To the Editor:
Since Kohgo et al1 developed in 1987 the first immunoassay for measuring the levels of truncated transferrin receptor in human serum (TFR) and reported their results in Blood in 1987, Cook et al2-5 have tested this parameter with their own enzyme-linked immunoassay in clinical hematology, particularly in iron deficiency, and have written excellent reviews on the subject. In 1990, Skikne et al6 reported in Blood the changes in TFR levels in healthy volunteers after repeated phlebotomies to induce an “experimental” iron store depletion and functional iron deficiency. They observed that TFR increased when serum ferritin and blood hemoglobin reached subnormal values. It has been shown that with the exception of disorders associated with enhanced erythropoiesis and very few types of neoplasias,7 an increase in TFR could be indicative of iron deficiency (ID).1-6 ,8
We evaluated the TFR levels and the TFR/ferritin ratio in patients in different stages of ID, from the earliest stages to fully expressed iron deficiency anemia, while assessing their usefulness in clinical ferropathology.9 The TFR level was measured in milligrams per liter by enzyme immunoassay with AMGEN Diagnostics reagents (R & D Systems, Minneapolis, MN). The serum ferritin (F ) was measured by immunoturbidimetric assay (Boehringer Mannheim Systems, Barcelona, Spain). Other laboratory data such as complete red blood cell counts which provided hemoglobin (HB), mean corpuscular volume (MCV), and mean corpuscular hemoglobin concentration (MCHC), as well as serum iron, total iron binding capacity, and transferrin saturation (SAT) were obtained by standard procedures.
A single determination of the serum TFR level was carried out in 439 hospital patients (152 males and 287 females, ages 14 to 94 years) who fulfilled at least one laboratory diagnosis criterion of ID. The subjects were classified according to stages of ID as follows. Group 1: 33 patients with depleted iron stores (DIS) with only low F; group 2: 55 patients with iron-deficient erythropoiesis without anemia (IDE), with low F and SAT; group 3: 29 patients with iron-deficient erythropoiesis without anemia but with additional hypochromia (HIDE), microcytosis (MIDE), or both (MHIDE); group 4: 149 patients with ID anemia (IDA) not yet fully expressed, either normocytic (NIDA), hypochromic (HIDA), or microcytic (MIDA); group 5: 173 patients with fully expressed typical microcytic and hypochromic iron deficiency anemia (MHIDA). Group 6 corresponded to 88 healthy controls, mainly staff members (36 males and 52 females, ages 17 to 90 years).
The TFR serum levels are shown in Table 1. Control reference levels ranged between 1.0 and 3.7 mg/L. These values are lower than those previously described with other assay methods,2-7 but they agree with those reported by Kralj et al10 using the same commercial reagents. According to other reports,3 ,5 neither sex-related (ANOVA: P = .523) nor age-related (correlation: r 2 = .085) variations in TFR were observed in the caseload (527 subjects, normals and patients). The TFR levels were different between all groups (ANOVA: P < .001) as were the TFR/F ratios (ANOVA: P < .001 in both sexes separately). No patient showed TFR values below 1 mg/L, the lowest reference value. A variable number of patients from different groups showed raised TFR levels above 3.7 mg/L, the highest reference value: this abnormal increase occurred in 12.1% of patients with DIS, in 18.2% with IDE, in all 4 patients with HIDE, in 41.2% of MIDE, in all 8 patients with MHIDE, in 54.7% of NIDA, in 75.7% of HIDA, in 96.8% of MIDA, and in all 173 patients with MHIDA. We observed a clear tendency toward a progressive increase in TFR levels, parallel to the peripheral erythroid ID changes. Only in the patients from group 5 (MHIDA) were the TFR levels significantly higher than those of the other groups, and only in the patients from group 4 did the levels significantly exceeded those of the control group 6 (Scheffé's test, in both analyses P ≤ .05).
The results of TFR/F ratio ranged from 7 to 57 and 75 to 24,000, respectively, in normal and in all ferropenic males, and from 11 to 257 and 117 to 30,000, respectively, in normal and in all ferropenic females. The mean TFR/F ratio was higher for females than for males, whether controls or patients (ANOVA: P = .0206). Just like the TFR levels, the ratio showed a similar trend, with values ranging between 75 and 280 and 239 and 24,000, respectively, for males of group 1 and 5, and between 117 and 1,600 and 409 and 30,000 for females of the same groups. However, only for females from group 5 was the ratio significantly higher than that for those from the other groups, whereas only for males from group 5 was it significantly higher than that for those from groups 1, 4, and 6 (Scheffé's test, in both analyses P ≤ .05); this difference could be attributed to the low volume of samples for males from group 2 (n = 11) and group 3 (n = 6). For all male patients the TFR/F ratio exceeded the highest reference value for male controls and was less consistent in females.
We conclude that, in practice, the increase of TFR level does not exceed the diagnostic value of a low serum F, either alone or with the typical diminution of SAT, HB, MCV, or MCHC. However, TFR serum levels are abnormally high in some patients with early ID during depletion of stores and plasma iron before any objective alteration is observed in the peripheral erythroid cells. The levels show a continuous increase accompanying the progressive expression of erythroid cell alterations. They exceed the highest reference value in 100% of patients with fully expressed ID (group 5), with microcytic and hypochromic ID anemia. Therefore, laboratory determination of TFR is superfluous in this type of patient. Nevertheless, the TFR level could be of diagnostic value in patients in the intermediate stages of ID, when the ferritin serum levels are equivocal, especially in young females and in older patients of both sexes. They could also be useful in differentiating between iron deficiency and other iron-related erythroid disorders. The TFR/F ratio exceeds the highest control level in males in all stages of ID. Further clinical evaluation of TFR levels and of TFR/F ratio is warranted.
- Copyright © 1997 American Society of Hematology