Under normal circumstances erythrocytes disappear from the blood because of senescence; their potential life span is limited by some process depending on age, possibly exhaustion of a necessary cell constituent, for instance catalase,96 or perhaps something resulting in a sharply inscreased mechanical fragility.112 As in other biologic phenomena, potential life spans of individual cells are distributed around a mean. In many cases, particularly in animals and in patients with certain types of anemia, there is an additional destructive process removing a definite per cent of cells daily, at random, and without regard for age or other characteristic. This decreases the average survival time of the cells without affecting their potential life span which is apparently an inborn characteristic. Only in the absence of random destructions is the average survival time equal to the average potential life span; otherwise it is shorter, but there is no simple relationship between the degree of shortening and the intensity of the destructive factor. Extrinsic agencies producing random destruction have been demonstrated in the serum of patients with acquired hemolytic anemia, nephritis, the anemias associated with lymphomas and other malignancies, and there is some evidence also for their presence in pernicious anemia. Instrinsic or inborn defects can also result in random destruction as in nocturnal hemoglobinuria, sickle cell anemia, and congenital hemolytic jaundice as well as certain cases of elliptocytosis. In other instances the potential life span is shortened leading to accelerated senescence; this probably occurs in pernicious anemia and may in the future be found elsewhere.
Distribution of life spans around a mean leads to the typical S-shaped curve seen when erythrocytes of approximately the same age are followed by tagging with some isotope, preferably radioiron. Reutilization of the isotope occurs in all such experiments, and this, as well as random destruction, has obscured the S-shaped portion, but it can usually be demonstrated. The rates of these processes can be estimated separately and values obtained for the average potential life span and the spread of individual values around it, as well as a measure of the intensity of the random destructive factor.
Errors in Ashby curve experiments are due in the first place to inadequate experimental methods. No experimental procedure can be considered satisfactory unless it is shown to give accurate counts of known mixtures of agglutinable and nonagglutinable cells at all ratios of these cells. Its the second place nonagglutinable cell counts are always obtained with reference to some magnitude of the recipient—blood volume or red cell mass. Since the blood volume is variable and since the hematocrit is different in different parts of the circulation, the red cell mass is the reference of choice, but long term changes in it should he determined and corrected for.
In the absence of random destruction, Ashby curves are linear; otherwise they show curvature, but their interpretation is much more difficult than are isotope experiments which must be relied on to establish the exact nature of the disappearance curves.