Prostate scanning: overexposed?10 February 2012
Research by Dr Jim Hu and Dr Hua-yin Yu of the Brigham and Women’s Hospital in Boston, US, calls into question the extent to which scanning is used in the treatment of prostate cancer.
Prostate cancer is the most common non-cutaneous malignancy and second most common cause of cancer-related death among men in the US. On a worldwide scale, it is the second most frequently diagnosed malignancy among men, lung cancer being the first, and is the sixth leading cause of cancer-related death. Given the economic and clinical impact of this disease, use of healthcare resources in the diagnosis and treatment for prostate cancer has come under scrutiny, both as an area with potential waste and for underutilisation of appropriate care.
Since the advent of prostate-specific antigen (PSA) screening, 92% of incident cases of prostate cancer are localised due to early disease detection, before the onset of any symptoms or clinical findings. Additionally, prostate cancer is a slow-growing disease and screening has led to the detection of large numbers of low-risk prostate cancers that may not otherwise have led to clinical manifestations for years, even decades.
This is particularly relevant in increasingly aging populations, where patients are living with multiple comorbid conditions that may prove to be of equal or greater threat to men's lifespans than their prostate cancer diagnosis. Therefore, patients and physicians are faced with complex treatment decisions between several competing therapeutic options, including surgery, radiation and active surveillance (observation with regular PSA testing and repeat biopsies to monitor for evidence of disease progression), as well as an array of other modalities that are under investigation.
There are multiple factors to take into consideration in the discussion of best management, including pretreatment PSA, clinical stage and grade. This data is used to assess the risk of metastatic disease and likelihood of disease-free survival, thus serving to assess whether further imaging is warranted to help guide treatment decisions.
The American Urological Association (AUA) created guidelines based on these pretreatment clinical factors, which provide a picture of a patient's risk of having metastatic disease, for which localised therapy would be of no oncologic benefit. The AUA recommends the use of computed tomography (CT) and radionuclide bone scan only for those men with high-risk disease: T3/T4 disease, Gleason grade 8 or higher, or PSA greater than 20ng/ml, with the European Association of Urology also recommending bone scans using the same parameters.
Similarly, the US National Comprehensive Cancer Network (NCCN) Guidelines recommends bone scans for T3/T4 disease or Gleason 8 disease, with exceptions for T1 or T2 disease where PSAs are greater than 20ng/ml and 10ng/ml respectively. CT recommendations are similar, and reserved for T3/T4 disease, or T1/T2 disease with >20% risk of lymph node involvement based on nomograms. PSA less than 20ng/ml has a high negative predictive value, with a less than 1% likelihood that radiologic findings will suggest metastatic disease.
While guidelines are created in an effort to improve the quality and consistency of patient care, assessing the degree of adherence to guidelines in actual practice is necessary to determine whether these proposed quality-improvement measures drive any clinical benefit. Guidelines are created based on existing high-level evidence and expert panel consensus, which are usually drawn from the experiences of high-volume tertiary referral centres (specialists). However, little is known about if practitioners in the community follow, or are even aware of these guidelines. Thus, research and efforts to improve quality may be futile in the absence of knowledge and adherence by those on the front line of prostate cancer diagnosis.
Unlike rare malignancies, which are often diagnosed and/or treated predominantly by specialists, the high incidence and ubiquity of prostate cancer makes it a 'community' disease that is diagnosed, treated and managed at all levels of the healthcare spectrum.
In an effort to evaluate population-based use of these imaging modalities for the staging of prostate cancer, we studied the US National Cancer Institute's cancer registry, the Surveillance Epidemiology and End Results (SEER) database, linked to the database for Medicare, which provides insurance coverage for 97% of patients over the age of 65 years, with SEER capturing 26% of all beneficiaries.
Using this data, we identified 30,183 men who were diagnosed with prostate cancer between 2004 and 2005, with continuous follow-up within Medicare through 2007. They were risk-stratified based on low (PSA =10, =T2a, =Gleason 6), intermediate (PSA 10> and =20, T2b or T2c, Gleason 7), or high-risk disease (PSA >20, =T3, =Gleason 8), with the high disease-risk group warranting staging imaging in accordance with AUA guidelines. This resulted in men with 32%, 43% and 25% with low, intermediate and high-risk disease, respectively. After adjusting for differences in baseline patient demographics (age, race, marital status, education level, income and population density), as well as comorbid conditions, we found that there were significant discrepancies between recommended care and actual practice.
First, CTs and bone scans were grossly underused for the high-risk group, with only 60% of men receiving the appropriate recommended care. The clinical implication of this finding is that 40% of men are not receiving adequate staging, and those with detectable metastatic disease may be subject to local therapies that will not adequately treat the disease, which should be treated with androgen deprivation therapy.
While this treatment is not curative and is fraught with metabolic, physical and constitutional side effects, such as weight gain, impotence, loss of libido, and hot flashes, the additional potential side effects associated with unwarranted radiotherapy or surgery may further compound decreases in patient quality of life.
Radiation is associated with early and late irritative bowel and urinary symptoms, as well as risks of impotence and constitutional effects. Surgery poses significant long-term risks of urinary incontinence and impotence, and immediate operative risks of bleeding and structural injury.
Second, imaging was inappropriately utilised in 34% and 48% of men with low and intermediate-risk prostate cancer respectively. In these men, the likelihood of detecting metastatic disease that will alter treatment plans is so low that its use proves not only to be unnecessary, but potentially harmful to the healthcare system at large in terms of cost and resource utilisation, as well as to patients themselves.
Of particular concern is the use of CT scans, where there is the well-recognised increased carcinogenic risk. While this risk increase is still very low, studies of the US population estimate that 1.5-2.0% of all malignancies may be attributable to radiation exposure from CT scans. Furthermore, with nearly all of these patients having negative studies, 50% of them went on to have radiotherapy, further increasing their dose exposure. Also, men who eventually underwent radiotherapy were more likely to have undergone imaging than those who went on to have surgery or active surveillance. Moreover, the additional imaging frequently leads to an array of incidental findings, leading practitioners to fruitlessly pursue further work-up of many clinically insignificant findings, increasing patient anxiety and straining healthcare resources.
This overuse has previously been found to occur even at the level of single high-volume referral centres, where there should be greater awareness of the clinical parameters by which such imaging is warranted. This may be a reflection of physicians practising defensive medicine in an increasingly litigious US healthcare climate. Other possible causes for these findings include differences in patient demand for further work-up as well as potential financial incentives for physicians to order more tests.
There were variations in imaging use based on patient demographics too. It is unknown if deviation between recommended care and actual practice is the result of practice variations between practitioners, specialities or region. Men in rural areas were more likely to undergo imaging than those in urban areas, possibly related to differences in awareness of practice guidelines among practitioners. Along this line, patients residing in areas with greater education levels were less likely to receive inappropriate imaging, suggesting greater awareness of recommended practice in these areas. Furthermore, men aged 75 years or older were the most likely to receive additional imaging, independent of their pretreatment risk.
Given the slow progression of low-risk prostate cancer relative to other malignancies, it is unlikely that men with a less-than-ten-year life expectancy will clinically benefit from treatment for the low-risk disease. Thus, in this population of older men, who are more likely to have competing comorbidities, aggressive diagnostic work-ups are less likely to alter treatment plans, and more importantly, unlikely to improve clinical outcomes.
There are increasing efforts throughout all medical fields to develop quality of care indicators by which to measure if appropriate care is being delivered, with the goal of not only improving compliance with recommended care, but also as standards by which to determine payment for services.
Currently, the use of bone scans in low-risk prostate cancer patients is already considered a negative performance quality indicator by the US federal agency, which determines Medicare reimbursement. Our study found that the use of imaging without indication resulted in nearly $3.5m in extraneous expenditures for the study population, which comprises only 26% of men over 65 years old; thus actual expenditures are several times higher than this figure. While the data reflects US prostate cancer care practice patterns, it is unknown what imaging practice patterns are worldwide in the staging of prostate cancer.
At many institutions, the role of gate-keeper to determine if imaging should be performed has fallen upon radiologists, who initially evaluate studies based on the clinical information provided. While the overuse of CT and bone scans for prostate cancer may be self evident to high-volume radiologists who interpret negative scan after negative scan for prostate cancer staging, increasing the awareness among all radiologists of the clinical rationale for or against prostate cancer imaging may better help them to screen out the unnecessary studies that contribute to needless patient radiation exposure and increased healthcare expenditures.
With increasing numbers of screening protocols being developed to curtail inappropriate use of radiologic technology, it behooves us to take on a similar approach to prostate cancer imaging. If proper protocols exist, it may serve not only to simplify the screening process for radiologists, but also act as continuing medical education for those practitioners responsible for determining whether these tests are necessary. Such measures may serve not only to improve quality of cancer care for patients, but also to benefit healthcare systems as a whole.
This article was first published in our sister publication Medical Imaging Technology.