SEARCH

SEARCH BY CITATION

Keywords:

  • autoimmune haemolytic disease;
  • hereditary spherocytosis;
  • anti-globulin test;
  • RBC survival studies.

The diagnosis of haemolytic anaemia is made when circulating red blood cell (RBC) survival is shortened. Today, the laboratory findings in haemolytic anaemia and the distinction between inherited and acquired forms are studied by preclinical medical students the world over. Although haemolytic anaemia is much less common than anaemia as a result of marrow failure or bleeding, general physicians are well familiar with the characteristic findings in the condition and, by the time the patient is seen by a haematologist, the diagnosis is usually already made.

Among the spherocytic haemolytic anaemias, hereditary spherocytosis is the most common inherited haemolytic anaemia and autoimmune haemolytic disease is the most common acquired form. The two diseases share many features that were known to clinicians in the early part of the twentieth century: jaundice and splenomegaly on physical examination, spherocytes and reticulocytosis on the blood film, increased osmotic fragility of RBCs, and clinical improvement following splenectomy. This constellation of clinical features was generally designated haemolytic or acholuric jaundice (icterus), acholuric referring to the characteristic absence of bile in the urine. A congenital or familial form of haemolytic anaemia was often diagnosed in children with a family history of jaundice, splenomegaly or haemolytic anaemia. However, spontaneous cases in children without a family history and the initial appearance of the disease in adults were common enough to cause confusion, as acquired cases were often diagnosed on the basis of acuity and absent family history in adult patients. No one seriously questioned the existence of familial or congenital forms of haemolytic anaemia, but many questioned the existence of acquired forms other than those obviously as a result of infection or toxic exposure.

It was not until the rediscovery of the principle of anti-globulin testing and its application to diagnosis of patients with spherocytic haemolytic anaemia that the existence of acquired spherocytic haemolytic anaemia was accepted and the means to easily distinguish the inherited and acquired forms was at hand. It is perhaps a unique aspect of these disorders that the underlying principles of our best current diagnostic tests, i.e. spherocytes on the blood film, the osmotic fragility test and the anti-globulin test, have been known for nearly a century or more and applied, with a few refinements, for over half a century. This review will trace the evolution of understanding of the spherocytic haemolytic anaemias from the earliest case descriptions to the middle of the twentieth century when it all came together.

Early cases

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References

Perhaps the first case of haemolytic anaemia in the medical literature was described by the physician Claudius Galen (Dreyfus, 1942). The patient, a slave of the emperor Marcus Aurelius, hunted snakes. He was bitten by a viper and, in spite of treatment with the ‘usual drugs’, his skin ‘turned the colour of a ripe leek.’ Galen prescribed theriaca, which cured the patient. Galen explained, ‘The spleen produces colours of this type, darker even than those produced by the liver. They are difficult to explain, but easy to recognize if you have seen them often.’

The first recognizable reports of haemolytic anaemia were undoubtedly cases of paroxysmal cold haemoglobinuria because of the dramatic presentation. Johannis Actuarius, a court physician at Constantinople during the late thirteenth century described azure, livid and black urine following chilling of certain individuals (Major, 1945). The syndrome was well known by physicians of the mid-nineteenth century. A relationship to jaundice was recognized and the absence of RBCs from the dark urine was noted. These patients were thought to suffer from a disorder of the kidneys.

Congenital or familial cases Towards the end of the nineteenth century, physicians began to note certain chronically icteric patients who had no bile in the urine, no evidence of liver disease and often, but not always, splenomegaly and a family history of jaundice. Vanlair and Masius (1871) described the case of a young woman who developed icterus, recurrent attacks of left upper quadrant abdominal pain and splenomegaly shortly after giving birth. The stools were not light coloured, but rather deeply pigmented. The patient's mother and sister were also icteric, and the sister's spleen was enlarged. The most remarkable aspect of this paper lies in their description of the blood findings. Although they made no mention of anaemia and had no concept of haemolysis as a pathological process, they unmistakably described spherocytes with remarkable clarity. The authors noted that some of the erythrocytes, which they called microcytes, were smaller than normal cells, 3–4 μm in diameter, spherical in shape (elle est parfaitement spherique) and the contours were completely smooth. Their drawing of the patient's blood cells and a companion drawing of normal cells is reproduced in Fig 1. With remarkable discernment, Vanlair and Masius stated, ‘The jaundice of our patient appears to be a peculiar type of icterus. The fact that the patient’s mother and sister had a slight jaundice and that the sister had an enlarged spleen may indicate that this condition is one disease entity.'

image

Figure 1. Reproduction of a camera lucida drawing from the classic paper of Vanlair and Masius (1871), de la Microcythemie. Left, sang microcytique from their propositus with microcythemia. Right, sang normal.

Download figure to PowerPoint

Murchison (1877), in his textbook, Clinical Lectures on Diseases of the Liver, describes a man who was ‘born yellow’ and who suffered from gout. His urine contained bile pigments and the liver was slightly enlarged. The patient's mother and brother also suffered from gout and jaundice. Unfortunately, there is no mention of their spleens or characteristics of their blood. Murchison was struck by the persistent and hereditary nature of the jaundice in the absence of obstruction of the bile duct.

Hunter (1888), who made important contributions to the recognition of pernicious anaemia, may have been the first to classify anaemias by mechanism, i.e. haemorrhage, decreased production and increased destruction. He classified pernicious anaemia and the experimental anaemia produced in animals by destructive agents such as pyrogallic acid as haemolytic. He distinguished haemorrhagic and haemogenic anaemias from haemolytic anaemias by the presence of ‘the small yellow microcytes, perfectly spherical in form and deep in colour which point to an anaemia of haemolytic origin.’

Wilson's cases (Wilson, 1890; Wilson & Stanley, 1893) almost certainly represent hereditary haemolytic anaemia. Six consanguinous relatives in four generations of a family exhibited pallor, enlarged spleens and icterus. Several of them described attacks of chills, fever and increasing jaundice, reminiscent of the crises of what would later be called haemolytic jaundice. Sir William Gull had seen the propositus, ‘Mrs. A.P.’ and two of her children in 1857; he thought they had ‘ague cake from catching malaria at Worthing.’ Thirty years later, Wilson was less certain, stating (Wilson, 1890), ‘…the more I see and think of these cases, the less do they remind me of malaria.’ Wilson measured RBC counts ranging from 3–4 million and haemoglobin levels ranging from 50–60% in three of the patients. The family tree of Mrs. A. P. (Wilson, 1890) is depicted in Fig 2.

image

Figure 2. The family tree of ‘Mrs. A. P.’Wilson (1890) studied six individuals from four generations of this consanguineous family. The sole surviving member of the family, ‘Mr R. T.’, was reported to have an abnormal osmotic fragility test when studied almost 40 years later (Campbell, 1926). Cases marked with an asterisk show enlarged spleen and sallow complexion.

Download figure to PowerPoint

‘Mrs. T’, the eldest daughter of Wilson's propositus, was sallow from the age of 5 or 6 years and exhibited splenomegaly from age 7 years. She died from complications of pregnancy and childbirth at age 34 years. Permission for a limited post-mortem examination was granted. The spleen weighed almost 2 lbs and measured 6·5 by 5 inches. Microscopically there was thickening of the capsule and trabeculae and engorgement with RBCs. The gall bladder was forgotten in the rush to complete the examination, a charming, but unfortunate, omission, given her history of occasional attacks of jaundice and pain that had been attributed to biliary concretions. Wilson attributed her death to ‘rapidly progressing anaemia, dependent upon an active haemolysis of splenic origin.’ By this time, Wilson was firmly convinced that his patients were not suffering from malaria. He thought that all the patients in this family were suffering from the same disease in spite of the differences in clinical history and pointed out that near relatives of these patients could suffer unknown from the condition.

Thirty-five years later, Dr Wilson introduced the only survivor of his six patients to Dr J. M. H. Campbell (1926). ‘Mr R. T.’ was the son of ‘Mrs. T.’ and the grandson of Wilson's propositus, ‘Mrs. A. P.’ Mr R. T. was 44 years old at the time and an active tennis player. He was deeply jaundiced and the spleen projected more than 4 inches below the costal margin. The haemoglobin measured 80%, the plasma indirect Van den Bergh test was positive and the urine contained urobilin, but no bile pigment. Most importantly, the RBC exhibited increased fragility in hypotonic saline, confirming the diagnosis of familial acholuric jaundice.

Minkowski (1900, 1905) reported important observations of two families with familial jaundice, splenomegaly and urobilinuria. The affected members of these families were in otherwise good health and one even became a physician. One of the patients died of pneumonia and, at post-mortem examination, Minkowski demonstrated the integrity of the liver, pigment gall stones and splenomegaly (1 kg), as well as siderosis of the kidneys. He stated ‘We are concerned with a peculiar inborn affection which is characterized by permanent jaundice, the continued presence of urobilin in the urine, hyperplasia of the spleen and increased iron in the kidneys. It is evidently hereditary in origin and does not appear to affect length of life…this affection constitutes a peculiar anomaly of the metabolism of the blood pigment, perhaps following a primary alteration of the spleen.’ The blood of Minkowski's patient was described as normal; he did not offer an opinion regarding the increased iron in the kidneys.

Minkowski was probably unaware of Wilson's patients, but his clinical descriptions and the post-mortem findings agree closely with the observations of Wilson. Around the same time, Bettman in Heidelberg and Krannhals of Riga (Dreyfus, 1942) published other accounts of anaemia in patients with familial jaundice. Krannhals measured RBC counts ranging from 2·8 to 3·5 million and noted polychromatophilia for the first time. He also commented on the fact that the stools were coloured in spite of the jaundice. He postulated that the RBCs may have been haemolysed by the bile pigments in the serum. Barlow & Shaw (1902) similarly described two patients with familial jaundice and anaemia; they also receive credit for the first description of chronic leg ulcers in this disease.

Osmotic fragility and reticulocytosis During the first decade of the new century, two important discoveries rounded out the clinical picture of haemolytic icterus: increased osmotic fragility of RBCs and reticulocytosis. Credit for priority in their description is generally given to the French investigator, Chauffard (Fig 3) and co-workers. Chauffard (1907) noted that the RBCs of several patients, but not those of normal subjects, were haemolysed by hypotonic saline. Almost 40 years earlier, Malassez (Dreyfus, 1942), who developed fluids for counting blood cells, had observed that RBCs from various patients lysed at different rates in 0·06% saline. In Chauffard's osmotic fragility test, RBCs were placed in a series of tubes containing successively decreasing concentrations of saline, not greatly different from current practice. The osmotic fragility was expressed as the concentration of saline at which haemolysis began and at which haemolysis was complete. Chauffard recognized that the liver was not at fault and that the disorder was as a result of haemolysis. He wrote, ‘Perhaps after this clinical and haematologic inquiry, the cause of the haemolytic theory could be considered as won.’ This observation finally enabled physicians to distinguish hepatic and haemolytic jaundice, as Ribbierre (Dreyfus, 1942) had recently (in 1903) demonstrated that red cells from patients with hepatic jaundice are resistant to osmotic stress.

image

Figure 3. Anatole Chauffard. From the United States National Library of Medicine, History of Medicine Division.

Download figure to PowerPoint

Of course, Chauffard and co-workers had discovered the in vitro pathophysiological expression of the spherical microcytes first described by Vanlair and Masius almost 40 years earlier. They were probably unaware of the work of these early investigators and they certainly made no association between microcytic spherical cells and increased osmotic fragility. That correlation was noted much later by Haden (1934). Naegli is often credited with first use of the term ‘spherocyte.’ However, according to Crosby (1952), two British army officers, Christophers and Bentley, were the first. They were assigned to India to study blackwater fever and made very careful descriptions of spherocytes in a monograph published in 1909. Naegli also proposed that the spherocyte was pathognomonic of congenital haemolytic icterus and was the result of abnormal erythropoiesis in the marrow. Dameshek & Schwartz (1940) wryly note, ‘Like many other statements made by Naegli, this was soon accepted as gospel…’ and the effect was to constrict thinking about haemolytic icterus for the next 15 or 20 years. In fact, many authorities began to doubt the existence of an acquired type of haemolytic icterus, regarding the disease as a variation on the congenital form.

About a year after his description of increased osmotic fragility in congenital haemolytic icterus, Chauffard & Fiessinger (1907) and Chauffard (1908) stained RBCs from patients with haemolytic icterus with Pappenheim's solution and noted large numbers of cells containing a peculiar basophilic granulation or reticulum, which they called ‘granular degeneration.’ Ehrlich (Dreyfus, 1942) had first described this special staining method in 1881 and noted increased numbers of reticular cells in anaemic patients. Vaughan (1903) noted these granular cells constituted about 1% of the RBCs in normal subjects. Chauffard had hoped to explain the anatomical lesion that underlay the increased fragility of the RBCs. What he actually discovered, or rediscovered, was the reticulocytosis that is now a hallmark of haemolytic anaemia. Chauffard's drawing of a blood smear stained with Pappenheim stain from a patient with familial haemolytic icterus (Chauffard, 1908) is shown in Fig 4.

image

Figure 4. Drawing of a blood smear (Pappenheim stain) as seen by Chauffard (1908). The granular appearing cells are reticulocytes from a patient with familial haemolytic icterus.

Download figure to PowerPoint

Acquired cases Georges Hayem (1898)) (Fig 5) is generally given credit for the first description of acquired haemolytic icterus. In his paper entitled ‘Sur une variete particulaire d’ictere chronique', he describes the several known varieties of jaundice and then continues, ‘Among these cases, I will alert you to a chronic jaundice very particular to certain dyspeptics that manifest themselves by a yellow chamois colouration of the skin and most notably of the palms of the hands. Above all, this jaundice has the unique characteristic of the absence of biliary pigment in the urine while blood serum clearly shows Gmelin’s reaction. I have observed several examples.' Hayem documented anaemia and splenomegaly in his patients, but considered the disease an affliction of the liver and placed great emphasis on the digestive disorders for which his patients had initially consulted him. He proposed the name ‘chronic infectious icterus of the paroxysmal type, with splenomegaly.’ His was the first clear description of the acholuric nature of the jaundice.

image

Figure 5. Georges Hayem. From the United States National Library of Medicine, History of Medicine Division.

Download figure to PowerPoint

At about the same time as Chauffard reported his cases of congenital haemolytic icterus, his colleagues Widal et al (1907a, b, c, 1909) (Fig 6) published a series of papers on a haemolytic icterus that was apparently neither congenital nor familial, that could appear gradually or suddenly during the course of various diseases or could be unassociated with any underlying disease. These cases were considered similar to those described by Hayem 10 years earlier. In contrast to Chauffard's patients, whom he described as more icteric than ill (icterique plutot que vraiment malade), Widal's patients were severely anaemic and ill. They were all adults who lacked a family history or past history of icterus. The patients exhibited reticulocytosis, but the alterations in the fragility test were less marked than in the congenital form. Autoagglutination was observed in some of these patients and Chauffard & Vincent (1909) later described isohaemolysins in the sera of two of Widal's patients. For a time thereafter, two types of haemolytic icterus were distinguished, the congenital or familial type of Chauffard and Minkowski (Dr Wilson's precedence notwithstanding), and the acquired type of Hayem and Widal.

image

Figure 6. Fernand Widal. From the United States National Library of Medicine, History of Medicine Division.

Download figure to PowerPoint

Splenectomy The postulated role of the spleen in the pathogenesis of the anaemia led to attempts to cure the disease by splenectomy. Reviews of haemolytic icterus by Tileston (1922), Campbell (1926) and Dameshek & Schwartz (1940) credit Micheli in 1911 with the first published report of a successful splenectomy, in a patient with the inherited form of the disease. However, priority probably goes to Sir Spencer Wells who operated successfully on a patient in 1887, subsequently reported by Dawson in 1914. The patient undoubtedly had familial haemolytic icterus, as two family members were affected, and the patient whose spleen was removed had an abnormal osmotic fragility test when studied by Dawson over 20 years after the operation. Elliot & Kanavel (1915) reported 47 splenectomized cases from the literature and one of their own. Of these, 16 were thought to be acquired cases, 23 were congenital and nine cases were unclassified. Only two patients died, one shortly after surgery, the other 6 weeks after surgery, of sepsis. The 46 surviving patients were relieved of their jaundice and presumed to be cured. Correction of the osmotic fragility test after surgery was the exception rather than the rule. Interestingly, their table of patients showed that only in acquired cases did the osmotic fragility correct after splenectomy, but the authors apparently failed to notice this correlation. Giffen (1917) reported nine curative splenectomies out of 12 attempted. Two patients died of operative complications and one patient with acquired disease initially responded to splenectomy, but relapsed after 2 years. In the last patient, the osmotic fragility test returned to normal during remission, in sharp contrast to the seven cured patients studied after splenectomy, whose osmotic fragility remained abnormal. Mayo (1917) further demonstrated the safety of the procedure, reporting one fatality in 19 operations. Thus, Tileston (1922) was able to recommended splenectomy for patients with congenital or familial haemolytic jaundice, if the symptoms warranted, and for primary cases of the acquired type.

Between the Wars

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References

World War I brought a halt to investigation and case reports of haemolytic icterus. After a hiatus of over a decade, Lederer (1925, 1930) rediscovered the acquired syndrome and published accounts of six cases manifesting acute onset, rapid course, fever, acholuric jaundice, splenomegaly, severe anaemia and leucocytosis. Several of these cases exhibited a remarkable response to blood transfusion. The blood film exhibited reticulocytosis, anisocytosis, poikilocytosis and microcytes. Platelet counts were generally normal. Fragility studies and serology were not done. Brill (1926) described another similar case of his own in which the patient was near death until he received blood transfusions, after which he promptly recovered. He cites a clinically similar case reported by Macintosh and Cleland from Australia in which the patient died without transfusion. Brill thought his case and the Australian case resembled the recently reported cases of Moschkowitz, which the latter called ‘acute pleichromic anaemia with hyaline thrombosis of terminal arterioles and capillaries’, and which we now call thrombotic thrombocytopenic purpura (TTP). Indeed, Moschkowitz thought his cases resembled those of Lederer, although Lederer (1930) disagreed. Lederer was probably right; the constellation of acute haemolysis, fever and remarkable response to blood transfusion could cause one to consider TTP, but the normal platelet counts in his patients militate against the diagnosis.

Strangely, neither Lederer nor Brill were aware of the contributions of the French and German investigators in the early part of the century. Lederer's ‘discovery’ was hailed as a new disease entity, in spite of the extensive review of haemolytic icterus by Tileston (1922), acknowledging the work of the European investigators just 3 years earlier. The cases of Lederer and Brill and other anecdotal reports were accepted as cases of acute haemolytic icterus and, for a time, the eponyms ‘Lederer’ or ‘Lederer–Brill’ were applied to these cases.

Dyke and Young (1938) published a series of cases of ‘macrocytic haemolytic anaemia’ characterized by splenomegaly and haemolytic anaemia with increased indirect bilirubin, increased RBC fragility, reticulocytosis and macrocytosis. Microcytes or spherocytes were not prominent. The patients exhibited either an acute fulminant course, or a protracted relapsing and remitting course. Splenectomy was of doubtful benefit for the four patients in whom it was performed. They distinguished their cases from the congenital cases of Chauffard, in that the latter exhibited prominent microcytosis and a predictable response to splenectomy. Lederer's acute cases were thought to be a subtype of their own. Others applied the eponym ‘Dyke–Young’ to the chronic macrocytic cases to distinguish them from the acute transient cases of the ‘Lederer–Brill’ type. We now understand that they were probably variations of one and the same disease.

Dameshek and Schwartz (1940) published a remarkable review of acquired haemolytic icterus. Ninety-six pages in length with 380 references from the world literature, it is the most extensive review of the subject extant, with the exception of whole books devoted to the topic. They identified 81 articles that described cases fitting their concept of acute (acquired) haemolytic icterus. Based on their own clinical observations of haemolysins in some patients, cases reported in the literature, including those of Chauffard and co-workers, and their own experiments involving injection of varying amounts of haemolytic serum into guinea pigs, they proposed that all cases of haemolytic icterus were as a result of haemolysins. The differences in clinical manifestations, ranging from mild congenital cases to fulminant acute haemoglobinurias, were accounted for by the dose of haemolysin. Figure 7 from their review illustrates their hypothesis.

image

Figure 7. The relationship between concentration of haemolysin and the various clinical syndromes of haemolysis, as conceived by Dameshek and Schwartz (1940). These investigators hypothesized that differences between apparent familial cases and acquired cases of haemolytic anaemia lay primarily in the blood concentration of haemolysin. [Used with permission. Dameshek, W. and Schwartz, S. O., Acute Hemolytic Anemia (Acquired Hemolytic Icterus, Acute Type), Medicine, volume 19, pp 231–327, copyright Lippincott, Williams and Wilkins, 1940.]

Download figure to PowerPoint

Thus, by the end of the 1930s, a clinical syndrome called haemolytic icterus or acholuric jaundice was well recognized. A number of eponyms had been applied to one or another variation of the clinical syndrome, but none gained widespread use. The congenital and frequently familial form was generally accepted. A presumably acquired form was well described, but not so well accepted, despite recent evidence to the contrary. Both types of cases held several clinical and haematological features in common. Both were characterized by clinical icterus and hyperbilirubinaemia without bilirubinuria. The spleen was usually palpable. Anaemia varied from mild to severe. Reticulocytosis and spherocytosis were present and the red cells exhibited increased fragility in hypotonic saline. Distinguishing features of the congenital and acquired forms lay in the presence or absence of family history, age of onset (younger age more probably congenital) and tempo of the disease, more acute types thought to be acquired. Serological studies for haemolysins and agglutinins were almost never reported. Splenectomy was consistently beneficial in the congenital and familial cases, but of inconsistent benefit in cases thought to be acquired. However, exceptions to all these distinctions were well described. What was sorely needed to distinguish congenital and acquired haemolytic icterus, if the latter indeed existed, was an understanding of aetiology and pathogenesis; it was soon to come.

World War II

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References

It was the War and the anticipated need for blood transfusions for large numbers of civilian and military casualties that brought together in Great Britain the talented and imaginative investigators who finally defined congenital and acquired haemolytic icterus. The distinctions between these two disorders are based on studies of RBC survival and on the anti-globulin test.

Studies of RBC survival The first means of studying RBC survival was developed by Winifred Ashby (1919), a predoctoral fellow at the Mayo clinic. At that time, it was known that the RBCs of group I (now group AB) were agglutinated by serum from all other groups and Group I serum had no agglutinating properties for the cells of other groups. RBCs from groups II and III (groups A and B) were mutually agglutinated by each other's sera. Group IV (group O) serum agglutinated the RBCs of the other groups, but Group IV RBCs were not agglutinable by those sera. As it was possible to transfuse patients with RBCs not of their own group, Ashby reasoned it would be possible to measure the lifespan of transfused RBCs by periodically taking specimens of a patient's blood after transfusion, quantitatively agglutinating the patient's cells from the specimen and then counting the remaining cells which were the transfused cells. She applied her method in several subjects and found that the lifespan of transfused RBCs in the circulation extended for 30 d or more.

Dacie & Mollison (1943) first applied Ashby's technique in patients with haemolytic anaemia over 20 years after her publication. They were able to show that normal RBCs transfused into patients with familial haemolytic anaemia survived normally, for approximately 100–120 d. The survival curves from their paper are shown in Fig 8. In sharp contrast, Loutit & Mollison (1946) noted that normal RBCs transfused into patients with acquired haemolytic anaemia exhibited markedly reduced survival. The RBC survival curve in one such patient is shown in Fig 9. As shown, following splenectomy, RBC survival improved.

image

Figure 8. Dacie and Mollison, using the Ashby technique, were the first to demonstrate that normal RBCs survive normally in patients with familial haemolytic anaemia. The figure shows survival of blood from normal donors after transfusion to six patients with familial haemolytic anaemia. Case 3 was an Rh-negative patient who was later found to have developed an alloantibody to Rh (Loutit & Mollison, 1946), accounting for the shortened survival of transfused Rh-positive RBC. Although not shown in the figure, survival in cases 2 and 5 was followed to completion and found to exceed 100 days in each case. The dotted lines indicate the limits of survival in a group of normal recipients (Mollison, unpublished observations). [Figure and legend used with permission. Dacie, J. V. & Mollison, P. L., Survival of normal erythrocytes after transfusion to patients with familial haemolytic anaemia (acholuric jaundice). The Lancet, volume i, May 1, 1943, pp 550–552, copyright by The Lancet Ltd, 1943].

Download figure to PowerPoint

image

Figure 9. Survival of transfused normal RBCs in case 5 of acquired acholuric jaundice. Dotted line, survival graph before remission. Continuous line, survival graph after remission, following splenectomy. Figure and legend reproduced with permission. Loutit, J. F. & Mollison, P.L., Haemolytic icterus (acholuric jaundice), congenital and acquired. Journal of Pathology and Bacteriology, volume 58, pp 711–728, copyright by John Wiley & Sons 1946.

Download figure to PowerPoint

Loutit & Mollison (1946) also transfused RBCs from patients with congenital and acquired haemolytic icterus into normal recipients and followed their survival. The RBCs from patients with congenital acholuric jaundice, including those from a patient who had undergone splenectomy, exhibited short survival. Cells from four patients with acquired disease exhibited normal survival. The RBC survival curves from these patients are shown in Fig 10. It is interesting to note the initial rapid destruction of RBCs followed by a normalization of the slope of the survival curve in one of the patients with acquired disease. As we now know, in acquired haemolytic anaemia, the red cells are coated with antibody. The initial rapid destruction was because of antibody mediated destruction of the cells; as the antibody eluted from the cells, they began to survive normally.

image

Figure 10. Survival in normal recipients of transfused RBCs from three cases of congenital acholuric jaundice (continuous lines) and four cases of acquired acholuric jaundice (dotted lines). Figure and legend reproduced with permission. Loutit, J. F. & Mollison, P. L., Haemolytic icterus (acholuric jaundice), congenital and acquired. Journal of Pathology and Bacteriology, volume 58, pp 711–728, copyright by John Wiley & Sons 1946.

Download figure to PowerPoint

These studies provided important insights into the nature of congenital and acquired haemolytic anaemia. For the first time, decreased RBC survival was demonstrated in both disorders. Furthermore, it was evident that an acquired form of the disease existed, with a different aetiology and pathogenesis from the congenital or familial form. Most importantly, it was shown that red cells from patients with congenital haemolytic icterus were intrinsically defective and those from patients with the acquired form were essentially normal. These findings gave rise to the concept of an extracellular defect in acquired haemolytic anaemia and an intracellular defect in the red cells of patients with congenital haemolytic anaemia. We now know that essentially all inherited haemolytic anaemias are as a result of intrinsically defective red cells (intracellular defect) and the RBCs in all acquired haemolytic anaemias, with the exception of paroxysmal nocturnal haemoglobinuria, are normal, the haemolysis resulting from extracellular causes.

The Ashby technique for measuring red cell survival was soon supplanted by the use of radioactive chromium-labelled cells. The latter technique is superior to the former in that it is less tedious to perform and patient cells may be studied in their own circulation. Studies using radioactive chromium generally confirmed those using the Ashby technique. In the 1950s, several investigators used 51Cr-labelled cells or the Ashby technique to demonstrate pathogenicity of autoantibodies (Hollander, 1954; Crowley & Bouroncle, 1956). Patients with acquired haemolytic anaemia whose autoantibodies exhibited specificity for antigens in the Rh system were studied. When these patients were transfused with their own labelled RBCs or Rh-compatible RBCs, decreased RBC survival was noted. However, RBCs lacking the antigen recognized by the autoantibody survived normally in the patients.

The anti-globulin (Coombs') test Several years before the War, a group of researchers under the auspices of the Medical Research Council began to work in earnest on the genetics of blood groups in the Galton Laboratory housed at University College, London. When the War began, the group was moved to a safer place in the Department of Pathology at Cambridge University, where they became known as the Galton Laboratory Serum Unit. Their charge was to select suitable human donors and to process and distribute their serum for use as anti-A and anti-B blood grouping sera.

At the time of the move, Dr Robert Race was investigating the immunology and genetics of the newly discovered rhesus blood groups. He had recently shown the existence of two forms of anti-Rh. One so-called ‘normal’ (complete) form directly agglutinated D-positive RBC in saline suspension. A variant form, known as ‘incomplete’ antibody, could not directly agglutinate D-positive cells, but when the cells were first incubated in serum containing the incomplete antibody, the cells could no longer be agglutinated by the complete antibody. The incomplete antibody was subsequently called ‘blocking’ antibody.

In 1944, Dr Race was joined in the laboratory by Dr Arthur E. Mourant and by a young veterinarian, Dr Robin R. A. Coombs (Fig 11). Dr Coombs began investigations on the complete and incomplete antibodies with Dr Race. In one of their first collaborative efforts, they observed that RBCs sensitized with either complete or incomplete antibody migrated at the same rate in an electrophoretic field and at a markedly different rate than native RBCs. This suggested to them that the incomplete antibody, like the complete antibody, was a globulin. It was shortly before July 1945 that, according to Mourant (1983), Dr Coombs conceived the principle of the anti-globulin reaction in a brilliant feat of intuition. Dr Coombs (1998) described a late night ride on an ill-lit train from London to Cambridge during which he was pondering the nature of the incomplete antibodies attached to RBCs, with visions of Ehrlich's side-chain theory running through his head. ‘In a flash I could see the globulin antibody on the red cells, and these cells should be agglutinated with an antibody to serum globulin, i.e. an anti-globulin. All the necessary thinking had been done.’ The principle of Dr Coombs' anti-globulin reaction and a representation of Ehrlich's side-chain theory depicting ‘agglutinins’ combining with bacilli, are shown in Fig 12.

image

Figure 11. Robin R. A. Coombs (Photograph by Lawrence E. Young, M.D., Fellows' Garden, Kings College, Cambridge University 1950).

Download figure to PowerPoint

image

Figure 12. Left, Ehrlich's side chain theory, formation of agglutinins and precipitins. Adapted from Kolmer, John A., A Practical Text-book of Infection, Immunity and Specific Therapy, p153, copyright by W. B. Saunders Company, Philadelphia & London 1915. Right, the principle of the direct anti-globulin reaction. Adapted from Packman, C. H. (March 2000) Immune Haemolytic Anaemia, in Embryonic Encyclopedia of Life Sciences, Nature Publishing Group, London: http://http://www.els.net. Dr R. R. A. Coombs was studying Ehrlich's side-chain theory, perhaps visualizing agglutinins combining with bacilli as represented in the left panel, when he suddenly envisaged the principle of the anti-globulin reaction. His first experiments demonstrated the indirect anti-globulin reaction. Subsequently, Coombs et al (1946) and Boorman et al (1946) demonstrated the direct anti-globulin reaction on RBC of patients with haemolytic disease of the newborn and acquired haemolytic icterus respectively.

Download figure to PowerPoint

As luck would have it, Dr Muriel Adair in a nearby laboratory had a stock of rabbit anti-human globulin sera. She generously provided Dr Coombs and his co-workers with a supply of these sera which they absorbed with human group AB Rh-positive RBCs. Rh-positive human RBCs were sensitized with the incomplete form of anti-Rh, washed and incubated with the absorbed rabbit anti-human globulin serum. The sensitized cells agglutinated in the anti-globulin serum and the appropriate controls were negative. The first account of what we now call the indirect anti-globulin test was published by Coombs et al (1945a) in a short paper and in a more complete form later the same year (Coombs et al, 1945b).

Just as the page proofs of the second paper were about to be mailed back to the publisher, Dr Mourant, a considerable linguist according to Dr Coombs, found a paper published in the German literature (Moreschi, 1908), describing enhancement of sensitized RBC agglutination with an anti-serum to serum. Dr Moreschi, who died of smallpox in 1921, thus had already discovered the principle of the anti-globulin test almost 40 years earlier. His work, unknown to the scientific community for all those years, was acknowledged in proof as an addendum. In 1953, Dr Coombs generously acknowledged the work of Dr Moreschi during a lecture in Rome and subsequently in a publication in the Italian journal, l'Informatore Medico. Tabelle 2 from Dr Moreschi's 1908 paper is shown (Table I).

Table I. Tabelle II, translated from Moreschi (1908), demonstrating the principle of the anti-globulin (Coombs) reaction.
Rabbit RBCs Goat immune serum or goat normal serum Rabbit precipitating serumAgglutination with
Immune SerumNormal Serum
  1. Rabbit RBCs were incubated with goat immune serum, washed and incubated with rabbit antibody to goat serum (precipitating serum). The RBCs agglutinated in a dose-dependent manner. The controls, lacking either goat immune serum or rabbit precipitating serum, showed no agglutination

1 cc0·005 cc0·0001 cc00
1 cc0·005 cc0·005 ccScant0
1 cc0·005 cc0·001 ccMarked0
1 cc0·005 cc0·005 ccVery marked0
1 cc 0·005 cc0·01 ccVery marked0
1 cc0·005 cc0·05 ccVery marked0
1 cc0·005 cc0·1 ccVery marked0
1 cc0·1 cc00
1 cc0·01 cc00
2 h room temperature.Cells centrifuged and washed with normal saline2 h, room temperature  

Coombs et al (1946) first applied the direct anti-globulin test to detect in vivo sensitization of cord RBCs from infants with haemolytic disease of the newborn. Cord RBCs from patients agglutinated when exposed to the anti-human globulin reagent, but cells from healthy babies did not agglutinate. Coombs & Mourant (1947) later showed that the protein fraction responsible for sensitization of RBCs was gamma globulin. They accomplished this by inhibiting the anti-globulin test with various serum protein fractions including gamma globulin. It was not until the 1950s that the role of non-gamma fractions of serum, i.e. complement, in RBC sensitization was described.

Drs Barbara Dodd and Kathleen Boorman of the South London Transfusion Centre visited Dr Race's laboratory at Cambridge in 1945. Dr Race described the anti-globulin reaction to them during that visit, before the work appeared in print. They returned to the Transfusion Centre, bearing not only the generous advice of Dr Race, but also a supply of rabbit anti-human serum. According to Dodd (1984), when they told Transfusion Centre director, John Loutit, of their visit, he developed a gleam in his eye. Boorman et al (1946) quickly collected RBCs from 17 patients with familial haemolytic icterus and from five patients with the acquired type. The five patients with acquired haemolytic anaemia all had positive direct anti-globulin reactions, whereas the 17 patients with the familial type had negative tests. They wrote, ‘It is suggested that this agglutination test will discriminate the congenital from the acquired form, and that it indicates that the acquired form is due to a process of immunization, whereas the congenital form is not.’

Thus, it was now possible to distinguish congenital from acquired haemolytic jaundice by means of a simple laboratory test. The idea did not catch on immediately, of course. Sturgis' textbook, Hematology, published in 1948 (Sturgis, 1948), makes no mention of anti-globulin testing or even the earlier transfusion experiments of Dacie and Mollison in the chapter on haemolytic anaemia. But by the early 1950s, the distinction was established in the literature by authorities such as Dacie, Young and Dameshek.

New names New understanding compelled the need for new names. The term ‘hereditary spherocytosis’ was suggested to replace familial haemolytic icterus or congenital acholuric jaundice by the Third Committee for the Clarification of the Nomenclature of Cells and Diseases of the Blood and Blood-forming Organs (1950), at their meeting in 1948. They reasoned that not all patients with this disorder presented with symptoms or signs at birth, nor did all patients have affected family members. Furthermore, not all patients were icteric. The new name nicely emphasized the genetic nature of the disease and its cardinal manifestation, the spherocyte. The usual name for acquired haemolytic anaemia, ‘autoimmune haemolytic disease’ was first used in print by Young et al (1951). Young & Miller (1953) subsequently published a classical summary of the clinical and laboratory differences between hereditary spherocytosis and autoimmune haemolytic disease, based on studies of their own cases.

Conclusion

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References

The men and women who, in the early days, contributed to our understanding of the spherocytic haemolytic anaemias were clinicians in the true sense of the word. They studied at the bedside and in clinical laboratories, using their minds, hands, eyes and ears; their most sophisticated instrument was a microscope. Information transmittal and retrieval were rudimentary at best; if the journals were available, the language was more probably foreign to the reader than not, either French, German or English. They made errors, but they also identified and corrected them, so that by mid-century, a foundation had been laid for the sophisticated studies that were to come. When viewed as a whole, the discoveries in the 75 years from Vanlair and Masius to the mid-twentieth century were a hard act to follow. Nonetheless, the second half of the twentieth century brought important new insights into the spherocytic haemolytic anaemias: increased diagnostic sophistication; identification of the various cryopathic haemolytic syndromes as autoimmune diseases; identification of unusual causes of spherocytosis including Zieve's syndrome, clostridial sepsis and Wilson's disease; the roles of RBC structure and biochemistry, autoantibodies, autoantigens, complement, cold agglutinins, haemolysins and the spleen in shortened RBC survival; therapy of autoimmune haemolytic disease; the concept and mechanisms of drug-immune haemolytic anaemia; RBC structure and its genetic regulation; and mechanisms of autoimmunity. The next 50 years will undoubtedly bring new understandings of pathogenesis at the molecular and genetic level, and new means of treatment, probably involving the sciences of stem cell transplantation and gene replacement therapy.

Acknowledgments

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References

I am grateful to the following individuals for assistance with this project: Lee Kaufmann and her staff at the Information Resource Center of Carolinas Healthcare System and Charlotte AHEC, Charlotte, NC: in particular Larry Keesee, Sarah Lowrance and Cynthia Skorlinski showed skill and persistence at locating and obtaining copies of ancient journal articles. Lisa Springer, M.D., and Anne Elizabeth Gerber, M.A., provided expert translations of German and French respectively. Mary Fultz, Department of Medical Arts, Carolinas Medical Center, adeptly reproduced the figures from the old articles. James Thweatt, Historical Collections, Eskind Biomedical Library of Vanderbilt University, provided the photograph of blood from the article of Vanlair and Masius.

References

  1. Top of page
  2. Early cases
  3. Between the Wars
  4. World War II
  5. Conclusion
  6. Acknowledgments
  7. References
  • Ashby, W. (1919) The determination of the length of life of transfused blood corpuscles in man. Journal of Experimental Medicine, 26, 267281.
  • Barlow, T. & Shaw, H.B. (1902) XXIX.-Inheritance of recurrent attacks of jaundice and of abdominal crises, with hepato-splenomegaly. Clinical Society Transactions, 35, 155163.
  • Boorman, K.E., Dodd, B.E., Loutit, J.F. (1946) Haemolytic icterus (acholuric jaundice) congenital and acquired. Lancet, i, 812814.
  • Brill, I.C. (1926) Acute febrile anemia. A new disease. Archives of Internal Medicine, 37, 245247.
  • Campbell, J.M.H. (1926) Early accounts of alcholuric jaundice and the subsequent history of Wilson's patients. Quarterly Journal of Medicine, 19, 323332.
  • Chauffard, A. (1907) Pathogenie de l'ictere congenital de l'adulte. La Semaine Medicale, 27, 2527.
  • Chauffard, A. (1908) Les icteres hemolytique. La Semaine Medicale, 28, 4952.
  • Chauffard, A. & Fiessinger, N. (1907) Ictere congenital hemolytique avec lesions globulaires. Societe Medicale Des Hospitaux de Paris, 24, 11691178.
  • Chauffard, A. & Vincent, C. (1909) Hemoglobinurie hemolysinique avec ictere polycholique aigu. La Semaine Medicale, 29, 601604.
  • Coombs, R.R.A. (1998) Historical note: past present and future of the antiglobulin test. Vox Sanguinis, 74, 6773.
  • Coombs, R.R.A. & Mourant, A.E. (1947) On certain properties of antisera prepared against human serum and its various protein fractions: Their use in the detection of sensitisation of human red cells with ‘incomplete’ Rh antibody, and on the nature of this antibody. Journal of Pathology and Bacteriology, 59, 105111.
  • Coombs, R.R.A., Mourant, A.E., Race, R.R. (1945a) Detection of weak and ‘incomplete’ Rh agglutinins: a new test. Lancet, ii, 1516.
  • Coombs, R.R.A., Mourant, A.E., Race, R.R. (1945b) A new test for the detection of weak and ‘incomplete’ Rh agglutinins. British Journal of Experimental Pathology, 26, 255266.
  • Coombs, R.R.A., Mourant, A.E., Race, R.R. (1946) In vivo isosensitisation of red cells in babies with haemolytic disease. Lancet, i, 264266.
  • Crosby, W.H. (1952) The pathogenesis of spherocytes and leptocytes (target cells). Blood, 7, 261274.
  • Crowley, L.V. & Bouroncle, B.A. (1956) Studies on the specificity of autoantibodies in acquired hemolytic anemia. Blood, 11, 700707.
  • Dacie, J.V. & Mollison, P.L. (1943) Survival of normal erythrocytes after transfusion to patients with familial haemolytic anaemia (acholuric jaundice). Lancet, i, 550552.
  • Dameshek, W. & Schwartz, S.O. (1940) Acute hemolytic anemia (acquired hemolytic icterus, acute type). Medicine, 19, 231327.
  • Dawson, B. (1914). Proceedings of the Royal Society of Medicine, 7, 8586.
  • Dodd, B.E. (1984) First tests with anti-human globulin on the red cells of patients suffering from haemolytic anaemia. Vox Sanguinis, 46, 183184.
  • Dreyfus, C. (1942) Chronic hemolytic jaundice. A historical study. Bulletin of the New England Medical Center, 4, 122128.
  • Dyke, S.C. & Young, F. (1938) Macrocytic haemolytic anaemia associated with increased red cell fragility. Lancet, ii, 817821.
  • Elliot, C.A. & Kanavel, A.B. (1915) Splenectomy for haemolytic icterus. Surgery, Gynecology and Obstetrics, 21, 2137.
  • Giffen, H.Z. (1917) Hemolytic jaundice: a review of 17 cases. Surgery, Gynecology and Obstetrics, 25, 152161.
  • Haden, R.L. (1934) The mechanism of the increased fragility of the erythrocytes in congenital hemolytic jaundice. American Journal of the Medical Sciences, 188, 441449.
  • Hayem, G. (1898) Sur une variete particuliere d'ictere chronique. La Presse Medicale, 6, 121122.
  • Hollander, L. (1954) Erythrocyte survival time in a case of acquired haemolytic anaemia. Vox Sanguinis, 4, 164.
  • Hunter, W. (1888) Is pernicious anaemia a special disease? Practitioner, 41, 81103.
  • Lederer, M. (1925) Form of acute hemolytic anemia-probably of infectious origin. American Journal of Medical Sciences, 170, 500510.
  • Lederer, M. (1930) Three additional cases of acute hemolytic (infectious) anemia. American Journal of the Medical Sciences, 179, 228237.
  • Loutit, J.F. & Mollison, P.L. (1946) Haemolytic icterus (acholuric jaundice), congenital and acquired. Journal of Pathology and Bacteriology, 58, 711728.
  • Major, R.H. (1945) Classic Descriptions of Disease, 3rd edn. Charles C. Thomas, Springfield.
  • Mayo, W.J. (1917) The relation of the spleen to certain obscure clinical phenomena. Medical Record, 92, 705711.
  • Minkowski, O. (1900) Ueber eine hereditare, unter dem bilde eines chronischen icterus mit urobilinurie, splenomegalie und nierensiderosis verflaufende affection. Verhandlungen Des Kongresses Fur Innere Medizin, 18, 316319.
  • Minkowski, O. (1905) Icterus und leberinsufficienz. Die Deutsche Klinik Am Eingang Des XX Jahrhunderts, 5, 651686.
  • Moreschi, C. (1908) Neue tatsachen uber blutkorperchenagglutination. Zentralbl Bacteriol Parasitenkd Infektkr Originale, 46, 4951.
  • Mourant, A.E. (1983) The discovery of the anti-globulin test. Vox Sanguinis, 45, 180183.
  • Murchison, C. (1877) Clinical Lectures on Diseases of the Liver. William Wood, New York.
  • Sturgis, C.C. (1948) Hematology. Charles C. Thomas, Springfield.
  • Third, Fourth and Fifth Reports of the Committee for Classification of the Nomenclature of Cells and Diseases of the Blood and Blood-forming Organs. (1950) American Journal of Clinical Pathology, 20, 562579.
  • Tileston, W. (1922) Hemolytic jaundice. Medicine, i, 355388.
  • Vanlair & Masius. (1871) De la microcythemie. Bulletin de l'Academie Royal de Medicin Belgique, 3e Series, 5, 515613 .
  • Vaughan, Jr, V.C. (1903) On the appearance of certain granules in the erythrocytes of man. Journal of Medical Research, 10, 342366.
  • Widal, F., Abrami, P., Brule, M. (1907a) Differenciation de plusiers types d'icteres hemolytiques. La Presse Medicale, 81, 641644.
  • Widal, F., Abrami, P., Brule, M. (1907b) Hemolyse par fragilite globulaire et hemolyse par action plasmatique. Bulletin Societe de Biologie, 63, 346349.
  • Widal, F., Abrami, P., Brule, M. (1907c) Pluralite d'origine des icteres hemolytiques (recherches cliniques et experimentales). Societe Medicale Des Hospitaux de Paris, 24, 13541367.
  • Widal, F., Abrami, P., Brule, M. (1909) Retrocession des symptoms cliniques et des troubles hematiques au cours des icteres hemolytiques acquis. Societe Medicale Des Hospitaux de Paris, 28, 7385.
  • Wilson, C. (1890) Some cases showing hereditary enlargement of the spleen. Transactions of the Clinical Society of London, 23, 162172.
  • Wilson, C. & Stanley, D. (1893) XXXIV. A sequel to some cases showing hereditary enlargement of the spleen. Transactions of the Clinical Society of London, 26, 163171.
  • Young, L.E. & Miller, G. (1953) Differentiation between congenital and acquired forms of hemolytic anemia. American Journal of Medical Science, 226, 664673.
  • Young, L.E., Miller, G., Christian, R.M. (1951) Clinical and laboratory observations on autoimmune hemolytic disease. Annals of Internal Medicine, 35, 507517.