Laboratory Diagnosis of Anemia: Are the Old and New Red Cell Parameters Useful in Classification and Treatment, How?

International Journal of Laboratory Hematology The Ofﬁcial journal of the International Society for Laboratory Hematology

REVIEW ARTICLE INTERNATIONAL JOURNAL OF LABORATORY HEMATOLOGY

Laboratory diagnosis of anemia: are the old and new red cell parameters useful in classiﬁcation and treatment, how? M. BUTTARELLO

Clinical Pathology Laboratory, SUMMARY Hospital of Adria, Adria (RO), Italy Introduction: Anemia is a global problem affecting the population in both developing and developed countries, and there is a debate on correspondence: Mauro Buttarello, Clinical which hemoglobin level limit should be used to define anemia in Pathology Laboratory, Hospital general population and particularly in the elderly. We present of Adria, ULSS 19 Piazza degli herein a laboratory approach to diagnosing the possible causes of Etruschi, 45011 Adria (RO), anemia based on traditional and new erythroid parameters. In this Italy. Tel.: 00393407719537; E-mail: [email protected] article, we provide practical diagnostic algorithms that address to differential diagnosis of anemia. Based on both morphological and kinetic classifications, three patterns were considered: microcytic, doi:10.1111/ijlh.12500 normocytic, and macrocytic. Methods: Main interest is on the clinical usefulness of old and new accepted for publication 4 April parameters such as mean cell volume (MCV), red blood cell distri- 2016 bution width (RDW), hypochromic and microcytic erythrocytes, Keywords immature reticulocyte fraction (IRF), and some reticulocyte indices Anemia, red cell indices, reticu- such as reticulocyte hemoglobin content and mean reticulocyte locytes, immature reticulocyte volume. The pathophysiologic basis is reviewed in terms of bone fraction, reticulocyte hemoglo- marrow erythropoiesis, evaluated by reticulocyte count (increased bin content, mean reticulocyte volume or normal/decreased) and IRF. The utility of reticulocyte indices in the diagnosis of iron-deficient erythropoiesis (absolute or functional) and in monitoring of response to treatment in nutritional anemia (iron and cobalamin) was also investigated. Results: For each parameter, the availability, the possible clinical applications, and the limitations were evaluated. A discussion on intraindividual biological variation and its implication on the usefulness of conventional reference intervals and in longitudinal monitoring of the patients was also reported. Conclusion: Red cell parameters and reticulocyte indices play an essential role in differential diagnosis of anemia and in its treatment. More efforts are needed in harmonizing parameters whose results are still too different when produced by different analyzers.

underestimate Hct in hypochromic red cells. In this last INTRODUCTION condition, the use of Hct rather than the more accurate Anemia is when blood hemoglobin (Hb) concentra- measured Hb can overestimate the diagnosis of anemia tion is below the lower limit of the reference interval in subjects with iron deficiency (7). Also, the optical- stated for age, sex, race, and altitude. The commonly based instruments with isovolumetric sphering provide a accepted lower limits for adult population are the falsely elevated Hct when they analyze sickle red cells WHO criteria suggested by an expert committee that cannot be sphered. A typical artifactual dissociation nearly 50 years ago: 130 g/L in men and 120 g/L in with all the automated analyzers between Hb result women, without the distinction between age and race (usually correct) and Hct (underestimated) is the pres- (1). The definition of anemia has attracted interest in ence of red blood cell (RBC) agglutinates. Because the recent years because epidemiologic studies suggest upper volumetric threshold to consider cells as RBC is that anemia may be associated with poor prognosis in between 200 and 300 fL according to the analyzer, the many different diseases, particularly among aged peo- large RBC clumps are not counted as RBC. This causes a ple. A recent large population survey based on WHO spuriously low RBC count and low Hct. In contrast, Hb criteria (NHANES-III) (2) showed that nearly ten per- is measured after RBC lysis and is unaffected by agglu- cent of men and woman older than 65 years were tinins. As a consequence, MCHC is abnormally high, anemic. These percentages rose to 26% in males and usually greater than 360 g/L. Moreover, also Hb can be 20% in females older than 85. It is not clear whether erroneously overestimated, although more rarely, in the difference in lower limits, justified in androgen- subjects with severe hypertriglyceridemia or receiving dependent age, should be continued after 65 years of an intravenous administration of fat emulsions, or with age. Many of these subjects were apparently healthy, high WBC counts, due to the excessive turbidity. and in most cases, clinical investigations did not uncover a specific cause of anemia. These results sug- OLD ERYTHROCYTE INDICES gest that somewhat lower limits than ‘normal’ might be used in the elderly. Nevertheless, the too easy Maxwell Wintrobe 80 years ago proposed the anemia acceptance of mild anemia as physiologic in the classification based on the mean cell volume (MCV) elderly runs the risk of ignoring an underlying dis- obtained by Hct/RBC ratio from the measurement of ease. There is a debate on which hemoglobin lower spun Hct and manual hemocytometric RBC count. The limit should be used to define anemia in general pop- MCHC was calculated as Hb/Hct ratio, where Hb was ulation and particularly in the elderly (3). Two differ- also based on manual measure (8). Of these two ent, relatively recent, large databases (NHANES-III indices, which allowed to classify the anemia as micro- and Scripps-Kaiser) (4, 5) in which the hemoglobin cytic, normocytic, and macrocytic based on MCV value determination was carried out with standardized auto- and hypochromic, normochromic or hyperchromic mated methods obtained a good agreement and new based on MCHC, only MCV has survived as key lower limits are proposed (6). It would seem that parameter for the classification of anemia with auto- these limits (5% of normal distribution) are 137 g/L mated hematology analyzers (Figures 1–3). With the in white men (20–59 years) and 132 g/L for men after data collected on a cell-by-cell basis, modern instru- the age of 60; the corresponding value for women is ments generate a histogram of erythrocytes size distri- 122 g/L independently of age. In Afro-Americans, bution. From this histogram, an index of these limits are lower: 129 g/L in younger men and heterogeneity referred to as red cell distribution width 127 g/L in men older than 60, while the correspond- (RDW) can be determined. This is almost always ing value for women is 115 g/L at all ages. expressed as percentage coefficient of variation and, For many practical approaches, a decrease in hemat- less frequently, as standard deviation. The possibility ocrit (Hct) is considered equivalent to a decreased of a quantitative, nonsubjective measurement of an hemoglobin concentration, but this simplification is not anisocytosis index has rewakened interest. Bessman always correct. All hematology impedance-based analyz- et al. (9) in the early 1980 proposed a classification of ers falsely overestimate Hct in erythrocytes with a high anemia based on both MCV and RDW: homogeneous mean hemoglobin concentration (MCHC) and (with normal RDW) and heterogeneous (with

MICROCYTIC ANEMIA NORMOCYTIC ANEMIA (MCV < 80 fL) (MCV 80 - 100 fL)

Reculocyte count Reculocyte count (20 - 100 X 109/L) (20 - 100 X 109/L)

Normal or decreased Increased Normal or decreased Increased

Reculocyte hemoglobin % Hypochromic RBC Reculocyte hemoglobin - Anemia of renal Markers suggesve of Markers not suggesve of content / reculocyte volume content / reculocyte volume insuﬃciency hemolysis: Haptoglobin, hemolysis (< 28 pg / < 100 fL) LDH, bilirubin (< 28 pg / < 100 fL) % Microcyc RBC - Anemia of chronic diseases - Acute infecons - Primary bone marrow disorders - Aplasc anemia Decreased Increased Decreased Peripheral blood smear Blood loss

- βthalassemia trait - Early response to iron - Iron deﬁciency treatment Other ﬁndings (not Fragmented erythrocytes - Chronic diseases - βthalassemia diagnosc): consider - Membrane defect: PNH - Enzymopathy (G6PDH, PK) - Hemoglobinopathy Microangiopathy Spherocytosis % Micro / % Hypo rao

Direct Coombs test posive Direct Coombs test negave Decreased Increased - Autoimmune hemolyc - Hereditary spherocytosis anemia - Cold agglunin disease Iron deﬁciency β-thalassemia trait

Biochemical markers: HbA2 /hemoglobin analysis ferrin, %TSAT, … Figure 2. Diagnostic algorithm for normocytic anemia

- Absolute iron deficiency - Funconal iron deficiency The mean hemoglobin content (MCH), which is - Chronic diseases strongly correlated with MCV, is calculated as Hb/RBC ratio. These last measurements with automated ana- Figure 1. Diagnostic algorithm for microcytic anemia. lyzers are more accurate and precise than MCV, which is derived by a direct measurement of a single cell size using different analytical methods (impedance increased RDW) erythrocyte population. The former with or without hydrodynamic focusing, or light scat- includes hypoproliferative anemia, marrow aplasia, tering). Moreover, MCV, different from MCH, is and thalassemia heterozygosity; the latter includes affected by preanalytical variables such as storage nutritional anemias (iron, cobalamin, and folic acid temperature and time. A dissociation between MCV deficiency) and sideroblastic anemia. Although this and MCH was recently described (13): high MCV and approach was largely accepted and RDW was added to low MCH. This was the case of macrocytic hypochro- routine analysis in many laboratories, numerous mic anemia, indicating the coexistence of both exceptions began to be observed. There is a wide distri- macrocytosis due to cobalamin/folate deficiency and bution of RDW values within a given disease, whose hypochromia due to hemoglobin E disease. Inappro- usefulness in differential diagnosis has decreased, but priately low MCH with a high MCV can be found also its utility as a general marker of abnormality has been in anemia due to B12 (or folate) and iron deficiency maintained. A further complication derives from the (or thalassemia). difference in reference intervals obtained with analyzers from different manufacturers (10–12). This is NEW RED CELL PARAMETERS explained by the different algorithms used to cut the tails of distribution, which is needed to eliminate Some hematology analyzers can quantitate the per- extreme values often due to artifacts. centage of hypochromic, microcytic, and more rarely

MACROCYTIC ANEMIA hyperchromic cells is useful in the diagnosis of sphe- (MCV > 100 fL) rocytosis either by hereditary or by immune hemolysis (21, 22). The main limit in the use of these Reculocyte count (20 - 100 X 109/L) parameters is due to the fact that they are affected by

Normal or decreased Increased temperature and storage time. In fact, erythrocytes in the samples stored at room temperature tend to pro- Immature reculocyte fracon Immature reculocyte fracon gressively swell, with a consequent reduction in cellular hemoglobin concentration and an increase in

Normal or decreased Increased Increased hypochromic and decrease in hyperchromic RBC%.

- Myelodysplasc - Hemorrhage EVALUATION OF BLOOD SMEAR syndromes - Hemolysis - Acute infecons Peripheral blood smear - Response to B12 / folate treatment All the red blood cell indices, although useful, are the representation of the mean and overall dispersion of - Round macrocytes - Oval macrocytes - No hypersegmentaon of - Hypersegmentaon of the erythroid cellular population and provide little neutrophils neutrophils information about speciﬁc red blood cell shapes or the presence of minor populations of abnormal cells. Nonmegaloblasc Megaloblasc macrocytosis macrocytosis Examination of blood smear for some speciﬁc shapes as reported in Figures 2 and 3 can provide a valuable

- Hypothyroidism information to aid in the diagnosis of the underlying -B or folate deﬁciency - Myelodysplasc syndromes 12 - Therapy (hydroxyurea, - Liver diseases disease. zidovudine, methotrexate, - Alcohol chronic abuse ..)

RETICULOCYTE AND IMMATURE Figure 3. Diagnostic algorithm for macrocytic anemia. RETICULOCYTE FRACTION

In addition to red cell indices and morphological crite- hyperchromic erythrocytes (Table 1). In iron-deficient ria, anemia may be classified by kinetic approach, that erythropoiesis, a greater fraction of RBC is hypochro- is, the degree of bone marrow response evaluated by mic rather than microcytic, and the microcytic/hypo- the reticulocyte count. The reticulocyte count is clini- chromic ratio shows the best diagnostic efficiency in cally important both for the pathophysiological classi- the differential diagnosis with beta-thalassemia trait fication of anemia (due to an inadequate production (14, 15). Many other discriminant algorithms have of erythrocytes by the bone marrow, in which case been recently proposed including conventional and there is a decreased number of reticulocytes, or to an new RBC parameters (15, 16). The results were in excessive loss or the destruction of erythrocytes, in general better than the traditional discriminant func- which there is an increase in reticulocyte count) and tions, but the performance of any index seems to for the early identification of the normalization of ery- depend on the geographical origin of the population thropoiesis by the marrow after therapeutic interven- in which it is applied (17). Recent studies have shown tion (iron, cobalamin, folic acid, ESAs, etc.), after that hypochromic erythrocytes are useful in identify- spontaneous or pharmacologically induced aplasia of ing iron-restricted erythropoiesis in anemic patients the marrow, or following bone marrow transplanta- treated with erythropoiesis-stimulating agents (ESAs), tion. However, the imprecision of the manual micro- particularly anemia of the chronic kidney disease or scopic method (coefficient of variation (CV) between in hemodialyzed patients. The response to ESAs is 68.6% at low concentration and 16% at high level) strictly dependent on iron availability and is limited (23) makes it almost useless mainly in severe reticulo- by iron deficiency that can be absolute or functional cytopenia. It does not allow for the observation of (i.e., limitation of bone marrow erythropoietic activity small but significant variations during the early recov- by the inability to mobilize the sufficient iron from ery of erythropoietic bone marrow activity, nor does body storage sites) (18–20). The presence of it clearly define the difference between normal and

Table 1. Old and new RBC parameters and their clinical applications

Proposed clinical Parameter Availability applications Limitations References

Old RBC parameters n.Jl a.Hem. Lab. Jnl. Int.

• Mean cell volume All the analyzers Anemia classiﬁcation Affected by preanalytical 8, 45 (MCV) (fL) based on morphological variables (storage approach temperature, time) • Mean cell hemoglobin All the analyzers Useful when hemoglobin Highly correlated with MCV 13, 46 content (MCH) (pg) synthesis is impaired as in iron deﬁciency anemia • Mean cell hemoglobin All the analyzers Increased in spherocytosis With some impedance 7, 47 concentration (MCHC) (g/ because of a reduced analyzers, the value is L) surface/volume ratio clamped around the mean • Red cell distribution All the analyzers Generic marker of Of little usefulness in the 9–12 width (RDW) (%) abnormality when differential diagnosis of increased anemia. Reference intervals are method dependent

New RBC parameters BUTTARELLO M.

• Percentage of Hypo% (Siemens Advia Assessment of iron Affected by preanalytical 18, 19, 48, 49 hypochromic red cells 2120); %Hypo-He availability (absolute or variables (storage (Sysmex XE/XN); % functional) for temperature, time). HPO (Abbott erythropoiesis. Related Reference intervals and AOAOYDANSSO ANEMIA OF DIAGNOSIS LABORATORY | Sapphire); LHD% to iron status in the last diagnostic thresholds are (Beckman-Coulter 3 months. method dependent. Limited LH/DxH 800). value in the presence of b- thalassemia. • Percentage of Hyper% (Siemens Diagnosis of hereditary/ Reference intervals and 21, 22 hyperchromic red cells Advia 2120); % immune spherocytosis diagnostic thresholds are Hyper-He (Sysmex method dependent. XE/XN); %HPR (Abbott Sapphire).

(continued) 5 6 .BUTTARELLO M.

Table 1. (Continued)

Proposed clinical ANEMIA OF DIAGNOSIS LABORATORY | Parameter Availability applications Limitations References

• Percentage of microcytic Micro% (Siemens Useful in combination Reference intervals and 14–17 red cells Advia 2120); % with other RBC diagnostic thresholds are micro-R (Sysmex XE/ parameters (mainly method dependent. XN); %MIC (Abbott hypochromic Sapphire). erythrocytes) to obtain discriminant indices for the differential diagnosis of microcytic anemia. Reticulocyte parameters

• Immature reticulocyte All the analyzers Classiﬁcation of anemias Reference intervals and 25–35 fraction (IRF) (fraction) and monitoring of diagnostic cutoff are treatment. Verify aplastic method dependent. anemia. • Reticulocyte mean CHr (Siemens Advia Diagnosis of iron-deﬁcient Limited value in the 35–38 hemoglobin content (pg) 2120); Ret-He erythropoiesis. Early presence of a-orb- (Sysmex XE/XN); monitoring the response thalassemia. MCHr (Abbott to iron therapies. Sapphire); RHE (Mindray BC 6800); RHCc (ABX-Horiba

© Pentra Nexus DX) 06Jh ie osLtd, Sons & Wiley John 2016 • Mean reticulocyte MCVr (Siemens Advia Diagnosis of iron-deﬁcient Affected by preanalytical 34, 37, 40 volume (fL) 2120); MCVR (Abbott erythropoiesis. Early variables (storage Sapphire); MVR monitoring of treatment temperature, time). (Mindray BC 6800); with B12 /folate/iron in Reference intervals strictly MRV (ABX-Horiba nutritional anemia. method dependent. Pentra Nexus DX); MRV (Beckman- Coulter LH/DxH 800). n.Jl a.Hem. Lab. Jnl. Int. M. BUTTARELLO | LABORATORY DIAGNOSIS OF ANEMIA 7

low reticulocyte levels. Automated analyzers represent with positive covariance corresponding to the healthy a revolution for this cell type using dyes to bind retic- subjects and to accelerated erythropoiesis; (iii) for ulocyte RNA and flow cytometers to perform rapid marked reticulocytosis, the covariance is negative, and objective counts. The possibility to analyze tens of suggesting a gradual deceleration of erythropoiesis thousands of cells per sample has reduced imprecision (28). Therefore, IRF and reticulocyte count may vary (CV between 25% at low concentration and 3.0% at in a concordant or independent way according to the high counts) (24) (M. Buttarello, personal observa- erythropoietic conditions and can be hypothesized the tions). Furthermore, using an absolute reticulocyte IRF as an index of acceleration and the absolute retic- count (expressed as the number of cells per unit of ulocyte count as a quantitative measure of the effec- volume: x 109/L) rather than proportion no longer tiveness of erythropoiesis (28, 29). This parameter is requires correction for reduced hemoglobin concentra- therefore useful in distinguishing (i) anemia charac- tions. A little complication derives from the differ- terized by an increase in erythropoiesis, like acquired ences in the reference intervals that are strictly hemolytic anemias or the loss of blood, which pro- method dependent (lower limit of the 95% interval duces an increase both in total reticulocytes and in between 19 and 30 9 109/L and upper between 85 IRF; (ii) anemias due to the reduced marrow produc- and 130 9 109/L) (24). What has created an addition (i.e., chronic renal disease) in which both values tional interest about automated reticulocyte analysis is are found to be decreased, and (iii) anemia of acute the availability of a new parameter called immature infections or myelodysplastic syndromes in which reticulocyte fraction (IRF) based on reticulocytes RNA there is a dissociation between total reticulocyte count content (25). Reticulocytes originate from orthochro- (reduced or normal) and the IRF which can be matic erythroblasts following ejection of the nucleus, increased (30–33). Other uses include monitoring the and they gradually mature, partly in the marrow therapy efficacy in nutritional anemia (e.g., cobal- (3 days on average) and partly in the peripheral blood amin, folates, and iron) because the increase in IRF (1 day). Reticulocytes gradually lose their RNA and precedes the increase in total reticulocyte count by ultimately become RNA-free red cells, while some several days. In subjects with iron-deficient anemia RNA-rich, more immature reticulocytes are found in treated with iv iron, it increased at day 1 and contin- relatively narrow proportion in the peripheral blood ued to increase until reaching the maximum value at of the healthy subjects. There are, however, various day 5 (34). This value was correlated with the ery- expressions according to the analyzer used, and thus, thropoietin concentration at the beginning of therapy the reference intervals are different (low level (M. Buttarello, data not published). The combination between 0.012 and 0.20 and high level between 0.14 of reticulocyte count >80 9 109/L with a reticulocyte and 0.40) (26, 27). Independently by the way on as it count/IRF ratio >7.7 is considered useful for the is produced, the IRF is an early and sensitive index of screening of trait and mild hereditary spherocytosis erythropoiesis; in fact, immature reticulocytes appear (35). in a larger proportion when red cell production increases. The IRF has a weak but significantly posi- RETICULOCYTE INDICES tive correlation with the absolute reticulocyte count, indicating that it is an additional useful parameter to The latest generation of hematology analyzers pro- evaluate the erythropoietic activity. The greatest vides some reticulocyte indices analogous to the clinical usefulness, especially in the classification of equivalent RBC indices (Table 1). Among these, anemia based on marrow response, is found using the most promising from a clinical point of view are two-dimensional matrices of IRF vs. the absolute retic- the hemoglobin content of reticulocyte and the mean ulocyte count (25, 26). With covariance analysis (AN- reticulocyte volume. The hemoglobin content, which COVA), it is possible to identify some well-differentiated directly reflects the synthesis of hemoglobin in mar- behaviors in certain areas of the matrices: (i) In retic- row precursors, is a measure of adequacy of iron ulocytopenia, there is no covariance, both in marrow availability (34, 36–38). This parameter is important aplasia and in early erythropoietic response; (ii) in because its reduction indicates iron-deficient erythro- normal or mild reticulocytosis, there are two subsets poiesis, even in conditions in which traditional

150 30 6 INTRAINDIVIDUAL BIOLOGICAL VARIATION

125 One source of variability of clinical laboratory results 25 4 (besides preanalytic and analytic variations) is the 100 intraindividual variation around the homeostatic set- Hb (g/L) CHr (pg) 20 2 ting point (41). Several studies have investigated this 75 RETICULOCYTES (%) biological variation and the results are similar even if carried out at different time points and in different 50 15 0 -1 21 42 63 84 105 126 147 168 189 210 231 252 geographical areas (42–44). The reported values varied

END (IV) IRON DAYS CH r START (OS) IRON THERAPY THERAPY Hb between 1.9 and 2.8% for Hb and Hct; between 0.6 START (IV) IRON RET % THERAPY and 1.3% for MCV; and between 5.8 and 9.5% for Ferr µg/L : 5.7 Ferr µg/L : 3.4 Ferr µg/L : 74 Ferr µg/L : 43 reticulocytes. From these values, an index of individu- Iron µmol/L : 3.58 Iron µmol/L : 3.22 Iron µmol/L : 12.0 Iron µmol/L : 8.24 TRF g/L: 3.71 TRF g/L: 3.50 TRF g/L: 3.41 TRF g/L: 2.73 ality as ratio of intraindividual to interindividual coefﬁcient of variation can be calculated. For the Figure 4. Response in hemoglobin concentration (Hb), above-mentioned parameters, it is between 0.19 and reticulocyte hemoglobin content (CHr), and 0.42 for Hb, between 0.17 and 0.27 for MCV, and reticulocyte% (Siemens Advia 120 analyzer) to between 0.18 and 0.30 for reticulocytes, according to intravenous (iv) iron administration in a patient the different studies. A low index of individuality unresponsive to oral iron therapy. Ferr: serum < ferritin concentration; Iron: serum iron ( 0.5) indicates that conventional reference intervals concentration; TRF: serum transferrin concentration. may be of little usefulness, especially when deciding if the change observed in a subject is clinically

biochemical markers such as ferritin and transferrin 120 saturation are inadequate (e.g., in inﬂammations or 40 150 5 anemia from a chronic disease), and besides, it is use- MCVr 110 CHr 140 ful for monitoring early response to intravenous iron RET % 4 35 Hb g/L therapy because it increases signiﬁcantly after only 100 – 130 48 72 h (Figure 4) (34, 36). Exceptions are heterozy- 3 Hb (gL)

90 gotes for beta-thalassemia whose reticulocyte hemo- 30 120 MCVr (fL)

globin content is found to be always reduced CHr (pg) 2 independently of iron stores (39). Low values of this 110 80 RETICULOCYTES (%) 25 index are indicative even in functional iron deﬁciency 1 100 which appears in patients treated with erythropoietin 70

(38). Few studies are available on the clinical useful- 20 90 0 60 -5 0 5 10 15 ness of mean reticulocyte volume. In subjects with DAYS depleted iron stores, this index increases rapidly fol- START OF IRON (IV) THERAPY START OF COBALAMIN (IV) THERAPY lowing iron therapy and decreases equally as rapidly with the development of iron-deficient erythropoiesis. The reticulocyte volume decreases dramatically and Figure 5. Changes in mean reticulocyte volume reticulocytes are smaller than the circulating RBCs, in (MCVr), reticulocyte hemoglobin content (CHr), reticulocyte % and hemoglobin concentration (Hb) nutritional macrocytosis after therapy with vitamin (Siemens Advia 120 analyzer) to intravenous B12 and/or folic acid (Figure 5) (37). The main limit cobalamin administration, followed by intravenous of the use of these indices is the difficulty to compare iron due to the functional iron deficiency numeric results obtained from the analyzers of differ- development. MCVr and CHr decreased significantly ent manufacturers (the lower limit of the 95% inter- after iv cobalamin therapy. The excessive decrease in CHr suggests an iron deficiency, promptly corrected val between 91 and 100 fL and the upper limit by iv iron administration. between 111 and 120 fL) (40).

remarkable. Useful in serial monitoring of physiologic CONCLUSIONS or pathologic conditions is the critical difference (called also reference change value: RCV), which Red cell parameters and reticulocyte indices play an defines the percentage change that should be essential role in the differential diagnosis of anemia exceeded (given the analytic and intraindividual bio- and in its treatment. More efforts are needed in har- logical variations) so that there is a significant differ- monizing parameters whose results are still too difference between two consecutive measurements. The ent when produced by different analyzers. Moreover, significant percentage changes (for probabilities of it should be remembered that despite the essential 95%) for the named red cell parameters are (43, 44) role of automation, microscopic control of pathologic as follows: between 6.22 and 6.82% for Hb and Hct, samples remains indispensable. between 2.35 and 3.12% for MCV, and between 36.7 and 41.7% for reticulocytes. Differences depend on COMPETING INTERESTS biological variation (42–44) and on the analytical variability of analyzers, and these differences are The author has no competing interest. smaller for instruments with smaller imprecision.

REFERENCES the red corpuscles. Arch Intern Med Hernandez Martinez P, Bartels PCM. Effi- 1934;54:256–80. cacy of advanced discriminating algorithms 1. Blanc B, Finch CA, Hallberg L, Herbert V, 9. Bessman JD, Gilmer PRjr, Gardner FH. for screening on iron-deficiency anemia Lawkowicz W, Layrisse M, Mollin DL, Improved classification of anemias by MCV and b-thalassemia trait. A multicenter Rachmilewitz M, Ramalingaswami V, San- and RDW. Am J Clin Pathol 1983;80: evaluation. Am J Clin Pathol 2012; chez-Medal M, Wintrobe MM. Nutritional 322–6. 138:300–4. anaemias. Report of a WHO Scientific 10. Van den Bossche J, Devreese K, Malfait R, 17. Hoffman JJML, Urrechaga E, Aguirre U. Group. World Health Organ Tech Rep Ser Van de Vyvere M, Wauters A, Neels H, De Discriminant indices for distinguishing 1968;405:1–40. Schouwer P. Reference intervals for a com- thalassemia and iron deficiency in 2. Guralnik JM, Eisenstaedt RS, Ferrucci L, plete blood count determined on different patients with microcytic anemia: a meta- Klein HG, Woodman RC. Prevalence of automated haematology analysers: ABX analysis. Clin Chem Lab Med 2015;53: anemia in persons 65 years and older in Pentra 120 retic, Coulter Gen-S, Sysmex SE 1883–94. the United States: evidence for a high rate 9500, Abbott Cell Dyn 4000, and Bayer 18. Buttarello M, Pajola R, Novello E, Rebes- of unexplained anemia. Blood 2004;104: ADVIA 120. Clin Chem Lab Med 2002;40: chini M, Cantaro S, Oliosi F, Naso A, Ple- 2263–8. 69–73. bani M. Diagnosis of iron deficiency in 3. Schrier SL. Hematology, ASH, and the ane- 11. Buttarello M, Plebani M. Automated blood patients undergoing hemodialysis. Am J mia of the aged. Editorial. Blood 2005;106: cell counts. State of the art. Am J Clin Clin Pathol 2010;133:949–54. 3341–2. Pathol 2008;130:104–16. 19. Macdougall IC, Cavill I, Hulme B, Bain B, 4. Cheng CK, Chan J, Cembrowski GS, Van 12. Lippi G, Pavesi F, Bardi M, Pipitone S. Lack McGregor E, McKay P, Sanders E, Coles Assendelft OW. Complete blood count ref- of harmonization of red blood cell distribu- GA, Williams JD. Detection of functional erence interval diagrams derived from tion width (RDW). Evaluation of four iron deficiency during erythropoietin treat- NHANES III: stratification by age, sex, and hematological analyzers. Clin Biochem ment: a new approach. Br Med J 1992;304: race. Lab Hematol 2004;10:42–53. 2014;47:1100–3. 225–6. 5. Waalen J, Felitti VJ, Gelbart T, Ho NJ, 13. Hazarika B. What is missed in an auto- 20. WayneThomas D, Hinchliffe RF, Briggs C, Beutler E. Penetrance of hemochromatosis. mated count? (Images in Hematology). Macdougall IC, Littlewood T, Cavill I, on Blood Cell Mol Dis 2002;29:418–32. Blood 2012;120:3374. behalf of British Committee foe Standards 6. Beutler E, Waalen J. The definition of ane- 14. d’Onofrio G, Zini G, Ricerca BM, Mancini in Haematology. Guideline for the labora- mia: what is the lower limit of normal of S, Mango G. Automated measurement of tory diagnosis of functional iron deficiency. the blood hemoglobin concentration? Per- red blood cell microcytosis and hypochro- Br J Haematol 2013;161:639–48. spective Blood 2006;107:1747–50. mia in iron deficiency and beta-thalassemia 21. Gilsanz F, Ricard P, Millan I. Diagnosis of 7. Mohandas N, Clark MR, Kissinger S, Bayer trait. Arch Pathol Lab Med 1992;116:84–9. hereditary spherocytosis with dual-angle C, Shohet SB. Inaccuracies associated with 15. Urrechaga E. Red blood cell microcytosis differential light scattering. Am J Clin the automated measurement of mean cell and hypochromia in the differential diag- Pathol 1993;100:119–22. hemoglobin concentration in dehydrated nosis of iron deficiency and beta tha- 22. Conway AM, Vora AJ, Hinchliffe RF. The cells. Blood 1980;56:125–8. lassemia trait. Int J Lab Hematol 2009;31: clinical relevance of an isolated increase in 8. Wintrobe MM. Anemia: classification and 528–34. the number of circulating hyperchromic treatment on the basis of differences in the 16. Schoorl M, Schoorl M, Linssen J, Martinez red blood cells. J Clin Pathol 2002;55: average volume and hemoglobin content of Villanueva M, Velasco NoGuera JA, 841–4.

23. Buttarello M, Bulian P, Venudo A, Rizzotti P. of reticulocyte maturation parameters in 42. Buttarello M, on behalf of the hematology Laboratory evaluation of the Miles H.3 auto- the differential diagnosis of macrocytic study group of the Italian Society of Labo- mated reticulocyte counter. A comparative anemia. Clin Lab Haematol 2003;25: ratory Medicine. Variabilita biologica dei study with manual reference method and 283–8. parametri ematologici. Riv Med Lab (JLM) Sysmex R-1000. Arch Pathol Lab Med 34. Buttarello M, Temporin V, Ceravolo R, Far- 2003;4(suppl 1):88–91. 1995;119:1141–8. ina G, Bulian P. The new reticulocyte 43. Zhang P, Tang H, Chen K, Chen Y, Xu D. 24. Buttarello M, Bulian P, Farina G, Temporin parameter (Ret-y) of the Sysmex XE 2100. Biological variations of hematologic param- V, Toffolo L, Trabuio E, Rizzotti P. Flow Its use in the diagnosis and monitoring of eters determined by UniCelDxH 800 hema- cytometry reticulocyte counting. Parallel posttreatment sideropenic anemia. Am J tology analyzer. Arch Pathol Lab Med evaluation of five fully automated analyz- Clin Pathol 2004;121:489–95. 2013;137:1106–10. ers: an NCCLS-ICSH approach. Am J Clin 35. Mullier F, Lainey E, Fenneteau O, Da Costa 44. Cappelletti P, Biasioli B, Buttarello M, Pathol 2001;115:100–11. L, Schillinger F, Bailey N, Cornet Y, Chate- Bulian P, Casolari B, Cenci A, Miconi V. 25. Davis BH. Immature reticulocyte fraction lain C, Dogne JM, Chatelain B. Additional Within and between subject biological vari- (IRF): by any name, a useful clinical erythrocytic and reticulocytic parameters ability in complete blood count (CBC), parameter of erythropoietic activity. Lab helpful for diagnosis of hereditary sphero- leukocyte differential count (LDC), and Hematol 1996;2:2–8. cytosis: results of a multicentre study. Ann reticulocytes (abstract). Proceeding of the 26. Buttarello M, Bulian P, Farina G, Petris Hematol 2011;90:759–68. XIX International Symposium on Techno- MG, Temporin V, Toffolo L. Five fully auto- 36. Brugnara C, Laufer MR, Friedman AR, logical Innovation in Laboratory Hematol- mated methods for performing immature Bridges K, Platt O. Reticulocyte hemoglobin ogy, ISLH. Amsterdam, the Netherlands; reticulocyte fraction. Comparison in diag- content (CHr): early indicator of iron defi- April 25-28, 2006: 327. nosis of bone marrow aplasia. Am J Clin ciency and response to therapy. Blood 45. Zini G. Stability of complete blood count Pathol 2002;117:871–9. 1994;83:3100–1. parameters with storage: toward defined 27. Piva E, Brugnara C, Chiandetti L, Plebani 37. d’Onofrio G, Chirillo R, Zini G, Caenaro G, specifications for different diagnostic appli- M. Automated reticulocyte counting: state Tommasi M, Micciulli G. Simultaneous cations. Editorial. Int J Lab Hematol of the art and clinical applications in the measurement of reticulocyte and red cell 2014;36:111–3. evaluation of erythropoiesis. Clin Chem indices in healthy subjects and patients 46. Brugnara C, Mohandas N. Red cell indices Lab Med 2010;48:1369–80. with microcytic and macrocytic anemia. in classification and treatment of anemias: 28. Buttarello M, Bulian P, Farina G, Temporin Blood 1995;85:818–23. from M.M. Wintrobe’s original 1934 classi- V, Toffolo L. Frazione di reticolociti imma- 38. Brugnara C. Reticulocyte cellular indices: a fication to the third millennium. Curr Opin turi (IRF) e conteggio reticolocitario asso- new approach in the diagnosis of anemias Hematol 2013;20:222–30. luto: analisi multivariata e studio della and monitoring of erythropoietic function. 47. Van Hove L, Schisano T, Brace L. Anemia covarianza in diversi subsets eritropoietici Crit Rev Clin Lab Sci 2000;37:93–130. diagnosis, classification, and monitoring (Abstract). Riv Med Lab (JLM) 2000;1 39. Noronha JFA, Grotto HZW. Measurement using Cell-Dyn technology reviewed for the (suppl 1):155. of reticulocyte and red blood cell indices in new millennium. Lab Hematol 2000;6:93– 29. d’Onofrio G, Kuse R, Foures C, Jou JM, patients with iron deficiency anemia and 108. Pradella M, Zini G. Reticulocytes in haema- b-thalassemia. Clin Chem Lab Med 48. Urrechaga E, Unceta M, Borque L, Esca- tological disorders. Clin Lab Haematol 2005;43:195–7. nero JF. Low hemoglobin density potential 1996;18(suppl 1):29–34. 40. Cappelletti P, Biasioli B, Buttarello M, marker of iron availability. Int J Lab Hema- 30. Tsuda I, Tatsumi N. Maturity of reticulo- Bulian P, Casolari B, Cenci A, Miconi V. tol 2012;34:47–51. cytes in various hematological disorders. Mean reticulocyte volume (MCVR): refer- 49. Hoffmann J, van den Broek N, Curver J. Eur J Haematol 1989;43:252–4. ence intervals and the need for standard- Reference intervals of extended erythrocyte 31. Davis B, Bigelow N. Automated reticulo- ization (abstract). Proceeding of the XIX and reticulocyte parameters. Clin Chem cyte analysis Clinical practice and associ- International Symposium on Technological Lab Med 2012;50:941–8. ated new parameters. Hematol/Oncol Clin Innovation in Laboratory Hematology, 50. Fraser CG, Wilkinson SP, Neville RG, Knox North Am 1994;8:617–30. ISLH. Amsterdam, the Netherlands; April JD, King JF, MacWalter RS. Biologic varia- 32. Chang C, Kass L. Clinical significance of 25-28, 2006: 327. tion of common hematologic laboratory immature reticulocyte fraction determined 41. Ricos C, Alvarez V, Cava F, Garcia-Lario JV, quantities in the elderly. Am J Clin Pathol by automated reticulocyte counting. Am J Hernandez H, Jmenez CV, Minchinela J, 1989;92:465–70. Clin Pathol 1997;108:69–73. Perich C, Simon M. Current databases on 33. Torres Gomez A, Casano J, Sanchez J, biological variation: pros, cons, and progress. Madrigal E, Blanco F, Alvarez MA. Utility Scand J Clin Lab Invest 1999;59:491–500.