Home  ■  Participation  ■  Assessments  ■  Epitopes  ■  Protocols  ■  Techniques  ■  Links

Tissue handling ■  Section handling  ■  Epitope retrieval  ■  Endogenous biotin
Primary antibody  ■  Visualization  ■  Chromogene  ■  Controls  ■  Recommended control tissue

Controls

What kind of controls are needed for clinical immunohistrochemistry ?

Two types of controls need to accompany patient’s samples: positive and negative controls. Generally, one positive control for each antibody used and one negative control for each type of tissue pre-treatment should be included. Negative controls are automatically achieved using an appropriate antibody panel.

 

 

What kind of tissue can I use for positive controls ?

The tissues for positive controls are previously characterized tissues with known levels of expression of a certain antigen. Benign tissues with high antigen expression are frequently used for positive controls. However, such tissues are not sufficient  for the clinical laboratory. Considering that the antigen content of many tumours will vary with the degree of tumour differentiation, false-negative results are frequent if we use controls with abundant antigen. The biological Stain commission (University of Rochester Medical Center, Rochester, NY) and several authors recommend suing control tissue with low levels of target antigen to ensure that the stain system has the sensitivity to label (stain) low antigen levels and reduce the risk of false-negative results (1-4).

Multi-tissue control blocks for immunohistochemistry are ideal for clinical laboratories. Several techniques have been used since Dr. Hector Battifora introduced such blocks in 1986 (5-10). The technique which uses skin punch biopsy needles is illustrated in Figure 1.

Generally, the tissues may be collected prospectively or retrospectively. It is easier to start with retrospectively collected tissues. However, prospective collecting of tissues will soon pay off. Record or collect tissues/biopsies which are large enough to be used for positive controls and have weak expression of antigens because these are more difficult to find.

The tissues for the controls have to be fixed the same way as tissues to be tested. We also routinely make multi-tissue blocks with B5-fixed tissues to be used for B5-fixed lymph node and bone marrow biopsies.

 

 

How many tissue samples I need to put in the multi-tissue block ?

It varies depending on the purpose of the block. Our laboratory uses several different multi-tissue blocks. One of them is an “epithelial/mesothelial control” with about 10 tissue fragments from various carcinomas and mesotheliomas (shown in Figure 1). The purpose of such multi-tissue blocks is to be used for all kinds of epithelial and mesothelial markers (AE1/AE3, Cam5.2, CK8, CK19, CK20, CK7, CK5/6, BerEP4, MOC31, calretinin, GCDFP-15, and many others). Another is a “melanoma” multi-tissue control which consists of 4 melanomas and one nevus. We use it for vimentin, S-100 protein, melanosoma specific antigen (HMB-45) and melan-A. There is no standard design of positive controls yet. If one keeps certain basic rules about how to make positive controls – the final design will be determined by a practicing pathologist who will prepare such positive controls to fit his/her practice needs. For instance, if you have busy breast pathology practice, you may want to make a multi-tissue block that will be used for estrogen and progesterone receptor analysis to include tumours that show strong, intermediate and weak expression of each antigen.

 

 

Is it always necessary to make multi-tissue blocks for positive controls ?

No, it is not. One can use just one kind of tissue/organ with known antigen expression. The most practical positive control is a normal appendix. For example, it is a perfect control for CD23 because follicular dendritic cells in the germinal centres show strong expression of CD23 and there should also be numerous cells in the mantle zone of the follicle which show very weak expression. It also controls for endogenous biotin because there are numerous epithelial cells in the same tissue. Besides having abundant expression of numerous antigens (actin, smooth muscle actin, CD45, CD20, CD3, CD4, CD8, CD5, CD21, CD23, CD35, AE1/AE3, Cam5.2, CK8, CK20, CD34, and many others), a normal appendix is a good positive control because it also has tissue with very low expression of many antigens. It is an ideal control for bcl-2 (positive in lymphoid tissue, negative in benign germinal centres, and gradient expression in crypts), CD10 (germinal centres, but also weak expression in the surface basement membrane), CD23 (follicular dendritic cells and mantle zone), CD45 (lymphocytes, monocytes, macrophages, and finally very weak expression in plasma cells in the lamina propria), and some others. Also, a cross section of the appendix is small which will leave plenty of room for patient’s sample. Use only one case of classical Hodgkin’s disease as your CD30/CD15 positive control if you are certain that malignant cells only weakly express CD30 and CD15. One section of the kidney is an excellent control for pan-cytokeratin. The expression varies in different parts of the tubules and Bowman’s capsule.

  

How do I find tissue with weak expression of certain antigens ?

It may be difficult to be certain that neoplastic tissues characterized in your laboratory are true “weak positives”. It may be a good idea to borrow few unstained sections from the positive controls from already established laboratory to make certain that your choices are correct.

 

 

What kind of tissue can I use for negative controls ?

Negative controls are prepared from patient’s sample/biopsy. Additional sections from the paraffin block that is used for specific staining are used for that purpose.

 

 

Is it really necessary to cut positive controls on each tested slide ?

Yes, it is. One finds out that this is true only when such practice is instituted in the laboratory. Occasionally, we find that the patient’s sample was negative, but so was the positive control on that slide. Sometimes it is not clear what happened because all other samples tested with the same antibody are positive. Most of the time certain antibody shows uniform results on all slides. However, some variability from slide to slide can be found. One would never be able to note such variability in the performance of the test if all slides do not have their own positive controls.

 

 

Selected references

1. Taylor CR. Immunomicroscopy: A diagnostic tool for the surgical pathologist. 2nd ed. Philadelphia, pa: Saunders;1994.

2. Elias JM. Immunohistopathology: A practical approach to diagnosis. Chicago, Ill: ASCP Press: 1990.

3. Taylor CA, Jones CM, Tandon A, et al. Quality control for immunohistochemistry. BioLink. 1992;1:1.

4. Elias TM, Gown AM, Nakamura RM, et al. Quality control in immunohistochemistry: report of a workshop sponsored by the Biological Stain Commission. Am J Clin Pathol. 1989;92:836.

5. Battifora H (1986). The multitumor (sausage) tissue block: Novel method for immunohistochemical antibody testing. Lab Invest 55:244- 248.

6. Battifora H and Mehta P (1990). The checkerboard tissue block: An improved multitissue control block. Lab Invest 63:722-724.

7. Enghardt HM, Aghassi BN, Bond JC, and Elson MD (1995). A simplified multitissue control block. J Histotechnol 18:51-55.

8. Miller RT (1 993). Multitumor "sandwich" blocks in immunohistochemistry: Simplified method of preparation and practical uses. Appl Immunohistochem 1:156-159.

9. Press MF, Bernstein L, Thomas PA, Meisner LF, Zhou JY, Ma Y, Hung G, Robinson RA, Harris C, ElNagger A, Slamon DJ, Peyrot M, Ross J, Phillips R, Wolman SR, and Flom KJ (1997). HER2/neu gene amplification by fluorescence in situ hybridization: Evaluation of archival specimens and utility as a marker of poor prognosis in node negative invasive breast carcinomas. J Clin Oncol 15:2894-2904.

10. Press MF, Hung G, Godolphin W, and Slamon DJ (1994). Sensitivity of HER2/neu antibodies in archival tissue samples: Potential source of error in immunohistochemical studies of oncogene expression. Cancer Res 54:2771-2777.

 

To the top

 

ET/AS

Last update 21-09-2005