Allergic purpura, including purpura due to foods, drugs and infections

Abstract

• 1. Purpura is due to a vascular lesion. Thrombocytopenia, when present, tends to increase the haemorrhagic tendency.

• 2. Platelets and capillary endothelium are antigenically related. An antibody which injures the platelets can also damage the capillary endothelium. Thrombocytopenia and capillary endothelial damage may, therefore, have a common cause.

• 3. Allergic purpura is of two types: (1) Purpura associated with an erythematous skin lesion, and also with joint and visceral symptoms: the Henoch-Schönlein syndrome. This syndrome is generally regarded as a manifestation of allergy but, apart from a very small proportion of cases which are undoubtedly due to hypersensitivity to foods, the cause of the syndrome is unknown and its allergic basis is entirely unproved. (2) True purpura in which the surrounding skin is normal. It may be due to an abnormal reaction to an infection or to a drug or, occasionally, to some other substance but is rarely if ever due to foods. Many infections and drugs can cause, in different individuals, both thrombocytopenic and athrombocytopenic purpura.

• 4. Cases of true purpura due to infections may be divided into those which occur at the height of an infection and those which occur during convalescence. The former type appears to be due to an abnormal susceptibility of the patient's tissues to the factors which normally cause thrombocytopenia and increased capillary fragility at the height of an infection. If this abnormal susceptibility is shown only by the vascular endothelium, athrombocytopenic purpura will result. If the megakaryocytes and platelets are also affected, the purpura will be thrombocytopenic. Purpura occurring during convalescence can best be explained on the assumption of an allergic basis, similar perhaps to that to which nephritis following streptococcal infections has been attributed.

• 5. The only example of purpura due to a drug which has been subjected to experimental analysis is thrombocytopenic purpura due to sedormid (allyl-isopropyl-acetyl-carbamide). Sedormid appears to act by combining with platelets, so conferring upon them the properties of a weak antigen. This antigen is formed whenever the drug is taken but it has the power of stimulating antibody formation in only a very small proportion of those taking the drug. Thrombocytopenia occurs only in those who develop the antibody and is due to lysis of the platelet-sedormid antigen by antibody and complement. The capillary lesion is probably produced in a similar way, the drug combining with the endothelial cells to form a further antigen which then reacts with the antibody which causes platelet lysis, this reaction producing the vascular lesion which plays such an important part in the development of purpura. It is not known how far these findings are applicable to purpura due to other drugs. An isolated observation on thrombocytopenic purpura due to quinine suggests that the mechanisms underlying sedormid purpura may be similar to those underlying thrombocytopenic purpura due to other drugs. In athrombocytopenic purpura due to drugs it is suggested that only the capillary endothelium combines with the drug to form an antigen, and therefore that it alone reacts with the antibody, this reaction causing the capillary lesion.