Who should be performing liver biopsies

Luddite: One who opposes technological change Examination of tissue by percutaneous liver biopsy is a valuable diagnostic tool in hepatology. There has been a quiet debate over many years about the best method of obtaining liver tissue, with advocates of routine use of ultrasound guidance claiming that it provides better tissue samples with lower risks than ‘blind’ biopsy, based largely on retrospective case series (1–5). However, there are notable Canadian dissenters (6) and the recent American Association for the Study of the Liver position paper on liver biopsy (7) did not recommend for or against the routine use of ultrasound guidance. The retrospective comparative review of liver biopsies performed blindly by gastroenterology personnel versus ultrasound-guided biopsies completed by interventional radiologists over the same time period reported in this issue of the Journal (8) (pages 425–430) documents and partially assesses a widespread trend to increasing use of the latter technique in one Canadian centre, at least for the grading and staging of chronic hepatitis C. This trend has also been documented at other centres (9–13). However, this latest review raises as many questions as it answers. The first obvious question is why the trend has developed at all. It is not clear whether the same caregivers have abandoned blind biopsy in favour of ultrasound-guided biopsies, or whether it is due in part to a shift in the background of the decision makers. It is quite understandable that if infectious disease specialists are increasingly caring for hepatitis C patients, they may request that radiologists perform the biopsies rather than involve gastroenterologists or hepatologists in the care of their patients. It is less understandable if trained gastroenterologists and hepatologists are abandoning a time-tested, safe and effective technique in favour of a more complex procedure in the belief that the latter is safer and provides better specimens. The results of the review do not support any definitive conclusion about increased safety or adequacy of ultrasound-guided biopsies. The blind biopsies performed in this study were of similar overall area and size as the ultrasound-guided biopsies, albeit with fewer complete portal tracts, but with more partial portal tracts – a discrepancy that is difficult to explain. The pathologist was probably not truly blinded to the technique used; the width and length of the specimens should expose the technique in many cases. There is no evidence that the blind biopsies were more prone to understage fibrosis – the greater fragmentation that resulted from the blind biopsies may be compensated for by the increased width of the specimens. There was a discrepancy in the experience level of the personnel that favoured the interventional radiologists performing the ultrasound-guided procedures, and a difference in the needles used. However, the rate of complications was higher in the ultrasound-guided group, although the study population was not large enough to reach any statistical conclusion regarding relative safety. As the authors note, complication rates have been previously shown to be related to operator experience and to the number of needle passes (12,13). It is, therefore, somewhat surprising to note that the experienced radiologists in this study routinely used two passes into the liver, and complicated the procedure further by routinely starting an intravenous and administering a sedative. A pervasive societal and medical belief that technological change is always advantageous undoubtedly contributes to a major degree in this trend. Even the language used seems designed to silence doubters – who, except for diehard luddites could argue against ‘advances’ or ‘enhanced’ technology such as contrast-enhanced ultrasound? The use of newer technologies that have replaced simpler ones – often without proof of increased safety or efficacy – is widespread and may be appropriate in some situations. Many hospital wards now use hand-held scanning devices (at a cost of $23,000 each) to measure residual bladder volume in a variety of sick patients, replacing in-out catheterization. The use of ultrasound guidance to place needles in critically precise locations, such as adjacent to the sciatic nerve for blocks, is one example of new technology whose usefulness is so self-evident that a trial is not warranted. Such is not the case in the placement of a needle in the liver. The priorities of gastroenterology trainees and misguided perceptions of the risks of liver biopsy, at times conveyed by luminal gastroenterologists, probably come into play in this trend. The costs should not, and probably are not, a major factor in deciding which technique to use. The ultrasound-guided biopsies are somewhat more costly (at least in Ontario, where this study was performed), as reflected by the Ontario Health Insurance Plan fee schedule ($107.65 versus $70.80 per procedure, respectively) excluding visit components, pharmacy costs and tray costs. As noninvasive assessments of liver fibrosis such as fibro-scanning improve, the need for biopsy by any technique will probably decrease. Given the universal consensus that experience with liver biopsy is important in obtaining adequate specimens and minimizing risks (14,15), it is vital to provide detailed training to both gastroenterologists and radiologists to ensure an adequate cadre of people who are experienced and comfortable performing the fewer biopsies that will be needed in the future. Trainees in both gastroenterology and interventional radiology need to decide early in their training whether they are going to perform liver biopsies and if so, strive to complete as many as possible under supervision during their training. For the present, it is clear that experienced gastroenterologists and hepatologists do not compromise their patient’s care by continuing to perform liver biopsies without formal ultrasound guidance. As the authors suggest, it may be appropriate to use a quick-look ultrasound scanner to locate the liver before a blind biopsy, although in this reviewer’s experience, it seldom changes the approach. To teach gastroenterology trainees to rely on ultrasound before performing a biopsy may be unwise because they may not have access to it in their future practices. Until there is a much needed head-to-head prospective comparative study of the two techniques using similar suction or cutting needles, gastroenterologists and hepatologists should continue to perform and teach the technique of blind liver biopsy. This debate about the best method, inevitably framed as being between old luddites who favour the blind biopsy technique versus young scientists who favour ultrasound guidance, should be viewed as between scientists of any age who demand evidence before adopting new technology versus those who blindly accept new technology as automatically more effective. From a broader perspective, the Darwinian biological principle that change is equivalent to improvement should not be applied blindly to technological evolution. If the pejorative term ‘luddite’ could be modified slightly to mean ‘one who demands evidence of superiority before routinely adopting newer technologies’, we should all be happy to be called luddites.