It can be comforting to be watched. At the swimming pool, it is reassuring for children and adults alike when there is a lifeguard on duty, eagle-eyed and ready to rescue those in distress. At home, modern parents can invest in remote video technologies to observe their kids even when Mom or Dad is out of the house or simply in another room. In the post-9/11 era, our intelligence services monitor an incessant stream of information for chatter about impending attacks, listening for a crucial signal in the noise in the hopes of intervening and averting another catastrophe.
Of course, not everyone sees such vigilance as wholly benevolent. Edward Snowden raised the alarm that our efforts to defeat terrorism through intensive monitoring may have built a bridge too far in terms of invading our privacy. In fiction, George Orwell foresaw our love affair with the black mirrors of our camera-enabled smartphones and computers as the telescreens by which Big Brother monitored the citizenry, while J.R.R. Tolkien depicted the evil Sauron as an all-seeing eye over Middle Earth.
In medicine, tension has arisen between the literal & figurative insights to be gained from radiology versus the costs that patients pay for these glimpses of their interior selves. One of my faculty supervisors during my oncology fellowship at the Mayo Clinic was Dr. Carrie Thompson, who taught me and other trainees to be judicious when ordering scans. A focus of her research has been how to reconcile the benefits of imaging against the scans’ potential to cause harms both measurable & immeasurable.
Dr. Thompson’s tumor-specific expertise is the treatment of lymphoma, a group of diseases in which it had been common to surveil patients who were in remission for evidence of relapse. It might seem intuitive that early detection of a resurgent cancer would improve outcomes, but substantive evidence is lacking that there is any significant benefit to a scan unveiling a malignancy before it is clinically evident. Thompson, along with other researchers, followed a group of 680 patients with diffuse large B cell lymphoma who had achieved a complete remission, of whom 20% later suffered a relapse; of those 112 patients, the majority of recurrences were apparent before scheduled clinical follow-up and only 9 patients had a relapse detected on scans before it was already known to them or their physician (this finding was validated in a cohort of French patients where there was a similarly low, less than 2% rate of a scan being the only alert of a returning lymphoma). Crucially, there was no difference in survival whether a lymphoma relapse was detected at regularly scheduled follow-up or between appointments. The authors concluded that they “were unable to identify a subgroup of patients where [radiographic] surveillance may be useful”.
Furthermore, scanning to detect malignancy is, in and of itself, not necessarily benign. In fact, there is worry within the oncology community that computed tomography (CT) might be inducing the very disease we are trying to prevent & detect. It has been estimated that up to 2% of all cancers in the US are caused by radiation from CTs, although critics of this statistic’s derivation have questioned the direct extrapolation of scans’ carcinogenic potential from the amount of ionizing radiation to which atomic-bomb survivors were exposed. Nonetheless, it is indisputably true that each CT carries a dose of milliCuries (a standard unit of radioactivity) at least a hundredfold that of a chest X-ray.
Compounding the risks of CT, many scans are performed with intravenous contrast given concurrently to enhance the images for the interpreting radiologist, and this iodinated dye can pose a risk to kidney function. In assessing less physiologic threats, Dr. Thompson has also documented the phenomenon of ‘scanxiety,’ a spike in apprehension around the time of imaging, even in long-term survivors with no pre-scan concern for active disease. Interviewing lymphoma patients in the midst of durable remissions, Thompson et al found “that anxiety symptoms worsen before a follow-up appointment [and] peak leading up to obtaining the results of the scan,” suggesting a negative emotional impact associated with the process of radiographic observation.
A psychological argument can be made that patients suffer tremendous worry in the absence of regular scans, and some do (especially if they have become accustomed to this form of longitudinal monitoring), but another counterpoint is the troubling problem of false-positives, whereby tests can beget tests and non-invasive scans can lead to invasive procedures. The value of any diagnostic tool in medicine can be assessed by sensitivity and specificity, which together establish the balance between correctly identifying when a problem is present while being certain when declaring its absence. Many scans are sensitive, affording us incredible anatomic detail of internal organs that, even 5 decades ago, might have required exploratory surgery to examine with similar perspicacity. But not all variants from normal anatomy are pathologic or cancerous; cross-sections through the liver, for instance, can detect non-malignant lesions like cysts or hemangiomas (anomalous blood vessels) that are not dangerous but which merit mention nonetheless. Such unintentional discoveries, known within medicine as ‘incidentalomas’, sometimes require further, serial scans to ensure stability or even biopsies to exclude the presence of cancer.
The sensitivity/specificity trade-off may be even more relevant when it comes to discussing the utility of PET (positron emission tomography) scans, which most often use a specially labeled sugar called fluorodeoxyglucose to map the body’s metabolism. Within the context of oncology, this form of functional imaging exploits the Warburg effect, the altered biochemistry of cancer cells that can cause them to consume glucose at a rate up to 200 times faster than normal tissues. The tremendous sensitivity of PET for cancer is a sufficiently appealing notion, at least in theory, that many patients will ask their physicians to order such scans, whether or not they are indicated in the detection of a certain tumor or in the monitoring of treatment response. Some organs have such inherently high glucose uptake that it can be difficult to discern a hypermetabolic cancer within their substance, and, perhaps even more confoundingly, the sugar tracer in PET scans can accumulate in foci of infection or inflammation, such as surgical sites in the process of healing. Finally, PET scans don’t obviate concerns about radiation exposure in patients already at risk for recurrent or second malignancies, as they involve nuclear isotopes whose three-dimensional distribution in the body is usually mapped against a CT performed at the same time.
ALARA is a principle of environmental health that aims to maximize safety by making radiation exposure as low as reasonably achievable. To this end, there are scanning modalities in oncology that do not incur any ionizing radiation at all. Magnetic resonance imaging (MRI) harnesses the power of large magnets to generate fields wherein the hydrogen atoms of the body align in discernible spatial patterns, particularly allowing distinctions between water and fat. Disadvantages include the relatively narrow aperture in the center of closed-bore MRI machines, which can exacerbate claustrophobia in some patients, as well as the need for long periods of stillness, lest the images be marred by motion artifact (which makes the lungs particularly hard to capture). Also MRI equipment and interpretation tend to be expensive, adding to the cost of care, especially if being deployed routinely. Ultrasound typically carries a lower pricetag and the imaging apparatus does not confine the patient (in fact, the necessary machinery is frequently portable and can be brought to the bedside in the hospital setting); again, there is no radiation exposure associated with sonography (hence, its obstetric use to monitor fetal growth during pregnancy) but not all organs can be penetrated or reliably reached by the sound waves emitted by the transducer.
In summary, radiology is its own sub-specialization within medicine, where doctors train for years to interpret the images that can yield a treasure trove of information about the internal architecture of the most complicated machine in the world: the human body. The work of radiologists is crucial to the practice of modern medicine, and not to be discounted or denigrated. Rather, the onus of responsibility lies with the doctor, and the oncologist especially, who is ordering the scans that the radiologist then studies. In doing so, we should carefully weigh the benefits of our enhanced perception (which sometimes quite literally endow us with the X-ray vision of Superman!) against the potential disadvantages of ‘scanxiety’, cumulative radiation exposure, false-positive findings, and increased financial expenditure in a world of finite healthcare resources. We hold the keys, but are we opening Pandora’s box?
Patients inhabit their bodies 24/7 and are best attuned to the afflictions that can arise therein, especially what Siddhartha Mukherjee so aptly named the emperor of all maladies. Intermittent visits to their doctors, whether or not those appointments are linked to the intelligence-gathering of scans, are less likely to detect recurrences than the patients’ own innate sense of normalcy and disturbances in that equilibrium. Patients are the best sentinels of their own health, and we should listen to them when they sound an alarm. In my opinion, there is less risk of crying wolf than there is of silencing the most important voice in the room.
 Thompson CA et al. Utility of Routine Post-Therapy Surveillance Imaging in Diffuse Large B-Cell Lymphoma. J Clin Oncol. 2014 Nov 1;32(31):3506-12.
 Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med. 2007 Nov 29;357(22):2277-84.
 Nagataki S. Computed tomography and radiation exposure. N Engl J Med. 2008 Feb 21;358(8):850-1; author reply 852-3.
 Thompson CA et al. Surveillance CT scans are a source of anxiety and fear of recurrence in long-term lymphoma survivors. Ann Oncol. 2010 Nov;21(11):2262-6.
Dr. Mark A. Lewis is an assistant professor in general & gastrointestinal medical oncology at the MD Anderson Cancer Center in Houston, Texas. He is double-boarded in hematology/oncology after completing a fellowship at the Mayo Clinic in Rochester, Minnesota, where he was chief fellow in that training program. He has a passionate interest in patient-doctor communication, including online dialogue, and moderates a Facebook group for patients with multiple endocrine neoplasia, a rare tumor syndrome that personally affects him and his son. He is also active on Twitter as @marklewismd. He is co-chair of the Social Media Working Group for SWOG, one of America’s largest cancer research cooperative groups, and co-chairs their Adolescent & Young Adult Cancer committee as well.