| THE
GRADING OF PROSTATE CANCER BIOPSIES
Dr Jane Melia Institute of Cancer Research, Royal Marsden Hospital. |
At the time any prostate cancer is diagnosed by biopsy the pathologist will see features in that cancer which puts its classification into a "good" or "bad" cancer. Because this has important implications for the patient's future it is important that all pathologists "sing from the same hymn sheet".
This study looked at the variability of reporting amongst various pathologists and found that those who specialized in prostate disease exhibited much greater concordance than "general" pathologists. Partially as a consequence of this research a national scheme is now in place to better educate pathologists about prostate cancer.
The above is a short research summary prepared by Prostate Research
Campain UK
Project 2000/11
A longer summary by Dr Melia of this important work follows:
For most prostate cancers, the diagnosis is made by taking prostatic biopsies, which are examined by a histopathologist. In addition the pathologist grades the cancer to help determine the prognosis(prospects for survival), so that the appropriate patient management is selected. However, many prostate cancers are now being detected at an early stage when they are very small, and grading of tissue from biopsy cores can be difficult, sometimes leading to differences in opinion of the grade between pathologists. The grading system widely used in the UK and elsewhere is that designed by Gleason and provides a sum score. It includes the grades of the two greatest areas of cancer in the biopsy specimens. The sum score values range from 2 (best prognosis) to 10 (worst prognosis), and appear in histology reports eg. 3+2=5, 4+4=8.
With support from the Prostate Research Campaign UK, a study was carried out to investigate the reproducibility of grading prostate cancer biopsies, and to identify areas of difficulty experienced by both urological and general pathologists in the UK. A panel of urological pathologists was formed: Dr. Parkinson (Chair), Professor Ball, Dr. Griffiths, Dr. Grigor, Dr. Harnden, Dr. Jarmulowicz, Dr.McWilliam, Professor Montironi and Dr. Moseley. Data collection and analysis was conducted by the Cancer Screening Evaluation Unit at the Institute of Cancer Research (ICR), Sutton.
Over a two year period, the panel of urological pathologists studied 81 slides from biopsy cores which were randomly selected from specimens originally diagnosed at their laboratories. The slides were selected to represent a mix of low, intermediate and high Sum scores. All slides were anonymised as the aim of the study was to study variation in grading between pathologists taking part in the study, and not to reclassify individual specimens.
A set of 20 slides at a time was circulated to each member of the panel. The pathologists recorded the Sum score of each slide on a standard form. They were not given the original grade of the slides. The forms were sent in confidence to ICR for data entry and analysis. Some slides were re-circulated to study consistency in the grades recorded by each pathologist. After each circulation the panel met in London to review the results and examine any slide with noticeable variation in grading under a multi-headed microscope.
Overall there was good agreement between the urological pathologists in the grading of the slides. Difficulty was found in the grading of low and one of the intermediate Sum scores. The results were similar or sometimes better than those reported from other countries.
Two groups of general pathologists, who overall had a smaller urological workload than the panel, were invited to attend two study days: 31 attended a meeting in Cambridge and 24 a meeting in Cardiff. At each meeting the pathologists were asked to grade each slide on a record sheet without conferring with the other pathologists. A teaching session on Gleason grading was given at lunchtime, and the pathologists were then asked to read slides again in the afternoon. The main finding was that there was poorer agreement in grading among the generalists than the specialists. Some improvement was seen after the teaching session, but it would be important to ensure that there is long-term benefit from any education measures. The results are similar to those found elsewhere such as the USA.
Our results come at time when important changes on the diagnosis and management of cancers are taking place within the NHS. A national scheme is being developed to monitor and educate pathologists about the diagnosis and grading of prostate cancer, and the organisation of cancer networks will provide increased opportunity for pathologists to share opinions on diagnosis and grade.
We are very grateful to the Prostate Research Campaign UK for supporting this work, and to all the general pathologists who took part in the study days.