52 patients (30 male, 22 female) with paroxysmal cold haemoglobinuria (PCH) were identified by critically examining the records of all cases with Donath-Landsteiner antibodies seen over a 37 year period. Although ages ranged from 1-82 years, PCH was much commoner in young children; the median age at presentation was 5 and the peak incidence, 0.4 per year per 100,000 of the population at risk, was in the 4 years and under group. 44 patients had acute transient PCH, 3 chronic non-syphilitic PCH and 1 chronic syphilitic PCH; 4, in whom the positive Donath-Landsteiner tests were incidental findings, could not be classified. Acute PCH typically presented in young children as sudden onset of malaise, haemoglobinuria and pallor, often associated with mild jaundice - all 30 patients who were 13 or younger had this type. There was usually a history of a recent viral type infection, most commonly of the upper respiratory tract. The occurrence of acute PCH had no obvious relation to exposure to cold. Dramatic and rapid falls in haemoglobin level were common, often accompanied initially by relative or absolute reticulocytopenia. The illness was severe, but the prognosis was generally good and the majority of patients had completely recovered within one month, some requiring no treatment. In approximately 68% of patients, blood transfusion was needed; the P blood group was not taken into account, but the patients were kept warm throughout. Steroids (usually prednisolone) were given in many cases; but since there was no evidence to support their benefit, it was recommended that they were stopped as soon as the diagnosis was confirmed. Chronic non-syphilitic PCH was much rarer; the patients had a characteristic presentation of severe systemic symptoms (paroxysms) and haemoglobinuria brought on by exposure to cold. The clinical manifestations varied in intensity between individuals; at the extreme, severe debility was experienced over many years. Warmth and avoidance of cold were an effective treatment, though in a severely afflicted patient, an attack could be precipitated by relatively little exposure to cold. With chronic syphilitic PCH there was the added need to treat the specific infection. The direct antiglobulin test was almost always positive (50 out of 51 cases tested), with C3d coating the red cells. The Donath-Landsteiner antibodies were of IgG class, but this was rarely demonstrated unless direct antiglobulin tests were carried out at 4 degrees C. The antibodies showed the classical anti-P specificity in 27 of the 30 patients tested; other specificities were unusual. Although acting much better as haemolysins, Donath-Landsteiner antibodies could also cause weak agglutination at room temperature. This was paralleled in vivo by predominantly intravascular haemolysis with an extravascular component. Diagnosis was usually easy when PCH was suspected, though in some patients Donath-Landsteiner tests did not become positive until more sensitive techniques involving papainised red cells or two-stage procedures were employe...

An understanding of the causes and circumstances under which posttransplant immune hemolysis arises is required for proper management. As more patients become long-term survivors of unrelated bone marrow and/or PBPC transplants, chimerism and complex serologic problems will become more common.

No abstract

Autoimmune hemolysis due to IgA antibodies alone in rare, with red cell destruction occurring through mechanisms similar to those for IgG. Most commonly, IgA acts synergistically with other immunoglobulins (usually IgG) and complement; the hemolysis may be severe. Whether IgA autoantibodies alone can activate complement remains controversial, but increasing evidence suggests that they can, possibly via the alternative pathway, and that this activation may result in intravascular hemolysis.

The provision of blood for transfusing patients whose sera contain red cell autoantibodies requires considerable expertise. Over 8 years, 3888 samples from 2149 patients were examined; the varying clinical presentation necessitated a flexible investigative approach. The autoantibodies showed evidence of blood group specificity in 706 patients (32.9%), usually within the Rh system for warm reacting antibodies, whereas cold antibodies were mostly anti-I. Concomitant alloantibodies were detected by noting varying reaction strengths during antibody investigation and compatibility testing, and by absorption techniques using autologous or selected allogenous red cells. Alloantibodies were found in 294 patients (13.7%); the most frequent were anti-E and anti-K. Compatibility tests were performed on SAG-M donor blood of suitable ABO group, similar Rh genotype, Kell negative and lacking antigens to any alloantibodies detected. All units of blood were incompatible by at least one technique and were issued as 'not compatible but considered suitable'. A total of 7052 units was issued for 1685 patients; no haemolytic reactions were reported. It was concluded that blood can be safely given to patients with autoantibodies, even in serologically complex cases, providing adequate investigations are carried out.

The effects and interrelationships of multiple immunoglobulin coating (i.e., increased red cell [RBC]-bound IgM and/or IgA in addition to IgG) were investigated in 404 patients with warm-reactive RBC autoantibodies on 590 occasions. Multiple immunoglobulins were detected by enzyme-linked direct antiglobulin tests in 218 samples (37%), but in only 87 (15%) by agglutination methods. Differences in populations were examined by chi-square, with p less than 0.005 being required for significance because of the multiple tests. Compared with IgG coating alone, multiple immunoglobulins were significantly associated with larger quantities (greater than 800 molecules/RBC) of IgG, multiple IgG subclasses, IgG3 and C3d bound to the cells, and with serum haptoglobin levels of less than 0.1 g per L. The latter association was still significant when higher levels of RBC-bound IgG and subclass pattern were taken into account. In samples with multiple immunoglobulin coating, there was no significant relationship (p greater than 0.05) between haptoglobins of less than 0.1 g per L and either C3d or multiple IgG subclasses. It was concluded that multiple immunoglobulin coating, even when undetected by agglutination methods, is a major cause of hemolysis: it is part of a more generalized autoimmune response and acts with other factors such as the quantity of bound IgG, the IgG subclass pattern, and complement; it also has an important hemolytic effect in its own right.

No abstract

No abstract