• A strong, distinct cytoplasmic and nuclear staining of the peripheral nerves (ganglion cells and axons) and macrophages in the appendix.
• A strong, distinct cytoplasmic and nuclear staining of the majority of the neoplastic cells of the epithelioid mesothelioma, whereas at least focally the neoplastic cells of the biphasic mesothelioma should be demonstrated.
• A moderate, distinct cytoplasmic and nuclear staining of the luteinized cells in the granulosa cell tumour and a strong reaction in the stromal component.
• A negative or only a focal staining of the tubular epithelium in the kidney.

87 laboratories participated in the assessment. At the assessment 16 achieved optimal marks (18 %), 33 good (38 %), 28 borderline (32 %) and 10 (12 %) poor marks.

The following Abs were used:
mAb clone DAKCalret1 (Dako, n=38)
mAb clone 5A5 (Novocastra, n=17)
mAb clone 2E7 (ImmunVision, n=1)
mAb clone CAL6 (Novocastra, n=1)
rmAb clone SP13 (NeoMarkers, n=2)
pAb 18-0211 (Zymed, n=16)
pAb 760-2700 (Ventana, n=7)
pAb 7699/4 (Swant, n=3)
Unknown Ab (n=1)

Optimal staining for CR in this assessment was obtained with the mAb clones DAKCalret1 (8/38) and 5A5 (3/17), and the pAbs 18-0211 (4/16) and 7699/4 (1/3).
All optimal protocols, irrespective of the Ab, were based on heat induced epitope retrieval (HIER) using Tris-EDTA/EGTA pH 9.
DAKCalret1 was used in the range of 1:25 – 1:500 depending on the total sensitivity of the protocol employed, or as a Ready-To-Use (RTU) Ab. Using these protocol settings 25 out of 34 laboratories (74%) produced a sufficient staining (optimal or good).
5A5 was used in the dilution 1:40. Using these protocol setting 4 out of 6 laboratories (67%) obtained a sufficient staining.
18-0211 was used in the range of 1:100- 1:1,500 depending on the total sensitivity of the protocol employed. Using these protocol settings 5 out of 6 (83%) laboratories produced a sufficient staining.
7699/4 was used in the dilution 1:2,000. 1 out of 3 laboratories (33%) using this pAb obtained a sufficient staining marked as optimal.

Comparing run 17 and run 19, the following pass rates are seen with Abs used by at least 3 participants:

When only protocols with HIER in an alkaline buffer as Tris-EDTA/EGTA pH 9 or equivalent as CC1, Ventana, BERS-2 Vision BioSystem and TRS S2367, Dako are included, the following pass rates are seen:

From the tables the overall pass rate increases compared to the general result obtained in run 17 and 19. Of particular interest, the pass rate is markedly higher using the mAb clone DakCalret1, 5A5 and the pAb 18-0211 with HIER in an alkaline buffer: In total 88 out of 123 protocols (72 %) were assessed as sufficient.

The most frequent causes of an insufficient staining were:
- Less successful primary Ab
- Insufficient HIER (typically with Citrate pH 6 as the heating buffer)
- Too low concentration of the primary Ab.

In this assessment (and in concordance with the CR assessment in run 17) the prevalent feature of an insufficient staining was a too weak or false negative staining of the granulosa cell tumour. The stromal component in the tumour was almost always positive, but only in the correctly calibrated protocols the neoplastic luteinized granulosa cells were demonstrated. The two mesotheliomas showed different levels of CR expression, as the epithelioid tumour was strongly positive - CR being demonstrated by almost all laboratories - whereas the biphasic mesothelioma expressing limited amounts of CR - only demonstrated by the laboratories with correctly calibrated protocols.

As observed in run 17 appendix can be used as control for CR. However to serve as a reliable quality indicator for the immunohistochemical demonstration of CR, the axons and ganglion cells in the Aurbach’s plexus in the appendix and the fat cells must show an intense reaction, while smooth muscle cells and epithelial cells shall remain negative.

In the CR assessment run 17 (2006) 82 laboratories participated out of which 44% (36 laboratories) obtained an insufficient mark. Each of these laboratories was given specific recommendation to improve their protocol. 29 laboratories, which obtained an insufficient result in run 17, submitted a new CR stain in run 19. 18 of these followed the recommendation, of which 12 improved to good or optimal (67 %). 11 laboratories did not follow the recommendation, and only 1 of these (9 %) obtained a sufficient staining in run 19.

However, the proportion of insufficient results has not been reduced from run 17 as 44% of the results still are insufficient. The main reasons seem to be a high proportion of laboratories using less successful antibodies and difficulties in calibrating and setting the Ab concentration of otherwise successful antibodies. This emphasizes the importance of identifying a robust control for CR. Currently appendix is the preferred choice.

Conclusion
The mAb clones DAKCalret1, 5A5 and the pAb 18-0211 are recommendable Abs for CR. HIER in an alkaline buffer is highly recommended for optimal performance with all 3 Abs. Appendix can be used as positive control: The nerves must be as strongly stained as possible, while no staining of the epithelial and smooth muscle cells shall be seen.