The North East Haematopathology Diagnostic Service, based at the Royal Victoria Infirmary in Newcastle upon Tyne, provides a regional service for expert assessment of lymphoproliferative and myeloproliferative disorders for a regional network of 14 NHS hospitals in the North East and Cumbria.
Provision of an extensive repertoire of immunohistochemistry is essential for this service, and methods employed must be suitable for successful use with material fixed and processed according to protocols that vary considerably between submitting hospitals.
Access to material representing a wide range of diagnoses, in a great variety of tissues, diversely processed for histology has provided us with an excellent opportunity to undertake a systematic, prospective evaluation of three new monoclonal antibodies, NCL-L-CD7-580, NCL-L-CD19-163 and NCL-L-CD30-591 (Novocastraª, Leica Biosystems, Newcastle, UK) alongside our standard antibody panels in regular use.
For a three month period, a new anti-CD19 monoclonal antibody (clone BT51E) was added to our standard basic lymphoid cell immunostaining panel (CD20, CD79a, CD2, CD3 and Ki67). A new anti-CD30 clone (JCM182) was added in to our “Hodgkin panel” which currently employs the CD30 antibody clone BerH2 plus monoclonal antibodies reactive with CD15, CD45, EBV-LMP1 and EMA. A new anti-CD7 clone (LP15) was added to our extended panel for investigation of putative T cell lymphomas (CD4, CD5, CD7 (CD7-272), CD8, CD43, CD45RO and CD246/ALK1).
Sections stained with the novel antibodies were scored semi-quantitatively by two assessors (BW, GM) at the time of histological evaluation of cases and scores also assigned on the same basis to sections stained with our regular antibodies.
Ideal staining scored 3, with scores of 2 and 1 being assigned to cases with mildly or markedly reduced intensity of staining. A score of 0 was given if staining failed completely. Scores were then analysed as a percentage of the maximum score possible for the number of cases assessed. During the 3-month period, 100 cases were accrued that required staining with at least one of the immunohistochemical panels.
B Cell Markers assessed according to tissue immunostained
This strategy provided a very thorough assessment of CD19 in comparison with our current standard B cell markers and also gave useful insight into the comparative performance of the new CD30 clone.
However, only a small number of putative T cell lymphomas were evaluated with the extended T cell phenotyping panel during the study period, so that information for this reagent is only anecdotal.
Scores were analysed as a percentage of the maximum score possible for the number of cases assessed and the results are shown in Figure 1.
For CD19, sufficient cases were available to analyse the results according to source for the six largest hospitals in the network. Bone marrow trephine specimens were omitted from this sub-analysis, since decalcification methods are known to vary between these hospitals. Scores were also analysed according to the tissue type and diagnosis, to evaluate whether these factors influenced the quality of staining. The CD19 antibody gave equivalent strong staining to that for CD20 or CD79a in many cases but was rather more variable overall (Figure 1).
This was not significantly influenced by variations in fixation and processing that exists between the different hospitals in our network. Variation apparently linked to the underlying diagnosis in fact largely reflected differences in tissue type, with decalcification of bone marrow trephine specimens (using either formic acid or chelation in different hospitals) having an adverse effect on staining quality (Figure 2A-C). Reduction in staining quality found with spleens is likely to reflect the slow and often suboptimal fixation achieved for splenectomy specimens; although it is unclear why there should be an adverse effect for tonsils.
The numerical scores do not do full justice to the results obtained, however, the staining for CD19 was crisp and easy to interpret in the many good preparations, staining a wider spectrum of B lineage cells than CD20 and subjectively easier to assess in many cases than CD79a.
One major advantage of the availability of this new CD19 antibody is that it creates greater comparability between flow cytometric immunostaining of fresh cells and fixed tissue immunohistochemistry.1 It can also provide an additional B cell surface marker for use in complex lymphoid proliferations and those with down-regulated CD20 expression following anti-CD20 immunotherapy.2
Current and new CD30 assessed according to diagnosis
For CD30, the new clone, JCM182, out-performed the current gold-standard BerH2 clone in almost all of the 25 cases studied (Figure 3, 4A & B). The interpretation was generally unequivocal although non-specialists should be aware that a population of medium-sized to large activated perifollicular cells, predominantly intra- and peri-follicular in distribution, express CD303 and these are more clearly apparent in sections immunostained with the new antibody, clone JCM182.
For CD7, our study included too few examples for anything more than anecdotal comment. However, the new antibody performed well (Figure 5A & B), giving an overall score of 93% in the 7 cases examined, in comparison with 73% scored by our current antibody. A major improvement in sections stained for CD7 with the new clone is the absence of the strong nucleolar staining that occurs commonly with clone CD7-272.
Prospective testing of three novel antibodies for immunohistochemistry alongside routine panels in a busy diagnostic laboratory provided a simple and cost-effective strategy to evaluate the performance of the CD19 reagent, because of the frequency and range of material with which it could be tested.
This kind of small scale study is difficult to justify and fund independently as “research” in the current UK research climate but is essential for appropriate validation of novel reagents as they are developed for potential diagnostic use. Our strategy was less successful for the CD30 and, in particular, CD7 reagents because we made less frequent use of immunostaining panels requiring these antibodies than we anticipated in the 3-month study period.
Retrospective analysis of a selected cohort of cases would be the method of choice for such antibodies not included in an immunostaining panel applied to a high proportion of routine cases.
- Dunphy CH (2004). Applications of flow cytometry and immunohistochemistry to diagnostic hematopathology. Archives of Pathology and Laboratory Medicine: 128 (9), 1004–1022.
- Foran JM, Norton AJ, Micallef INM, et al (2001). Loss of CD20 expression following treatment with rituximab (chimaeric monoclonal anti-CD20): a retrospective cohort analysis. British Journal of Haematology: 114 (4), 881–883.
- Segal GH, Kjeldsberg CR, Smith GP, et al (1994). CD30 antigen expression in florid immunoblastic proliferations. A clinicopathologic study of 14 cases. American Journal of Clinical Pathology: 102 (3), 292–298.