Statistics from Altmetric.com
I read with great interest the article by Egger and colleagues (Gut 2003;52:18–23) evaluating laser induced fluorescence endoscopy (LIFE) and methylene blue (MB) directed biopsies for detection of dysplasia in Barrett’s oesophagus.
As the authors point out, there have been no fully published studies to date on this much talked about procedure. The authors found that LIFE and MB had limited accuracy, as did standard random biopsy. Although LIFE and MB detected a total of five cases of high grade dyplasia and 11 cases of low grade not seen on four quadrant random biopsy (4QB), they concluded that these methods are “not capable of increasing the diagnostic accuracy or replacing standard four quadrant biopsies”.
How could these data lead to this conclusion? The authors discount all but one high grade and seven low grade lesions detected by LIFE or MB because they were “within the 4QB protocol”. It was assumed by the authors that these sites would have been biopsied by random techniques had it not already been sampled with AF or MB. Given that the biopsies were standard 7 mm forceps, that dysplasia can be very focally distributed, and the area included within the 4QB covers two linear centimetres, it is difficult to assume that this exact site would have been biopsied with a random technique. This assumption, if incorrect, would result in underestimation of the value of LIFE or MB.
In addition, the authors further discounted the one remaining high grade dysplasia site and four more low grade sites because they occurred in patients with known cancer who presumably would be treated for the cancer regardless. There is little doubt that detection of low or even high grade dysplasia has little relevance if a cancer is already known. The main group of patients where LIFE, MB, and other advanced techniques should be applied are those without macroscopically evident tumours and cancer. Discounting LIFE and MB for this reason may further underestimate its value.
If we do not discount these cases then LIFE and MB appear to compliment 4QB for the detection of dysplasia, with each technique independently detecting dysplastic sites that the other missed.
I agree that LIFE and MB remain controversial and applaud the authors for publication of this first LIFE trial. Given the constraints of the study however, it may be premature to proclaim these techniques incapable. More well conducted studies are clearly needed. The field of imaging technologies is also evolving rapidly and new and better techniques are constantly on the near horizon.