still dividing, and they assess for urinary obstruction resulting from

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 73 • NUMBER 6 JUNE 2006 585

Downloaded from www.ccjm.org on September 29, 2021. For personal use only. All other uses require permission. RADIATION DOSE LOCKWOOD AND COLLEAGUES

suspected renal or ureteral calculi. If CT is breast cancer, have a risk of dying from breast absolutely necessary, then a single, low-dose cancer two to four times that of women with- CT scan of the abdomen and pelvis is pre- out this risk factor. She therefore stands to ferred. If urography is indicated, conventional benefit even more. excretory urography is likely to entail a lower Also worth mentioning to such patients is radiation dose than CT urography. that mammography is one of the most tightly If a pregnant woman requires imaging, regulated diagnostic tests, and the radiation specific procedures will minimize fetal expo- doses used are very small (TABLE 1) and are get- sure, including lead shielding of the abdomen ting smaller with technical advances. and pelvis and low-dose techniques. In short, the expected benefit of screening Consider nuclear medicine studies that mammography in patients such as this far use radiopharmaceuticals with a lower radia- exceeds the risks. tion dose. For example, most of the effective This patient’s concerns should remind us dose of a dual-isotope cardiac stress test comes that exposure to ionizing radiation is associ- from the thallium. A two-stage study with ated with small but real risks, that children, technetium uses one third of the dose. young adults, and pregnant women are espe- Minimize imaging of the young. Risks cially vulnerable, and that exposure in diag- from radiation exposure are higher in children nostic imaging should be kept as low as pos- and young adults, as these patients are likely sible while still answering the clinical ques- to survive the latent period of cancer develop- tion. ment. Medical physicists in hospital depart- Avoid studies that do not influence ments of radiology can provide specific infor- patient care, such as plain radiography for sus- mation about radiation doses of common diag- pected rib and coccyx fractures, and lumbar nostic procedures. spine radiography in a patient without radicu- lopathy, which uses an exceptionally high ■ SUGGESTED READING effective dose for a plain radiographic study Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of (TABLE 1). radiation-induced fatal cancer from pediatric CT. AJR Am J Consider alternatives to ionizing radia- Roentgenol 2001; 176:289–296. tion. Ultrasonography and MRI as yet have no Bushberg JT. The Essential Physics of , 2nd ed. practically demonstrated adverse effects. Also, Philadelphia: Lippincott Williams & Wilkins, 2002. direct visualization by endoscopy or laryn- Bushong SC. Radiologic Science for Technologists. St. Louis: goscopy can often answer a clinical question Mosby, 1997. without any radiation. Cardenosa G. Breast Imaging: the Core Curriculum. Consider whether follow-up diagnostic Philadelphia: Lippincott Williams & Wilkins, 2004. radiologic studies are truly necessary and Committee to Assess Health Risks from Exposure to Low Levels what the appropriate follow-up interval of Ionizing Radiation, National Research Council. Health risks should be. Doubling the follow-up interval for from exposure to low levels of ionizing radiation: BEIR VII phase 2 (Free Executive Summary). www.nap.edu/catalog/11340.html. regular examinations halves the cumulative effective dose. Dawson P. Patient dose in multislice CT: why is it increasing and does it matter? Br J Radiol 2004; 77(Spec No 1):s10–s13. When in doubt, consult with a medical physicist or radiologist. Kalra MK, Maher MM, Toth TL, et al. Strategies for CT radia- tion dose optimization. Radiology 2004; 230:619–628.

Nicklas AH, Baker ME. Imaging strategies in the pregnant can- ■ WHAT TO TELL THIS PATIENT cer patient. Semin Oncol 2000; 27:623–632.

Saha GB. Fundamentals of Nuclear Pharmacy, 5th ed. New One in every eight women will develop breast York: Springer, 2004: 201. cancer, and one in every 30 will die of it. Toppenberg KS, Hill DA, Miller DP. Safety of radiographic Clinical trials have shown that screening imaging during pregnancy. Am Fam Physician 1999; mammography is associated with a 20% to 59:1813–1818. 40% reduction in the rate of death from breast ADDRESS: David Einstein, MD, Department of cancer. Women such as our hypothetical Diagnostic Radiology, Hb6, Cleveland Clinic, 9500 Euclid patient, with a first-degree relative who had Avenue, Cleveland, OH 44195.

586 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 73 • NUMBER 6 JUNE 2006

Downloaded from www.ccjm.org on September 29, 2021. For personal use only. All other uses require permission.