Balancing Risk-benefit in Pediatric Nuclear Medicine Far from Childs Play
In pediatric nuclear medicine, balancing clinical diagnostic imaging with radiation risk remains a challenge. Professional societies in the U.S. and Europe have stepped up efforts to minimize administered activity, based on updated dosage cards and guidelines. Medical physicists recognize that while these protocols are helpful, there is room for improvement. Using simulations and other tools, they are discovering strategies to reduce ionizing radiation without compromising image quality in children.
“We can’t lose sight that we are doing the study because we are fairly confident that we will give physicians useful information that will help them treat the patient,” says Frederic H. Fahey, DSc, director of nuclear medical physics at Children’s Hospital in Boston. “On the other hand, we need to keep in mind that there may be a small risk associated with it.”
Children are more sensitive to radiation than adults, and have a longer life span in which a radiation-induced cancer potentially can arise. But establishing long-term risk from imaging studies is challenging, particularly in pediatric patients, says Adam M. Alessio, PhD, of the radiology department at Seattle Children’s Hospital. It is difficult to estimate radiation dosimetry in children because of their variable organ sizes, morphology and biokinetics. He describes evidence of long-term risk from low-level radiation exposure as weak, with coarse estimates based on evaluations such as the Life Span Study of survivors of the Hiroshima and Nagasaki bombings and assessments for nuclear regulatory agencies.
“It is a theory of potential risk,” Alessio says. “There has never been an example where an imaging study was directly linked to future cancer. We are making an educated guess that it might happen later on, but it is a tenuous connection.”
Concern over the cumulative effect of radiation exposure from sources such as tracers used in medical imaging has grown in recent years with the increased use of diagnostic scans. In a 2000 report to the United Nations General Assembly, the U.N.’s scientific committee reported that the use of ionizing radiation for diagnostics was widespread. On average, 37 million nuclear medicine examinations are annually performed worldwide, with the number of procedures in the U.S. increasing almost three-fold between 1980 and 2006 (Radiology 2009;253:520-531). The National Council on Radiation Protection and Measurements reported in 2009 that Americans were exposed to more than seven times as much ionizing radiation from medical procedures in 2006 than in the early 1980s.
In the U.S., imaging procedures in children may be frequent, according to one population-based study that looked at the use of low-dose ionizing radiation in children enrolled in five large healthcare markets. The study found that 42.5 percent of insured children had at least one procedure in the three-year study period. In this patient population, the authors calculated that the average child will receive seven procedures before reaching age 18. The majority of the procedures were plain radiography, though, and nuclear medicine accounted for only 0.9 percent of the procedures (Arch Pediatr Adolesc Med; online Jan. 3, 2011).
“It was not the best idea to scale by adult activity,” says Michael Lassmann, PhD, head of the nuclear medicine department at the University of Wuerzburg and chair of the EANM dosimetry committee from 2001 to 2008. “Different European countries had different diagnostic reference levels.” As a result, a child in one European country who weighed the same as a counterpart in another European country could receive a different—and sometimes larger—dose.
Jacobs et al recommended using three tracer-dependent cards: one for tracers for renal studies, one for iodine-labeled tracers for thyroid studies and one for the remaining tracers (Eur J Nucl Med Mol Imaging 2005;32[5]:581-588). In 2007, the EANM dosimetry and pediatrics committees endorsed the three tracer-dependent dosage card protocol for 95 radiopharmaceuticals. They added minimum recommended administered activities in 39 procedures frequently performed in children to guarantee a minimum standard of imaging quality throughout Europe (Eur J Nucl Med Mol Imaging 2007;34:796-798).
Europe was not alone in practice variability. Fahey and colleagues found that radiopharmaceutical dosimetry also varies greatly across institutions in North America. In a 2008 survey of dosimetry practices for 16 nuclear medicines in 13 pediatric hospitals in the U.S. and Canada, they found minimum administered activity varied by as much as a factor of 20 and maximum activity in children older than 12 months on average varied by a factor of three, and in one case varied by a factor of 10. The results highlighted the need for standards (J Nuc Med 2008;49[6]:1024-1026).
In response, the Society of Nuclear Medicine, the Pediatric Imaging Council, the Society for Pediatric Radiology and the American College of Radiology formed a working group to create consensus guidelines for administered doses in children. The 2010 North American guidelines recommended dose based on body weight in nine commonly used pharmaceuticals (J Nuc Med 2008;52[2]:318-322).
The guidelines and dosage card are generally in concordance but vary in some aspects because of differing practices. The guidelines, for instance, recommend slightly lowered administered activities for infants and tiny children.
The North American consensus group noted that local practice may vary, depending on the patient population, choice of equipment and software and physician preference. “I don’t know that you can standardize to exactly the same administered activity at every site, but you can get much less variability,” says Fahey, who along with Alessio was among the consensus group members. The authors, most hailing from pediatric hospitals, also wanted to provide guidance to general hospitals that do not perform these procedures routinely.
To test the notion, they conducted 14 whole-body 18F-FDG PET/CT scans on 13 pediatric patients following the protocols for administered activity and fixed acquisition durations: three minutes per field of view (FOV) if the patient weighed less than 22 kg and five minutes per FOV if the patient weighed more than 22 kg. From each exam, they truncated the data into shorter duration sets to simulate shorter scans, for instance, four minutes per FOV, three minutes per FOV, down to one minute per FOV.
The volumetric PET/CT images were randomized, blindly reviewed and scored. Examinations using the initial protocol were all graded as adequate. For larger patients, durations of three minutes per FOV were considered adequate as well, with no loss of diagnostic utility. Alessio et al argued that if duration is used as a surrogate for dose, then dose could be reduced by 40 percent if patients were scanned for five minutes per FOV (J Nucl Med 2011;52:1028-1034). “This gave us some confidence that we could lower dose a little bit or acquire a little bit longer, if necessary, and supported the use of the consensus guidelines,” Alessio says.
Just as Jacobs and his colleagues posited alternatives to weight-based adjustment, George Sgouros, PhD, director of the radiopharmaceutical dosimetry section at Johns Hopkins University School of Medicine in Baltimore, and colleagues have shown that adjusting by body type also may provide an opportunity to refine dose. In an approach that is similar in spirit to Jacobs et al, Sgouros and his colleagues devised a virtual reality study to obtain simulated SPECT images from two types of pediatric patients of the same weight: one 10-year old tall, thin girl and one 10-year-old short, stout girl. Using anthropomorphic phantoms that were derived from true CT anatomy for patients and sophisticated algorithms for simulated administered activity and imaging, they showed that they could obtain adequate image quality in the tall, thin girl using half the standard mass-based administered activity that was needed for similar results in the short, stout girl (J Nucl Med 2011;52:1923-1929).
“These are fully validated results that point to the idea that even if you have weight-based adjustment, there is room for improvement if you take into account other factors,” Sgouros says. The study focused on the labeling agent technetium-99m dimercaptosuccinic acid, which can be used to assess renal function in children. But the methodology could be extended to other radiopharmaceuticals for tables categorizing pediatric patients by weight, sex, age and morphology, which could, for instance, be incorporated into future guidelines.
This virtual reality approach offers an alternative to experiments such as clinical trials, which in children are sometimes ethically questionable and often lack the support of pediatric patients’ parents and physicians. “This reflects a trend, not just in medical imaging but in general, as we have more and more sophisticated computational tools and more data become available,” Sgouros says.
Lassmann categorizes the Sgouros team’s approach as potentially game changing for pediatric nuclear medicine, “provided you have the resources to do all the calculations.” Pat B. Zanzonico, PhD, a medical physicist at Memorial Sloan-Kettering Cancer Center in New York City and chair of a committee that reviews experimental protocols involving ionizing radiation, also is enthusiastic about the potential for computer simulations for optimizing effective dose in children.
“With the rigor achieved in the current study,” Zanzonico wrote in an editorial accompanying the Sgouros paper, “virtual reality is indeed better than the real thing” (J Nucl Med 2011;52:1845-1847).
Zanzonico adds that physicians and parents of pediatric patients should be critical about adopting new techniques that supersede well-established clinical procedures. “On the other hand, if we want to move the field forward and optimize the risk-benefit ratio, this approach is the best way to do so precisely because it avoids having to study live patients,” he says. “It may not eliminate the need for clinical validation altogether, but it expedites the process.”
Defining risk
The benefit of administering radioactive agents for procedures such as PET or SPECT derives from the information gained through imaging studies. Higher doses provide better images but exposing pediatric patients to more radiation that potentially may trigger detrimental cellular changes and eventual cancer risk (although the risk has only been modeled and has never been proven). Lower doses reduce the potential cancer risk but may produce inadequate images, compromising the diagnostic process and possibly requiring follow-up tests that again may expose patients to ionizing radiation. But the advantages of appropriately administered pediatric nuclear medicine are huge, medical physicists say.“We can’t lose sight that we are doing the study because we are fairly confident that we will give physicians useful information that will help them treat the patient,” says Frederic H. Fahey, DSc, director of nuclear medical physics at Children’s Hospital in Boston. “On the other hand, we need to keep in mind that there may be a small risk associated with it.”
Children are more sensitive to radiation than adults, and have a longer life span in which a radiation-induced cancer potentially can arise. But establishing long-term risk from imaging studies is challenging, particularly in pediatric patients, says Adam M. Alessio, PhD, of the radiology department at Seattle Children’s Hospital. It is difficult to estimate radiation dosimetry in children because of their variable organ sizes, morphology and biokinetics. He describes evidence of long-term risk from low-level radiation exposure as weak, with coarse estimates based on evaluations such as the Life Span Study of survivors of the Hiroshima and Nagasaki bombings and assessments for nuclear regulatory agencies.
“It is a theory of potential risk,” Alessio says. “There has never been an example where an imaging study was directly linked to future cancer. We are making an educated guess that it might happen later on, but it is a tenuous connection.”
Concern over the cumulative effect of radiation exposure from sources such as tracers used in medical imaging has grown in recent years with the increased use of diagnostic scans. In a 2000 report to the United Nations General Assembly, the U.N.’s scientific committee reported that the use of ionizing radiation for diagnostics was widespread. On average, 37 million nuclear medicine examinations are annually performed worldwide, with the number of procedures in the U.S. increasing almost three-fold between 1980 and 2006 (Radiology 2009;253:520-531). The National Council on Radiation Protection and Measurements reported in 2009 that Americans were exposed to more than seven times as much ionizing radiation from medical procedures in 2006 than in the early 1980s.
In the U.S., imaging procedures in children may be frequent, according to one population-based study that looked at the use of low-dose ionizing radiation in children enrolled in five large healthcare markets. The study found that 42.5 percent of insured children had at least one procedure in the three-year study period. In this patient population, the authors calculated that the average child will receive seven procedures before reaching age 18. The majority of the procedures were plain radiography, though, and nuclear medicine accounted for only 0.9 percent of the procedures (Arch Pediatr Adolesc Med; online Jan. 3, 2011).
 |
| A newborn phantom, one of a library of 3D pediatric models developed at the University of Florida in Gainesville is used to simulate radiopharmaceutical activity distributions. Source: Wesley E. Bolch, PhD |
A better way
In 2004, Filip Jacobs, PhD, of the nuclear medicine department at Ghent University Hospital in Ghent, Belgium, and colleagues proposed that the European Association of Nuclear Medicine (EANM) change its pediatric dosage card from a weight-dependent fraction of the dose administered to adults to a tracer-dependent protocol. The goal was to harmonize the country-by-country variability in administered activities in children throughout Europe.“It was not the best idea to scale by adult activity,” says Michael Lassmann, PhD, head of the nuclear medicine department at the University of Wuerzburg and chair of the EANM dosimetry committee from 2001 to 2008. “Different European countries had different diagnostic reference levels.” As a result, a child in one European country who weighed the same as a counterpart in another European country could receive a different—and sometimes larger—dose.
Jacobs et al recommended using three tracer-dependent cards: one for tracers for renal studies, one for iodine-labeled tracers for thyroid studies and one for the remaining tracers (Eur J Nucl Med Mol Imaging 2005;32[5]:581-588). In 2007, the EANM dosimetry and pediatrics committees endorsed the three tracer-dependent dosage card protocol for 95 radiopharmaceuticals. They added minimum recommended administered activities in 39 procedures frequently performed in children to guarantee a minimum standard of imaging quality throughout Europe (Eur J Nucl Med Mol Imaging 2007;34:796-798).
Europe was not alone in practice variability. Fahey and colleagues found that radiopharmaceutical dosimetry also varies greatly across institutions in North America. In a 2008 survey of dosimetry practices for 16 nuclear medicines in 13 pediatric hospitals in the U.S. and Canada, they found minimum administered activity varied by as much as a factor of 20 and maximum activity in children older than 12 months on average varied by a factor of three, and in one case varied by a factor of 10. The results highlighted the need for standards (J Nuc Med 2008;49[6]:1024-1026).
In response, the Society of Nuclear Medicine, the Pediatric Imaging Council, the Society for Pediatric Radiology and the American College of Radiology formed a working group to create consensus guidelines for administered doses in children. The 2010 North American guidelines recommended dose based on body weight in nine commonly used pharmaceuticals (J Nuc Med 2008;52[2]:318-322).
The guidelines and dosage card are generally in concordance but vary in some aspects because of differing practices. The guidelines, for instance, recommend slightly lowered administered activities for infants and tiny children.
The North American consensus group noted that local practice may vary, depending on the patient population, choice of equipment and software and physician preference. “I don’t know that you can standardize to exactly the same administered activity at every site, but you can get much less variability,” says Fahey, who along with Alessio was among the consensus group members. The authors, most hailing from pediatric hospitals, also wanted to provide guidance to general hospitals that do not perform these procedures routinely.
Beyond weight-based dose
One challenge in pediatric molecular imaging is keeping the child comfortable and still to avoid motion artifacts. “All of these nuclear medicine studies take 30 minutes to an hour, and that is a lot of time for a child to stay still for the camera to take images,” Alessio says. Noting that 18F-FDG pediatric dosing and acquisition durations were based on an extrapolation from adult guidelines, he and colleagues wondered if scan durations might be reduced without losing the diagnostic utility of the scan, and if dose might also be reduced.To test the notion, they conducted 14 whole-body 18F-FDG PET/CT scans on 13 pediatric patients following the protocols for administered activity and fixed acquisition durations: three minutes per field of view (FOV) if the patient weighed less than 22 kg and five minutes per FOV if the patient weighed more than 22 kg. From each exam, they truncated the data into shorter duration sets to simulate shorter scans, for instance, four minutes per FOV, three minutes per FOV, down to one minute per FOV.
The volumetric PET/CT images were randomized, blindly reviewed and scored. Examinations using the initial protocol were all graded as adequate. For larger patients, durations of three minutes per FOV were considered adequate as well, with no loss of diagnostic utility. Alessio et al argued that if duration is used as a surrogate for dose, then dose could be reduced by 40 percent if patients were scanned for five minutes per FOV (J Nucl Med 2011;52:1028-1034). “This gave us some confidence that we could lower dose a little bit or acquire a little bit longer, if necessary, and supported the use of the consensus guidelines,” Alessio says.
Just as Jacobs and his colleagues posited alternatives to weight-based adjustment, George Sgouros, PhD, director of the radiopharmaceutical dosimetry section at Johns Hopkins University School of Medicine in Baltimore, and colleagues have shown that adjusting by body type also may provide an opportunity to refine dose. In an approach that is similar in spirit to Jacobs et al, Sgouros and his colleagues devised a virtual reality study to obtain simulated SPECT images from two types of pediatric patients of the same weight: one 10-year old tall, thin girl and one 10-year-old short, stout girl. Using anthropomorphic phantoms that were derived from true CT anatomy for patients and sophisticated algorithms for simulated administered activity and imaging, they showed that they could obtain adequate image quality in the tall, thin girl using half the standard mass-based administered activity that was needed for similar results in the short, stout girl (J Nucl Med 2011;52:1923-1929).
“These are fully validated results that point to the idea that even if you have weight-based adjustment, there is room for improvement if you take into account other factors,” Sgouros says. The study focused on the labeling agent technetium-99m dimercaptosuccinic acid, which can be used to assess renal function in children. But the methodology could be extended to other radiopharmaceuticals for tables categorizing pediatric patients by weight, sex, age and morphology, which could, for instance, be incorporated into future guidelines.
This virtual reality approach offers an alternative to experiments such as clinical trials, which in children are sometimes ethically questionable and often lack the support of pediatric patients’ parents and physicians. “This reflects a trend, not just in medical imaging but in general, as we have more and more sophisticated computational tools and more data become available,” Sgouros says.
Lassmann categorizes the Sgouros team’s approach as potentially game changing for pediatric nuclear medicine, “provided you have the resources to do all the calculations.” Pat B. Zanzonico, PhD, a medical physicist at Memorial Sloan-Kettering Cancer Center in New York City and chair of a committee that reviews experimental protocols involving ionizing radiation, also is enthusiastic about the potential for computer simulations for optimizing effective dose in children.
“With the rigor achieved in the current study,” Zanzonico wrote in an editorial accompanying the Sgouros paper, “virtual reality is indeed better than the real thing” (J Nucl Med 2011;52:1845-1847).
Zanzonico adds that physicians and parents of pediatric patients should be critical about adopting new techniques that supersede well-established clinical procedures. “On the other hand, if we want to move the field forward and optimize the risk-benefit ratio, this approach is the best way to do so precisely because it avoids having to study live patients,” he says. “It may not eliminate the need for clinical validation altogether, but it expedites the process.”
| Putting Risk in Perspective |
| In the tug of war between risk and benefit, benefit often gets the short end of the stick in terms of awareness. Putting risk and benefit into perspective may alleviate concerns for pediatric patients and their parents. “Whenever we talk about risk, we have to put it in the context of benefit,” says Frederic H. Fahey, DSc, director of nuclear medical physics at Children’s Hospital in Boston. “To be able to give patients and their families an answer and discuss it with them relieves them and makes them feel more comfortable to move forward with the test if it is necessary.” Fahey and colleagues offer these tips for effectively communicating about radiation risk in pediatric nuclear medicine (J Nucl Med Technol 2012;40:13-24):
|