|Home | About | Journals | Submit | Contact Us | Français|
An examination of a blood smear may be requested by physicians or initiated by laboratory staff as a diagnostic aid when a lymphoproliferative disorder or mononucleosis is suspected.1 The classification of lymphocyte disorders is complex because there are various manifestations of these disorders.2,3 Three major groups can be distinguished: reactive lymphocytosis, premalignant neoplastic disorders of lymphocytes and neoplastic disorders of lymphocytes.4 Lymphocyte disorders may be accompanied by abnormal lymphocyte morphology; therefore it is important to recognise and report abnormal lymphocytes, such as atypical lymphocytes and lymphoblasts. The ability of individual technicians, however, to recognise abnormal lymphocytes is consistently quite poor.5 There are no standardised definitions regarding the morphology of the various cells, and interpretation is based on individual experience and dependent on the availability of additional clinical information. Often, transitional forms between lymphocytes and plasma cells are seen in the blood of patients with viral infections. These cells are variously known as atypical lymphocytes, lymphocytoid plasma cells or plasmacytoid lymphocytes.6 The so‐called atypical lymphocyte is a non‐neoplastic lymphocyte seen in the peripheral blood, and appears to be a non‐specific response to stress from a variety of disorders.7 Small lymphocytes become larger in size and become capable of dividing. These atypical lymphocytes vary in morphology and surface markers.8,9 Hoagland developed criteria for diagnosis of infectious mononucleosis comprising 50% of the white blood cell differential, with atypical lymphocytes accounting for 10% of the total white blood cell count.10 The morphology of lymphoblasts is even more complicated, as morphological appearance varies from small cells with scanty cytoplasm and coarse chromatin to large cells with an abundance of vacuolated cytoplasm and fine chromatin.11
The morphological assessment of the peripheral blood smear remains a valued diagnostic tool,1 despite the considerable interobserver variation in interpretation. New techniques such as automated recognition systems may reduce the microscopic review of blood films in the near future.12 Nevertheless, the relevance of the morphological evaluation of abnormal lymphocytes can be questioned, considering the large variation in reported results.
In this letter we show the general inconsistency in reporting whether an abnormal lymphocyte is (non‐neoplastic) reactive, neoplastic or an artefact.
From a blood smear of an orthopaedic patient (female, age 3 years) with a lymphocytosis which was normal for a child of that age, 100 white blood cells were randomly selected, microphotographed and processed in a PowerPoint presentation. This PowerPoint presentation was sent to 157 different hospital laboratories in The Netherlands and a leucocyte differential was requested. All participants were asked to differentiate the lymphocytes into normal lymphocytes, atypical lymphocytes, plasma cells, prolymphocytes or (lympho)blasts. Participants did not have any information of the patient.
All results were processed in Excel and for each laboratory and individual technician the mean differential was calculated. For each lymphocyte subtype the degree of concordance was calculated.
Of the 157 laboratories, 114 responded (73%) with a total of 671 individual results. The overall mean lymphocyte count (including sub‐cell types) was 56%. For 7 cells there was a concordance of >90% for normal lymphocytes (fig 1A1A),), which means that more than 90% of the individual observers called these cells normal lymphocytes. For the other 49 lymphocytes there was no concordance in subtyping. In the majority, lymphocytes (29 cells) were classified as normal or atypical (fig 1B1B).). One cell (fig 1C1C)) was classified as a normal lymphocyte or a plasma cell. Nine cells were classified as normal or atypical or prolymphocyte (fig 1D1D)) and two cells as normal or atypical or plasma cell (fig 1E1E).). For seven cells, four different subtypes were mentioned (fig 1F,G1F,G),), and to one cell, five different assessments were attributed (fig 1H1H).
One cell was shown twice in the PowerPoint presentation: in 210/671 the same cell was classified by the same observer as another subtype.
Even in the age of molecular analysis and sophisticated laboratory tests, the blood smear remains an important diagnostic tool.1 The recognition of abnormal lymphocytes in blood smears can contribute to a rapid diagnosis of various diseases, both reactive and (premalignant) neoplastic, and enables rapid therapeutic intervention. Some rare diseases may be diagnosed faster by a critical review of the lymphocytes, such as I‐cell disease.13 The need for recognition of abnormal lymphocytes should not be underestimated. The morphology of lymphocytes is complex; cell size, amount and colour of cytoplasm, as well as shape and chromatin structure of the nucleus, must be combined to come to a good characterisation of the lymphocyte. Diagnosis of infectious mononucleosis may be based on Hoagland's criteria. However, previous studies have questioned the percentage of patients with serological evidence using these criteria.14 Unfortunately a uniform definition of abnormal lymphocytes is lacking. A confused terminology is used for abnormal lymphocytes, for example, variant lymphocytes, atypical lymphocytes and even combinations of several cells are used (e.g. monocytoid lymphocytes and plasmacytoid lymphocytes). In this study we showed the variability in classification of abnormal lymphocytes and would recommend abnormal lymphocytes should be interpreted and reported with care. One cell was shown twice in the PowerPoint presentation (fig 22),), and remarkably, 31% of the morphologists were not able to reproduce their previous classification. This illustrates the problem of uniform and consistent reporting of the subtypes of abnormal lymphocytes even more clearly.
In spite of highly qualified technologists who spend large amount of time on manual microscopy, clinicians often base their diagnosis and treatment on other information. This is because they have access to physical findings, radiographs, cultures etc.15 We recommend provision of additional clinical information, which should lead to a better interpretation of blood cell morphology. However, additional information from automated haematology analysers could lead to overestimation of abnormal lymphocytes.5
We are grateful to the many technicians and institutions that performed the blood cell morphology. We also would like to thank the SKML (Dutch Society of Quality Assurance in Medical Laboratory Diagnostics) for their support.
Competing interests: None.