If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Correspondence: Address to Carlo Brugnara, MD, Department of Laboratory Medicine, Boston Children’s Hospital, 300 Longwood Avenue, Bader 760, Boston, MA 02115.
To assess the accuracy of a simplified approach for the diagnosis of iron deficiency anemia (IDA) based on the complete blood cell count (CBC) and reticulocyte analysis.
Patients and Methods
Five hundred fifty-six consecutive, nonselected patients referred for diagnosis and/or treatment of anemia were included in this diagnostic study to compare the performance of reticulocyte hemoglobin equivalent (RET-He) versus traditional biochemical markers for diagnosis and treatment of IDA. Complete blood count, serum ferritin, iron, and transferrin saturation were performed as clinically indicated. Reticulocyte hemoglobin equivalent was measured with a Sysmex XN-450 analyzer on the residual CBC sample. The study period was from September 20, 2017, through and including November 15, 2018.
Results
Patients (N=556) were studied at baseline, of whom 150 were subsequently treated with intravenous iron. Receiver operating characteristic analysis yielded an RET-He cut-off of 30.7 pg to identify IDA (area under curve, 0.733; 95% CI, 0.692 to 0.775), with 68.2% sensitivity and 69.7% specificity. Patients (n=240) were seen at follow-up, with 57 treated and 183 not treated with intravenous iron. Responsiveness was defined as a hemoglobin increase of ≥1.0 g: a combination of RET-He <28.5 pg and hemoglobin value <10.3 g/dL had 84% sensitivity and 78% specificity as response predictor (area under the curve, 0.749; 95% CI, 0.622 to 0.875).
Conclusion
Data from CBC and RET-He can identify patients with IDA, determine need for and responsiveness to intravenous iron, and reduce time for therapeutic decisions. Limitations of this study are uncontrolled design, its single-site and retrospective nature, and that it requires prospective validation.
Anemia affects more than 30% of the world’s population, with iron deficiency the overwhelmingly most common cause. Whether absolute because of blood loss and/or iron sequestration because of underlying morbidity, the need for repletion, especially in females, is a formidable medical issue.
The diagnosis of iron deficient states traditionally combines biochemical parameters such as serum ferritin, iron, total iron binding capacity, and transferrin saturation (TSAT), with changes in red cell indices and hemoglobin (Hb) concentrations. In recent years, reticulocyte hemoglobin content has emerged as an additional parameter which is helpful in identifying iron deficient erythropoiesis, especially when traditional biochemical parameters are noninformative.
For the first time, reticulocyte hemoglobin content allowed quantification of the Hb contained in the reticulocyte pool and its changes in conditions of altered erythropoiesis.
Reticulocyte Hb content is a direct assessment of the functional availability of iron to the erythropoietic tissue and is an important indicator of functional iron deficiency, a condition in which tissue iron is not readily available for incorporation into the bone marrow due to excessive erythropoietic stimulation or sequestration of iron into stores. Because of the shorter life span of reticulocytes than mature red cells (a few days vs ∼120 days), measurements of reticulocyte Hb can identify iron deficiency before the distinguishing changes in red cell parameters become apparent. Assessment of iron status is particularly relevant for patients treated with erythropoietic stimulating agents, as iron parameters are important predictors of bone marrow functional response.
Efficacy of different dosing regimens for recombinant human erythropoietin in a simulated perisurgical setting: the importance of iron availability in optimizing response.
A rapid estimate of iron availability and need, or lack thereof, would help in avoidance of overtreatment and iron overload, which is a concern in chronic kidney disease.
Reticulocyte Hb can be measured by two methods, the originally developed method known as reticulocyte Hb content (CHr) which is available on Siemens analyzers,
and the latter developed method known as reticulocyte Hb equivalent (RET-He) which is available on Sysmex analyzers. Acceptable correlation has been demonstrated between RET-He and CHr
As intravenous iron has been moved to the front line in a host of conditions associated with iron deficiency, such as abnormal uterine bleeding, inflammatory bowel disease, after gastric bypass, mid and late stages of pregnancy, angiodysplasia, and chronic kidney disease, this study examines how RET-He performs in diagnosing iron deficiency in an outpatient setting and how it further identifies not only the requirement for intravenous iron supplementation, but also the appropriateness of the therapeutic response and the need, or lack thereof, for further iron replacement.
Methods
Study Design, Methodology, and Enrollment
Patients referred for anemia management were considered eligible for this study if they had clinically indicated samples collected for complete blood count (CBC), iron parameters including serum ferritin, iron, total iron binding capacity, and TSAT. All samples were collected after fasting. The study period was from September 20, 2017, through and including November 15, 2018. All samples for chemistry were performed at LabCorp and Quest Diagnostics as dictated by insurance. Complete blood counts were performed in the office on a Medonics M Series hematology analyzer. A Sysmex XN-450 hematology analyzer was used for parallel determination of CBC and RET-He on the leftover material from the CBC. The person performing the test on the Sysmex instrument had no access to either clinical information or other test results.
Consecutive, non-selected patients (N=556) referred for anemia, were included in this study. Of these, 150 were selected for, and received intravenous iron, which was administered as a complete replacement dose of 1000 mg in one or two visits. Iron deficiency was diagnosed based on either a low serum ferritin (<30 ng/mL), and/or a low TSAT (< 20%). The decision to treat was based on these parameters and clinical considerations. Patients with missing data were not included in the study.
Patients (n=240) were studied at a second follow-up visit, including 57 who had received intravenous iron between the first and second visit.
Statistical Analysis
All continuous data are presented using medians, interquartile ranges (IQRs), and/or full ranges, and all categorical data are presented using frequencies and percentages. The correlation between parameters was assessed using the Spearman nonparametric rank correlation coefficient rho. Comparisons of continuous clinical and laboratory data between visits were performed using the nonparametric Wilcoxon signed rank test for paired data. The comparison of independent groups, including the comparison of the top versus the bottom quartile of the distributions of RET-He and mean corpuscular Hb (MCH) minus RET-He, were performed using the Wilcoxon rank sum test. Receiver operating characteristic (ROC) curve analysis was implemented to determine the value of RET-He in identifying iron deficiency as well as the ability of RET-He and Hb in predicting response to intravenous iron treatment, with results presented as area under the curve (AUC) with corresponding 95% CI and P value. Multivariable median regression modeling was used to assess the independent associations between changes in Hb and the other parameters among patients receiving intravenous iron. Median regression modeling was used due to non-normality of the delta Hb variable. All statistical analyses were conducted in Stata (version 15.0, StataCorp LLC., College Station, TX). A two-tailed alpha level of 0.05 was used to determine statistical significance.
Institutional Review Board Approval
No informed consent was required for this study because sample and data were de-identified and no additional blood was drawn. The research CBC and reticulocyte analysis were performed on leftover material from the clinically indicated CBC sample. The study was performed with approval of the Boston Children’s Hospital Committee on Clinical Investigation (Protocol I#: IRB-P00031521 and NS06-07-0354).
Results
Clinical and laboratory data on 556 patients were collected. The median age was 66 years (Supplemental Table 1, available online at http://www.mayoclinicproceedings.org) and the majority were female (n=409, 73.6%). The median value of RET-He was 31.0 pg/cell (range, 12.1 to 40.3 pg/cell). RET-He was positively correlated with Hb concentration (rho=0.365; P<.001) (Supplemental Figure 1, available online at http://www.mayoclinicproceedings.org), mean corpuscular volume (MCV) (rho=0.576; P>.001), MCH (rho=0.777; P<.001) (Supplemental Figure 1), serum iron (rho=0.526; P<.001), and TSAT (rho=0.492; P<.001). Based on either a serum ferritin less than 30 ng/mL and/or a TSAT less than 20%, 241 of 556 (43.4%) patients were diagnosed as iron deficient. Anemia (defined as Hb<13 g/dL for males and Hb<12 g/dL for females) was present in 139 of 241 (57.7%) of the iron-deficient patients. One hundred fifty-one of 241 (62.7%) patients were anemic with an Hb cut-off of 12.5 g/dL, and 64 of 241 (26.6%) with an Hb cut-off of 10.5 g/dL.
To assess hematologic and biochemical characteristics associated with decreased RET-He, we compared the bottom and top quartiles in the RET-He distribution. As shown in Table 1, significant differences were observed between the highest and the lowest quartiles of RET-He. The lowest quartile comprises RET-He values less than 28.6 pg, and identifies a group with reduced Hb concentration, MCV, MCH, lower ferritin, serum iron, and TSAT, with similar reticulocyte counts. The majority of ferritin values in the low RET-He quartile were above the lower limit of normal, with only 25% being below 28 ng/mL. Transferrin saturation data also encompassed both abnormal and normal values, with only 207 of 556 (37.2%) subjects having a TSAT less than 20%. In the top RET-He quartile, a value greater than 33.1 pg identifies patients with normal or borderline-reduced Hb values, normal MCV and MCH, and normal/high serum ferritin, serum iron, and TSAT. Thus, a low RET-He (<28.6 pg) is associated with iron-deficient erythropoiesis and/or iron deficient anemia, whereas the classic biochemical markers are noninformative in a substantial fraction of these patients.
Table 1Comparison of Baseline Parameters Between Top Versus Bottom Quartiles for RET-He
When iron deficiency anemia develops, the RET-He value is lower than the value of MCH because reticulocytes are the earliest indicators of reduced Hb synthesis, exhibiting changes much earlier than the mature red cell population, which has a 120-day life span. We report in Supplemental Table 2 (available online at http://www.mayoclinicproceedings.org) data on the quartile exhibiting MCH values greater than that of RET-He (difference >0.1pg): this subgroup has reduced Hb, MCH, serum iron, ferritin, and TSAT with no changes in MCV or absolute reticulocyte count. In this group, 43 subjects (30.7%) have ferritin greater than 30% and TSAT greater than 20%. Thus, when reticulocyte parameters are indicative of ongoing development of iron-deficient erythropoiesis, hematological markers such as MCV and biochemical markers such as serum ferritin and TSAT are not diagnostically helpful in a substantial proportion of cases.
Despite the limitations of the biochemical markers outlined above, we performed ROC analysis assessing the value of RET-He in identifying iron deficiency as defined by serum ferritin less than 30 ng/mL or transferrin saturation less than 20%. Receiver operating characteristic analysis shows a reasonable performance for RET-He (AUC, 0.733; 95% CI, 0.692 to 0.775), with a cut-off value of <30.7 pg yielding 68.2% sensitivity and 69.7% specificity. Using both Hb and RET-He in a multivariable ROC analysis does not provide an improved AUC as compared with just using RET-He (AUC, 0.605 vs 0.733, respectively).
Of the 556 patients for whom we collected baseline samples, 150 were selected for and received intravenous iron based on low serum ferritin, low TSAT, Hb value, and clinical considerations. Table 2 presents baseline data for the 150 patients selected to receive intravenous iron and 406 patients who did not. Those selected for intravenous iron had significantly lower Hb values, and lower MCV, MCH, RET-He, serum ferritin, iron, and TSAT, confirming that the selection criteria for intravenous iron administration were appropriate.
Table 2Comparison of Baseline Laboratory Parameters Among Patients Selected for Intravenous Therapy and Untreated Patients
Follow-up samples were collected for 240 of 556 subjects studied at baseline, including 57 of 150 treated with intravenous iron and 183 of 406 who received no intravenous iron between visits 1 and 2.
Table 3 presents data for first and second visits for the 57 patients who were treated with intravenous iron and for the 194 who were not. The 57 selected for intravenous iron had lower Hb, MCH, RET-He, ferritin, and TSAT compared with 194 patients who were deemed to not require therapy. Importantly, RET-He values were 28.3 pg (IQR, 24.2 to 30.2 pg) in the intravenous iron group versus 32.1 pg (IQR, 29.7 to 34.2 pg) in the no–intravenous iron group.
Table 3Comparison of First and Second Visit Laboratory Parameters Among Patients With and Without Intravenous Iron Therapy
Analyses were performed only using patients with 2 visits. Data are presented as median (interquartile range). Changes in parameters were assessed using the Wilcoxon sign-rank test for paired data.
Parameter
Visit 1
Visit 2
P
Among patients receiving intravenous iron therapy associated with first visit
a Hb, hemoglobin; MCH, mean corpuscular hemoglobin; MCV, mean corpuscular volume; RET-He, reticulocyte hemoglobin equivalent; TIBC, transferrin and iron-binding capacity; TSAT, transferrin saturation.
b Analyses were performed only using patients with 2 visits. Data are presented as median (interquartile range). Changes in parameters were assessed using the Wilcoxon sign-rank test for paired data.
Intravenous iron administration was associated with significant increases in Hb, MCV, MCH, RET-He, serum ferritin, iron, and TSAT, whereas in the no–intravenous iron cohort there was a small reduction in RET-He and small increases in MCV and MCH, with no significant variations in Hb and in the other parameters (Table 3). Serum ferritin was below 30 ng/mL in 18 of 57 (31.6%) of the patients requiring intravenous iron and in 19 of 183 (10.4%) of those not requiring iron at visit 1. These values changed to 4 of 57 (7.0%) (P=.002) and 23 of 183 (12.6%) (P=.623) at visit 2, respectively.
Regression analysis for Hb response following intravenous iron showed that baseline RET-He values are predictive of Hb response, with every unitary increase in RET-He corresponding to a blunting of the Hb change by -0.19 g/dl (95% CI, -0.27 to -0.11; P<.001). Changes in RE-He associated with intravenous iron administration are also predictive of the Hb response, with every additional unit increase in RET-He corresponding to a 0.21 g/dL increase in Hb (95% CI, 0.13 to 0.28; P<.009). Receiver operating curve analysis for the capability to predict Hb response among the 57 patients receiving intravenous iron shows that a value of baseline RET-He less than 28.5 pg together with a baseline Hb value less than 10.3 g/dL provide the highest Youden’s index for predicting Hb response greater than or equal to 1.0 g/dL, with sensitivity of 84% and specificity of 78%. The Figure and Table 4 present data for the 21 of 57 patients who had RET-He less than 28.5 pg and Hb less than 10.3 g/dL versus the 36 of 57 who did not. Serum ferritin is a poor predictor of Hb response to intravenous iron with an AUC of 0.572 (95% CI, 0.419 to 0.725; P=.726). Therapeutic responsiveness is blunted in the anemia of chronic disease (ACD).
Using a surrogate definition for ACD based on serum ferritin values greater than 100 ng/mL and reduced transferrin and iron-binding capacity (TIBC) (<250 μg/dL)
; 7 of 57 (12.3%) patients fit these criteria. In this subgroup, intravenous iron therapy did not produce significant changes in either Hb (baseline median Hb, 10.2 g/dL; IQR, 9.1 to 10.6 g/dL; and visit 2 Hb, 9.8 g/dL; IQR, 9.4 to 10.8 g/dL; P=.706) or RET-He (baseline RET-He, 30.6 pg; IQR, 28 to 31.8 pg; visit 2 RET-He, 29.4; IQR, 24.9 to 31.6 pg; P=.551).
FigureHemoglobin (Hb) response to intravenous iron therapy. The change in Hb value between baseline and follow-up visits is plotted for subjects who had baseline reticulocyte hemoglobin equivalent (RET-He) less than 28.5 pg and Hb less than 10.3 g/dL (n=21) versus subjects who had RET-He greater than or equal to 28.5 pg or Hb greater than or equal to 10.3 g/dL (n=36). Medians and interquartile ranges are denoted. The Wilcoxon rank sum test was implemented to compare these two groups regarding changes in Hb following intravenous iron therapy (P=.003).
In a cohort of patients referred to an outpatient hematology practice for either diagnosis or treatment of anemia, we have tested the concept that identification of iron deficiency and the selection of patients most likely to benefit from intravenous iron administration can be best accomplished with the use of hematologic biomarkers which can be obtained as part of the CBC on an AutoAnalyzer rather than relying on traditional biochemical markers which have a much longer turnaround time. Our data indicate that RET-He can be used for this purpose and that a value lower than 30.7 pg can identify iron deficiency with a 68.2% sensitivity and 69.7% specificity. Reduced RET-He values indicate the presence of an iron-deficient state in patients for whom the traditional biochemical parameters are often not informative (Table 1). Similarly, an approach which emphasized the reduction of the Hb content of reticulocytes in contrast to the Hb content of the mature red cells also exhibits significant power in identifying iron deficiency, with many not having yet exhibited the expected reductions in MCV and biochemical markers (Supplemental Table 2). When Hb response is defined as an Hb increased greater than 1.0 g/dL, an RET-He value below 28.5 pg can predict Hb response with sensitivity of 84% and specificity of 78% when patients have a Hb values less than 10.3 g/dL. The severity of iron-deficient anemia is reflected by the extent of reduction in RET-He (Supplemental Figure 1, bottom panel); thus, it is not surprising that responsiveness to intravenous iron can be predicted by a reduction in RET-He values, and that the optimal RET-He value for predicting therapeutic responsiveness (28.5 pg) is lower than that for simply diagnosing iron deficiency (30.7 pg).
Several papers have provided evidence for the clinical value of reticulocyte Hb measurement in identifying iron-deficient states. Earlier studies were performed using CHr, a parameter provided by Siemens type instruments in both adults
Later studies have used the RET-He parameter generated by Sysmex instruments, which have shown a reasonable correlation with CHr despite the different methodology.
The cut-off value of 30.7 pg for RET-He for diagnosing iron deficiency we have determined in this work is similar to values reported by other investigators: 30.6-pg cut-off value determined by Buttarello et al
Evaluation of the hypochromic erythrocyte and reticulocyte hemoglobin content provided by the Sysmex XE-5000 analyzer in diagnosis of iron deficiency erythropoiesis.
In 209 patients affected by either solid tumors or hematological malignancies, an RET-He threshold of 31 to 33 pg was found to have a high negative predictive value (98.5%) to rule out iron deficiency.
Contrary to parameters such as hematocrit and MCV, the RET-He parameter is not affected by storage and it is stable at room temperature and in cold storage (4ºC to 8°C) for up to 24 hours according to manufacturer’s guidelines and up to 72 hours in independent studies.
Limited information is available about the value of RET-He in predicting response to iron therapy. In our study, an RET-He value less than 28.5 pg identified responsiveness to iron therapy with 84% sensitivity and 78% specificity when Hb is less than 10.3 g/dL. In patients undergoing chronic dialysis, an RET-He less than 30.7 pg identified response to intravenous iron with 78.7% sensitivity and 87.2% specificity.
Response to oral iron therapy was associated with either baseline RET-He or RET-He increase at 1 week (81% sensitivity, 86% specificity) in a cohort of 38 patients treated with 200 mg ferrofumaraat/day for 6 weeks.
Therapeutic responsiveness is reduced in the ACD; in the 7 of 57 patients meeting this definition, Hb response to intravenous iron was blunted and RET-He did not change significantly. Functional iron deficiency is characterized by reduced iron availability despite normal iron stores. None of the 57 patients met criteria for this condition using a TSAT less than 20% and serum ferritin greater than 100 ng/mL.
These data become especially poignant with the advent of four formulations of intravenous iron which permit a complete or near-complete replacement dose in a single visit over 15 to 60 minutes. In a cohort of 888 oral iron-intolerant, iron-deficient patients who received 1288 doses of 1000 mg of low molecular weight iron dextran (LMWID) in 1 hour, 76% were reported to have an Hb increase of 1 to 2 g in 4 weeks. No serious adverse events were noted.
Safety and efficacy of rapidly administered (one hour) one gram of low molecular weight iron dextran (INFeD) for the treatment of iron deficient anemia.
In both studies, ferric carboxymaltose was associated with a more rapid increase in Hb. No serious adverse events were observed.
Additional data supporting the use of total dose infusion has been published with ferumoxytol. Current labeling of ferumoxytol recommends 510 mg as a 15-minute infusion be administered over two visits on different days. Sixty patients were treated with 1020 mg of ferumoxytol in a single 15- to 20-minute infusion, which is consistent with previously published evidence, more than 80% achieved a 1 to 2 g Hb increment without any observed side effects.
These data were recently corroborated in another cohort of 184 consecutive, nonselected iron-deficient patients who received 1020 mg of ferumoxytol in 30 minutes.
Ferumoxytol (1020 mg) administered as two 510-mg 15-minute infusions 1 week apart was compared with 1500 mg of ferric carboxymaltose administered exactly the same way.
No difference in safety was reported by a blinded independent adverse event adjudication committee and both formulations were associated with greater than 80% observed Hb increment of 1 to 2 g at 5 weeks.
A prospective, multi-center, randomized comparison of iron isomaltoside 1000 versus iron sucrose in patients with iron deficiency anemia; the FERWON-IDA trial.
The newest formulation allowing a complete replacement dose in 15 minutes is ferric derisomaltose (FD) (formerly known as iron isomaltoside). One thousand four hundred eighty-seven iron-deficient patients were randomized 2:1 to either 500 or 1000 mg of FD administered intravenously over 15 minutes or 4 to 5 infusions of 200 mg of iron sucrose.
A prospective, multi-center, randomized comparison of iron isomaltoside 1000 versus iron sucrose in patients with iron deficiency anemia; the FERWON-IDA trial.
Ferric derisomaltose administration was noted to have a more rapid Hb response without an increase in adverse events compared with iron sucrose. A retrospective analysis compared 213 pregnant women at different stages of pregnancy and gravidity who received a single infusion of FD matched 1:1 to an equal number of age-matched, pregnant women with similar delivery dates and parity who did not. No severe adverse events were observed and the investigators concluded that a single high dose (up to 1500 mg) of FD is safe, effective, and convenient.
Study Limitations
This study has several limitations. It is retrospective in nature and performed in a single site. Response to intravenous iron was not evaluated via randomization. It did not actually test an RET-He–based strategy versus the conventional approach, which would be required to fully validate this approach. Not all patients in need of intravenous iron are identified, requiring results of standard iron parameters. However, this limitation is mitigated by this requirement in current practice.
Conclusion
The present data show that abnormally low RET-He values identify iron-deficient states and the need for iron replacement, obviating delays to obtain standard iron parameters. Baseline and changes in RET-He also associate with Hb response. Given the enormous prevalence of iron deficiency in the general population, the use of the RET-He informs on need for iron replacement, or lack thereof, and represents an increase in convenience for patient and physician as definitive therapy can be administered on the day of the initial visit.
Acknowledgments
The authors thank the Sysmex Scientific Group (Sysmex America Inc) for their technical help and support of the study, and Tarsha White for curating the data collection.
Efficacy of different dosing regimens for recombinant human erythropoietin in a simulated perisurgical setting: the importance of iron availability in optimizing response.
Evaluation of the hypochromic erythrocyte and reticulocyte hemoglobin content provided by the Sysmex XE-5000 analyzer in diagnosis of iron deficiency erythropoiesis.
Safety and efficacy of rapidly administered (one hour) one gram of low molecular weight iron dextran (INFeD) for the treatment of iron deficient anemia.
A prospective, multi-center, randomized comparison of iron isomaltoside 1000 versus iron sucrose in patients with iron deficiency anemia; the FERWON-IDA trial.
Grant Support: The study was supported by an unrestricted grant from Sysmex America Inc . for the present study. The sponsor had no role in the design, conduct, or interpretation of the analysis, in the management of data, or in the decision to submit the manuscript for publication.
Potential Competing Interests: Dr Auerbach has received research funding for data management by Sysmex America, Inc . Dr Brugnara has an ongoing consulting agreement with Sysmex America, Inc. Mr Staffa reports no potential competing interests.