Magnetic Resonance Imaging and Computed Tomography May Carry Similar (but Very Low) Risks of Carcinogenesis

Abstract

In comparing magnetic resonance imaging (MRI) and computed tomography (CT), it is often stated thata significant advantage of MRI is its lack of ionizing radiation, which may contribute to carcinogenesis. This article questions whether this concern is justified. First, the basis for the linear no-threshold (LNT) hypothesis, which suggests a carcinogenic potential of CT, is now considered weak. Historically, the foundation of the LNT model was not grounded in robust scientific data, and the epidemiological evidence indicating an increased cancer incidence in children undergoing CTscans is not scientifically reliable. Both MRI and CT examinations elevate reactive oxygen species, leading to DNA damage, chromosomal aberrations, and micronucleus formation. Following a single scan with either MRI or CT, the γH2AX index in human lymphocytes increased to a similar degree; however, the small amounts of DNA damage produced are efficiently repaired, thus not resulting in carcinogenesis. While it may be argued that ionizing radiation induces more complex DNA damage than electromagnetic fields, no studies have compared the quality of DNA damage between MRI and CT at clinically relevant exposure levels. Considering the effective defense mechanisms of living organisms, infrequent MRI and CT examinations do not appear to pose an increased risk of carcinogenesis.

Magnetic resonance imaging (MRI) and computed tomography (CT) are two major indispensable modalities in diagnostic radiology, each with its own advantages and disadvantages. Significant overdoses of radiation from CT, particularly brain perfusion CT, can result in skin burns and hair loss. ¹ In 0.2% of subjects across five U.S. healthcare markets, cumulative doses from diagnostic imaging―primarily CT and nuclear imaging―exceeded 50 mSv. ² Consequently, it is often said that an advantage of MRI is no usage of ionizing radiation; there is concern that even low doses (<100-200 mGy) from standard CT scans may contribute to carcinogenesis. This article raises the question of whether this concern is warranted. The apprehension regarding carcinogenesis from CT is rooted in the linear no-threshold (LNT) hypothesis, which posits that even minimal doses of radiation carry some risk of developing cancer. However, recent works by Prof. Calabrese uncovered the falsehood of the LNT theory foundation in the mid 1900’s; the model was not based on the scientific data available at that time and was largely a construct of Hermann J. Muller, a Nobel Prize winner, and his followers.^3,4 Specifically, existing data from that era indicated the presence of a threshold for radiation effects. In many subsequent laboratory, epidemiological, and volunteer-based human studies, no harmful effects of low-dose radiation have been observed―and in some cases, a hormetic response has been noted, ⁵ although some studies have suggested negative biological effects, particularly in animals with radiosensitive genetic backgrounds. ⁶ Nevertheless, the International Commission on Radiological Protection (ICRP) and the National Council on Radiation Protection and Measurements (NCRP) continue to endorse the LNT model for regulatory and protection purposes.^7,8

In MRI, subjects are positioned in strong static, time-varying gradient, and pulsed radiofrequency electromagnetic fields. Radio-waves (radio-frequency pulses) are added to the magnetic field to cause hydrogen nuclei to resonate. These processes are quite unphysiological. According to the International Commission on Non-Ionizing Radiation Protection (ICNIRP), occupational exposure to a static magnetic field is recommended not to exceed 2 T for head and body and 8 T for extremities, and the recommended level for the public is 400 mT. Nevertheless, magnetic fields of 3 T or higher are being increasingly used in MRI examination. Although the allowable duration of exposure is not defined, it is considered that a longer exposure would be more harmful, ⁹ and approximately 0.5- to 1-h examination time of MRI is not considered short. It has been reported that reactive oxygen species (ROS) including highly toxic hydroxyradicals are increased and DNA damage, chromosome aberration, and micronucleus induction occur following exposure to magnetic fields.^9,10 These events are known to occur even at ∼100 mT. In 5 out of 8 studies that investigated genetic damage from MRI in human lymphocytes using clinically relevant scanners or pulse sequences (1.5-7.2 T for various exposure durations), significant increases in DNA damage were observed at certain time points following exposure. ⁹ However, less attention has been paid to the carcinogenic potential of MRI compared to X rays, probably because there have been no disasters like Hiroshima and Nagasaki originating from magnetic field exposure and no provocators like Hermann Muller regarding the harm of electromagnetic fields. Epidemiological studies examining the carcinogenic risk associated with electromagnetic fields have typically involved relatively small sample sizes and various confounding factors. Consequently, the International Agency for Research on Cancer (IARC; https://monographs.iarc.who.int/list-of-classifications) classifies static magnetic fields as Group 3 (not classifiable as to its carcinogenicity to humans), while extremely low-frequency magnetic fields and radiofrequency electromagnetic fields are classified as Group 2B (possibly carcinogenic to humans).

The grounds for carcinogenic potential of standard CT are now considered scarce. DNA damage produced by low-dose radiation is readily and almost completely repaired, and by receiving low-dose stimuli, DNA repair capacity is increased. ⁵ If increased DNA repair capacity outweighs the level of residual DNA damages produced by low-dose irradiation, carcinogenesis may even be decreased. In addition to the increase in the DNA repair capacity, various other reactions that stimulate living organisms have been demonstrated, including induction of radioprotective substances, such as superoxide dismutase and glutathione, and activation of the immune system. Indeed, many recent studies suggest the beneficial effects of low-dose radiation or at least the presence of a threshold in the radiation dose-response curves. ⁵

Some epidemiological studies claimed that CT during childhood was associated with increased incidences of leukemia, brain tumors, and other cancers. ¹¹ However, the flaws in those studies have been clearly demonstrated. In children undergoing CT in 2013 at a university hospital, the reason for CT examination was the presence of congenital anomaly in as many as 32% of the children. ¹² Since the patients with a congenital anomaly carry a 300% increased risk of developing cancer, the observed increase of about 24% in cancer incidence in the epidemiological study was considerably low. ¹² As every clinician knows that healthy children never undergo CT, comparing the cancer incidence in children with and without CT is not reasonable. Study designs of such epidemiological studies do not seem to be sound. Newer epidemiological studies have found a correlation between radiation doses from CT and the incidence of hematological malignancies ¹³ ; however, this correlation appears to reflect the fact that children with more severe pathological conditions require CT examinations more frequently.

Radiation from electromagnetic fields and radio-waves are categorized as non-ionizing radiation, while X rays are ionizing radiation. Although both low-dose radiation and electromagnetic fields can produce DNA damage, it is reported that ionizing radiation causes more complex DNA damage such as clustered ones; these complex damages are more difficult to repair and may be more likely to lead to carcinogenesis, suggesting that the quality of DNA damage is more important than the quantity as measured by the γH2AX assay. ⁹ However, this observation may be likely to apply to high-dose irradiation and it is unknown how such complex DNA damage occurs at low-dose irradiation. Moreover, most of such damages were efficiently repaired within 18 hours after X-ray irradiation, compared to Fe-ion irradiation. ¹⁴

The hypothesis that very frequent MRI and CT examinations pose similar carcinogenic risks is not supported by concrete data, as no study has directly compared the genetic damage resulting from these two types of examinations. In a study investigating DNA double-strand breaks in human subjects immediately following cardiac MRI, the median number of γH2AX foci per lymphocyte was 0.066 before the MRI and 0.190 after (P < .05). ¹⁵ On the other hand, in another study examining DNA damage after a chest CT scan (with a median effective dose of 5.0 mGy), the median number of γH2AX foci per lymphocyte was 0.11 before the CT and 0.16 immediately after (P < .001). ¹⁶ Further research is warranted on this issue. However, it should be noted that increased γH2AX indices returned to the control levels within 24 hours after radiation doses exceeding 150 mGy. ¹⁷

Human and other living beings experience various oxidative stresses including radiation, electromagnetic fields, radio-waves, and ultraviolets, and after a hundred million years of evolution, every living organism has acquired capacities to deal with these stresses. Living organisms are activated by adequate amounts of stresses, and without any stress, defensive capacity of the organisms would become lowered. When activation of defense mechanisms outweighs the harms produced by stress, hormesis would occur. Otherwise, small oxidative stresses induced by low-dose irradiation and non-excessive electromagnetic field exposure and increased responses to the stresses may counterbalance in the body, so CT and MRI examinations performed infrequently may have no harm. However, if CT and MRI examinations are conducted too frequently, particularly with an excessive number of scans during a single examination, the risk of cancer may be slightly elevated. Since the specific threshold associated with increased cancer risk is unknown, it is advisable to avoid overuse and to utilize these imaging techniques when there is a clear benefit to the individual.