Background
Grading of prostatic needle biopsies has undergone several refinements in the last decade. First, Epstein suggested that a diagnosis of Gleason score (GS) 2 + 2 = 4 cancer should not be made on the needle biopsies, because subsequent radical specimens showed upgrading in virtually all cases [
1]. Next, worst Gleason score (WGS) was shown superior to overall Gleason score (OGS) in predicting the final GS of the radical specimen, yielding fewer cases of unwanted upgrading events [
2]. The third major adaptation was made in the consensus conference of International Society of Urological Pathology 2005, leading to a refinement called modified GS [
3]. In that scheme, any aggressive cancer seen on the needle biopsies should be recorded and incorporated to the GS, even if present in small amount.
Worst Gleason score (WGS) is recommended for individually processed biopsies by ISUP 2005 consensus conference [
3]. In the case of pooled biopsies, the exact number of biopsies is sometimes difficult to know due to tissue fragmentation and/or overlapping of the biopsies, and thus, WGS cannot be reliably assessed [
3].
According to a recent survey among European pathology laboratories, approximately one half of the participants use individually processed biopsies, while the others immerse multiple biopsies per formalin container without special identification tags (Lars Egevad, personal communication). Individually processed biopsies allow clinicians to localize the histopathological findings to the anatomic biopsy site. In addition, when the biopsy cores are individually embedded in paraffin blocks, a separate GS can be assessed for each biopsy, and the worst of them is usually reported to the clinicians to guide the treatment. Instead, the uropathologists did not reach consensus whether to use worst or overall GS in the case when multiple cancer-containing biopsies are pooled to one formalin container without identification tags [
3].
A few studies comparing OGS and WGS have been published and only one of them after the ISUP conference [
4]. In three studies WGS at any biopsy site was better than OGS at predicting the pathological T-stage and GS in radical prostatovesiculectomy specimens [
2,
4,
5] whereas in one study, OGS performed better in predicting progression-free survival in patients treated with radiotherapy [
6].
Our earlier study analyzing biopsies from endocrine-treated patients indicated that OGS was the strongest independent prognosticator of all histopathological parameters [
7]. Gleason score assessment according to ISUP 2005, using the most aggressive pattern as a secondary Gleason grade even when it is present in only a small area, yielded the best prognostic classification using groupings < 7(4 + 3), 7(4 + 3)-8, and 9-10. In the present study, we examined whether the WGS in a single biopsy core would improve prognostic accuracy when compared with OGS. We also evaluated the prognostic value of compound Gleason score from the original pathology reports before the ISUP 2005 era.
Discussion
The refinements of the ISUP 2005 consensus conference on Gleason scoring of needle biopsies have generally yielded better prognostic accuracy [
10]. Our results indicate that modified Gleason scores according to the ISUP 2005 system are higher than compound GS's from 1999-2003, and this upgrading is associated with improved prognostic accuracy. Moreover, the results suggest, that OGS may be a slightly stronger or at least equally adequate predictor of PSA progression than WGS, when assessed from pooled biopsies.
A major implication of the revised 2005 ISUP guidelines has been the mandate to integrate the most aggressive tertiary patterns to secondary in needle biopsy scoring, even when the pattern is limited to a small area. A recent webmicroscope-based concordance study about Gleason grading of GS 6-8 by the European Network of Uropathologists suggested that general pathologists are starting to overgrade the experts (Lars Egevad, personal communication). Because Gleason grading is subjective, it is not difficult to detect some glandular fusion, and to interpret them as secondary Gleason pattern 4. Due to aforementioned issues, a fraction of cancers previously graded as GS 3 + 3 = 6 would nowadays end up with GS 3 + 4 = 7. Thus, it has been suggested that changing definitions shift the cut-off between low-grade and high grade cancers from 3 + 4 to 4 + 3 [
11,
12]. The results of the present study are consistent with that.
According to the 2005 ISUP consensus conference, the highest (worst) GS should not be assessed from biopsies immersed in the same formalin container ("pooled biopsies") due to tissue fragmentation [
3]. When all six biopsies from one lobe are formalin-fixed in the same container, they may become fragmented or overlap when embedded, disturbing the attempt to assess the WGS of the individual needle biopsies. To avoid this, some laboratories choose to ink pooled cores different colors and thus be more specific about sites and fragmentation. On the other hand, WGS was recently shown to be a better predictor of the histopathological findings from subsequent radical prostatectomy specimens [
4]. In our study, the WGS was assessed from the needle biopsies of one prostate lobe embedded in one paraffin block. Because of this, it is possible that our WGS results were biased by tissue fragmentation. However, in the majority of the cases (
n = 193/236, 82%), the WGS was equal to the OGS. If there were a bias due to fragmentation, we should expect more cases with WGS > OGS.
A major problem when multiple biopsies are stored in one container is that the exact locus information of the biopsies is lost unless site identifiers are used. Another disadvantage is that it is harder to keep all the biopsies in the same plane of section, but this can be avoided in eg. by using foam plastic inside the cassettes. The locus information is essential when considering targeted brachytherapy or cryotherapy in focal carcinomas. Moreover, the anatomic localization of carcinoma foci is useful when planning nerve-sparing radical prostatectomy and to avoid side effects from external-beam radiotherapy. The problems associated with placing multiple biopsies in one container can be overcome by immersing one core biopsy per formalin-container, which is quite laborious for all the participants: the urologist, laboratory technicians, and pathologist. Two major advantages of embedding multiple needle biopsy cores in one paraffin block are the reduced workload and the ability to analyze immunohistochemical stainings from all the biopsies at once, when deemed necessary.
There are a few limitations in the present study. First, although PSA progression works as a surrogate end-point for progressive prostate cancer, it does not necessarily correlate specifically with cancer or overall survival. Due to the small number of deaths in our series, we cannot conclude that OGS was a better prognostic factor in terms of death as a hard end-point. To address this question, a longer follow-up is needed. Second, CGS was not re-evaluated in the present study; instead it was obtained from the original pathology reports, which limits the value of this comparison. Third, the study contained a limited number of cores and the number of cases in which WGS > OGS was rather low (n = 43). Therefore, it is not surprising that the WGS and OGS yield similar results.
Acknowledgements
We wish to thank Ms. Mariitta Vakkuri and Ms. Riitta Vaalavuo for their skillful technical assistance. The work was supported by the Academy of Finland, the Cancer Society of Finland, the Reino Lahtikari Foundation, the Sigrid Juselius Foundation and the Competitive Research Funding of the Pirkanmaa Hospital District.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
This study has been designed by PK, TV, and TTT. The needle biopsy samples have been analyzed by TTT. Manuscript has been written by TTT, PK and JI. VT is responsible for the digitalization of the images and virtual microscope system. TLJT has acquired the clinical database of the patients. TV is responsible for the statistical analyses and finalization of the manuscript. Conclusions have been drawn mainly by TTT, JI, TV and PK. All authors read and approved the final manuscript.