快樂小藥師 Im pharmacist nichts glücklich

疑似綠茶引起的紫斑症

George D. Liatsos, Antonios Moulakakis, Ioannis Ketikoglou and Stella Klonari

Purpose. A case of a patient who developed thrombotic thrombocytopenic purpura (TTP) after consuming a weight-loss product containing green tea is reported.

Summary. A 38-year-old, 68-kg Caucasian woman arrived at the emergency department with a one-week history of malaise, fatigue, and petechiae of the skin. She had no symptoms of infection and denied illegal drug use. Her medical history included hypothyroidism, for which she was treated with levothyroxine 150 µg daily for the past four years. She reported that she had been using a green tea preparation for the two months before admission to lose body weight. The daily preparation contained 200 mg of green tea extract 5:1, equivalent to 1 g of natural green tea. On clinical examination, the patient appeared acutely ill and was afebrile, with pallor, petechiae, and purpura of the extremities. Laboratory test results at the time of admission revealed that the patient had anemia and marked thrombocytopenia. A peripheral blood smear demonstrated a feature of microangiopathic hemolytic anemia. Immunoglobulin G autoantibodies against ADAM metallopeptidase with thrombospondin type 1 motif, 13 were detected. On hospital day 3, the patient appeared confused and exhibited aphasia that was initially transient but then recurrent. Brain computerized tomography did not exhibit focal pathology. Over the next few days, her neurologic symptoms subsided and her platelet count and hematocrit value gradually increased. Plasmapheresis was performed (12 procedures). Corticosteroid treatment was also initiated. After 20 days of hospitalization, the patient was discharged.

Conclusion. A 38-year-old woman developed TTP after consuming a weight-loss product containing green tea extract for two months.

Case report

A 38-year-old, 68-kg Caucasian woman arrived at the emergency department with a one-week history of malaise, fatigue, and petechiae of the skin. She had no symptoms of infection and denied illegal drug use. Her medical history included hypothyroidism treated with levothyroxine 150 µg daily for the past four years. She reported that she was using a green tea preparation for the two months before admission to lose body weight. The daily preparation contained 200 mg of green tea extract 5:1 (the ratio of the catechins contained in the supplement product), equivalent to 1 g of natural green tea. The preparation, manufactured by a pharmaceutical company approved by the national Greek drug organization, was supplied by a pharmacy. There were no indications that other substances were contained in the preparation. The product was discontinued at the time of admission.

On clinical examination, the patient appeared acutely ill and was afebrile, with pallor, petechiae, and purpura of the extremities. Laboratory test results at the time of admission revealed that the patient had anemia and marked thrombocytopenia. A peripheral blood smear showed microangiopathic hemolytic anemia with fragmented red blood cells (schistocytes ++), reticulocytosis, polychromasia, and circulating nucleated red blood cells. Renal function and coagulation test values were within normal limits. Direct and indirect Coombs’ test results were negative. Results of detection tests for cryoglobulin, autoimmune disease autoantibodies, and bacterial and viral antibodies (including anti-HIV antibodies) and screening tests for paroxysmal nocturnal hemoglobulinuria were negative. Serum levels of human chorionic gonadotropin were within normal limits (nonpregnant). Immunoglobulin G (IgG) autoantibody against ADAM metallopeptidase with thrombospondin type 1 motif, 13 (ADAMTS13) was detected with an enzyme-linked immunosorbent assay (65 IU/mL; negative values are <15 IU/mL). The activity of von Willebrand’s factor (vWF) cleaving protease was not measured.

On hospital day 3, the patient appeared confused, with aphasia that was initially transient but then recurrent. Computerized tomography of the brain did not reveal focal pathology. Over the next few days, the patient’s neurologic symptoms subsided, and her platelet count and hematocrit value gradually increased. She was treated with plasmapheresis (12 procedures). Corticosteriod treatment was initiated with high-dose i.v. methylprednisolone (1 g daily) for the first three days, followed by i.v. prednisone (75 mg daily) for one week. The patient was given oral prednisone (60 mg daily) for two weeks, and the dosage was tapered over the following three weeks. Her laboratory test values had improved at the time of discharge from the hospital. She sustained remission over the next few months of outpatient follow-up.

Discussion

Over the past decade, numerous discoveries have revolutionized our understanding of TTP. Normally, multimeric vWF adheres to endothelial cells, platelets adhere to vWF through platelet membrane glycoprotein Ib, and platelet-rich thrombi are cleaved by ADAMTS13. In TTP, due to deficient ADAMTS13 activity, vWF-dependent platelet accumulation continues, causing microvascular thrombosis and subsequent hemolysis.

It is well-known that the ADAMTS enzymes, a disintegrin and metalloproteinase with thrombospondin type 1 repeats, are related to the matrix metalloproteinases (MMPs). Extensive research has been conducted on catechins’ anti-metalloproteinase activity. Makimura et al. Demonstrated that GTCs can inhibit collagenase derived from both eukaryotic and prokaryotic cells.

Biochemical and experimental data in several animal and human models have suggested extended MMP inhibition by GTCs, including inhibition of MMP-1, MMP-2, MMP-7, MMP-9, and MMP-12, that results in potential beneficial effects of GTCs. In their attempt to gain insight into the mechanism of inhibition, Garbisa et al. reported that MMP-2 and MMP-9 inhibition by epigallocatechin gallate (EGCG), the most abundant component of GTCs, was zinc-chelation independent, dose dependent, and noncompetitive. They also found that tumor cell invasion was reduced by 50%, with EGCG concentrations equivalent to those in the plasma of moderate green tea drinkers, while tissue metalloproteinase inhibitor (TIMP)-1 and TIMP-2 were upregulated in the presence of higher EGCG concentrations.

Further investigation provided more understanding of the molecular mechanism of inhibition, suggesting that EGCG may enhance pro-MMP-2 binding and activated MMP-2 binding to TIMP-2. Moreover, inhibition of ADAMTS1, ADAMTS4, and ADAMTS5, the proteinases ("aggrecanases") responsible for the breakdown of cartilage aggrecan, which is a major constituent of articular cartilage, can prevent aggrecan breakdown in osteoarthritic cartilage. GTCs have been shown to inhibit aggrecan and collagen degradation in vitro.

Finally, Vankemmelbeke et al. suggested that inhibition of aggrecan and collagen degradation by GTCs is due to a direct inhibitory effect of GTCs on the activity of human recombinant ADAMTS1, ADAMTS4, and ADAMTS5. The mechanism of inhibition was not defined, but inhibition was demonstrated to be reversible and not due to zinc sequestration. The Vankemmelbeke et al. report is the only one published to date that suggests that human ADAMTS enzymes may be inhibited by GTCs.

Patients with idiopathic TTP usually have severe ADAMTS13 deficiency, often caused by inhibitory autoantibodies that neutralize its activity or accelerate its cleavage in vivo. In secondary TTP, which has been associated with pregnancy, stem-cell transplantation, autoimmune diseases, HIV infection, cancer, or drugs (e.g., cyclosporine, quinine, mitomycin, gemcitabine, pento- statin, carmustine, tacrolimus), severe ADAMTS13 deficiency and autoantibodies are less common. However, in patients who develop secondary TTP associated with ticlopidine or clopidogrel, two widely prescribed antiplatelet agents, ADAMTS13 deficiency (more likely for ticlopidine) and detectable inhibitory serum levels of IgG autoantibody may be observed.

EGCG has been shown to inhibit platelet aggregation in rats ex vivo and in humans in vitro by lowering the activity of platelet phospholipase A₂, cyclooxygenase, and thromboxane A₂ and by inhibiting platelet aggregation induced by the carboxyl terminal peptide of thrombospondin-1, thus exhibiting antiplatelet activity.

Our patient exhibited a typical TTP syndrome with a high titer of anti-ADAMTS13 IgG autoantibody that resolved with corticosteroids and plasma exchanges. The patient’s two-month history of consumption of green tea before TTP onset led us to suspect a possible causative relationship. The scientific data mentioned above concerning catechins’ extended antimetalloproteinase activity, including inhibition of MMP-1, MMP-2, MMP-7, MMP-9, and MMP-12, as well as their effect on human metalloproteinases ADAMTS1, ADAMTS4, and ADAMTS5, may support our hypothesis that GTCs inhibit another metalloproteinase, ADAMTS13, with the production of ADAMTS13 IgG autoantibody. Our patient’s ADAMTS13 IgG values exceeded normal values by fourfold, resulting in TTP. As discussed above, autoantibodies against ADAMTS13 are less common in secondary TTP with the exception of secondary TTP associated with antiplatelet agents. Because EGCG exhibits definite antiplatelet activity, we further hypothesize that possible ADAMTS13 inhibition by GTCs via production of anti-ADAMTS13 autoantibodies may follow a pattern that resembles the pathogenesis of antiplatelet agents ticlopidine- and clopidogrel-associated TTP. This generates a unique hypothesis for innovations in clinical research concerning a widely consumed herbal tea. Indeed, experimental data confirming or rejecting this theory could significantly advance the understanding of TTP’s etiology.

No similar case has been reported in the English-language literature to date. This case scored a 3 on the Naranjo et al. scale of causality, classifying as "possible" the likelihood that green tea caused the patient’s TTP. It is unknown why no other reports have indicated an association between green tea consumption and TTP. There could be two explanations for this. Induction of TTP by GTCs could be an idiosyncratic drug reaction, meaning that its occurrence is infrequent, unpredictable, and dose independent and that it might occur at any time during or shortly after exposure to the tea. The other explanation is that green tea preparations are considered harmless, even among physicians; thus, even if other similar incidents had occurred, a causative connection would not have been considered. Therefore, it is important to inform clinicians about this possible, rare, but severe adverse effect of green tea consumption.