Background

Although prostate cancer is a common disease (over 500 000 cases are diagnosed in North America and Europe each year), there is considerable variation in its natural history.1 Many cancers are indolent and will never become life threatening while other cases, with an initially similar appearance, may rapidly become life threatening. Unfortunately, established prognostic factors (Gleason score, T stage, blood prostate-specific antigen, cancer volume) are inadequate, and it is important to identify new biomarkers that can predict more precisely the clinical behaviour of individual patients. We have previously reported methods that can be used to test biomarkers in formalin-fixed needle biopsies taken from patients with prostate cancer at the time of diagnosis. In each method, we reorientate the biopsies into a vertical orientation allowing construction of biopsy TMAs and subsequent analysis of the markers in cross-sections taken from each biopsy TMA. In our original technique, a maximum of 20 biopsy specimens could be included within each TMA.2 A particular problem with this method is that the cutting out of small regular ‘checkers’ of wax containing a segment of the biopsy on one face, which is required for the construction of the biopsy TMA, is technically demanding and time-consuming. To address this problem, we recently presented a simplified technique for cutting wax checkers that allowed construction of biopsy TMAs containing 54–72 checkers.3 The technique can be used to examine both IHC and FISH based biomarkers.3 These approaches have been validated in studies of the biomarker Ki67 in biopsy specimens from active surveillance patients,4 for the biomarkers Hif-1α, VEGF, osteopontin, MDM2, p53 and Bcl-2 in biopsy specimens from radiotherapy patients.5 6 Here, we present substantial improvements of this procedure that allows more rapid construction of TMAs containing up to 104 biopsy cores.

Technical development

Prostate biopsy samples for these studies were collected from men with untreated, localised (clinical stage T1/2a, Gleason score ≤3+4; prostate-specific antigen <15; ≤50% positive cores) prostate cancer who were managed in a prospective active surveillance study at the Royal Marsden Hospital NHS Foundation Trust. The methods for selecting samples for biopsy TMA construction from patients with prostate cancer entered into active surveillance have been described previously.4 All patients gave their written consent to take part in the active surveillance study, which was approved by the local research ethics committee.

The new procedure consists of two steps. In the first step, biopsy donor blocks were constructed using previously reported methods,2 but with only five biopsy checkers per block (figure 1A–F). The height of the ‘checker’ was specifically kept at around 4 mm (figure 1A,B) as previously reported.3 A variation in the new procedure was that the end of the core was painted with a red dot to identify the core end after embedding (figure 1C), and the side of the checker opposing the biopsy core was painted blue (figure 1D) for orientation. A finished biopsy ‘checker’ (figure 1E) and donor block (after paraffin wax embedding) (figure 1F) are shown.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Construction of biopsy donor blocks. For clarity, the formalin-fixed prostate cancer needle biopsy has been coloured green. A malignant portion of the biopsy was marked and represented here by a black rectangle (A). The biopsy was then cut from its original block with a longitudinal length of about 4 mm to create a ‘checker’ (B). The cancerous end of the biopsy ‘checker’ was coloured red (C), and the opposing side to the biopsy was coloured blue (D) for orientation. A finished checker (E) and donor block after embedding (F) are pictured.

In the second step, we use the conventional TMA construction techniques described by Kononen et al7 to punch out the needle biopsies from the donor block and reset them in a recipient wax block. A hollow needle punch was pushed into the donor block using an MTA1 Manual Tissue Arrayer (Beecher Instruments, Silver Spring, Maryland), to produce a 1.5 mm diameter core with 4 mm depth (figure 2A) that encompassed a biopsy specimen: the red marker dot (figure 1F) was used in each case to define the position that was punched. Each donor core was then transferred to a 1.5 mm diameter, 4 mm deep hole that had been punched into a recipient wax block (figure 2B). In total, a maximum of 104 cores (8×13) could be arrayed in each recipient block. Usually a number of blank spaces were left in each recipient block defining a unique pattern for block identification. This transfer procedure was usually highly efficient: a detailed examination of the many of the donor blocks following transfer failed to identify residual tissue. The only problems encountered were with very curved biopsies specimens that were not suitable for arraying using this procedure. Preparation and sectioning of biopsy TMA blocks were carried out exactly as described previously.2 An example biopsy TMA block constructed by this method is shown in figure 2C, and the corresponding H&E section is shown in figure 2D. The method was initially used to construct biopsy TMAs in a 13×8 format (104 cores), but this created brittle block edges, and we consider the 91 (13×7) pattern to represent the most robust format. We have now used this procedure to construct 12 TMAs from around 1000 biopsies selected from Active Surveillance and MRC RT01 Radiotherapy trials4–6 including the coring and resetting of all biopsies in low-density TMAs constructed using our original method.2

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Biopsy tissue microarray (TMA) construction. An MTA1 manual tissue arrayer was used to punch out a biopsy core from a donor block (A). The biopsy core was then transferred to a recipient wax block (B). (C) Example of a finished biopsy TMA with (D) corresponding H&E. A black cross denotes the position of blank spaces.