Run 22 (General module) was accomplished January to April 2008. 123 laboratories participated and a total number of 600 slides was assessed. The overall distribution of marks were: optimal 34%, good 34%, borderline 22%, and poor 10%. This is an average distribution but less satisfactory than in run 21, probably due to more challenging markers.

A short summary of the results is given below. Click on the epitope name to see the complete assessments for each marker.

CD3: The proportion of sufficient results has now increased to 84 %, but many labs can improve their results by a more efficeint HIER and higher concentration of the primary Ab. Using a tonsil as control, the dispersed T-cells in the germinal centre must show a strong and distinct staining.

CD15: The proportion of sufficient stains increased slightly to 66 %, which is not satisfactory but in part may be ascribed to many new participants. Among labs changing their protocol according to our recommendations, 73 % improved (while those ignoring the recommendation improved in 15 %). As in previous runs, clone MMA performs better than clone C3D-1. Even though the latter may give optimal results, it seems less robust than MM1. The new clone Carb-3 looks promising but is still used by only few labs.

CDX2: In this first CDX2 assessment the clone CDX2-88 appears to be more sensitive and robust than clone AMT28. However, clone CDX2-88 is an ascites format giving the so-called Mouse ascites Golgi reaction with blood type A tissue. The producer is encouraged change the production method to a supernatant format to reduce this potential source of error. Pancreas should be used for control, not colon or appendix!

CGA: While mAb clones LK2H10 and LK2H10+PHE5 and pAb A0430 (Dako) all give sufficient stains in more than 70 % of the labs, clone DAK-A3 have given no sufficient results among 27 stains! The labs using DAK-A3 are encouraged to change clone, and the producer/vendor (Dako) encouraged to withdraw the product from the market.

MSH2: Even thoug there are several good Abs, only 20 % of the labs got an optimal result. This is mainly due to too low concentration of the Ab and insufficient HIER. LAbs can improve by calibrating their protocols on a tonsil: A strong nuclear staining should be seen not only in the proliferating germinal centre cells but also in the dormant lymphocytes of the mantle zone.

SYP: While mAb clones 27G12 and Snp88 give sufficient results in almost all stains, clone SY38 have in the two latest runs only given a sufficient result in 14 % of the stains. The labs using SY38 are encouraged to change clone, and the producers/vendors (Dako and others) encouraged to withdraw the product from the market. In stains based on clone Snp88 an unwanted cytoplasmic reaction is frequently seen in tissues from patients with blood group A. This is due to Snp88 being an ascites format. The producer is encouraged change the production method to a supernatant format to reduce this potential source of error.

7 April 2008
Mogens Vyberg
Scheme manager

Run 21 (General module) was accomplished September to November 2007. 124 laboratories participated and a total number of 650 slides was assessed. The overall distribution of marks were: optimal 46%, good 31%, borderline 15%, and poor 8%. Thus the proportion of optimal stains is still going in the right direction. About 2% were inappropriate, particularly because several laboratories selected a wrong antibody to detect alpha smooth muscle actin.

A short summary of the results is given below. Click on the epitope name to see the complete assessments for each marker.

CD20: The proportion of sufficient stains is satisfactory: 87%. However, many laboratories could improve by simply increasing the concentration of the primary Ab. HIER is mandatory: citrate buffer pH 6 tended to give a more crisp membrane reaction than alkaline buffers. Tonsil is an appropriate control: The mantle zone B-cells and the germinal centre B-cells must show a strong reaction. No other cells should stain.

CD79a: Clones JCB117 and SP18 may give very good results provided efficient HIER and proper calibration of the primary Ab. On the other hand, laboratories should not stock clone HM57 for human tissue! Appendix and tonsil are appropriate controls: The germinal centre B-cells must show a moderate to strong staining reaction.

CD117: pAb A4502 (Dako) is still the only Ab giving optimal results in a proportion of stains but the number of stains based on other Abs is too low for a conclusion. With A4502, HIER gives better results than no retrieval. The current assessment has revealed a hitherto unrecognized lot-to-lot variation (see assessment for details). Appendix is an appropriate control: The Cajal cells in muscularis propria must show a distinct reaction, while the smooth muscle cells  should be negative.

CD138: Optimal results can be obtained with several clones (MI-15, B-B4, BC/B-B4, B-A38) provided HIER and a proper calibration of the primary Ab. Laboratories using clone 5F7 do not get sufficient results and are encouraged to change clone. Tonsil is an appropriate control: activated germinal centre B-cells, plasma cells and squamous epithelial cells must show a distinct membranous reaction.
 

Figure: Using mAb clone 5F7 for CD138 a laboratory got poor marks due to weak/false negative reactions (A). According to NordiQC advise clone 5F7 was changed. In the reassessment optimal marks were obtained with clone Mi15 (B). 1. Germinal centre, 2. Squamous epithelium, 3. Diffuse large B-cell lymphoma, activated type, 4. Plasmacytoma.

Desmin (DES): In spite of 78% sufficient results more than half of the labs can improve their staining, primarily by increasing the concentration of the primary Ab. HIER is mandatory. Appendix is an appropriate control: not only the muscular layers but also the arteriolar walls must show the strongest possible staining reaction without staining of the enterocytes.

Run B4 (Breast module) was accomplished in parallel with run 21.

HER-2: As in previous runs, the FDA approved systems Herceptest and Pathway performed much better than in-house (home made) systems. However, only 56% of the sufficient stains could be used directly to decide amplification status, and only 43% of the laboratories both had a sufficient staining and an interpretation in concordance with the NordiQC assessors.

Progesterone receptor (PR): Nine different Abs were used, five of them gave optimal results in a fair proportion of cases with clones 1E2, PgR636 and 16 being the most robust. In almost half of the protocols, improvement can be obtained by increasing the Ab concentration and prolong heating time. Uterine cervix is an appropriate control: not only the stromal cells but also the basal squamous cells and the columnar epithelial cells must show a strong nuclear reaction.

30 November 2007 Mogens Vyberg Scheme manager

Fig. Uterine cervix stained for progesterone receptor: Nuclear reaction must be obtained in epithelial cells (left). In insufficient protocols, a strong staining of stromal cells can still be obtained, while the epithelial cells are false negative (right).

Run 20 (General module) was accomplished April to June 2007. About 110 laboratories participated and a total number of more than 600 slides was assessed. The overall distribution of marks were: optimal 31%, good 39%, borderline 24%, and poor 7%. About 4% were inappropriate, particularly because several laboratories selected a wrong antibody to detect low molecular weight cytokeratin (CK-LMW).

A short summary of the results is given below. Click on the epitope name to see the complete assessments for each marker.

CK-LMW: The proportion of sufficient results has been increased from 46% in run 16 to 67% in the current run. NordiQC recommendations for protocol optimization resulted in a marked improvement of the performance. However, detection of CK-LMW still causes difficulties for many laboratories, mainly when the old clones 35BH11 and cam5.2 are used. Laboratories stocking these should consider a change to clone DC10 or C51. For the latter HIER should be applied in an alkaline buffer. Include liver tissue as control: virtually all liver cells must be stained.

CK-Pan: The proportion of sufficient results increased slightly (from 58% to 62%). A major reason for suboptimal stains is the demasking: of 24 laboratories using proteolytic pretreatment for AE1/AE3, only two obtained a good staining and none were optimal. In spite of producer recommendations, HIER is mandatory!

MLA: There are still too many borderline and poor stains, because the protocols are too insensitive. Clone A103 as a Ready-To-Use reagent is particularly insensitive (none of seven stains were sufficient) and cannot be recommended. When specific recommendations given to the participating laboratories are followed, this has a marked impact on the performance.

MSA: 97% of the laboratories obtained a sufficient stain, indicating that clone HMB-45 is very robust.

p53: Only 19% obtained optimal marks and 49% good, mainly due to improper calibration of the antibody dilution. This may be a difficult task, as no normal tissue can be used for control.

S-100: Insufficient results seem in many laboratories to be due to lack of awareness of optimal positive and negative control material giving difficulties in finding the right pre-treatment and dilution of the primary Ab. The Langerhans cells in the skin appear to be good controls.

Run B3 (Breast module) was accomplished in parallel with run 20.

ER: The proportion of sufficient stains is 84%, almost unchanged from two previous runs. The lobular breast carcinoma included in the multitissue block still is a hindrance for the optimal result in many laboratories. A more efficient demasking, careful Ab calibration and choice of a biotin free visualization system are the clues.

HER-2: It was encouraging that all laboratories using FDA approved systems this time adhered to the protocols. On the other hand there are still too many laboratories using poorly calibrated in-house systems. Furthermore, the capability to score the stains correctly lacks very much behind and more training is strongly advised.

1 July 2007

Mogens Vyberg

Scheme manager

Run 19 (General module) was accomplished January to April 2007. About 100 laboratories participated with a total of 530 stained slides (including 35 reassessments). Podoplanin/D2-40 was stocked by only 29 labs.

A short summary of the results is given below. Click on the epitopes to see the complete assessments for each marker.

Overall, 27% of the stains in run 19 were assessed as optimal, 30% good, 32% borderline, and 11% poor. As in all previous runs, major reasons for insufficient results were:

• inefficient HIER

• too dilute primary antibody

• less successful antibody

Particularly staining for TTF1 produced a large proportion of borderline and poor results: 72 out of 74 using clone 8G7G3/1 were insufficient, mainly due to false negative reaction in the carcinoid tumour, which stained with clone SPT24. With the latter clone 60% of the stains were optimal and 16% good. However, using SPT24 the laboratories should be aware that TTF1 is not as specific for lung and thyroid adenocarcinomas as previously believed. It is important for diagnostic purposes that TTF1 is included in a panel.

The overall laboratory performance for CD23 decreased, in part because of the application of ready-to-use antibodies that are not appropriately calibrated by the producers. The Ab concentration in the RTU format must be related to the total sensitivity of the protocol employed, and this sensitivity has to be defined in a clinical setting and validated by the producer as well as the laboratories.

As regards CR, CyD1 and Ki67 the combination of efficient HIER, proper Ab selection and calibration of the Ab dilution is emphasized.

Marked improvements were seen in laboratories adjusting their protocols (fully or partly) according to NordiQC recommendations in a previous run, typically about 70% of the laboratories have obtained sufficient marks. However, too many laboratories continue to use insufficient protocols in spite of specific advice for optimization!

From 2007 NordiQC will contact producers of antibodies that have been less successful in a least two assessments and have been used by at least four participants.

For run 19, the following less successful antibodies are listed:

CD23, clone MHM6 and 1B12 (ready-to-use format)

CR, pAb 760-2700

CyD1, clones DCS6 and P2D11F11

TTF1, clone 8G7G3/1

The producers will be requested to inform the costumers about problems with their antibodies as revealed in the NordiQC assessments and either to publish protocols for an improved performance or consider withdrawal of the antibodies from the market.

NordiQC welcome a new assessor, biotechnologist Ole Nielsen, Odense University Hospital. Ole Nielsen has worked with immunohistochemical protocol optimization for many years and given multiple courses in this field.

01 April 2007
Mogens Vyberg
Scheme manager

Run 18 (General module) was accomplished August to December 2006. About 100 laboratories participated with a total of about 500 stained slides. A short summary of the results is given below. Click on the epitope names to see the complete assessments for each marker.

As in run 15, staining of immunoglobulin epitopes caused most troubles. For both IgK and IgM, the insufficient results were prevailing. The laboratories getting suboptimal marks are encouraged to read the assessment description and recommendations carefully. A sufficient staining can be obtained, when simple steps are followed: heat induced epitope retrieval, best Ab selection, and careful calibration of the Ab dilution.

The laboratories' capability of detecting chromogranin A (CGA) improves: The proportion of sufficient results have increased from 39 % in run 9 (2003) to 70 % in the present run! During the two previous runs, specific recommendations have been given to a total of 68 laboratories with insufficient staining results. In 46 of the laboratories, the advice of changes in the protocols have been followed, subsequently giving sufficient staining in 42 (91 %). 22 of the laboratories that apparently ignored the advices, obtained sufficient marks in the subsequent run in 5 cases (23%) .

Neurofilament protein (NFP) was included for the first time. Remarkably, 52 out of 56 laboratories chose the mAb clone 2F11, even though this clone only detects phosphorylated filaments that rarely are represented in neuroendocrine and low differentiated neuronal tumours. The laboratories are encouraged to read the assessment description and recommendations as regards selection of the right clones for staining tumours. In the current run, the purpose of NFP was not explicit, even though the inclusion of tumours in the slides was a hint. mAb clone 2F11 is a good antibody for detecting well differentiated neurons but will be considered inappropriate for tumour diagnostics in future assessments.

In order to detect synaptophysin (SYP) the laboratories applied twelve different Abs. Only five of them gave optimal results in this assessment. As for six others, they were used by few participants and no firm conclusion can be drawn, until the Abs have been compared in optimized protocols. In contrast, mAb clone SY38, one of the oldest on the market, were used by eleven laboratories, none of whom obtained an optimal staining result. This is in accordance with studies carried out in Aalborg showing clone SY38 to be less sensitive than three other Abs (clones 27G12 and Snp88 and pAb A0010). The  study will soon be published (Vyberg M, Ulhøi BP, Teglbjærg PS: Neuronal features of Oligodendrogliomas. Histopathology, in press).

Finally in the general module, terminal deoxynucleotidyl transferase (TdT) was included for the first time. 93% obtained optimal or good marks indicating that TdT staining is a robust method.

Run B2 (Breast module) had 81 participants. Staining for progesterone gave sufficient results in 75%, the insufficient stains were typically due to insufficient heat induced epitope retrieval and too dilute Ab.

HER-2 immunohistochemistry assessment disclosed unexpected reaction patterns in some cell cultures and one of the carcinomas. Even though the results should be interpreted with care, there is a clear tendency to improvement. However, the home made systems still lacks behind HercepTest.

HER-2 FISH/CISH revealed a high concordance as regards amplification versus non-amplification. However, analyses of slides and protocols were cumbersome and did not give sufficient information for advising the laboratories.

Reassessments

54 slides were submitted for assessment of new stains where the primary stains had been assessed as insufficient in run 17. Grouped together, 15 of the new slides were marked as optimal (28%), 25 as good(46%), 8 as borderline (15 %), and 6 as poor (11 %).

Thus, 74 % improved their marks to either good or optimal. However the improvement rates were depending on the epitopes detected. The highest improvement rate (89%) showed calretinin and CD5, whereas none of three improved their Melan A stain.

1 and 7 December 2006
Mogens Vyberg
Scheme manager

Frontpage figure in December (click on photo for magnification): In run 17, 14 labs. stained for CD5 using clone CD5/54/F6, none obtained an optimal result. An insufficient result from one lab.: a) The T-cells in the tonsil are moderately stained, but the subpopulation of B-cells in the mantle zone expected to be positive are virtually unstained. b) The mantle cell lymphoma shows a weak and indistinct reaction. The lab. requested reassessment in run 18 after changing to clone 4C7 and obtained optimal marks: c) The B-cell subpopulation in the tonsil is now distinctly stained, though the reaction is weak compared to the strongly stained T-cells. d) The mantle cell lymphoma is moderately-strongly positive.

Run 17 & B1 was accomplished April to June 2006. 100 laboratories participated with a total of 633 stained slides (including reassessments), the highest number of slides since NordiQC was established. Click on ALK, bcl-6, Calretinin, CD5, Cyclin D1, Ep-CAM, ER, and HER-2 to see the general results for each of these markers.

The overall marks were as follows: optimal 35%, good 31%, borderline 17%, and poor 17%.

For HER-2, Ep-CAM, Calretinin, Cyclin D1, almost half of the stains were deemed insufficient (borderline or poor; 49%, 46%, 44% and 41%, respectively). For CD5 and ER 34% and 25%, respectively, were insufficient, while for bcl-6 and ALK, 13% and 7%, respectively, were insufficient.

As in previous runs, it is fairly apparent from multiple protocol analyses what is needed to obtain optimal stains, particularly the selection of a good antibody and a protocol with efficient epitope retrieval. At the same time, the causes of insufficient results are revealed in almost all cases.

It is important that the participants study the general results of the assessments carefully to be able to benefit fully from the suggestions for improvement, which are given with the individual results in case of insufficient stains. Inspiration can also be found in the recommended protocols.

For some of the markers previously included in the assessments, the proportion of insufficient stains unfortunately increased! There are several reasons for this: 1) New participants for the first time staining for the markers included, 2) The multitissue blocks now include tumours that are more demanding or challenging, e.g. because of fewer epitopes or higher content of endogenous biotin. In this context it is satisfying to see that laboratories changing their protocols according to NordiQC recommendations after previous runs improve their staining results considerably. However, some participants apparently ignore the recommendations given and continue to use less successful antibodies and inappropriate retrieval techniques. Most of them continue to get insufficient results.

NordiQC has until now only has given suggestions to the participants, e.g. as regards choice of antibody and retrieval method. However, based on the rather firm and reproducible results, NordiQC will now also contact producers and vendors to have less successful (in fact poor performing) antibodies withdrawn from the market and misleading protocol recommendations changed.

The most worrisome result in this run was the low score for HER-2. Just like in the pilot run in 2005, optimal results could only be obtained with a FDA approved kit (in run B1 only Herceptest), following the guidelines given by the system manufactures.

Laboratories using more or less home-made HER-2 protocols should change their system immediately.

Run 16 was accomplished January to April 2006. 102 laboratories participated with a total of 567 stained slides (including reassessments) -  the highest number of laboratories and slides since NordiQC was established. Click on AMACR/P504S, CD10, CK-LMW, CK-HMW, Melan A and p63  to see the general results for each of the markers.

The overall marks were as follows: optimal 39%, good 27%, borderline 17%, and poor 13%, while in 4% of the stains inappropriate antibodies were used. The proportion of insufficient (borderline or poor) stains was highest for Melan A (68%), mainly because of difficulties in the demonstration of Melan A in the desmoplastic malignant melanoma when using less sensitive protocols. Obviously the calibration of protocols in many laboratories have been carried based out on tissues with a relatively high expression of epitopes.  Protocols giving optimal results are available at the website, the participants are encouraged to study the photos and the recommended protocols.

For CK-LMW, there were also many insufficient results (54%). Here the main reason appeared to be the use of less successful antibodies.  Most likely, these laboratories could improve their results simply by using either clone DC10 or C51 with HIER. Out of 7 labs. that changed their CK-LMW protocols after Run 9 according to the NordiQC advice, 5 improved their marks, while improvement was not seen for any of 9 labs. that did not follow the advice.

Run 15 was accomplished September to November 2005. 92 laboratories participated with a total of 407 stained slides. Click on CA125, CD45, IgK, IgL, pan-CK and WT1  to see the general results for each of the markers.

The overall marks were as follows: optimal 24%, good 27%, borderline 22%, and poor 25%, while in 2% of the cases inappropriate antibodies were used. The proportion of insufficient (borderline or poor) stains was rather high, particularly due to unsatisfactory results from immunoglobulin light chain staining. However, there were also several protocols giving optimal results, the participants are encouraged to study the recommended protocols.

The most frequent causes of insufficient stains were generally (as in previous runs, and often in combination):

• Insufficient calibration of the primary antibody concentration (appearing to a be cardinal step particularly for the light chains)
• Missing, insufficient, inappropriate or excessive epitope retrieval (e.g., laboratories staining for pan-cytokeratin following the package inserts and instructions provided by the producers obtained insufficient results in part due to erroneous recommendations for epitope retrieval).
• Less successful antibodies

In the HER-2 pilot run 68 laboratories submitted stains. At the assessment 28 achieved optimal marks (41 %), 22 good (32 %), 8 borderline (12 %) and 10 (15 %) poor marks. An optimal staining result in combination with a correct interpretation was seen in only 17 out of 57 laboratories (30%). The preliminary results indicate that the FDA approved immunohistochemical HER-2 systems Herceptest and Pathway seems to be more reliable than in-house methods, and that training in the interpretation of HER-2 stains is needed for a large number of the laboratories.

NordiQC has decided to establish a separate breast module from 2006, opening 1 April with 2 runs each catering for four markers including HER-2, ER, PgR, and another relevant breast marker to be defined for each run.

Run 14 was accomplished April to June 2005. 91 laboratories participated with a total of 457 stained slides.

Click on CD3, CD4, CD8, CD15, CD117 and PLAP  to see the general results for each of the markers.

The overall marks were as follows: optimal 32%, good 43%, borderline 15%, and poor 10%. The proportion of insufficient (borderline or poor) stains is approximately the same as in previous run, but cannot be compared directly due to new markers and new laboratories.

Most interesting is it to see the development for CD15 and CD117 staining in laboratories that have stained for these markers the second time:

33 laboratories with an insufficient CD15 staining in run 10 submitted a new CD15 stain in run 14. 17 out of the 33 laboratories followed the recommendations, and 12 of these (71 %) improved their score from insufficient to either good or optimal. 16 laboratories did not follow the recommendations and only 2 (12 %) improved.

19 laboratories with an insufficient CD117 staining in run 7 submitted a new CD117 stain in run 14. 16 out of these followed the recommendations and 15 of these (94 %) improved from insufficient to either good or optimal. 3 laboratories did not follow the recommendations and only one improved.

The most frequent causes of insufficient stains were generally (as in previous runs, and often in combination):
- Too low (or - rarely - too high) concentration of the primary antibody
- Missing or insufficient (or - rarely - excessive) epitope retrieval
- Inappropriate choice of primary Ab.

With CD4, clone 1F6, an inappropriate blocking of endogenous peroxidase was a likely cause of insufficient result. However, this will be analyzed in a new run in 2006.

By appointment with UK-NEQAS, NordiQC plan to take over quality assurance of HER2 staining from 2006.

A test run free of charge is included with Run 15 opening 01 September.

 

- ■ Anvendt Immunhistokemi ("Applied Immunohistochemistry" - in Danish only): Update and ordering: Click on Epitopes

April 2005

Run 13 was accomplished January to April 2005. Click on Bcl-2, CGA, ER, GFAP, MLH1, and MSH2 to see the general results for each of the markers.

92 laboratories participated with a total of 368 stains. Marks for the total material were as follows:

Optimal: 36%, good: 38%, borderline: 19%, and poor: 6%.

The largest proportion of sufficient (i.e., optimal or good) stains was achieved with bcl-2 (93%), the smallest with MSH2 (43%).

The main causes for insufficient staining were generally the same as in all previous runs:

● too diluted primary antibody (in few cases too concentrated antibody)

● insufficient, missing or inappropriate epitope retrieval

● inappropriate primary antibody.

The most interesting - and inspiring - observation was the impact of recommendations previously given for improving the protocols for estrogen receptor alpha (ER), which was included for the third time, and chromogranin A (CGA), which was included for the second time.

ER is described on the front page. As regards CGA, the proportion of insufficient staining was reduced from to 61 % in run 9 to 36 % in run 13.  43 laboratories, which obtained an insufficient result in run 9 submitted a new CGA-stain in run 13. 28 of the 43 laboratories followed the recommendations given, and 25 of these (89%) improved the score from insufficient to either good or optimal, while 3 were still insufficient. 15 laboratories did not follow the recommendations. 3 of these (20%) obtained a good but not optimal result in run 13, whereas 12 still had an insufficient staining.

Please read the assessments of Run 13 carefully and compare the general descriptions with your individual assessment score, as well as comments and recommendations accompanying insufficient results.

In case of good marks, comments are rarely given. However, you may contact NordiQC for an explanation, if you wish us to give suggestions for optimization.

The origin of uploaded protocols giving optimal staining results are published, encouraging technicians and pathologists to communicate directly when needed. If a participating laboratory wish to remain anonymous, NordiQC should be informed by e-mail.

If you haven't got your individual score per e-mail or are in doubt, do not hesitate to send us an e-mail.

December 2004

Run 12 was accomplished September to December 2004. Click on CD99, CEA, CK5, PAP, PSA, and vimentin to see the general results for each of the markers.

79 laboratories participated with a total of 372 stainings. Marks for the total material of stainings were as follows:

Optimal: 53%, good: 29%, borderline: 13%, and poor: 5%.

The largest proportion of sufficient (i.e., optimal or good) stains was achieved with vimentin (94%), the smallest with CD99 (44%).

However, 15 CEA stains were excluded in the assessment because of the use of antibodies that we considered  inappropriate due to cross reaction with CEA-like proteins. Consequently we could not assess the true reactivity to CEA in these cases.

The main causes for insufficient staining were generally the same as in all previous runs - often in combination:

June 2004

78 laboratories participated with a total of 327 stainings. Marks for the total material of stainings were as follows: optimal: 37%, good: 32%, borderline: 17%, and poor: 14%. The general quality appears slightly improved compared to previous runs. The largest proportion of sufficient (i.e., optimal or good) stainings were achieved with CD30 (92%), the smallest with factor VIII (42%).

The main causes for insufficient stainings are generally the same as in all previous runs - often in combination:

▪ Insufficient epitope retrieval, typically when citrate pH 6 is used as buffer instead of Tris/EDTA pH 9

▪ Inappropriate epitope retrieval, e.g., use of proteolytic pre-treatment for CD30, CD31 and CD68, which should all be retrieved with HIER

▪ Too dilute antibodies (for hCG also too concentrated antibodies)

▪ Inappropriate antibodies.

Via e-mail (about 22 June 2004) all participants are informed about their individual scores. Particularly in case of insufficient stainings, explanation of the probable causes and suggestions for improvement are given.

If you have not received your score, please contact us at nordiqc@nja.dk

Run 12 opens 1 September: The markers are CD99, carcinoembryonic antigen (CEA), cytokeratin (CK) 5 (5/6), prostate specific antigen (PSA), and vimentin. Furthermore, prostate specific acid phosphatase is included as an optional marker.

A new protocol form has been constructed, it will be available when run 12 opens. It is password protected. Password will be sent to all participants in the middle of August:

April 2004

In Run 10 accomplished January to April 2004, 80 laboratories participated with a total of 376 stainings.

Marks for the total material of stainings were as follows: optimal: 34%, good: 28%, borderline: 22%, and poor: 17%. These percentages are almost identical with those of Run 9 (in spite of different antibodies and tissues). Noteworthy, the largest proportion of optimal stainings were achieved with progesterone receptor, the smallest with estrogen receptor. CD15 revealed the largest proportion of poor stainings.

Run 10 comprised Alpha smooth muscle actin, CD15, Epithelial membrane antigen, Estrogen receptor alpha and Progesterone receptor. Click on one of these to see the general results.

The main causes for insufficient stainings are the same as in all previous runs: Insufficient epitope retrieval, too dilute antibodies and inappropriate antibodies. Following the advices from NordiQC, it is possible to improve the staining quality. For example, those laboratories that changed their ER-protocol from Run 8 to Run 10 according to the advices improved their score in 10 out of 13 cases, while laboratories that did not follow the advices improved in only 3 out of 12 cases.

Via e-mail (1 April 2004) all participants are informed about their individual scores. Particularly in case of insufficient stainings, explanation of the probably causes and suggestions for improvement are given.

If you have not received your score, please contact us at nordiqc@nja.dk

Previously we announced that HER/neu would be included as an extra, optional marker in Run 12. However, UK-NEQAS has this year established four HER2/neu runs, in which all Danish laboratories as well as a number of other Nordic laboratories participate. Therefore, NordiQC postpone the inclusion of this marker.

Scheme 2004

Run 10 (1 January): ASMA, CD15, EMA; ER, PR

Run 11 (1 April): CD14/CD68, CD30, CD31, FVIIIrag, and HCG.

Run 12 (1 September): CD99, CK5/6, PSA, PSAP, and vimentin.

Photo: Assessing Run 10 stainings, from the left: Søren Nielsen, Heikki Helin, Mogens Vyberg, Tomas Seidal and Bjørn Risberg.

December 2003

In Run 9 accomplished September to December 2003, 72 laboratories participated with a total of 304 stainings comprising CD23, Cyclin D1, Chromogranin A, LMW Cytokeratin and Thyroid transcription factor-1. Click on one of these to see the results.

At the assessment, marks for the total material of stainings were as follows: optimal: 31%, good: 24%, borderline: 27%, and poor: 19%.

Compared to Run 8, the proportion of optimal staining was the same, while the proportions of borderline and poor stainings were considerably larger, because many laboratories have had difficulties in producing optimal stainings: While the CD23 stainings were marked optimal or good in 76%, the Chromogranin A stainings were only marked optimal or good in 38% and Cyclin D1 stainings in 48%. As in all previous runs, more than 2/3 of the protocols may be improved. In the present run, protocol changes appear to be urgently needed in about 40%.

It was possible to identify the probable causes of insufficient stainings in the large majority of cases allowing individual advices for improvement.

The scores are based on the NordiQC slides only. We request the control stainings from the laboratories for comparison in case of an insufficient staining of a NordiQC slide merely to improve the basis for advices.

Via e-mail (1 December) all participants are informed about their individual score. Particularly in case of insufficient stainings, explanation of the probably cause and suggestions for improvement are given. If you have not received your score, please contact us at nordiqc@nja.dk

Previous member of the core group Emina Torlakovic left Norway for a position outside the Nordic countries. As a new member per 1 January 2004 we welcome Bjørn Risberg.

Scheme 2004:

Run 10 opens for enrolment 1 January (deadline for protocol submission 15 January), comprising Alpha smooth muscle actin (ASMA), CD15, Epithelial membrane antigen (EMA), Estrogen receptor (ER), Progesteron receptor (PR).

Run 11 (1 April): CD14/CD68, CD30, CD31, FVIIIrag, and HCG.

Run 12 (1 September): CD99, CK5/6, PSA, PSAP, and vimentin. Supplementary: HER2/neu.

Photo: The core group assessing in Run 9.

July 2003

In Run 8 accomplished April to June 2003, 74 laboratories participated with a total of 350 stainings comprising CD5, pan cytokeratin (CK), CK7, CK20, and Estrogen receptor (alpha).

At the assessment, marks for the total material of stainings were as follows (see diagram): optimal: 32%, good: 36%, borderline: 21%, and poor: 11%.

This means that more than 2/3 of the protocols may be improved and that protocol changes are urgently needed in about 1/3. There were, however, marked variations between quality of the various stainings.

E.g., 90% of the CK20 stainings were marked optimal or good, while only 45% of the ER-stainings got these marks.

It should be emphasised that the scores are based on NordiQC slides only. We request the control stainings from the laboratories for comparison in case of insufficient staining of a NordiQC slide, to improve the basis for advices.

Via e-mail (4 July) all participants are informed about their individual score. In case of insufficient stainings, explanation of the probably cause and suggestions for improvement are given.

There were five laboratories, which after a borderline or poor result of one to three stainings in Run 7 requested slides for restaining after protocol improvement. Their new stainings were re-assessed, and all but one were marked optimal or good.

Run 9 opens for enrolment 1 September (deadline 15 September), comprising Cyclin D1, CD23, Chromogranin A, CK8/LMW-CK, TTF-1 (see details in Participation).

During 2004 we include new markers as well as some of those previously stained for (marked with *) that gave the worst results in previous runs. The following are planned:

Run 10: ER*, PR*, Her-2-neu, EMA, CD15*.

Run 11: CD31, FVIIIrag, CD30*, HCG, CD14/CD68*.

Run 12: PSA, PSAP, CD99, CK5/6, vimentin*.

April 2003

To see the detailed results of Run 7, click on CD34, CD117, MSA-HMB45, Melan-A and S-100 protein. To see the complete list of assessments from all runs, click on Assessments. The individual scores of Run 7 will be e-mailed to the participants about 15 April. At the same time, examples of good protocols from Run 7 will be available, click on the epitopes above or on Protocols.

Run 8, comprising the following epitopes: CD5, Cytokeratin (pan-CK), CK7, CK20, and Estrogen receptor (alpha) is now open for registred participants. Attend Run 8 not later than 15 April by filling out the protocol forms, click on Participation

Run 9, comprising Cyclin D1, CD23, Chromogranin A, CK8, and TTF-1, opens 1 September.

NordiQC invites all Nordic laboratories dealing with immunohistochemistry to take part. An Invitation letter were sent to the scientific pathology organizations and mailed to all participants from previous runs about 1 January. For the fiscal year 2003 the charge is DKK 5.500 (~ € 740) covering all expenses from participation in the NordiQC quality scheme including the assessment of 15 stainings during 2003 (5 epitopes in each of three runs).

For Nordic laboratories, which did not reach to participate in Run 7, we offer participation in Run 8 and 9 for a charge of DKK 4.000 (~ € 550).

December 2002

In Run 6 (the second open NordiQC assessment scheme) accomplished autumn 2002, 65 Nordic laboratories participated. A total of 252 stainings were submitted and assessed. The marks were as follows:

optimal: 38%, acceptable: 37%, borderline: 17%, poor: 7%.

The results show that almost 2/3 of the protocols may still be improved and that protocol changes is urgently needed in about 1/4. The results are very close to those of run 5 (spring 2002). However, a direct comparison is not possible, as many new laboratories have attended.

To see the detailed results of Run 6, click on CD10, CD20, CD79a, Epithelial antigen Ber-EP4 and Calretinin. To see the complete list of assessments from all runs, click on Assessments. The individual scores of Run 6 will be e-mailed to the participants about 15 December. Examples of good protocols from Run 6 are now available, click on the epitopes above or on Protocols

The fact that more than half of all Nordic immunohistochemical laboratories have participated in the last run indicates a great interest in NordiQC way of quality work. The core group has now decided to establish NordiQC as a permanent organization. To finance the work, however, a charge from the participating laboratories is necessary.

For 2003 the charge is DKK 5.500 (~ € 740) covering all expenses from participation in the NordiQC quality scheme including the assessment of 15 stainings during 2003 (5 epitopes in each of three runs).

Run 7, January 2003, comprises the following epitopes: CD34, CD117, S-100 protein (beta), melanosome antigen (HMB-45) and Melan-A (MART-1).

Run 8, April 2003, comprises the following epitopes: CD5, Cytokeratin (pan-), CK7, CK20, and Estrogen receptor (alpha).

Run 9, September 2003, comprises the following epitopes: Cyclin D1, CD23, Chromogranin A, CK8, and TTF-1.

NordiQC invites all Nordic laboratories to take part. An Invitation letter (pdf-format: Adobe Acrobat Reader needed) will be sent to the scientific pathology organizations and mailed to all participants from previous runs about 1 January. You can attend the next run now by clicking on Participation.

August 2002

The first open NordiQC assessment scheme, Run 5, accomplished spring 2002, was a great success with 47 laboratories participating. A total of 168 stainings were assessed. Of these, 59 (35%) were marked optimal, 67 (40%) acceptable, 33 (20%) borderline, and 9 (5%) poor by the assessors. The individual scores have been e-mailed to all participants (if you have not received it, please contact NordiQC).

Click on Assessments to see the results of the CD3, Desmin, Ki-67 and Synaptophysin stainings.

NordiQC now opens Run 6, again inviting all Nordic laboratories to take part. We think, and hope, that many will take the opportunity to introduce this quality assurance, which can contribute to better performance in the field of immunohistochemistry. An Invitation letter has been sent to the scientific pathology organizations and will be mailed to all previous participants about September 1st.

The following five stainings will be assessed in Run 6: CD10, CD20, CD79a, Calretinin, and Epithelial antigen Ber EP-4. Click on Participation to get the instructions and protocol form. You can participate in one or more assessments without any obligations. There will be no charge during 2002.

May 2002

Run 5, the first open NordiQC Assessment is now completed. 45 Nordic laboratories submitted one or more stainings (Unfortunately, the Finnish laboratories were not informed in proper time, and only one Finnish laboratory participated).

A total of 168 stainings have been assessed. Of these, 59 (35%) were marked optimal, 67 (40%) acceptable, 33 (20%) borderline, and 9 (5%) poor by the assessors.

The major results are presented at the website. All laboratories are anonymous with the exception of those that provided examples of good protocols giving optimal staining results. In the good protocols the name of a contact person and her or his e-mail address is included. This is done to encourage communication directly between the laboratories. Of cause, further information can also be given by the core group.

Click on Assessments to see the first results of Run 5 comprising CD3, Desmin, Ki-67 and Synaptophysin. In the following weeks, we will continue the analyses of stainings and protocols to add further details. Any comments to the methods and results are most welcome.

In a few weeks, all participants will receive an e-mail with the assessment of their individual staining results.

Run 6 is planned to start in August 2002. All participants and the scientific societies will be notified.