Effects of single dose compared with three days'

Aims-To investigate the effects of longer term corticosteroid treatment on circulating lymphocyte subsets. Methods-Prednisolone (20 mg daily) was given to 12 healthy volunteers in a single morning dose for three days. Circulating lymphocyte subsets were measured by flow cytometry after whole blood lysis. Results-Seven hours after the first dose of prednisolone there was a significant fall in absolute numbers of lymphocytes, T cells, CD4+ and CD8+ cells, and B cells. The percentage ofT cells fell significantly, due to a fall in percentage of CD4+ cells. In contrast to the seven hour findings, at 72 hours there was a significant rise in absolute numbers of lymphocytes, T cells, CD4 +, CD8 +, and B cells. This trend was already apparent by 24 hours. The percentage of CD4+ cells was significantly raised at 72 hours, while that of CD8 + cells had fallen significantly. The percentage of natural killer cells had fallen at 72 hours; that of B cells remained increased at 72 hours. Conclusions-These findings show that corticosteroid treatment causes significant changes in lymphocyte subsets, and that such changes must be considered when designing studies of lymphocyte subsets during illness. (7 Clin Pathol 1993;46:1089-1092) Rheumatology Research Unit, Addenbrooke's Hospital, Cambridge CB2 2QQ G D Pountain B L Hazleman Department of Clinical Inmmunology MT Keogan D L Brown Correspondence to: Dr G Pountain Accepted for publication 29 July 1993 In a study of T cell subsets in polymyalgia rheumatica and giant cell arteritis (PMR/GCA)' we observed that once prednisolone treatment had been started, total numbers ofT cells rose and the percentage of CD8+ cells fell significantly. Lymphocyte subsets were normal before treatment compared with age matched controls. This suggested that the prednisolone, not the PMR/GCA, was responsible for the T cell changes. We therefore decided to investigate further the effects of prednisolone treatment, using healthy volunteers. Some of the short-term effects of corticosteriods on lymphocyte subsets have been documented. Yu2 and Fauci 3 showed lymphopenia to be maximal 4-6 hours after a single dose of corticosteroids given to healthy volunteers. With the advent of monoclonal antibodies and flow cytometry, ten Berge et al 4 documented reduced T cells, particularly OKT4 + cells, at six hours after a single dose of prednisolone given to normal volunteers compared with controls. They also showed that by 24 hours this effect had disappeared with a slight "rebound effect"-OKT4 + and OKT8 + cells were slightly increased by 24 hours, though this was not significant. This study did not look at the effects of longer term administration of prednisolone. Similarly, Tonnesen et all infused cortisol into healthy volunteers for five hours and showed reduced lymphocytes, OKT3 +, OKT4 + and OKT8 + cells by two hours compared with controls. These changes persisted 15 minutes after the cortisol infusion had been discontinued, but were not monitored after that time. Hogevold et a16 gave high dose methylprednisolone preoperatively and four and 12 hours after total hip replacement. After this short duration of corticosteroids, they showed reduced total T cells and helper and suppressor cells, compared with control patients, at 20 hours ( = eight hours after the last dose of methylprednisolone). The effects on T cells of longer term corticosteroid administration have not been documented in healthy people, although changes occurring in patients treated with corticosteroids have been described. Ferrari et al7 reported increased lymphocytes, including increased absolute numbers of T cells, CD4 + and CD8+ cells, after four weeks of corticosteroid treatment in idiopathic thrombocytopenic purpura (ITP). The percentages of CD4 + and CD8 + cells were not significantly changed. In this study, therefore, the longer term effects of corticosteroid seem to be quite different from the acute effects a few hours after a single dose. Fauci8 described the effects of seven days of cortisone administration in guinea pigs, as well as the acute effects of a single dose of hydrocortisone. At both stages there was a fall in lymphocyte and T cell numbers. The observation of reduced lymphocytes and T cells four hours after the hydrocortisone corresponds to the acute effects seen in human volunteers.2-5 But the seven day effects in Fauci's guinea pigs8 were at odds with the effects in Ferrari's patients with ITP, where after four weeks of corticosteroids lymphocytes and T cells had risen.7 The difference between these studies might be a species effect, or the effects of corticosteroids in patients might have been modified by the disease process itself. Although the nature of the acute changes 1089 on Jne 5, 2022 by gest. P rocted by coright. http/jcp.bm jcom / J C lin P ahol: frst pulished as 10.113.46.12.1089 on 1 D ecem er 193. D ow nladed fom Pountain, Keogan, Hazleman, Brown Effects on circulating lymphocyte subsets ofprednisolone EC 20 mgfor three days in healthy acute postdose effects on T cell subsets to be volunteers (n = 12) Y auepsds fet nTcl ust ob maximal around seven hours after a dose of (55 hours) 20 mg prednisolone enteric coated (EC), Baseline 7 hours 24 hours 48 hours (n = 4) 72 hours while in the longer term maximal effects were Lymphocytes (x 109/1) 2 09 1-02** 2-53 2 65 (2 35) 3.39** seen after three to four days of prednisolone Total T cells (x 109/l) 1-59 0-58** 1-84 2-00 (1-71) 2.54** CD4+ cells (x 109/1) 1-06 0-36** 1 30 1-42 (1 01) 161** adminstration. We therefore used a three day CD8+ cells (x 109l) 0-58 0-41** 067 059 (069) 0-83** period of corticosteroid treatment in the volActivatedTcells 0 11 007 NS 0-17 0-13 0-14 NS ( x 109/1) (n = 9) (p = 0 75) (p = 0 063) unteer group with blood tests taken over four NK cells ( x 109/1) 0.32 0-35 NS 0 33 0-23 (0 45) 0-34 NS days. B cells ( x 10'/1) 0 23 0-15* 0-27 0 34 (0 32) 0-53** Each volunteer had baseline blood samples T cells 72-5 61* 75-5 76 (70 5) 75 NS taken at 0900 hours and then took pred(% of lymphocytes) * E CD4+ cells 455 36* 48 495 (435) 51.5* nisolone EC 20 mg daily orally at 0900 hours (% of lymphocytes) for three days. Further blood samples were CD8+ cells 26-5 29 NS 25 24 (25-5) 24.5* taken at seven hours after the mntial dose and (% of lymphocytes) Activated T cells 4 6* 6 7 6 NS again at 24, 48, and 72 hours-24 hours after (% of lymphocytes) (n = 9) Natural killer cells 16 5 20* 12 5 9 5 (16) 9.5** the latest prednsolone dose, to avoid the (% of lymphocytes) acute postdose effects. Four volunteers also B cells 10-5 13* 12-5 13 (12-5) 14-5** had blood samples taken at 55 hours-seven (% of lymphocytes) CD4+ *CD8+ ratio 1-83 1-45 NS 2-00 2-10 (1-95) 2l10** hours after the third dose of prednisolone. (Results expressed as medians with two tailed p values for the comparison with baseline data All blood samples were processed within (Wilcoxon's rank sum test); **p < 0 01; *p < 0 05; NS = not significant.) six hours of venesection. Total white cell counts and lymphocyte counts were measured using routine methods in the in the few hours after a single dose of cortihaematology department at Addenbrooke's costeroid has been documented in the work Hospital. T cell subsets were analysed by flow described above,2 6 there is little information cytometry using a whole blood lysis techabout the effects of longer term corticosteroid nique. Total T cell numbers were measured administration in people, and in particular we using anti-CD3 (Leu 4). Activated T cells were unable to find any study of these longer were those CD3 + cells which coexpressed term effects in the absence of disease. We anti-HLA-DR. CD4 + cells were defined by therefore studied a group of healthy volundual staining with anti-CD3 (Leu 4) and teers taking prednisolone. anti-CD4 (Leu 3), while CD8 + cells were defined using anti-CD3 with anti-CD8 (Leu 2). Natural killer cells were CD3expressing Methods CD16/56 (Leu lIc + 19). B cells were meaTwelve healthy volunteers were recruited sured using anti-CD19. All monoclonal antifrom among senior medical staff. Eight were bodies were purchased from Becton men and the ages ranged from 31 to 50 years. Dickinson (Oxford, England) from the None had contraindications to corticosteroid Simultest range. Aliquots of blood were administration and all gave informed consent. incubated with antibody pairs for dual stainA pilot study in one volunteer showed the ing for 15 minutes at room temperature. Figure 1 Lymphocytes and subsets in 12 healthy volunteers, in relation to doses ofprednisolone EC 20 mg (4): 0-in blood samples taken immediately before the next prednisolone dose and 24 hours after the previous dose; 0-in blood samples taken 7 hours after the initial prednisolone dose. Results expressed as medians + interquartile range. 4 a) 0