Distinguishing iron deficiency anemia from thalassemia by the red blood cell lifespan…
Distinguishing iron deficiency anemia from thalassemia by the red blood cell lifespan with a simple CO breath test: a pilotstudy
Mei-qing Lei1, Ling-feng Sun1,Xian-sheng Luo, Xiao-yang Yang, Feng Yu, Xiao-xia Chen and Zhi-ming Wang
Department of Hematology in Haikou Municipal People’s Hospital, Affiliated Haikou Hospital Xiangya School of Medicine Central South University, Haikou, 570208, People’s Republic of China
1 Mei-qing Lei and Ling-feng Sun contributed equally to this paper. They are co-first authors.
E-mail: wzm201806@163 .com
Keywords: microcytic hypochromic anemia, iron deficiency anemia, thalassemia, differential diagnosis, red blood cell lifespan
Abstract
Background: Extant indices for distinguishing between iron deficiency anemia (IDA) and thalassemia (Thal) have substantial practical limitations. The aim of this pilot study was to assess the predictive value of red blood cell lifespan (RBCLS), as determined by an automated CO breath test analysis approach, in the differential diagnosis of these two common forms of microcytic hypochromic anemia (MHA). Methods: RBCLS measurements were conducted in 35 healthy controls (HCs) and 114 patients diagnosed with MHA (IDA, N = 59; and Thal, N = 55) with ELS TESTER that provides a direct RBCLS value read-out. RBCLS between IDA and Thal was compared and evaluated by referring to normal cut-off from the instrument. Results: Compared with that in HCs, RBCLS in IDA and Thal groups was shortened; and median RBCLS was shorter in the Thal group than that in IDA group (33 d versus 79 d,p < 0.001). The median RBCLS in IDA patients with chronic gastrointestinal (GI) bleeding was shorter than that those without GI bleeding (38 d versus 100 d,p < 0.001). Using 75 d as a cut-off, RBCLS had a sensitivity of 96.4% and a specificity of 50.8% for detecting Thal. When GI bleeding patients were excluded from the IDA group, discriminant effciency of RBCLS was further improved. Conclusions: MHA with a normal RBCLS is suggestive of IDA, whereas MHA with a significantly shortened RBCLS without signs of chronic GI bleeding is suggestive of Thal.
Introduction
Iron deficiency anemia (IDA) and thalassemia (Thal) are common causes of microcytic hypochromic ane mia (MHA). Both are highly prevalent in the Mediter ranean basin, southeast Asia, and southern China [l, 2]. In IDA, there is an insufficiency of iron stores that results in reduced hemoglobin (Hb) production, whereas Thal is caused by ineffective erythropoiesis due to a globin chain synthesis imbalance caused by an Hb gene abnormality [3-5]. Definite diagnoses can be made based on a detailed medical history and labora tory tests (i.e. serum ferritin, genetic, and bone marrow smear with Prussian blue). However, due to limited healthcare awareness and resources in most Thal-endemic regions, IDA and Thal diagnoses may be missed [6]. Therefore, a simple and rapid screening test is needed to enable timely initial assessments in busy clinical practices.
Numerous discriminating indices have been pro-posed to distinguish Thal from IDA, including red blood cell (RBC) count, RBC distribution width, the Sirdah index, and the % microcytic/% hypochromic ratio index, and so on [7-11]. None of these metrics are sufficiently reliable to be officially recommended for routine clinic practice . Recently, a confocal and atomic force microscopy technique has been reported
to distinguish IDA from Thal well, but the expensive instruments and subjective morphological recogni
tion make it untenable as a preliminary screening tool [12].
RBC lifespan (RBCLS) refers to the average period of time that RBCs survive in circulation after being released from bone marrow. A shortened RBCLS is a fundamental characteristic, and thus gold standard diagnostic criterion of hemolytic anemia [13]. Because Thal is a chronic hemolytic disease, and IDA is not, it should be possible to discriminate between them with an RBCLS measurement. In fact, patients with Thal have been shown to have significantly shortened RBCLSs by the classic 51Cr labeling technique [14 16]. Unfortunately, although the 51Cr labeling techr reliable, it is not well suited for routine clinical practice due to its cumbersome procedure, radiation exposure, and its taking several weeks to complete. Based on the knowledge that endogenous carbon monoxide (CO) originates mainly from degraded RBCs, a team lead by Levitt developed a simple, rapid, and accurate RBCLS estimation technique with a CO breath test [17, 18].Recently, an automated instrument based on Levitt’s principle has become available, facilitating RBCLS data collection in clinical settings [19]. Hence, the aim of this pilot study was to determine whether RBCLS assessment with this automated CO breath test can be served as a differentiation index for the two major types of MHA: IDA and Thal.
Materials and methods
Study subjects
35 non smoking healthy controls (HCs) [16 males and 19 females; mean age ± standard deviation (SD), 27.0 ± 17.8 years] and 114 non-smoking patients with MHA were enrolled in this study at Haikou People Hospital between September 2017 and July 2018. The group of patients with MHA consisted of 59 with IDA (15 males and 44 females; mean age ± SD,46.8 ± 20.5 years) and 55 with Thal (25 males and 30 females; mean age ± SD, 25.0 ± 21.5 years). All 149 subjects met the age criterion for the CO breath test (≥ 7 years old). The exclusion criteria were: acute disease or emergent medical needs [e.g. acute gastro-intestinal (GI) bleeding]; currently taking hemolytic drugs (e.g. ribavirin); severe chronic cardiopulmonary disease; a blood transfusion within 3 weeks of the study; and active or severe passive smoking (or similar heavy air contamination exposure) within 24 h before the breath test.
Of the 59 patients with IDA, 23 had chronic GI tract blood loss (GI bleeding subgroup). The remain ing 36 IDA patients (Non-GI bleeding subgroup) con sisted of 8 women of childbearing age with menstrual blood loss, 8 patients with hemorrhoids hemorrhage, 18 patients with iron malabsorption, and 2 patients with reduced dietary iron intake.
Of the 55 patients with Thal including 15 α-Thal and 40 β-Thal, 9 Thal patients were with concomitant iron deficiency.
MHA was defined as anemia (Hb < 110 gl1) with the following RBC index levels: mean corpus cular volume < 80 fl， mean corpuscular hemoglobin < 27 pg, and mean corpuscular Hb concentration < 320 gl1. IDA was confirmed on the basis of a serum ferritin level < 15 µgl1 and bone marrow iron staining (Prussian blue reaction) negativ ity [20]. Determining the cause of iron deficiency would require a medical history combined with fur-ther examination, such as endoscopy and imaging, etc [21]. Thal was confirmed by Hb electrophoresis and genetic analysis [22]. Thal with iron deficiency was diagnosed based on the simultaneous fulfilment of the diagnostic criteria of both diseases.
The protocol for this study was approved by the Review Boards of Haikou People’s Hospital. This experiment was carried out in accordance with the Declaration of Helsinki. Prior to study participation, written informed consent was obtained from each adult subject and from a parent of each juvenile subject.
RBCLS measurement
An automated Levitt’s CO breath test machine was used to determine RBCLS. The test is based on the principle that endogenous CO in the breath originates mainly （~70%) from heme oxidation that occurs during Hb degradation following RBC rupture. Mean RBCLS equates to the total capacity of CO from Hb divided by the CO quantity released per day. Thus, RBCLS values were calculated as follows:
where [Hb] is the Hb concentration in g/ml, endo Pco is the endogenous alveolar CO concentration in ppm. Itwas determined with a routine peripheral blood test, and endo Pco was determined by subtracting atmo spheric Pco from alveolar Pco. Alveolar Pco was determined via non-dispersive infrared spectroscopy by the ELS TESTER (Seekya Biotec Co. Ltd, Shenzhen, China), an automated instrument that assays paired alveolar and air gas samples and uses those measure ments to calculate RBCLS according to Levitt’s formula.
The procedure of the test is quite simple. Briefly, alveolar air samples were collected in the morning without a fasting requirement. After a deep inspira tion, each subject held his or her breath for 10 s, and then exhaled into a collection system through a mouthpiece. Atmospheric samples were then col lected immediately after breath sampling. Alveolar air and atmospheric samples were stored at room temper ature and analyzed within 5 d of collection. Peripheral vein blood samples were collected for routine Hb mea surement on the same day that the air samples were collected. The instrument operator connects the paired air-alveolar gas samples to inlets, enters the subject’s blood Hb data, and then presses a start but ton that initiates the aforementioned series of auto mated measurements. The results are reported in 15 min. A normal RBCLS range was considered to be 126 ± 26 d with 95% confidence range of 75-177 d [19].
Table 1.RBCLS of HC, IDA, and Thal groups and analyzed subgroups.
Statistical methods
All data were analyzed in SPSS 19.0. The distribution of the continuous variables was calculated using Mann Whitney U test. p < 0.05 was considered to be statistically significant.
Results
Median RBCLS values for the 35 non-smoking HC subjects and 114 patients with MHA (including IDA and Thal groups) are reported in the table 1 and illustrated for individual subjects data in the figure 1. The median RBCLS values obtained for the group of 59 patients with IDA and the group of 55 patients with Thal were significantly lower than that obtained for the HC group (79 d, 33 d versus 132 d; both p < 0.001). The Thal group had a significantly shorter median RBCLS than the IDA group (79 d versus 33 d, p < 0.001). Moreover, RBCLS in the 23 patients in the chronic GI bleeding IDA subgroup was similar to that in the Thal group (38 d versus 33 d, p > 0.05) and significantly shorter than that in the 36 patients with non-GI bleeding IDA (38 d versus 100 d, p < 0.001). The median RBCLS of 9 patients with comorbid Thal and IDA was not significantly different from that of 46 patients with Thal alone (51 d versus 30 d, p > 0.05).
All 35 HCs (100%) had an RBCLS above the 75 d normal baseline value, whereas only half of the IDA patients (30/ 59, 50.8%) and very few Thal patients (2/ 55, 3.6%) did. The proportion of Thal patients with RBCLS values below 75 d was significantly greater than the proportion of IDA patients with RBCLS valuesabove75 d (p < 0.001).
A 75 d RBCLS cut-off valuewas adopted for differ entiating among the 114 MHA cases. Employing ≥75 d as a putative diagnostic criterion for IDA, our RBCLS results provided a sensitivity of 50.8% (30/ 59) and specificity of 96.4% (53/55), with a positive predictive value of 94.0% (30/ 30 + 2), a negative predictive value of 67.1% (53/26 + 53), and an accuracy level of 72.8% (30 + 53/ 114). On the other hand, using a RBCLS ≤ 75 d criterion as an indicator for Thal, we obtained commensurate reliability figures (accuracy, 72.8%; sensitivity, 96.4%; specificity,50.8%; positive predictive value, 67.1%; and negative predictive value, 94.0%). When GI bleeding IDA patients were excluded from the analysis, accuracy for Thal was improved to 91.2% (30 + 53/91).
Discussion
Differentiation between IDA and Thal has important clinical implications because these two diseases have entirely different causes, prognoses, and treatments. As a consequence of the differing pathophysiological mechanisms of IDA and Thal, RBCLS tends to be shorter in patients with Thal than in patients with IDA. This study introduces a RBCLS-based method for distinguishing these two patient categories. Using 75 d as a cut-off value, we obtained a differential accuracy of 72.8% in a cohort of 114 MHA patients with a 96.4% sensitivity level for Thal detection and a 50.8% specificity level for IDA detection. These results indicate that RBCLS measurement has value in the differential diagnosis of IDA and Thal.
Compared with classical standard labeling meth ods for determining RBCLS, Levitt’s CO breath test yielded a similar mean value (126 ± 26 d), but with the advantages of being quick, noninvasive, and inex pensive [19]. Consistent with an accelerated ery throcyte destruction rate, patients with hemolytic anemia exhibit a significantly shortened RBCLS [17-19]. In our study, the RBCLS values obtained for all HCs were within the normal range, whereas 96.4% of our Thal patients had an RBCLS below 75 d, with the two remaining Thal patients having RBCLS values close to the baseline level (76 d and 79 d), which may be attributed to mild Thal with mild hemolytic ane mia. RBC counts were previously reported to be higher in mild Thal than in IDA [8]. Because RBC count and RBCLS are altered in opposite direction in mild Thal and IDA, the ratio between these two values may be a superior discriminant to any single one in these two diseases.
Notably, we obtained a relatively wide RBCLS dis tribution range for the IDA group. The broadness of this range can be attributed to GI bleeding: all patients in the GI bleeding subgroup of the IDA group had sub normal RBCLS durations, whereas most of the patients in the non-GI bleeding subgroup had normal RBCLS durations. This distinction arises from the etiology of IDA, variance in severity of iron deficiency, and the breath test method itself. With chronic GI bleeding, damaged RBCs stay in the intestine for extended periods of time metabolizing CO, thereby lowering RBCLS. Conversely, non-GI bleeding IDA patients have a normal RBCLS because of rapid blood exclusion (e.g. menstrual loss) or chronic iron insuffi-ciency due to iron malabsorption or insufficient iron intake. Furthermore, severe iron deficiency impairs RBC membrane durability, which may affect RBCLS [23-26]. In this sense, RBCLS data may be useful for determining iron deficiency cause.
The very high sensitivity of our CO breath test RBCLS data for distinguishing Thal from IDA (96.4%) is consequent to the generally shortened RBCLS in patients with Thal together with the relatively normal RBCLS in patients with IDA (and thus, good specificity for detection of IDA).A shortened RBCLS in the con text of chronic GI bleeding reduces the test’s sensitivity for IDA and specificity for Thal. When the GI bleeding subgroup was excluded from the IDA group, the acc uracy of our RBCLS results for distinguishing Thal from IDA increased markedly from 72.8% to 91.2%. Therefore, reliable prediction of Thal based on a shor tened RBCLS must first exclude chronic GI bleeding, which can be easily identified with a fecal occult blood test and medical history (including elderly people, patients taking antithrombotic medication, and patients with parasite infections). To our knowledge, our study is the first to report RBCLS as a distinguish - ing parameter between IDA and Thal.
The major limitations of this study were its small sample size and the lack of a validation cohort from our or another institute. These limitations might reduce the robustness of the clinical findings. Not-withstanding, the results of this preliminary study are of great interest and worthy of further verification .
Conclusions
The presence of MHA with normal range RBCLS is suggestive of IDA, whereas Thal is highly indicated by a significantly shortened RBCLS without evidence of GI bleeding.
Acknowledgments
The authors thank Professor Hou-de Zhang [19] for his professional advice on our research design and manuscript.
Disclosure
The authors declare that we have no conflict of interest.
ORCID iDs
Zhi-ming Wang https://orcid.org/0000-0002-4011-5051
References
[1] Yao H eta/ 20 14Thespectrum of alpha and beta-thalassemia mutations of the Li people in Hainan province of China Blood Cells Mal. Dis. 53 16-20
[2] De Sanctis V et al 2017 Beta-thalassemia distribution in theold world: an ancient disease seen from a historical standpoint Mediterr J. Hematology Infect Dis.9 e2017018
[3] Camaschella C 2015 Iron-deficiencyanemia N. Engl. J.Med. 372 1832-43
[4] Peyrin-Biroulet L, Williet N and Cacoub P 2015 Guidelines on the diagnosis and treatment of iron deficiency across indications: a systematic review Am.J. Clin.Nutrition 102 1585-94
[5] Birgens H and Ljung R 2007 The thalassaemia syndromes Scand J. Clin. Lab. Investt. 67 11-25
[6] Brugnara C 2003 Iron deficiency and erythropoiesis: new diagnostic approaches Clin. Chem.49 1573-8
[7] Hoffmann JJ，Urrechaga E and Aguirre U 2015 Discriminant indices for distinguishing thalassemiaand iron deficiency in patients with microcytic anemia: a meta analysis Clin. Chem. Lab.Med. 53 1883-94
[8] Vehapoglu A, Ozgurhan G, Demir A D, Uzuner S, NursoyM A, Turkmen S and Kacan A 2014 Hematologicalindices for differential diagnosis of beta thalassemia trait and iron deficiency anemia Anemia 2014 576738
[9] Urrechaga E 2009 Red blood cell microcytosis and hypochromia in thedifferential diagnosis of iron deficiency and beta-thalassaemia trait lnt.J. Lab.Hematology 31 528-34
[10] Janel A, Roszyk L, Rapatel C, Mareynat G, Berger MG and Serre-Sapin A F 2011 Proposal of a score combining red blood cell indices for early differentiation of beta-thalassemia minor from iron deficiency anemia Hematology 16 123-7
[11] Novak R W 1987 Red blood cell distribution width in pediatric microcytic anemias Pediatrics 80 251-4
[12] Tariq S, Bilal M, Shahzad S, Firdous S,Aziz U and Ahmed M 2017 Diagnosis of thalassemia and iron deficiency anemia using confocal and atomic force microscopy Laser Phys.Lett. 14115703
[13] Franco R S 2012 Measurement of Mred cell lifespan and aging Transfusion Med.Hematology 39 302-7
[14] Hillcoat B L and Waters A H 1962 The survival of 51Cr labelled autotransfused red cells in patients with thalassaemia Australas.Ann.Med. 11 55-8
[15] Bernini L, Latte B, Siniscalco M, Piomelli S, Spada U,Adinolfi M and Mollison P L 1964 Survival of 51Cr-labelled red cells in subjects with thalassaemia trait or G6pd deficiency or both abnormalities Br. J.Haematolgy 10 171-80
[16] Kasfiki A G, Antipas SE, Dimitriou P A, Gritzali FA and Melissinos K G 1982 Mathematical analysis of 51Cr-labelled red cell survival curves in congenital haemolytic anaemias Eur.J. Nucl.Med. 7 181-3
[17] Strocchi A, Schwartz S, Ellefson M, Engel RR, Medina A and Levitt M D 1992 A simple carbon monoxide breath test to estimate erythrocyte turnover J. Lab. Clin.Med. 120 392-9
[18] Furne J K, Springfield J R,Ho S B and Levitt M D 2003 Simplification of the end-alveolar carbon monoxide technique to assess erythrocyte survival ]. Lab. Clin. Med. 142 52-7
[19] Zhang H D et al 2018 Human erythrocyte lifespan measured by Levitt’s CO breath test with newly developed automatic instrument J. Breath Res. 12 036003
[20] De Franceschi L, Iolascon A, Taher A and Cappellini M D 2017 Clinical management of iron deficiency anemia in adults: systemic review on advances in diagnosis and treatment Eur. J.Intern. Med.42 16-23
[21] Brim H,Shahnazi A, Nouraie M, Badurdeen D, Laiyemo A 0, HaidaryT, Afsari A and Ashktorab H 2018 Gastrointestinal lesions in Africa n American patients with iron deficiency anemia Clin.Med.Insights Gastroenterol. 11 1179552218778627
[22] Ryan K et al 2010 Significant haemoglobinopathies: guidelines for screening and diagnosis Br.J.Haematology 149 35-49
[23] Diez-Ewald M and Layrisse M 1968 Mechanismsof hemolysis in iron deficiency anemia Further Stud. Blood 32 884-94
[24] Anderson C, Aronson I and Jacobs P 1999 Erythrocyte deformability is reduced and fragility increased by iron deficiency Hematology 4 457-60
[25] Vaya A, Simo M, Santaolaria M, Todoli J and Aznar J 2005 Red blood cell deformability in iron deficiency anaemia Clin.Hemorheology Microcirculation 33 75-80
[26] Nagababu E, Gulyani S, Earley CJ, Cutler RG,Mattson M P and Rifkind J M 2008 Iron-deficiency anaemia enhances red blood cell oxidative stress Free Radie Res. 42 824-9
