Cataract Risk in a Cohort of U.S. Radiologic Technologists Performing Nuclear Medicine Procedures

Study results indicated an increased risk of cataract in a population of medical workers performing nuclear medicine procedures who were followed up during the period from the completion of the third cohort survey in the years 2003—2005 to the completion of the fourth survey in the years 2012—2013.

Abstract

Purpose

To estimate the risk of cataract in a cohort of nuclear medicine (NM) radiologic technologists on the basis of their work histories and radiation protection practices.

Materials and Methods

In the years 2003—2005 and 2012—2013, 42 545 radiologic technologists from a U.S. prospective study completed questionnaires in which they provided information regarding their work histories and cataract histories. Cox proportional hazards models, stratified according to birth-year cohort (born before 1940 or born in 1940 or later) and adjusted for age, sex, and race, were used to estimate hazard ratios (HRs) for the risk of cataract in radiologic technologists according to NM work history practices according to decade.

Results

During the follow-up period (mean follow-up, 7½ years), 7137 incident cataracts were reported. A significantly increased risk of cataract (HR, 1.08; 95% confidence interval [CI]: 1.03, 1.14) was observed among workers who performed an NM procedure at least once—as opposed to never. Risks of cataract were increased in the group who had performed a diagnostic (HR, 1.07; 95% CI: 1.01, 1.12) or therapeutic (HR, 1.10; 95% CI: 1.04, 1.17) NM procedure. Risks were higher for those who had first performed diagnostic NM procedures in the 1980s to early 2000s (HR, 1.30; 95% CI: 1.08, 1.58) and those who had performed therapeutic NM procedures in the 1970s (HR, 1.11; 95% CI: 1.01, 1.23) and in the 1980s to early 2000s (HR, 1.14; 95% CI: 1.02, 1.29). With the exception of a significantly increased risk associated with performing therapeutic NM procedures without shielding the radiation source in the 1980s (HR, 1.32; 95% CI: 1.04, 1.67), analyses revealed no association between cataract risk and specific radiation protection technique used.

Conclusion

An increased risk of cataract was observed among U.S. radiologic technologists who had performed an NM procedure at least once. This association should be examined in future studies incorporating estimated lens doses.

^© RSNA, 2017

Introduction

Diagnostic and therapeutic nuclear medicine (NM) procedures involve the use of ionizing radiation to assess organ function and treat patients with a wide range of diseases. The number of NM procedures performed in the United States has increased dramatically during the past few decades, from 7 million in the early 1980s to 18 million in 2006 (1). This trend is largely being driven by an increasing use of high-radiation-dose procedures, including cardiac imaging (2) and positron emission tomography (PET) (3).

Some of the radionuclides used for organ imaging in NM practices are high-energy, highly penetrating gamma emitters (4). In contrast to professionals who work with standard radiologic procedures, those who work with NM procedures cannot avoid being in close contact with radioactive pharmaceuticals when they prepare and/or administer injections and during the imaging process. Although the annual mean effective radiation dose has been reported to be 0.7 mSv worldwide (5), the effective dose can reach several millisieverts, depending on the procedure type and the method and behavior used in performing the procedure (5–8). However, radiation exposure levels can be reduced by using shielding devices and/or reducing the time spent in close proximity to patients who are undergoing a procedure involving radiation. Notable variations in work practices and radiation protection techniques can result in heterogeneous levels of occupational exposure from NM procedures (5).

The potential health effects of occupational radiation exposure on medical staff have been investigated in cohort studies (9–14); however, sparse information is available for NM workers. In 2000, there were approximately 120 000 NM workers worldwide. Although they represented only 5% of all medically exposed workers, they reportedly received 10% of the collective radiation dose (5). To our knowledge, only one study (15) has evaluated the long-term health risks in this population. Kitahara et al (15) observed an increased risk of lung cancer mortality in association with performing NM procedures in 22 000 U.S. radiologic technologists who reported having performed a NM procedure at least once and an increased risk of squamous cell carcinoma among those who reported having performed diagnostic radionuclide procedures. Cataract outcome was not investigated.

Cataract was identified early on as a late effect of ionizing radiation, and for more than 50 years it has been considered a tissue reaction (formerly considered a deterministic effect) with a threshold of 2 Gy. Recent epidemiologic studies have yielded new evidence of radiation-induced cataract at lower doses (16), leading the International Commission on Radiological Protection to lower the assumed threshold to 0.5 Gy (17). However, investigators in some literature reviews focused on more recent epidemiologic and biologic study findings (18,19) have questioned whether the data are still consistent with the theory that cataract is a tissue reaction and have described potential stochastic phenomena linked to DNA damage as an explanation for the increased risks at lower radiation exposure levels. In a previous analysis and follow-up investigation (20) conducted from 1983 to 2004 and involving 35 705 U.S. radiologic technologists from the current study, personal medical radiation exposure, defined as radiation therapy to the head and neck or exposure to a large number (>15) of radiographic procedures, was found to be associated with an increased risk of cataract. However, the estimated occupational radiation exposure dose was not significantly associated with this risk.

In the current study, we focused on occupational histories and more recently collected radiation exposure information (including the use of radiation protection techniques) according to decade worked, as well as follow-up data from the U.S. Radiologic Technologists cohort study (20). We estimated the risk of cataract in NM technologists, taking into account their work histories and radiation protection practices.

Materials and Methods

U.S. Radiologic Technologist Cohort

The U.S. Radiologic Technologist cohort study was established under a collaboration among the National Cancer Institute, the University of Minnesota, and the American Registry of Radiologic Technologists.

A description of the study population and methods can be found on the U.S. Radiologic Technologists website (https://study.nci.nih.gov.study.nci.nih.gov/). Briefly, 146 022 radiologic technologists who were U.S. residents and had been certified for at least 2 years in the years 1926–1982 were identified from the American Registry of Radiologic Technologists records. Passive follow-up of this cohort to obtain vital status data has been based on linkage to yearly American Registry of Radiologic Technologists recertification records and Social Security Administration database records for those who did not recertify. Technologists identified as deceased, presumed to be deceased, or with unknown vital status were linked to the National Death Index (NDI Plus, U.S. Centers for Disease Control and Prevention, Atlanta, Ga) to confirm their vital status and obtain their causes of death.

Additional information was acquired from a series of questionnaires, which were mailed and completed in the years 1983–1989, 1994–1998, 2003–2005, and 2012–2013. The questionnaires included questions regarding work history, demographic characteristics, lifestyle, and other disease risk factors; personal diagnostic and therapeutic medical procedures performed; and health outcomes, including age at the time of cataract diagnosis or cataract surgery. This study was approved by the institutional review boards of the National Cancer Institute and the University of Minnesota.

Study Population and Follow-up

The baseline questionnaire (mailed and completed in years 1983–1989) for the present study was completed by 73 613 radiologic technologists, who were then followed up and sent a subsequent survey. Of these 73 613 technologists, 51 291 completed the third and fourth questionnaires. After the exclusion of 8746 of the 51 291 technologists (17%) because they reported receiving a diagnosis of cataract at the time of or before the third survey, the study population included 42 545 technologists. Participants were followed up from the date on which the third survey was completed to the date on which cataract was diagnosed or the date on which the fourth survey was completed (during 2012–2013).

Exposure Assessment

The work history reported by individuals on the third questionnaire (in years 2003–2005) referred to the worker’s first year of employment as a radiologic technologist and specific questions about performing diagnostic and therapeutic NM, fluoroscopically guided interventional procedures, and radiographic procedures. For specific time periods (before 1940, 1940–1949, 1950–1959, 1960–1969, 1970–1979, and 1980 to present), questions were asked about the frequency of performing NM procedures per week, the specific tasks performed, and the radiation protection measures taken, including the amount of time wearing a lead apron and standing in close proximity to the patient and the use of a protective shield around the radioactive source and afterloading devices to transfer the radioactive source from the safe to the patient. No information regarding the use of lead glasses by technologists was available from the third survey. Because few participants reported working with NM procedures before 1950, responses related to the two earliest time periods (before 1940 and 1940–1949) were combined with those related to the 1950–1959 period in the analysis. Technologists were classified as having ever worked in NM if they reported having performed at least one NM procedure per week in any decade. Other baseline information evaluated included race, education, marital status, smoking status, eye color, solar ultraviolet radiation exposure, medical history, and medical exposure to radiation.

Statistical Analyses

With use of Cox proportional hazards regression, hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) were calculated to evaluate the risk of cataract in radiologic technologists who ever, as compared with never, performed NM procedures. All models were fitted with age as the time scale metric to control for age, stratified according to birth-year cohort (born before 1940 or born in 1940 or later) to control for secular trends, and adjusted for sex and race. Additional adjustments were made for potential confounders known to be associated with cataract, such as smoking, body mass index, eye color, ultraviolet radiation exposure, diabetes (21), personal history of medical exposure to radiation (22–25), and occupational radiation exposure related to fluoroscopically guided interventional radiology practice (26–28). Decade-specific risks were assessed according to frequency of performing procedures per week and use of radiation protection measures. Missing or inapplicable values for categorical variables were removed from the specific analyses. P values for trend were calculated by using continuous data, when available, and excluding missing or inapplicable values.

Statistical significance was estimated by using likelihood ratio tests. Scaled Schoenfeld residuals (29) were examined to assess the proportional hazards assumption over the age time scale. All analyses were two sided and performed by using software (SAS, version 9.3; SAS Institute, Cary, NC).

Results

The demographic characteristics of the 42 545 technologists who were cataract free at completion of the third survey (in years 2003–2005) and who completed the fourth survey (in years 2012–2013) are summarized in Table 1. Of the 42 545 technologists, 34 096 (80%) were women, 40 694 (96%) were white, and 36 540 (86%) attended more than 12 years of education. The mean ages at completion of the third and fourth surveys were 55 and 62½ years, respectively, with a mean follow-up of 7½ years between the two questionnaires. Of the 42 545 technologists in the study population, 40 725 (96%) reported having performed a radiographic procedure at least once and 10 434 (25%) reported having performed a fluoroscopically guided interventional radiology procedure at least once. Thirty percent (n = 12 964) of the technologists reported having worked at least once a week with NM procedures. Compared with the technologists who never worked with NM, those who did were more likely to be men and older and were less likely to have completed only 2 years of radiologic technologist training.

Table 1.

Characteristics of U.S. Radiologic Technologists Who Were Cataract Free at Completion of the Third Survey and Who Completed the Fourth Survey, Stratified according to NM Work History

Note.—Data are numbers of radiologic technologists, with percentages in parentheses. The third survey was completed in the years 2003–2005, and the fourth survey was completed in the years 2012–2013. BMI = body mass index.

*Included 3754 workers with unknown status.

There were a total of 7137 self-reported diagnoses of cataract, and 2458 of these diagnoses were in technologists who reported performing an NM procedure at least once (Table 2). The main results were adjusted for sex and race because additional adjustments for other known risk factors did not modify the estimates (data not shown). Having ever worked with any NM procedure was associated with a significantly increased risk of cataract (HR, 1.08; 95% CI: 1.03, 1.14).

Table 2.

Cataract Frequency and Surgery in U.S. Radiologic Technologists according to NM Work History Characteristics (n = 42 545)

Note.—Missing answers were due to unknown answers or nonapplicable questions.

*Data are numbers of cases of a self-reported diagnosis of cataract in the population of radiologic technologists in the given subgroup.

^†HRs were calculated by using Cox proportional hazards models fitted with age as the time scale metric, stratified according to birth-year cohort (born before 1940 or born in 1940 or later), and adjusted for sex and race. Numbers in parentheses are 95% CIs. Ref = reference.

^‡Data are numbers of cases of cataract removal surgery in the population of radiologic technologists with self-diagnosed cataract in the given subgroup.

^§Statistically significant.

A modest increase in the risk of developing cataract was apparent for technologists who had ever performed diagnostic NM procedures (HR, 1.07; 95% CI: 1.01, 1.12) as compared with those who had never performed them. However, the risks were higher for technologists who first performed diagnostic NM procedures in the 1980s (HR, 1.30; 95% CI: 1.08, 1.58) than for those who started performing these procedures in the 1950s (P < .001 for the trend). Among the technologists who had ever performed diagnostic NM procedures, cataract risk was not associated with number of performed diagnostic procedures or number of prepared or eluted radiopharmaceutical kits. Cataract risk was not associated with the use of radiation protection measures, including shielding the radioactive source during administration of the radioactive material and standing at least 3 feet away from the patient during the procedure (Table 3).

Table 3.

Frequency of Cataract in U.S. Radiologic Technologists Who Ever Performed NM Procedures, Stratified according to Work History Characteristics for Each Decade Worked

Note.—Unknown or nonapplicable answers were removed from this specific analyses.

*Data are numbers of cases of cataract in the population of radiologic technologists in the given subgroup who worked in the given decade.

^†HRs were calculated by using Cox proportional hazards models fitted with age as the time scale metric, stratified according to birth-year cohort (born before 1940 or born in 1940 or later), and adjusted for sex and race. Numbers in parentheses are 95% CIs. Ref = reference.

^‡Statistically significant.

Having ever performed a therapeutic NM procedure was associated with a significantly increased risk of developing cataract (HR, 1.10; 95% CI: 1.04, 1.17), and the risk increased significantly by decade worked (P value for trend = .01). Among the technologists who had performed a therapeutic NM procedure at least once, we observed significantly increased risks for those who performed one or more therapeutic NM procedures per week in the 1970s and 1980s (Table 3). With the exception of a significantly increased risk related to the lack of radioisotope source shielding during therapeutic procedures in the 1980s, cataract risk was not associated with specific activities or radiation protection measures.

Similar findings were observed when the analysis was limited to cases of cataracts that were surgically removed (n = 2472) (Table 4). However, owing to the small number of surgeries reported, some of the risk data were less precise. The risks were also similar when we included cataract diagnoses that were rendered before the third survey was conducted (data not shown).

Table 4.

Cataract Surgery in U.S. Radiologic Technologists Who Ever Performed NM Procedures, Stratified according to Work History Characteristics for Each Decade Worked

Note.—Unknown or nonapplicable answers were removed from this specific analyses.

*Data are numbers of cases of cataract surgery in the population of radiologic technologists in the given subgroup who worked in the given decade.

^†HRs were calculated by using Cox proportional hazards models fitted with age as the time scale metric, stratified according to birth-year cohort (born before 1940 or born in 1940 or later), and adjusted for sex and race. Numbers in parentheses are 95% CIs. Ref = reference.

^‡Statistically significant.

Discussion

In this study, we found an increased risk of cataract in a population of medical workers who performed NM procedures and who were followed up from the time of completion of the third cohort survey in the years 2003–2005 to the time of completion of the fourth survey in the years 2012–2013. Increased risks were observed in those technologists who conducted diagnostic and therapeutic NM procedures, as compared with those who did not perform NM procedures. With the exception of the reduced risk associated with shielding the radioactive source when performing therapeutic NM procedures in the 1980s, no association between cataract risk and either use of radiation protection techniques or usual practices for handling radionuclides was observed for those technologists who performed NM procedures. Adjustments for known cataract risk factors, personal radiation exposure from radiographic or other medical procedures, or other occupational radiation exposure did not modify the risk estimates.

Although an increased risk of cataract among medical workers who were occupationally exposed to radiation has been previously reported (19), mainly among those who performed fluoroscopically guided interventional procedures, to our knowledge such a risk increase has never been reported for NM workers. The radiation doses to the eyes of NM workers are expected to be low because only photon emitters of low energy and/or beta emitters with energy levels higher than 0.7 MeV penetrate deep enough to reach the lens. However, among NM workers, estimated annual radiation doses to the eyes have ranged from 0.9 to 25.0 mGy, depending on the type of procedure performed (6,30–32); these radiation doses could contribute to a relatively high cumulative dose after several decades of work. Increases in radiopharmaceutical production, iodine 131 therapy administration, and PET/computed tomography use in recent decades may have contributed to higher eye lens doses in technologists who perform NM procedures (30,31). In addition, in recent reviews in the epidemiologic literature, there has been discussion of a potential effect on cataract development from radiation doses lower than the previously accepted threshold of 0.5 Gy (18,19).

We observed modest increases in cataract risk in association with ever—as opposed to never—performing diagnostic (7% risk increase) and therapeutic (10% risk increase) NM procedures. No trend was observed with increasing number of procedures performed. The increase in cataract risk was smaller than that reported for interventional radiologists and cardiologists (2.5%–5.7% higher risk compared with that reported for unexposed groups) (27–29,34). However, radiation doses to the eye can be much higher in interventional physicians—sometimes cumulating to more than 1 Gy during all years worked (27,34).

The reported trends of increased cataract risk for workers performing NM procedures during the more recent decades (1970s and 1980s) may reflect the possibility that these technologists were still performing NM work in the year 2000 and afterward, when the use of higher-radiation-dose procedures such as PET and cardiac scintigraphy were increasing—and potentially resulting in higher cumulative doses to the eye (30,31). Selection bias linked to the exclusion of cataract diagnoses before the study period was ruled out in a sensitivity analysis that included cases that occurred before the third survey (data not shown).

Not shielding the radioactive source when performing therapeutic procedures, which leads to increased radiation exposure to the technologists, was associated with an increased risk of cataract in the 1980s. Risk was not associated with other work practices or radiation safety measures used. In addition, the technologists may have had a tendency to overreport their use of radiation protection tools and may have had difficulty remembering other work practices used in the distant past. However, the collection of occupational exposure data before the assessment of outcomes minimized potential recall bias.

These results should be interpreted with consideration of the following limitations: The cataract diagnoses were self-reported, and the types of cataract were not available. However, surgically removed cataracts—especially those reported by medically trained workers—should be reported with greater accuracy (35). Sensitivity analyses restricted to cataract extractions yielded similar findings, suggesting that any misclassifications were minimal. This study was mainly limited by a lack of individual occupational dosimetry estimates specific to NM technologists. Individual occupational dosimetry reconstructions have been performed for all study participants on the basis of literature review, work history, and dosimetric measurements (36). However, procedure-specific exposures from the differing energy levels emitted by the radiopharmaceuticals used in NM procedures were not taken into account when the estimated doses were calculated. An updated organ dose reconstruction effort for technologists who work with NM procedures is under way but not yet complete. In addition, in the current analysis, we did not consider the changes in NM work history practices during follow-up, which may have introduced some bias; however, the direction of that bias is unknown.

The strengths of this study include the large size of the technologist cohort and the collection of detailed data on work history practices and radiation protection measures used according to decade. The information collected from the comprehensive questionnaires enabled us to rule out confounding bias from known risk factors for cataract (21–25).

In conclusion, the association between occupational exposure of medical and other radiation workers and cataract frequency has been widely reported in studies conducted during the past several years, with new findings from epidemiologic studies supporting increasingly lower radiation dose thresholds (17). Nevertheless, our finding of an increased risk of cataract in a population of NM workers should be interpreted with caution, as the exposure categories are based only on reports of working with these procedures and specific types of cataract were not available. To confirm and quantify this relationship, the estimation of individual radiation doses to the eye lens from both occupational and personal medical sources should be a key focus of further study. Specific information about performing relatively high-dose NM procedures and assessment according to cataract type are also needed.

Advances in Knowledge

• ■ A significantly increased risk of cataract (hazard ratio, 1.08; 95% confidence interval: 1.03, 1.14) was observed among U.S. radiologic technologists who had performed a nuclear medicine (NM) procedure at least once, as compared with those who had never performed them.

• ■ Significantly increased risks of cataract were observed for those radiologic technologists who had performed a diagnostic (7% risk increase) or therapeutic (10% risk increase) NM procedure at least once, as opposed to never.

• ■ No strong association between cataract risk and either number of diagnostic or therapeutic NM procedures performed or specific radiation protection technique used was observed.