Blood count changes due to mononuclear cell apheresis.

Complete blood count and CFU-GM count of 41 healthy donors was determined immediately before and immediately after mononuclear cell (MNC) apheresis. Red blood cell count dropped from a mean of 5.2 to 4.9 x 10(12)/l, the difference being significant (p less than 0.001). Platelet count was reduced from a mean of 219 to 187 x 10(9)/l, the difference was once again significant (p less than 0.005). Polymorphonuclear cell count did not change significantly, while MNC count was reduced substantially (from mean 3.1 to 2.4 x 10(9)/l, p less than 0.001). The difference between pre- and postapheresis CFU-GM count (median 11.3 versus 8.8 x 10(3)/l respectively) was on the borderline of statistical significance (p = 0.05).     

Incidence of peripheral blood eosinophilia and the threshold eosinophile count for indicating hypereosinophilia-associated diseases

BACKGROUND: A definite threshold of the peripheral blood eosinophile count that indicates the presence of hypereosinophilia-associated diseases has not yet been determined. METHODS: The threshold eosinophile count at which cases of hypereosinophilia-associated diseases (n = 25) can be differentiated from those of bronchial asthma (n = 101) was determined. Then, the incidences of eosinophile counts greater than 1.0 x 10(9)/l or the threshold level, were studied by analysis of 43,805 samples sent to the laboratory, and the diseases associated with the increased counts were determined. RESULTS: The eosinophile count in cases of hypereosinophilia-associated diseases and in those of bronchial asthma were 10.967 +/- 1.680 x 10(9)/l and 0.574 +/- 0.045 x 10(9)/l, respectively (P < 0.001); the threshold was 2.052 x 10(9)/l. The percentages of samples with an eosinophile count of more than 1.0 x 10(9)/l and 2.052 x 10(9)/l were 0.6% and 0.1%, respectively; the latter comprised of 41 samples from 24 patients including eight with hypereosinophilia-associated diseases. The patients with hypereosinophilia-associated diseases had a significantly higher count, and a higher incidence of counts of more than 2.052 x 10(9)/l, than others, including patients with malignancies and symptoms conventionally referred as "atopic diseases". CONCLUSION: Hypereosinophilia-associated diseases are associated with a very high eosinophile count of more than 2.052 x 10(9)/l, which was observed rarely.     

Effectiveness of interferon alfa in different stages of thrombocythemia (preliminary report)]

In 6 women aged 38 to 68 years with thrombocythaemia during chronic myeloid leukaemia (4 cases), myelofibrosis (1 case), and idiopathic thrombocythaemia (1 case) the effects of recombinant human alpha-interferon (Intron A, rh IFN alpha -2b, Schering) were studied. The drug was given to all patients subcutaneously in one daily dose of 3 x 10(6) u, every day for 3 weeks, and then in the same doses twice weekly for 2 weeks (5 cases) and for 14 weeks (1 case). Intron A caused in all cases a fall of peripheral blood platelet count by 37% to 65.5% (mean 50%) in relation to the initial count (532 - 1,453 x 10(9)/l). The fall of the platelet count occurred usually after 7-10 days of this treatment, and the lowest count was noted usually after 24 days (10 to 42 days). During the treatment in 4 cases the peripheral leucocyte count dropped as well by 20-70%. In no cases exacerbation of chronic myeloid leukaemia was noted, and in the patient with myelofibrosis the enlarged spleen shrunk somewhat. These results of treatment and follow-up of patients with thrombocythaemia treated with Intron A indicate a significant although short-lasting effect of platelet count fall limited, however, to the time of the treatment. Side effects of the drug included mainly febrile conditions, myalgia and arthralgia.     

Cytomegalovirus (CMV) infection, CD4+ lymphocyte counts and the development of AIDS in HIV-1-infected haemophiliac patients.

After a maximum of 11 years (median 8.3 years) from the time of HIV seroconversion, 25 out of 59 (42%) of CMV-seropositive haemophiliacs had progressed to AIDS, as opposed to eight out of 50 (16%) CMV seronegatives. The age-adjusted relative risk for AIDS among CMV seropositives was 2.4 (P = 0.03). In order to determine how this adverse effect is mediated, the mean rate of decline in serial CD4+ lymphocyte counts was studied. CD4+ lymphocyte counts tended to decline more rapidly in CMV seropositives than in seronegatives (-0.087 x 10(9)/l per annum versus -0.082 x 10(9)/l per annum), but this difference did not reach statistical significance. The average CD4+ lymphocyte count at the time of HIV seroconversion was estimated to be similar in CMV seropositives and negatives, because in HIV-1-negative haemophiliacs the CD4+ counts were virtually identical, after adjustment for age (0.94 x 10(9)/l and 0.97 x 10(9)/l, respectively). The median CD4+ cell count at which AIDS developed was higher in the CMV-seropositive group (0.07 x 10(9)/l) than in the seronegative group (0.04 x 10(9)/l), but this difference did not reach statistical significance. We conclude from these findings that the adverse effect of CMV is not wholly mediated via a more rapid loss of CD4+ cells. We discuss other processes that may be mediated by CMV, such as a functional deficiency of residual CD4+ cells, or dissemination of HIV in other organs, which may be important in determining the earlier onset of AIDS among CMV-seropositive subjects.     

Clinical haematology of the great bustard (Otis tarda)

The haematological parameters of healthy great bustards (Otis tarda L.) have been determined. The values obtained were red cell count (3.0 x 10(12) +/- 0.2 x 10(12/)1), white cell count (33.0 x 10(9) +/- 2.6 x 10(9)/1), haematocrit value (0.51 +/- 0.01 1/1), haemoglobin (13.0 +/- 0.3 g/dl), mean corpuscular volume (178.7 +/- 12.5 fl), mean cell haemoglobin concentration (25.0 +/- 0.6 g/dl), mean corpuscular haemoglobin (42.5 +/- 3.2 pg), differential white cell count: heterophils (22.5 x 10(9) +/- 0.7 x 10(9)/1), lymphocytes (6.0 x 10(9)+/-0.7 x 10(9)/1), eosinophils (2.7 x 10(9) +/- 0.3 x 10(9)/1) and monocytes (1.8 x 10(9)+/-0.2 x 10(9)/1).     

Simultaneous presentation of sarcoidosis and acute myeloid leukaemia: predisposition to pulmonary haemorrhage.

A case of coexistent acute myeloid leukaemia and sarcoidosis is reported. This was complicated by recurrent pulmonary haemorrhage during reinduction and consolidation chemotherapy. A review of published papers on malignancy and sarcoidosis, in particular acute leukaemia is given. The outcome of most cases of acute leukaemia and sarcoidosis is poor with respiratory complications a frequent cause of death in this group. It is proposed that modifications to treatment to avoid pulmonary toxicity and maintenance of platelet counts above 40 x 10(9)/l are warranted to reduce the risk of this complication.     

The incidence of pseudothrombocytopenia in automatic blood analyzers

The aim of the present work was to undertake an assessment of the incidence of pseudothrombocytopenia (PTCP) in patients referred for evaluation of thrombocytopenia in an outpatient hematology clinic. METHODS: Prospective assessment of 60 consecutive cases with platelet count < 100 x 10(9)/l in a hematology clinic during a 2-year period. RESULTS: PTCP was the second most common cause for low platelet count, with an incidence of 17%. Platelet count of patients with PTCP at presentation was 42 +/- 22 x 10(9)/l, and when re-analyzed on fresh samples, 208 +/- 39 x 10(9)/l. The relatively high prevalence of pseudothrombocytopenia in our series was due to a lack of microscopic inspection of the blood smear in the primary care laboratories and considerable delay in sample processing. CONCLUSIONS: PTCP should be considered in the assessment of low platelet count. While decreasing the transfer time of blood specimens may decrease PTCP incidence, microscopic inspection of the blood smear may avoid erroneous diagnosis of thrombocytopenia.     

National comparative audit of the use of platelet transfusions in the UK

The objective of this national audit was to examine the use of platelet transfusions against audit standards developed from national guidelines. Hospitals were asked to provide data on 40 consecutive patients receiving platelet transfusions (15 haematology patients, 10 cardiac, 10 critical care and five in other clinical specialties). One hundred and eighty-seven UK hospitals participated, including 168/263 (64%) hospitals in England. A total of 4421 patients receiving platelet transfusions were audited. The reason for transfusion was documented in the medical records for 93% of transfusions and 57% were used for prophylaxis (in the absence of bleeding). Overall 3726/4421 (84%) of the transfusions were evaluable and 43% (1601/3726) were found to be non-compliant with the audit standards. A major non-compliance was failure to measure the platelet count before transfusion (29% of transfusions). Other non-compliances included the use of platelet transfusion in the absence of bleeding in 11% of cardiac surgery patients receiving platelet transfusions, the use of a threshold platelet count more than 10 x 10(9)/L for 60% of prophylactic platelet transfusions in haematology patients without risk factors indicating the need for a higher threshold, and a threshold platelet count more than 30 x 10(9)/L for 59% of prophylactic platelet transfusions in critical care. The reasons for the high rate of non-compliance were not explored in this audit, but this is a topic worthy of further study. The main recommendations were that hospitals should ensure there are written local guidelines for platelet transfusions, clinicians must be provided with training about their appropriate use, and hospitals should carry out regular audits of practice. More research should be carried out to develop the evidence base for the use of platelet transfusions, more detailed guidelines should be developed for platelet transfusions in critical care and cardiac surgery, and the audit should be repeated in about three years.     

A rapid and accurate closed-tube immunoassay for platelets on an automated hematology analyzer

Accurate and precise platelet counts are important for patients with severe thrombocytopenia or who are receiving chemotherapy. We developed a novel flow cytometric analysis of platelets that may be particularly valuable for assessing the necessity for platelet transfusions. This ImmunoPlt (CD61) assay is based in part on CD61 monoclonal antibody labeling and has been automated and implemented on the CELL-DYN 4000 hematology analyzer. It is well suited for thrombocytopenic specimens, since it reduces interference by nonplatelet particles. It takes less than 5 minutes from closed-tube aspiration to report. Data for more than 350 thrombocytopenic specimens demonstrate that the ImmunoPlt (CD61) assay is more accurate than the optical scatter or the impedance count for specimens with platelet counts between 1 and 60 x 10(3)/microL (1 and 60 x 10(9)/L). The ImmunoPlt (CD61) assay is more precise than the optical scatter or the impedance count for specimens with platelet counts between 1 and 50 x 10(3)/microL (1 and 50 x 10(9)/L).     

Lymphocyte subsets in healthy adult Kuwaiti Arabs with relative benign ethnic neutropenia

Relative and absolute neutropenia is frequently seen in the healthy adult Kuwaiti Arab population. Fluorescent monoclonal antibody labelling followed by flow cytometry was used to determine the lymphocyte subsets in 48 normal healthy individuals in the Kuwaiti adult population (24 males and 24 females, age 17-59 years) with relative or absolute neutropenia, and this was compared to age-matched controls (64 males and 63 females). The mean haemoglobin levels were 13.6+/-1.5 and 13.7+/-1.5 g/dl in the neutropenic and control groups, respectively. White blood cell counts, absolute neutrophil and lymphocyte counts in neutropenic individuals with the corresponding reference range, taken from the control subjects (in parenthesis) were: WBC, 6.7+/-1.6 x 10(9)/l (4-10.4 x 10(9)/l), neutrophils, 2.7+/-0.8 x 10(9)/l (1.87-6.63 x 10(9)/l), lymphocytes, 3.3+/-0.9 x 10(9)/l (1.4-3.62 x 10(9)/l). Absolute values of lymphocytes, CD2+, CD3+, CD19+, CD4+, CD8+, HLADR+ and CD45RA+ cells were significantly higher in the neutropenic group. The range of values with the corresponding reference ranges, in parenthesis, were: CD2+, 1.61-4.30 x 10(9)/l (0.95-2.99 x 10(9)/l), CD3+, 1.37-4.16 x 10(9)/l (0.83-2.71 x 10(9)/l), CD19+, 0.16-1.09 x 10(9)/l (0.05-0.61 x 10(9)/l), CD4+, 0.70-2.89 x 10(9)/l (0.45-1.65 x 10(9)/l), CD8+, 0.57-1.80 x 10(9)/l (0.29-1.17 x 10(9)/l), HLADR+ 0.27-1.74 x 10(9)/l (0.02-0.62 x 10(9)/l), CD45RA, 0.90-4.63 x 10(9)/l (0.34-2.05 x 10(9)/l), respectively. The levels of natural killer cells, CD56+ cells were significantly lower compared to controls while the values of memory T lymphocytes, CD45RO+ were comparable to controls. These results indicate that difference in the leukocyte subpopulations may also be indicative of differences in the lymphocyte subpopulations and that reference ranges for these cell types in healthy neutropenic and non-neutropenic individuals should be established.     

Role of CD8+ in late opportunistic infections of patients with AIDS.

We studied the relationship of CD4+ and CD8+ depletion to initial and late opportunistic infections in 62 patients with AIDS. The mean interval between initial and late infections was 12.2 months. Geometric mean (and 95% confidence intervals) of T-cell counts at the diagnosis of each infection were: (Pneumocystis carinii pneumonia) CD4+ 0.051 (0.044-0.058) x 10(9)/l, CD8+ 0.561 (0.476-0.646) x 10(9)/l; (cytomegalovirus retinitis) CD4+ 0.025 (0.019-0.031) x 10(9)/l, CD8+ 0.333 (0.183-0.483) x 10(9)/l. Mycobacterium avium-intracellulare bacteraemia closely followed cytomegalovirus dissemination. Most patients were free from late opportunistic infections caused by disseminated cytomegalovirus and M. avium-intracellulare until CD8+ declined below 0.500 x 10(9)/l. Zidovudine improved CD4+, but less so CD8+, and similarly enhanced the survival of patients treated in 1985-1990 and 1991.     

Pseudoplatelets: a retrospective study of their incidence and interference with platelet counting

AIMS: Spurious platelet counts can be found in acute leukaemias, as a result of the fragmentation of blood cells. Microscopic examination of a blood smear should be performed to detect the presence of these so called pseudoplatelets. When present, the platelet count should be corrected because of the important clinical consequences that a lower platelet count may have in these patients. METHODS: K(3)EDTA anticoagulated blood was measured on an automated blood cell counter, and a blood smear was made and stained according the May Grünwald-Giemsa method for microscopic observation. A 500 cell/particle differentiation was performed and the automated platelet count was corrected. RESULTS: The incidence of pseudoplatelets in 169 patients with acute leukaemia was studied. Pseudoplatelets were detected in 43 patients (25.4%), and seven patients (4.1%) were re-classified as having a major bleeding risk (platelet count, < 15 x 10(9)/litre). CONCLUSIONS: Platelets should be determined morphologically in patients with acute leukaemia and a routine screening method for the detection of pseudoplatelets should be developed.     

Laparoscopic splenectomy for immune thrombocytopenic purpura in patients with severe refractory thrombocytopenia

Laparoscopic splenectomy (LS) for immune thrombocytopenic purpura has a success rate of 70-90%. This invasive procedure is sometimes required for patients with refractory thrombocytopenia. However, no data on their outcome are available. If the hematologic response is inadequate, its value vis-à-vis the risk associated with major surgery is open to question. Twelve of 110 patients who underwent LS in our institution had a platelet count lower than 20 x 10(9)/l (mean 6.6 x 10(9)/l). Nine patients (75%) had a good-to-excellent hematologic response. The complication rate was 33%. LS in patients with very low platelet counts is feasible and yields a response rate similar to the average of patients with immune thrombocytopenic purpura. However, it bears a higher perioperative morbidity compared with LS in patients with higher counts.     

Chronic myelomonocytic leukemia developed 2 years after the onset of immune thrombocytopenic purpura like syndrome

An 80-year old man was diagnosed as having immune thrombocytopenic purpura based on epistaxis, purpura and by the platelet count 8 x 10(9)/l. Prednisolone and gamma globulin were administered and the platelet count had been kept around 50 x 10(9)/l during his follow up. Two years from the onset of immune thrombocytopenic purpura he was admitted because of leukocytosis (79 x 10(9)/l with 79% monocytes), anemia and thrombocytopenia. Hypercellular bone marrow with dysplasia of three lineages was observed. In the bone marrow cytogenic analysis, a -6, clonal cytogenic abnormality was observed. 45XY, der(6), t(6;6)(q16;q23). He was diagnosed as having chronic myelomonocytic leukemia. This is a difficult case in which it was diagnosed as refractory thrombocytopenia as a subgroup of myelodysplastic syndrome, rather than immune thrombocytopenic purpura. which might have preceded the development of chronic myelomonocytic leukemia.     

Infection and transfusion therapy in acute leukaemia

Granulocytopenia is the single most important risk factor for infection in patients with acute leukaemia. There are limitations to the effective prophylaxis of infection in granulocytopenic patients, but practical measures include the management of the patient in a private hospital room, the requirement of all medical personnel and visitors to wash their hands carefully and to wear masks, restricting the patient to a low-bacteria diet devoid of fresh fruit, vegetables and salads, and the administration of oral antimicrobial agents for gastrointestinal decontamination. When fever develops, empirical therapy with a combination of an aminoglycoside plus an antipseudomonal beta-lactam should be started promptly. A double beta-lactam combination of cefoperazone or ceftazidime plus piperacillin can be substituted if nephrotoxicity is a concern. The addition of empirical intravenous amphotericin may be useful in patients who remain febrile and granulocytopenic on broad-spectrum antibiotics, especially if surveillance cultures indicate fungal colonization. Amphotericin is also the most reliable agent for the treatment of established fungal infections. Acyclovir is not recommended for prophylaxis in acute leukaemia patients but should be reserved for the treatment of well-documented and clinically significant herpes simplex viral infections. During periods of remission, most patients with AML remain free of infection except when they become granulocytopenic again during intensification or consolidation chemotherapy. On the other hand, children with ALL in remission may experience frequent infections unrelated to granulocytopenia as a consequence of their maintenance chemotherapy. Pneumocystis carinii, varicella zoster, and other viruses are common pathogens. Trimethoprim-sulphamethoxazole is effective prophylaxis against Pneumocystis carinii pneumonia in patients with ALL, while intravenous acyclovir is the drug of choice for treatment of varicella zoster infection. Transfusion therapy in the acute leukaemia patient is guided by the patient's peripheral blood counts and degree of sensitization to blood products. Generally, packed red blood cells are given in order to maintain the haematocrit at greater than 30%, while random-donor platelets are administered to keep the platelet count at greater than 20 X 10(9)/l. If refractoriness to platelet transfusions develops, HLA-matched platelets from family members or selected unrelated donors can be used. Similarly, washed or filtered red blood cells may be given to patients with previous and recurrent non-haemolytic febrile reactions to red blood cell transfusions.(ABSTRACT TRUNCATED AT 400 WORDS)     

Neonatal alloimmune thrombocytopenia due to anti-HPA-2b (anti-Koa)

Most severe cases of neonatal alloimmune thrombocytopenia (NAIT) are due to anti-HPA-1a (anti-PlA1) antibodies. We report a case of NAIT due to anti-HPA-2b that resulted in in utero intracranial hemorrhage.A 33-year-old G2P1A0 Caucasian woman had a routine ultrasound at 34 weeks. The fetus appeared to have a left hemispheric hematoma. IVIG, 1g/kg, was started immediately and administered weekly until delivery. One day after receiving the first dose of IVIG, fetal platelet count was 18 x 10(9)/L, and Hb was 116 g/L. Eleven mL of matched platelets compatible by monoclonal antibody immobilization of platelet antigens (MAIPA) assay were transfused in utero, raising the platelet count to 62 x 10(9)/L. Repeat transfusions were done later that week and 1 week later, with pretransfusion counts of 19 x 10(9)/L and 16 x 10(9)/L, respectively. Delivery by C section was done at 35.5 weeks, after the third platelet transfusion. Platelet count at birth was 77 x 10(9)/L. Drainage of the hematoma was performed after transfusion. Testing with a solid phase ELISA revealed reactivity against GP1b/IX. MAIPA testing after platelet treatment with the protease inhibitor leupeptin demonstrated the presence of anti-HPA-2b. On PCR-SSP the mother was HPA-2a homozygous, the father was HPA-2a/2b. Antibodies against the HPA-2b antigen located on the GP1b/IX complex have been reported in rare cases of NAIT. Testing is complicated by proteolytic degradation of the antigen-bearing fragment. Compatible platelets are easily found since approximately 85 percent of donors are HPA-2a/2a.     

Risk of underestimating platelet count because of the increase of mean platelet volume at the end of pregnancy

We report the case of a pregnant woman for whom the platelet count (77 x 10(9)/L) was underestimated by Coulter STKS analyzer during the third trimester because of large platelets. The microscopic counting of platelets revealed an isolated thrombocytopenia (120 x 10(9)/L). When not pregnant, the patient has low but normal platelet count (155 x 10(9)/L) with high mean platelet volume (MPV > 12 fL). This case report recalls that concomitantly to the decrease in platelet count, the MPV significantly increases at the end of pregnancy. This poorly known phenomenon does not impair platelet count by blood cell analyzers in as much as the platelet volume is in the range of measurement but may be responsible for underestimation of the platelet count if the MPV is already high before pregnancy. We describe how to detect this anomaly and propose simple guidelines for thrombocytopenia in normal pregnancy.     

CD3+ CD8+ T cell lymphocytosis masking B cell leukaemia.

A patient with CD3, CD8 positive lymphocytosis presented with features consistent with T cell chronic lymphocytic leukaemia/proliferations of large granular lymphocytes. The marrow and blood lymphoid populations (19.4 x 10(9)/l) contained more than 80% CD3 and CD8 positive cells with no evidence of a monotypic B cell population. A biopsy specimen of a vasculitic rash showed a diffuse infiltrate of CD3, CD8 positive cells into the upper dermis, consistent with T cell lymphocytic disease. After follow up for two years without treatment the blood lymphocyte count was 53 x 10(9)/l and was composed of cytologically small lymphocytes. A monoclonal SIg M D k lymphoid population (more than 90%) was demonstrable in sample blood and marrow aspirate. Gene rearrangement studies carried out on DNA extracted from peripheral blood lymphocytes at presentation and at two year follow up exhibited JH and Ck immunoglobulin gene rearrangement but no rearrangement of T cell receptor TcR gamma and beta genes. It is thought that this is the first well documented case of an aggressive CD8 positive lymphocytosis preceding, or in response to, an underlying B cell neoplasm.     

Evaluation of platelet indexes in patients with aortic aneurysm

The purpose of this study was to investigate whether platelet indexes [platelet count, mean platelet volume (MPV), platelet-large cell ratio (P-LCR), and platelet distribution width (PDW)] could serve as diagnostic tools to evaluate the potential significance of platelet heterogeneity on thrombus formation. Blood samples were obtained from 54 patients with aortic aneurysm (AA; mean age 73 years; 40 males, 14 females), from 120 age-matched controls (AC; mean age 74 years; 61 males, 59 females), and from 38 young controls (YC; mean age 30 years; 20 males, 18 females). We measured the blood platelet indexes using an automated counter, as well as plasma D-dimer and thrombin-antithrombin III complex (TAT) using ELISA. The AA patients were divided into two groups, group A (platelet count more than 150 x 10(9)/l, n = 36) and group B (platelet count below 150 x 10(9)/l, n = 18). There was no difference in the platelet count between AC and YC. However, P-LCR was significantly higher (p = 0.0113) in AC. MPV and PDW were also higher, but not significantly so. The platelet count was not different between group A and AC. MPV (p = 0.0024 and <0.0001, respectively), P-LCR (p < 0.0012 and <0.0001, respectively) and PDW (p = 0.0006 and 0.0005, respectively) were significantly lower in group A than in group B and AC. The platelet indexes were significantly lower in the 54 AA patients than in the AC. There were significant inverse relationships between the platelet count and other indexes in the AC and 54 AA patients; however, no relationships were observed in group A, B and YC. The D-dimer and TAT levels were significantly higher in group B than in groups A and YC. In conclusion, these findings suggest that large platelets decrease rather than increase in patients with AA.     

Laboratory control values for CD4 and CD8 T lymphocytes. Implications for HIV-1 diagnosis.

With the advent of standard flow cytometric methods using two-colour fluorescence on samples of whole blood, it is possible to establish the ranges of CD3, CD4 and CD8 T lymphocyte subsets in the routine laboratory, and also to assist the definition of HIV-1-related deviations from these normal values. In 676 HIV-1-seronegative individuals the lymphocyte subset percentages and absolute counts were determined. The samples taken mostly in the morning. The groups included heterosexual controls, people with various clotting disorders but without lymphocyte abnormalities as well as seronegative homosexual men as the appropriate controls for the HIV-1-infected groups. The stability of CD4% and CD8% values was demonstrated throughout life, and in children CD4 values less than 25% could be regarded as abnormal. The absolute counts of all T cell subsets decreased from birth until the age of 10 years. In adolescents and adults the absolute numbers (mean +/- s.d.) of lymphocytes, CD3, CD4 and CD8 cells were 1.90 +/- 0.55, 1.45 +/- 0.46, 0.83 +/- 0.29 and 0.56 +/- 0.23 x 10(9)/l, respectively. In patients with haemophilia A and B the mean values did not differ significantly. In homosexual men higher CD8 levels were seen compared with heterosexual men and 27% had an inverted CD4/CD8 ratio but mostly without CD4 lymphopenia (CD4 less than 0.4 x 10(9)/l). However, some healthy uninfected people were 'physiologically' lymphopenic without having inverted CD4/CD8 ratios. When the variations 'within persons' were studied longitudinally over a 5-year period, the absolute CD4 counts tended to be fixed at different levels. As a marked contrast, over 60% of asymptomatic HIV-1+ patients exhibited low CD4 counts less than 0.4 x 10(9)/l together with inverted CD4/CD8 ratios. Such combined changes among the heterosexual and HIV-1-seronegative homosexual groups were as rare as 1.4% and 3%, respectively. For this reason, when the lymphocyte tests show less than 0.4 x 10(9)/l CD4 count and a CD4/CD8 ratio of less than unity, the individuals need to be investigated further for chronicity of this disorder, the signs of viral infections such as HIV-1 and other causes of immunodeficiency.