Instituto Nacional de Enfermedades Respiratorias, Mexico City
http://www.100md.com
《美国呼吸和危急护理医学》
Departments of Pediatric Pulmonology and Allergology and Pediatric Radiology, Erasmus MC–Sophia Rotterdam, Rotterdam, The Netherlands
James Hogg iCAPTURE Centre for Cardiovascular and Pulmonary Research, St. Paul's Hospital
Pulmonary Division, Department of Medicine, University of British Columbia
Subdivision of Radiation Physics, Department of Radiology, Vancouver General Hospital, Vancouver, British Columbia
Department of Radiology and Diagnostic Imaging, University of Alberta, Edmonton, Alberta, Canada
Department of Cardiovascular and Respiratory Medicine, Shiga University of Medical Science, Shiga, Japan
ABSTRACT
Rationale: Low-dose radiation from computed tomography (CT) may increase the risk of certain cancers, especially in children.
Objective: We sought to estimate the excess all-cause and cancer-specific mortality, which may be associated with repeated CT scanning of patients with cystic fibrosis (CF).
Methods: The radiation dose was calculated for a published CF surveillance CT scanning protocol of biennial CT scans, and the risk per scan was estimated using atom-bomb survivor data. A computational model was developed to calculate the excess mortality in a CF cohort associated with radiation from the CT scan and to evaluate the effects of background survival, scanning interval, and level of CT radiation used. The model assumed that there would be no survival benefits associated with repeated surveillance CT scanning.
Results: The average radiation dose for the published CT protocol was 1 mSv. Survival reduction associated with annual scans from age 2 yr until death was approximately 1 mo and 2 yr for CF cohorts, with a median survival of 26 and 50 yr, respectively. Corresponding cumulative cancer mortality was approximately 2 and 13% at age 40 and 65 yr, respectively. Biennial CT scanning reduced all-cause and cumulative cancer mortality by half.
Conclusion: Routine lifelong annual CT scans carry a low risk of radiation-induced mortality in CF. However, as the overall survival increases for patients with CF, the risk of radiation-induced mortality may modestly increase. These data indicate that radiation dose must be considered in routine CT imaging strategies for patients with CF, to ensure that benefits outweigh the risks.
Key Words: cancer mortality computational model computed tomography cystic fibrosis radiation
For the justification of clinical protocols and research proposals that include computed tomography (CT), the absolute risk associated with radiation from CT scans has to be weighed against potential or proven benefits. The major risk associated with the low-dose radiation delivered during chest CT scanning is chromosomal damage that can cause hematologic cancers within 5 yr (1–3) and solid cancers in more than 30 yr (3, 4) after the exposure. Firm data determining the absolute cancer mortality risk of CT radiation dose do not exist (1, 5); however, recent data derived from atom bomb survivors enables an estimate of cancer mortality after exposure to radiation (2, 4, 6, 7). Because of the substantial impact of CT imaging on patient management, CT use is increasing (5, 8), even in children and infants who have been shown to be more sensitive to radiation compared with adults (1, 2, 4, 8). Therefore, concerns have been raised regarding the routine use of surveillance CT scans to document disease progression in children with cystic fibrosis (CF) (9–13).
Patients with CF develop progressive, irreversible, structural lung damage early after birth leading to a reduced life expectancy (14–21). Traditionally, chest radiographs, together with spirometric measurements and clinical assessment, have been used to monitor disease progression and interventions in patients with CF. More recently, studies have shown that the structural lung damage in these patients could be detected earlier using chest CT compared with chest radiographs (22, 23). Therefore, certain Dutch (10) and Swedish (24) CF centers are now using thoracic CT scanning every second and every third year, respectively, for surveillance in lieu of annual chest radiography. The main rationale for this change is that CT scans may be more sensitive in detecting subtle but clinically salient disease progression than the current gold standard, FEV1 (10, 24). This perspective may stimulate the routine use of chest CT examinations for surveillance purposes in CF. The argument in favor of routine CT scanning in CF is that earlier disease detection would lead to more aggressive treatment (16, 25) and to prolonged survival. The major concern associated with CT scanning—namely, cancers arising from low-dose radiation exposure emitted during the procedure—has been minimal since there is a significant lag time between exposure and cancer occurrence and, traditionally, patients with CF did not live long enough to develop radiation-related cancers. Thus, the assumption has been that the potential survival advantage of earlier detection would outweigh the potential radiation risks of serial CT scanning.
However, with rapid medical progress over the past half century, the median survival of patients with CF has increased linearly from 1 yr in 1940 to 35 yr in 2004. Further improvements in life expectancy are expected over the next three decades (26). Because the risk of radiation-associated cancers increases with increased longevity, the potential harm from routine CT scanning is likely to be amplified in the future for patients with CF. To date and to our knowledge, there have been no studies that have quantitatively estimated the risk of harm associated with routine CT scanning in CF.
The aim of our study was to develop a computational model to estimate the excess risk for all-cause mortality and cancer-specific causes of mortality associated with a clinical protocol that promotes lifelong use of surveillance chest CT scanning in CF. We hypothesized that the risk associated with lifelong routine CT scanning in CF would be low, but that with continued improvements in survival in patients with CF over the coming decades, the risk associated with routine CT scanning may increase. Part of this study has been published in the form of an abstract (27).
METHODS
CT Dose Calculation
The radiation dose per CT scan was calculated for a published CF protocol (10, 11). Briefly, CT scans were obtained on a GE Prospeed SX scanner (General Electric Medical Systems, Milwaukee, WI) as 1.0-mm-thick images at 10-mm intervals using 120 kV and 160 mA/s (children younger than 9 yr, 120 mA/s). The dose was calculated in milli-Sieverts (mSv) using the imPACT CT dosimetry calculator (National Radiological Protection Board, Harwell, UK) (28) and corrected to pediatric values (29).
Computational Model
We designed a simulation model based on Markov modeling (Figure 1) (30). The mortality in the intervention cohort was broken down to all-cause CF mortality and mortality due to radiation-associated hematologic cancers (2) and due to radiation-associated solid cancers (3); these estimates were obtained from the Biological Effects of Ionizing Radiation committee report VII. The mortality in the control cohort was all-cause CF mortality. We developed seven variations of the model (Table 1). The first four models used data from 1990 when the median survival was approximately 26 yr (31). The fifth model used similar assumptions except for the overall CF survival, which was assumed to be 32 yr (the median survival in 2003 was 23 yr) (16). In addition, we extrapolated CF survival to 2030 when the median survival is expected to reach 50 yr (sixth model) (16, 32). The seventh model was the same as the sixth model, except we used the strategy of scanning these individuals only until the age of 18 yr.
We used 6-mo cycle lengths to establish accurate transitional state probabilities. For each cycle, new probabilities of death were imputed into the models. Patients who died during a cycle were censored from further analysis. Survivors were passed through another cycle wherein a new set of probabilities of death was applied. To determine the robustness of our data, multivariate sensitivity analyses were performed in which simultaneous adjustments for relevant covariates were made. We subjected all probabilities to a 10% variation around the probabilities of death using a triangular distribution (33). We sampled all variables on the basis of their distribution and produced 100,000 sample sets in a Monte Carlo simulation. All modeled simulations were conducted using Data Pro (TreeAge Software, Inc., Williamstown, MA). A more detailed explanation of Markov modeling and the triangular distribution is provided in the online supplement.
RESULTS
Dose Calculation
CT dose was calculated for 58 children with CF aged 9.9 ± 3.9 yr, mean ± SD (range, 3.5–17.3 yr) (10). Their height and weight were 1.4 ± 0.2 (1.0–1.8) m and 32.1 ± 12.3 (14.8–61.9) kg, respectively. The number of images per CT scan was 24 ± 4 (16–34). The calculated dose per CT scan was 1.0 ± 0.3 (0.5–1.9) mSv. For model 1, we used 1 mSv per CT scan.
Survival and Cancer Mortality after Repetitive CT Scanning
The reduction in median survival was 1 mo in patients with CF who had a median survival of 26 yr and received annual CT scans (1 mSv) from age 2 yr and on. An increased median survival to 32 yr resulted in a survival reduction of 2 mo. However, when the median CF survival increased to 50 yr, the survival reduction increased to 2 yr for both men and women. If CT scans in these patients were used for early detection of disease progression until the age of 18 yr, the survival reduction was 1 yr (Figure 2). When the time interval between the CT scans was increased to 2 yr, the risk decreased by a factor of 2. A fivefold decrease in the dose per CT scan (0.2 mSv and 5 mSv) resulted in approximately a fivefold decrease in mortality, whereas a fivefold increase in the dose per scan increased mortality by fivefold. For all modeled scenarios, men had a larger reduction in survival than did women (Table 2).
The survival reduction was driven by an excess number of deaths from hematologic and solid cancers (Table 3). When the median CF survival was relatively short, the cumulative risk of all cancer deaths was between 1 and 2% by age 40 yr for both sexes at an exposure level of 1 mSv. However, when CF survival increased to a median of 50 yr (models 6 and 7), the combined cumulative mortality from hematologic and solid cancers was approximately 13% for men and women when annual CT scans were used from the age of 2 yr and onwards. The risk decreased to approximately 7% when CT scans were discontinued at age 18 yr.
DISCUSSION
This study estimated the excess mortality associated with low-dose radiation from routine lifelong chest CT scanning in patients with CF. We used a cohort of patients with CF and varied the radiation dose per scan, scanning interval, and background CF survival to determine the survival effects of using different scanning strategies. An important assumption in our model was that CT scans would not provide clinical benefits that would improve survival in patients with CF. We made this assumption for two reasons. First, there is a paucity of studies with estimates of survival benefits related to routine CT scanning in CF. Second, the major objective of our study was to estimate the risk posed by low-dose radiation related to CT scanning, which has not been previously studied in the CF population. This question has both clinical and public health policy importance as a growing number of CF centers (9, 10) have adopted a strategy to use routine CT scans every second or third year in managing patients with CF during childhood.
We found that in the models in which the radiation dose per CT scan was 1 mSv and in which the overall expected median CF survival was 32 yr or less, the risks imposed by low-dose radiation from routine CT scanning were low. However, when the expected survival of the nonscanned (control) population increased to a median of 50 yr, the risk associated with lifelong routine CT scans became substantial. For instance, in model 6, the difference in median survival between those scanned and nonscanned was 2 yr. This reduction was driven largely by a 12% excess mortality from solid cancers. When CT scanning was used only for early disease detection (model 7), the excess mortality from solid cancers was greater than 6%. According to the latest projections based on data from the United States (16, 32), the median CF survival is expected to increase to 50 yr by 2030. Under this scenario, patients with CF born today may be expected to live to age 50 yr and beyond, which makes the current study highly relevant. Given the potential reduction in survival associated with radiation exposure from routine CT scans, it would be important to further demonstrate the clinical benefits related to this approach to justify the use of routine lifelong CT scanning in patients with CF. Our data may also be germane and applicable to other clinical settings in which routine CT scanning is used. In our models, the minimum number of CT scans per patients was 17 (model 7) and at age 65 yr, the cumulative cancer mortality associated with this strategy was approximately 7%. Notably, the present study showed that biennial CT scanning could reduce the risk by half compared with the strategy of using annual CT scans. Moreover, by reducing the radiation dose fivefold, the risks can also be substantially reduced. This shows that it is important to consider the dose and the frequency of CT scanning for routine clinical purposes. A full-lung volumetric CT scan of the chest would expose the subject to approximately 5 mSv (model 4). Three inspiratory and three expiratory images would expose the subject to 0.2 mSv (model 3). A fivefold reduction in the milliamperes per second would be another option to reduce the radiation dose to approximately 0.2 mSv. The findings from this study highlight the importance of adopting novel imaging procedures for surveillance purposes only when the burden of radiation exposure is below acceptable risk levels and when the potential adverse effects of the procedures are likely to be outweighed by their benefits.
There are several limitations to our study. First, our data may have overestimated the mortality risks related to low-dose radiation exposure since the atom-bomb survivor data were obtained before the advent of newer therapies to treat hematologic cancers, which have improved the survival of such patients (34). Second, we did not consider the negative effects of radiation on noncancer mortality, like heart and blood vessel disease, since there is no evidence of an increased risk for such diseases at the doses that are emitted during CT scanning (4). Third, we did not take costs or potential benefits of CT scanning into account since we aimed to estimate only to estimate the potential mortality risks associated with low-dose radiation exposure from routine CT scanning. Consideration of costs, in addition to safety issues, will be important in formulating a coherent public health policy in the use of routine CT scans in CF. Fourth, the data we used to model cancer risks were obtained from the general population, not from the CF population. It is, however, reassuring that several studies have shown that the cancer risks in CF are comparable to the normal population (35). Fifth, our data should be interpreted in the context of other sources of radiation. In a Swiss survey, the dose of a lateral plus anteroposterior chest radiograph was about 0.17 mSv and of volumetric chest CT scan was 9.0 mSv (36). The dose for this scan is 10 times the dose of a high-resolution CT scan (36). Therefore the dose for a high-resolution CT would be five times the dose of chest radiographs. Interestingly, 140 to 350 transatlantic flying hours in a subsonic aircraft or 60 to 150 transatlantic flying hours in a Concorde results in a radiation exposure of 1 mSv (37). Finally, natural background radiation in the United States due to cosmic radiation, natural radioactivity, and domestic radon results in an average exposure of approximately 3 mSv/yr. In sum, although our models used the best radiation risk estimates currently available (3), these data, nevertheless, must be interpreted cautiously since there is a substantial degree of uncertainty in the radiation risk estimates.
In conclusion, our model indicated that the risk of routine lifelong annual CT scanning in CF is low but will increase as the general survival of patients with CF improves. Our data urge caution in using CT scans for lifelong surveillance purposes in any disease state unless there is a clear understanding of the benefits. There are certain strategies by which radiation exposure from CT scans can be reduced. Additional investigations in dose reduction as well as the optimal timing of routine CT scans are needed to further mitigate the risk of radiation-induced morbidity and mortality. Our computational model may be used to determine the impact of technologic improvements on lifelong risks and whether protocols are sufficient to keep lifetime exposure within acceptable limits.
Acknowledgments
The authors thank Peter D. Pare for his guidance throughout this project.
FOOTNOTES
Supported by the British Columbia Lung Association and the Canadian Institute of Health Research/Michael Smith foundation (P.A.d.J.). H.O.C. is a Parker B. Francis Fellow in Pulmonary Research. D.D.S. holds a Canada Research Chair (COPD) and a Michael Smith/St. Paul's Hospital Foundation Professorship in COPD.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.200505-810OC on October 27, 2005
Conflict of Interest Statement: P.A.d.J. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.R.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Y.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.H.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.A.W.M.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.O.C. received $2,500 in 2002 and ￡1,500 in 2003 for serving on an advisory board for GlaxoSmithKline (GSK). In addition. he is a coinvestigator on two multicenter studies sponsored by GSK and has received travel expenses to attend meetings related to the project. A percentage of his salary between 2003 and 2006 ($15,000/yr) derives from contract funds provided to a colleague, Peter D. Pare, by GSK for the development of validated methods to measure emphysema and airway disease using computed tomography. D.D.S. has received honoraria for speaking engagements from AstraZeneca in 2003 ($4,000) and in 2004 ($3,000) and from GSK in 2003 ($4,000) and in 2004 ($8,000). He has also received unrestricted research funding as either the principal investigator or co-principal investigator from GSK in 2002 for $100,000, in 2003 for $80,000 and in 2004 for $1.5 million. He has also received $3,500 from GSK for consultancy work.
REFERENCES
Brenner DJ. Estimating cancer risks from pediatric CT: going from the qualitative to the quantitative. Pediatr Radiol 2002;32:228–233. [Discussion, 242–244.]
Pierce DA, Shimizu Y, Preston DL, Vaeth M, Mabuchi K. Studies of the mortality of atomic bomb survivors: report 12, part I. Cancer: 1950–1990. Radiat Res 1996;146:1–27.
Biological Effects of Ionizing Radiation report VII. Committee to assess health risks from exposure of low levels of ionizing radiation. National Academy of Sciences. Available from: http://www.nap.edu/openbook/030909156X/html/index.html (accessed September 15, 2005).
Preston DL, Shimizu Y, Pierce DA, Suyama A, Mabuchi K. Studies of mortality of atomic bomb survivors: report 13: solid cancer and noncancer disease mortality: 1950–1997. Radiat Res 2003;160:381–407.
Berrington de Gonzalez A, Darby S. Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet 2004;363:345–351.
Puskin JS, Nelson NS, Nussbaum RH, Pierce DA, Preston DL. Risks from low doses of radiation. Science 1996;272:631–635.
Pierce DA, Preston DL. Risks from low doses of radiation. Science 1996;272:632–633.
Mayo JR, Aldrich J, Muller NL. Radiation exposure at chest CT: a statement of the Fleischner Society. Radiology 2003;228:15–21.
Dakin CJ, Pereira JK, Henry RL, Wang H, Morton JR. Relationship between sputum inflammatory markers, lung function, and lung pathology on high-resolution computed tomography in children with cystic fibrosis. Pediatr Pulmonol 2002;33:475–482.
de Jong PA, Nakano Y, Lequin MH, Mayo JR, Woods R, Pare PD, Tiddens HA. Progressive damage on high resolution computed tomography despite stable lung function in cystic fibrosis. Eur Respir J 2004;23:93–97.
de Jong PA, Ottink MD, Robben SG, Lequin MH, Hop WC, Hendriks JJ, Pare PD, Tiddens HA. Pulmonary disease assessment in cystic fibrosis: comparison of CT scoring systems and value of bronchial and arterial dimension measurements. Radiology 2004;231:434–439.
de Jong PA, Lequin MH, Mayo JR, Pare PD, Tiddens H. Re: progressive damage on high-resolution computed tomography. Eur Respir J 2004;24:1071–1072.
Rawlings D, Tennant D, Furness J. Progressive damage on high-resolution computed tomography. Eur Respir J 2004;24:1071.
Bedrossian CW, Greenberg SD, Singer DB, Hansen JJ, Rosenberg HS. The lung in cystic fibrosis: a quantitative study including prevalence of pathologic findings among different age groups. Hum Pathol 1976;7:195–204.
Chow CW, Landau LI, Taussig LM. Bronchial mucous glands in the newborn with cystic fibrosis. Eur J Pediatr 1982;139:240–243.
Gibson RL, Burns JL, Ramsey BW. Pathophysiology and management of pulmonary infections in cystic fibrosis. Am J Respir Crit Care Med 2003;168:918–951.
Tiddens HA. Detecting early structural lung damage in cystic fibrosis. Pediatr Pulmonol 2002;34:228–231.
Long FR, Williams RS, Castile RG. Structural airway abnormalities in infants and young children with cystic fibrosis. J Pediatr 2004;144:154–161.
Brody AS, Klein JS, Molina PL, Quan J, Bean JA, Wilmott RW. High-resolution computed tomography in young patients with cystic fibrosis: distribution of abnormalities and correlation with pulmonary function tests. J Pediatr 2004;145:32–38.
Goris ML, Zhu HJ, Blankenberg F, Chan F, Robinson TE. An automated approach to quantitative air trapping measurements in mild cystic fibrosis. Chest 2003;123:1655–1663.
Bonnel AS, Song SM, Kesavarju K, Newaskar M, Paxton CJ, Bloch DA, Moss RB, Robinson TE. Quantitative air-trapping analysis in children with mild cystic fibrosis lung disease. Pediatr Pulmonol 2004;38:396–405.
Hansell DM, Strickland B. High-resolution computed tomography in pulmonary cystic fibrosis. Br J Radiol 1989;62:1–5.
Stiglbauer R, Schurawitzki H, Eichler I, Vergesslich KA, Gotz M. High resolution CT in children with cystic fibrosis. Acta Radiol 1992;33:548–553.
de Jong PA, Lindblad A, Rubin L, Hop WC, de Jongste JC, Brink M, Tiddens HA. Longitudinal evaluation of lung structure on computed tomography scans and lung function in children and adults with cystic fibrosis. Thorax 2005 Oct 21 [Epub ahead of print].
Fiel SB, FitzSimmons SC, Schidlow DV. Evolving demographics of cystic fibrosis. Am J Respir Crit Care Med 1994;15:349–355.
Fiel SB, Golmohammadi K, Schidlow DV. Evolving demographics of cystic fibrosis . Proc Am Thorac Soc 2005;2:A576.
de Jong PA, Golmohammadi K, Nakano Y, Mayo JR, Lequin MH, Tiddens HA, Aldrich J, Pare PD, Sin DD. Modelling the mortality risk associated with radiation exposure from cumulative computed tomography in cystic fibrosis . Proc Am Thorac Soc 2005;2:A576.
Jones D, Shrimpton PC. Survey of CT practice in the UK: normalised organ doses for x-ray computed tomography calculated using Monte Carlo techniques. Harwell, UK: National Radiological Protection Board; 1991. Available from: www.impactscan.org/ctdosimetry.htm (accessed March 15, 2005).
Khursheed A, Hillier MC, Shrimpton PC, Wall BF. Influence of patient age on normalized effective doses calculated for CT examinations. Br J Radiol 2002;75:819–830.
Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322–338.
Kulich M, Rosenfeld M, Goss C, Wilmott RW. Improved survival among young patients with cystic fibrosis. J Pediatr 2003;142:631–636.
Fogarty A, Hubbard R, Britton J. International comparison of median age at death from cystic fibrosis. Chest 2000;117:1656–1660.
Evans M, Hastings N, Peacock B. Triangular distributions. In: Statistical distributions, 3rd ed. New York: Wiley; 2000. pp. 187–188.
Sporn M. The war on cancer. Lancet 1996;347:1377–1381.
Neglia JP, FitzSimmons SC, Maisonneuve P, Schoni MH, Schoni-Affolter F, Corey M, Lowenfels AB. The risk of cancer among patients with cystic fibrosis. Cystic Fibrosis and Cancer Study Group. N Engl J Med 1995;332:494–499.
Aroua A, Bize R, Buchillier-Decka I, Vader JP, Valley JP, Schnyder P. X-ray imaging of the chest in Switzerland in 1998: a natiowide survey. Eur Radiol 2003;13:1250–1259.
Bottollier-Depois JF, Chau Q, Bouisset P, Kerlau G, Plawinski L, Lebaron-Jacobs L. Assessing exposure to cosmic radiation on board aircraft. Adv Space Res 2003;32:59–66.
Rosenfeld M, Davis R, FitzSimmons SC, Pepe M, Ramsey B. Gender gap in cystic fibrosis mortality. Am J Epidemiol 1997;145:794–803.(Pim A. de Jong, John R. Mayo, Kamran Gol)
James Hogg iCAPTURE Centre for Cardiovascular and Pulmonary Research, St. Paul's Hospital
Pulmonary Division, Department of Medicine, University of British Columbia
Subdivision of Radiation Physics, Department of Radiology, Vancouver General Hospital, Vancouver, British Columbia
Department of Radiology and Diagnostic Imaging, University of Alberta, Edmonton, Alberta, Canada
Department of Cardiovascular and Respiratory Medicine, Shiga University of Medical Science, Shiga, Japan
ABSTRACT
Rationale: Low-dose radiation from computed tomography (CT) may increase the risk of certain cancers, especially in children.
Objective: We sought to estimate the excess all-cause and cancer-specific mortality, which may be associated with repeated CT scanning of patients with cystic fibrosis (CF).
Methods: The radiation dose was calculated for a published CF surveillance CT scanning protocol of biennial CT scans, and the risk per scan was estimated using atom-bomb survivor data. A computational model was developed to calculate the excess mortality in a CF cohort associated with radiation from the CT scan and to evaluate the effects of background survival, scanning interval, and level of CT radiation used. The model assumed that there would be no survival benefits associated with repeated surveillance CT scanning.
Results: The average radiation dose for the published CT protocol was 1 mSv. Survival reduction associated with annual scans from age 2 yr until death was approximately 1 mo and 2 yr for CF cohorts, with a median survival of 26 and 50 yr, respectively. Corresponding cumulative cancer mortality was approximately 2 and 13% at age 40 and 65 yr, respectively. Biennial CT scanning reduced all-cause and cumulative cancer mortality by half.
Conclusion: Routine lifelong annual CT scans carry a low risk of radiation-induced mortality in CF. However, as the overall survival increases for patients with CF, the risk of radiation-induced mortality may modestly increase. These data indicate that radiation dose must be considered in routine CT imaging strategies for patients with CF, to ensure that benefits outweigh the risks.
Key Words: cancer mortality computational model computed tomography cystic fibrosis radiation
For the justification of clinical protocols and research proposals that include computed tomography (CT), the absolute risk associated with radiation from CT scans has to be weighed against potential or proven benefits. The major risk associated with the low-dose radiation delivered during chest CT scanning is chromosomal damage that can cause hematologic cancers within 5 yr (1–3) and solid cancers in more than 30 yr (3, 4) after the exposure. Firm data determining the absolute cancer mortality risk of CT radiation dose do not exist (1, 5); however, recent data derived from atom bomb survivors enables an estimate of cancer mortality after exposure to radiation (2, 4, 6, 7). Because of the substantial impact of CT imaging on patient management, CT use is increasing (5, 8), even in children and infants who have been shown to be more sensitive to radiation compared with adults (1, 2, 4, 8). Therefore, concerns have been raised regarding the routine use of surveillance CT scans to document disease progression in children with cystic fibrosis (CF) (9–13).
Patients with CF develop progressive, irreversible, structural lung damage early after birth leading to a reduced life expectancy (14–21). Traditionally, chest radiographs, together with spirometric measurements and clinical assessment, have been used to monitor disease progression and interventions in patients with CF. More recently, studies have shown that the structural lung damage in these patients could be detected earlier using chest CT compared with chest radiographs (22, 23). Therefore, certain Dutch (10) and Swedish (24) CF centers are now using thoracic CT scanning every second and every third year, respectively, for surveillance in lieu of annual chest radiography. The main rationale for this change is that CT scans may be more sensitive in detecting subtle but clinically salient disease progression than the current gold standard, FEV1 (10, 24). This perspective may stimulate the routine use of chest CT examinations for surveillance purposes in CF. The argument in favor of routine CT scanning in CF is that earlier disease detection would lead to more aggressive treatment (16, 25) and to prolonged survival. The major concern associated with CT scanning—namely, cancers arising from low-dose radiation exposure emitted during the procedure—has been minimal since there is a significant lag time between exposure and cancer occurrence and, traditionally, patients with CF did not live long enough to develop radiation-related cancers. Thus, the assumption has been that the potential survival advantage of earlier detection would outweigh the potential radiation risks of serial CT scanning.
However, with rapid medical progress over the past half century, the median survival of patients with CF has increased linearly from 1 yr in 1940 to 35 yr in 2004. Further improvements in life expectancy are expected over the next three decades (26). Because the risk of radiation-associated cancers increases with increased longevity, the potential harm from routine CT scanning is likely to be amplified in the future for patients with CF. To date and to our knowledge, there have been no studies that have quantitatively estimated the risk of harm associated with routine CT scanning in CF.
The aim of our study was to develop a computational model to estimate the excess risk for all-cause mortality and cancer-specific causes of mortality associated with a clinical protocol that promotes lifelong use of surveillance chest CT scanning in CF. We hypothesized that the risk associated with lifelong routine CT scanning in CF would be low, but that with continued improvements in survival in patients with CF over the coming decades, the risk associated with routine CT scanning may increase. Part of this study has been published in the form of an abstract (27).
METHODS
CT Dose Calculation
The radiation dose per CT scan was calculated for a published CF protocol (10, 11). Briefly, CT scans were obtained on a GE Prospeed SX scanner (General Electric Medical Systems, Milwaukee, WI) as 1.0-mm-thick images at 10-mm intervals using 120 kV and 160 mA/s (children younger than 9 yr, 120 mA/s). The dose was calculated in milli-Sieverts (mSv) using the imPACT CT dosimetry calculator (National Radiological Protection Board, Harwell, UK) (28) and corrected to pediatric values (29).
Computational Model
We designed a simulation model based on Markov modeling (Figure 1) (30). The mortality in the intervention cohort was broken down to all-cause CF mortality and mortality due to radiation-associated hematologic cancers (2) and due to radiation-associated solid cancers (3); these estimates were obtained from the Biological Effects of Ionizing Radiation committee report VII. The mortality in the control cohort was all-cause CF mortality. We developed seven variations of the model (Table 1). The first four models used data from 1990 when the median survival was approximately 26 yr (31). The fifth model used similar assumptions except for the overall CF survival, which was assumed to be 32 yr (the median survival in 2003 was 23 yr) (16). In addition, we extrapolated CF survival to 2030 when the median survival is expected to reach 50 yr (sixth model) (16, 32). The seventh model was the same as the sixth model, except we used the strategy of scanning these individuals only until the age of 18 yr.
We used 6-mo cycle lengths to establish accurate transitional state probabilities. For each cycle, new probabilities of death were imputed into the models. Patients who died during a cycle were censored from further analysis. Survivors were passed through another cycle wherein a new set of probabilities of death was applied. To determine the robustness of our data, multivariate sensitivity analyses were performed in which simultaneous adjustments for relevant covariates were made. We subjected all probabilities to a 10% variation around the probabilities of death using a triangular distribution (33). We sampled all variables on the basis of their distribution and produced 100,000 sample sets in a Monte Carlo simulation. All modeled simulations were conducted using Data Pro (TreeAge Software, Inc., Williamstown, MA). A more detailed explanation of Markov modeling and the triangular distribution is provided in the online supplement.
RESULTS
Dose Calculation
CT dose was calculated for 58 children with CF aged 9.9 ± 3.9 yr, mean ± SD (range, 3.5–17.3 yr) (10). Their height and weight were 1.4 ± 0.2 (1.0–1.8) m and 32.1 ± 12.3 (14.8–61.9) kg, respectively. The number of images per CT scan was 24 ± 4 (16–34). The calculated dose per CT scan was 1.0 ± 0.3 (0.5–1.9) mSv. For model 1, we used 1 mSv per CT scan.
Survival and Cancer Mortality after Repetitive CT Scanning
The reduction in median survival was 1 mo in patients with CF who had a median survival of 26 yr and received annual CT scans (1 mSv) from age 2 yr and on. An increased median survival to 32 yr resulted in a survival reduction of 2 mo. However, when the median CF survival increased to 50 yr, the survival reduction increased to 2 yr for both men and women. If CT scans in these patients were used for early detection of disease progression until the age of 18 yr, the survival reduction was 1 yr (Figure 2). When the time interval between the CT scans was increased to 2 yr, the risk decreased by a factor of 2. A fivefold decrease in the dose per CT scan (0.2 mSv and 5 mSv) resulted in approximately a fivefold decrease in mortality, whereas a fivefold increase in the dose per scan increased mortality by fivefold. For all modeled scenarios, men had a larger reduction in survival than did women (Table 2).
The survival reduction was driven by an excess number of deaths from hematologic and solid cancers (Table 3). When the median CF survival was relatively short, the cumulative risk of all cancer deaths was between 1 and 2% by age 40 yr for both sexes at an exposure level of 1 mSv. However, when CF survival increased to a median of 50 yr (models 6 and 7), the combined cumulative mortality from hematologic and solid cancers was approximately 13% for men and women when annual CT scans were used from the age of 2 yr and onwards. The risk decreased to approximately 7% when CT scans were discontinued at age 18 yr.
DISCUSSION
This study estimated the excess mortality associated with low-dose radiation from routine lifelong chest CT scanning in patients with CF. We used a cohort of patients with CF and varied the radiation dose per scan, scanning interval, and background CF survival to determine the survival effects of using different scanning strategies. An important assumption in our model was that CT scans would not provide clinical benefits that would improve survival in patients with CF. We made this assumption for two reasons. First, there is a paucity of studies with estimates of survival benefits related to routine CT scanning in CF. Second, the major objective of our study was to estimate the risk posed by low-dose radiation related to CT scanning, which has not been previously studied in the CF population. This question has both clinical and public health policy importance as a growing number of CF centers (9, 10) have adopted a strategy to use routine CT scans every second or third year in managing patients with CF during childhood.
We found that in the models in which the radiation dose per CT scan was 1 mSv and in which the overall expected median CF survival was 32 yr or less, the risks imposed by low-dose radiation from routine CT scanning were low. However, when the expected survival of the nonscanned (control) population increased to a median of 50 yr, the risk associated with lifelong routine CT scans became substantial. For instance, in model 6, the difference in median survival between those scanned and nonscanned was 2 yr. This reduction was driven largely by a 12% excess mortality from solid cancers. When CT scanning was used only for early disease detection (model 7), the excess mortality from solid cancers was greater than 6%. According to the latest projections based on data from the United States (16, 32), the median CF survival is expected to increase to 50 yr by 2030. Under this scenario, patients with CF born today may be expected to live to age 50 yr and beyond, which makes the current study highly relevant. Given the potential reduction in survival associated with radiation exposure from routine CT scans, it would be important to further demonstrate the clinical benefits related to this approach to justify the use of routine lifelong CT scanning in patients with CF. Our data may also be germane and applicable to other clinical settings in which routine CT scanning is used. In our models, the minimum number of CT scans per patients was 17 (model 7) and at age 65 yr, the cumulative cancer mortality associated with this strategy was approximately 7%. Notably, the present study showed that biennial CT scanning could reduce the risk by half compared with the strategy of using annual CT scans. Moreover, by reducing the radiation dose fivefold, the risks can also be substantially reduced. This shows that it is important to consider the dose and the frequency of CT scanning for routine clinical purposes. A full-lung volumetric CT scan of the chest would expose the subject to approximately 5 mSv (model 4). Three inspiratory and three expiratory images would expose the subject to 0.2 mSv (model 3). A fivefold reduction in the milliamperes per second would be another option to reduce the radiation dose to approximately 0.2 mSv. The findings from this study highlight the importance of adopting novel imaging procedures for surveillance purposes only when the burden of radiation exposure is below acceptable risk levels and when the potential adverse effects of the procedures are likely to be outweighed by their benefits.
There are several limitations to our study. First, our data may have overestimated the mortality risks related to low-dose radiation exposure since the atom-bomb survivor data were obtained before the advent of newer therapies to treat hematologic cancers, which have improved the survival of such patients (34). Second, we did not consider the negative effects of radiation on noncancer mortality, like heart and blood vessel disease, since there is no evidence of an increased risk for such diseases at the doses that are emitted during CT scanning (4). Third, we did not take costs or potential benefits of CT scanning into account since we aimed to estimate only to estimate the potential mortality risks associated with low-dose radiation exposure from routine CT scanning. Consideration of costs, in addition to safety issues, will be important in formulating a coherent public health policy in the use of routine CT scans in CF. Fourth, the data we used to model cancer risks were obtained from the general population, not from the CF population. It is, however, reassuring that several studies have shown that the cancer risks in CF are comparable to the normal population (35). Fifth, our data should be interpreted in the context of other sources of radiation. In a Swiss survey, the dose of a lateral plus anteroposterior chest radiograph was about 0.17 mSv and of volumetric chest CT scan was 9.0 mSv (36). The dose for this scan is 10 times the dose of a high-resolution CT scan (36). Therefore the dose for a high-resolution CT would be five times the dose of chest radiographs. Interestingly, 140 to 350 transatlantic flying hours in a subsonic aircraft or 60 to 150 transatlantic flying hours in a Concorde results in a radiation exposure of 1 mSv (37). Finally, natural background radiation in the United States due to cosmic radiation, natural radioactivity, and domestic radon results in an average exposure of approximately 3 mSv/yr. In sum, although our models used the best radiation risk estimates currently available (3), these data, nevertheless, must be interpreted cautiously since there is a substantial degree of uncertainty in the radiation risk estimates.
In conclusion, our model indicated that the risk of routine lifelong annual CT scanning in CF is low but will increase as the general survival of patients with CF improves. Our data urge caution in using CT scans for lifelong surveillance purposes in any disease state unless there is a clear understanding of the benefits. There are certain strategies by which radiation exposure from CT scans can be reduced. Additional investigations in dose reduction as well as the optimal timing of routine CT scans are needed to further mitigate the risk of radiation-induced morbidity and mortality. Our computational model may be used to determine the impact of technologic improvements on lifelong risks and whether protocols are sufficient to keep lifetime exposure within acceptable limits.
Acknowledgments
The authors thank Peter D. Pare for his guidance throughout this project.
FOOTNOTES
Supported by the British Columbia Lung Association and the Canadian Institute of Health Research/Michael Smith foundation (P.A.d.J.). H.O.C. is a Parker B. Francis Fellow in Pulmonary Research. D.D.S. holds a Canada Research Chair (COPD) and a Michael Smith/St. Paul's Hospital Foundation Professorship in COPD.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.200505-810OC on October 27, 2005
Conflict of Interest Statement: P.A.d.J. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.R.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Y.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.H.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.A.W.M.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.O.C. received $2,500 in 2002 and ￡1,500 in 2003 for serving on an advisory board for GlaxoSmithKline (GSK). In addition. he is a coinvestigator on two multicenter studies sponsored by GSK and has received travel expenses to attend meetings related to the project. A percentage of his salary between 2003 and 2006 ($15,000/yr) derives from contract funds provided to a colleague, Peter D. Pare, by GSK for the development of validated methods to measure emphysema and airway disease using computed tomography. D.D.S. has received honoraria for speaking engagements from AstraZeneca in 2003 ($4,000) and in 2004 ($3,000) and from GSK in 2003 ($4,000) and in 2004 ($8,000). He has also received unrestricted research funding as either the principal investigator or co-principal investigator from GSK in 2002 for $100,000, in 2003 for $80,000 and in 2004 for $1.5 million. He has also received $3,500 from GSK for consultancy work.
REFERENCES
Brenner DJ. Estimating cancer risks from pediatric CT: going from the qualitative to the quantitative. Pediatr Radiol 2002;32:228–233. [Discussion, 242–244.]
Pierce DA, Shimizu Y, Preston DL, Vaeth M, Mabuchi K. Studies of the mortality of atomic bomb survivors: report 12, part I. Cancer: 1950–1990. Radiat Res 1996;146:1–27.
Biological Effects of Ionizing Radiation report VII. Committee to assess health risks from exposure of low levels of ionizing radiation. National Academy of Sciences. Available from: http://www.nap.edu/openbook/030909156X/html/index.html (accessed September 15, 2005).
Preston DL, Shimizu Y, Pierce DA, Suyama A, Mabuchi K. Studies of mortality of atomic bomb survivors: report 13: solid cancer and noncancer disease mortality: 1950–1997. Radiat Res 2003;160:381–407.
Berrington de Gonzalez A, Darby S. Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet 2004;363:345–351.
Puskin JS, Nelson NS, Nussbaum RH, Pierce DA, Preston DL. Risks from low doses of radiation. Science 1996;272:631–635.
Pierce DA, Preston DL. Risks from low doses of radiation. Science 1996;272:632–633.
Mayo JR, Aldrich J, Muller NL. Radiation exposure at chest CT: a statement of the Fleischner Society. Radiology 2003;228:15–21.
Dakin CJ, Pereira JK, Henry RL, Wang H, Morton JR. Relationship between sputum inflammatory markers, lung function, and lung pathology on high-resolution computed tomography in children with cystic fibrosis. Pediatr Pulmonol 2002;33:475–482.
de Jong PA, Nakano Y, Lequin MH, Mayo JR, Woods R, Pare PD, Tiddens HA. Progressive damage on high resolution computed tomography despite stable lung function in cystic fibrosis. Eur Respir J 2004;23:93–97.
de Jong PA, Ottink MD, Robben SG, Lequin MH, Hop WC, Hendriks JJ, Pare PD, Tiddens HA. Pulmonary disease assessment in cystic fibrosis: comparison of CT scoring systems and value of bronchial and arterial dimension measurements. Radiology 2004;231:434–439.
de Jong PA, Lequin MH, Mayo JR, Pare PD, Tiddens H. Re: progressive damage on high-resolution computed tomography. Eur Respir J 2004;24:1071–1072.
Rawlings D, Tennant D, Furness J. Progressive damage on high-resolution computed tomography. Eur Respir J 2004;24:1071.
Bedrossian CW, Greenberg SD, Singer DB, Hansen JJ, Rosenberg HS. The lung in cystic fibrosis: a quantitative study including prevalence of pathologic findings among different age groups. Hum Pathol 1976;7:195–204.
Chow CW, Landau LI, Taussig LM. Bronchial mucous glands in the newborn with cystic fibrosis. Eur J Pediatr 1982;139:240–243.
Gibson RL, Burns JL, Ramsey BW. Pathophysiology and management of pulmonary infections in cystic fibrosis. Am J Respir Crit Care Med 2003;168:918–951.
Tiddens HA. Detecting early structural lung damage in cystic fibrosis. Pediatr Pulmonol 2002;34:228–231.
Long FR, Williams RS, Castile RG. Structural airway abnormalities in infants and young children with cystic fibrosis. J Pediatr 2004;144:154–161.
Brody AS, Klein JS, Molina PL, Quan J, Bean JA, Wilmott RW. High-resolution computed tomography in young patients with cystic fibrosis: distribution of abnormalities and correlation with pulmonary function tests. J Pediatr 2004;145:32–38.
Goris ML, Zhu HJ, Blankenberg F, Chan F, Robinson TE. An automated approach to quantitative air trapping measurements in mild cystic fibrosis. Chest 2003;123:1655–1663.
Bonnel AS, Song SM, Kesavarju K, Newaskar M, Paxton CJ, Bloch DA, Moss RB, Robinson TE. Quantitative air-trapping analysis in children with mild cystic fibrosis lung disease. Pediatr Pulmonol 2004;38:396–405.
Hansell DM, Strickland B. High-resolution computed tomography in pulmonary cystic fibrosis. Br J Radiol 1989;62:1–5.
Stiglbauer R, Schurawitzki H, Eichler I, Vergesslich KA, Gotz M. High resolution CT in children with cystic fibrosis. Acta Radiol 1992;33:548–553.
de Jong PA, Lindblad A, Rubin L, Hop WC, de Jongste JC, Brink M, Tiddens HA. Longitudinal evaluation of lung structure on computed tomography scans and lung function in children and adults with cystic fibrosis. Thorax 2005 Oct 21 [Epub ahead of print].
Fiel SB, FitzSimmons SC, Schidlow DV. Evolving demographics of cystic fibrosis. Am J Respir Crit Care Med 1994;15:349–355.
Fiel SB, Golmohammadi K, Schidlow DV. Evolving demographics of cystic fibrosis . Proc Am Thorac Soc 2005;2:A576.
de Jong PA, Golmohammadi K, Nakano Y, Mayo JR, Lequin MH, Tiddens HA, Aldrich J, Pare PD, Sin DD. Modelling the mortality risk associated with radiation exposure from cumulative computed tomography in cystic fibrosis . Proc Am Thorac Soc 2005;2:A576.
Jones D, Shrimpton PC. Survey of CT practice in the UK: normalised organ doses for x-ray computed tomography calculated using Monte Carlo techniques. Harwell, UK: National Radiological Protection Board; 1991. Available from: www.impactscan.org/ctdosimetry.htm (accessed March 15, 2005).
Khursheed A, Hillier MC, Shrimpton PC, Wall BF. Influence of patient age on normalized effective doses calculated for CT examinations. Br J Radiol 2002;75:819–830.
Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322–338.
Kulich M, Rosenfeld M, Goss C, Wilmott RW. Improved survival among young patients with cystic fibrosis. J Pediatr 2003;142:631–636.
Fogarty A, Hubbard R, Britton J. International comparison of median age at death from cystic fibrosis. Chest 2000;117:1656–1660.
Evans M, Hastings N, Peacock B. Triangular distributions. In: Statistical distributions, 3rd ed. New York: Wiley; 2000. pp. 187–188.
Sporn M. The war on cancer. Lancet 1996;347:1377–1381.
Neglia JP, FitzSimmons SC, Maisonneuve P, Schoni MH, Schoni-Affolter F, Corey M, Lowenfels AB. The risk of cancer among patients with cystic fibrosis. Cystic Fibrosis and Cancer Study Group. N Engl J Med 1995;332:494–499.
Aroua A, Bize R, Buchillier-Decka I, Vader JP, Valley JP, Schnyder P. X-ray imaging of the chest in Switzerland in 1998: a natiowide survey. Eur Radiol 2003;13:1250–1259.
Bottollier-Depois JF, Chau Q, Bouisset P, Kerlau G, Plawinski L, Lebaron-Jacobs L. Assessing exposure to cosmic radiation on board aircraft. Adv Space Res 2003;32:59–66.
Rosenfeld M, Davis R, FitzSimmons SC, Pepe M, Ramsey B. Gender gap in cystic fibrosis mortality. Am J Epidemiol 1997;145:794–803.(Pim A. de Jong, John R. Mayo, Kamran Gol)