Hemolysis is the disruption of the erythrocyte membrane with release of hemoglobin and other intra- cellular components into the surrounding serum or plasma. Intracellular components released in high con- centrations from erythrocytes include enzymes such as lactate dehydrogenase (LDH) and aspartate amino- transferase (AST), as well as electrolytes such as potas- sium and magnesium. In cases of massive hemolysis, the release of intracellular fluid from erythrocytes can even result in the dilution of serum or plasma analytes, such as sodium, that are usually present in low con- centrations within erythrocytes.
Normally, serum and plasma contain very low con- centrations of free hemoglobin, with plasma containing less than 2 mg/dL and serum less than 5 mg/dL.
Visual detection of hemolysis does not occur until the free hemoglobin concentration is greater than 20 70 mg/dL. The ability to visually detect hemoglo- bin in serum or plasma is affected by the concentration of other compounds present in the sample. For exam- ple, serum or plasma from patients who are jaundiced can mask hemolysis when hemolysis is estimated by visual means. Thus, samples with moderate hemolysis and free hemoglobin concentrations of 100 150 mg/
dL may go undetected when visual assessment of sam- ples is performed[10].
Hemolysis can be caused by a variety of mechan- isms, including physical disruption of cells and by immunological and chemical means. Hemolysis can be divided into in vivo hemolysis and in vitro hemolysis.
When red cell lysis occurs within the body with subse- quent release of intracellular components into the plasma, this is termed in vivo hemolysis. In contrast, when lysis of cells following the collection of blood occurs, it is termedin vitrohemolysis. Common causes of in vivo and in vitro hemolysis are shown in Table 5.1.
In Vivo Hemolysis
In vivo hemolysis can be categorized on the basis of where red cell destruction occurs. Intravascular hemolysis occurs when erythrocytes are destroyed while the cells are still within the vascular system,
54 5. HEMOLYSIS, LIPEMIA, AND HIGH BILIRUBIN
whereas extravascular hemolysis is due to destruc- tion of red cells by the phagocytic system in the liver, spleen, or bone marrow. It is important to rec- ognize and differentiate between in vivo and in vitro hemolysis because analytes such as potassium or LDH that might not be reported in samples showing in vitro hemolysis should be reported in samples with in vivo hemolysis. For example, increased potas- sium in a sample with in vitro hemolysis can be assumed to be artifactual and not representative of the patient’s true potassium concentration. However, increased potassium measured in a patient with in vivo hemolysis represents the true intravascular potassium in the patient.
In vivo intravascular hemolysis is typically charac- terized by increased LDH, decreased haptoglobin, and increased urine hemoglobin concentration. The increased LDH is the result of release of the enzyme
from erythrocytes, whereas the decreased haptoglobin is the result of binding of haptoglobin to free hemoglo- bin. Hemoglobin, which is a tetramer, is rapidly bro- ken down in plasma to dimers, with the resultant hemoglobin haptoglobin complex being rapidly cleared. The half-life of the hemoglobin haptoglobin complex is approximately 10 30 min due to rapid elimination by the monocyte macrophage system, whereas the half-life of free haptoglobin is 5 days[11].
Haptoglobin becomes saturated when free hemoglobin concentrations exceed approximately 150 mg/dL.
Plasma hemoglobin not bound to haptoglobin is read- ily filtered through the glomerulus and will be excreted in the urine, resulting in hemoglobinuria.
Note that low haptoglobin concentrations without in vivohemolysis may be seen in newborns and young children and in those individuals with haptoglobin deficiency[12].
The free hemoglobin will produce a positive reac- tion for heme protein when measured using a urine dipstick. The free hemoglobin dimers that remain in circulation are oxidized to form methemoglobin, which dissociates to produce free heme and globin chains. The oxidized free heme binds to hemopexin and is removed from circulation by the liver, spleen, and bone marrow. Heme is further metabolized to eventually form unconjugated bilirubin. Measurement of hemopexin, methemoglobin, and unconjugated bili- rubin can be helpful in the differentiation of in vivo intravascular or extravascular hemolysis from in vitro hemolysis.In vivohemolysis due to immunohemolytic causes such as ABO transfusion reactions, paroxysmal cold hemoglobinuria, and idiopathic autoimmune hemolytic anemia or due to use of certain drugs can result in further hemolysis of sensitized erythrocytes during the process of blood collection, clot formation, and centrifugation. Thus, the collection of a sample anticoagulated with heparin can help eliminate fur- ther hemolysis that might occur in vitro during clot formation. Recommended criteria have been pub- lished for helping differentiate in vivo from in vitro hemolysis[13]:
• Collect both serum and plasma samples.
• Because anticoagulation of blood helps minimize in vitrohemolysis, perform all tests on plasma wheneverin vivohemolysis is suspected.
• Measure hemoglobin and potassium concentrations and LDH activity in the serum and plasma
specimens.
• Specimens with increased LDH activity and hemoglobin concentrations but normal potassium concentrations indicatein vivohemolysis.
• Measure haptoglobin, hemopexin, and the reticulocyte count to confirmin vivohemolysis.
TABLE 5.1 Potential Causes ofin Vivoandin VitroHemolysis
IN VIVO HEMOLYSIS Extravascular hemolysis
Enzyme deficiencies (e.g., glucose-6-phosphate dehydrogenase deficiency)
Hemoglobinopathies (e.g., sickle cell, thalassemia)
Erythrocyte membrane defects (e.g., hereditary spherocytosis) Infection (e.g.,Bartonella,Babesia, malaria)
Autoimmune hemolytic anemia Other (e.g., hypersplenism, liver disease) Intravascular hemolysis
Microangiopathy (e.g., prosthetic heart valve, thrombotic thrombocytopenic purpura)
Transfusion reaction
Infection (e.g., sepsis, severe malaria) Paroxysmal cold hemoglobinuria Paroxysmal nocturnal hemoglobinuria IN VITRO HEMOLYSIS
Excessive aspiration force (blood drawn too vigorously, especially through small or superficial veins)
Catheter partially obstructed
Blood forced into the tube from syringe Specimen frozen
Mechanical damage (e.g., shaking, excessive force in pneumatic tubes)
Delay in analysis
Source: Data from Garby and Noyes[11].
CASE REPORT1 A 58-year-old female was admitted with a diagnosis of paroxysmal nocturnal hemoglobin- uria (PNH). Results of her physical examination were normal. Laboratory results obtained on days 1, 4, 5, and 7 of her admission are as follows:
Test (Reference Interval)
Specimen Day 1 Day 4 Day 5 Day 7
Potassium (3.2 4.6 mmol/L)
P 4.6 3.9
S 4.9 5.5 5.7 4.6
Lactate dehydrogenase (133 248 U/L)
P 2830 3020
S 2630 3170 3000 3305
Hemoglobin (0 50 mg/L)
P 52 162 132
S 297 301 221 210
Haptoglobin (410 2100 mg/L)
S ,150 ,150 ,150 ,150
Reticulocytes (0.5 1.5%)
10.6 9.5
P, plasma; S, serum.
Results for the plasma samples collected from this patient demonstrated increased plasma LDH and hemoglobin and a normal potassium concentration.
These findings, along with low serum haptoglobin and increased reticulocyte count, strongly suggest in vivo hemolysis. Of interest was the finding of higher serum potassium and hemoglobin concentrations and LDH activity compared with plasma. These findings suggest that further lysis of erythrocytes occurred during the process of clotting of serum samples, with subsequent release of potassium, LDH, and hemoglobin into the serum. Thus, this patient hadin vivo hemolysis due to PNH. In addition, there is concomitantin vitrohemoly- sis following the collection of blood into tubes not con- taining anticoagulant, with further lysis of erythrocytes occurring during the process of clot formation. It may be advantageous to collect blood from patients with immunohemolytic anemia into tubes containing an anti- coagulant such as heparin in order to prevent further hemolysis of cells that might occur during clotting.
In Vitro Hemolysis
Poor specimen quality is recognized as the most fre- quent source of errors in the pre-analytical phase of
testing. Among samples submitted to the clinical labo- ratory for testing and that are deemed unsuitable for analysis, in vitrohemolysis accounts for approximately 40 70% of the cases [1,14,15]. In vitro hemolysis can occur at a variety of stages, including during phlebot- omy, sample handling and processing, and storage. In addition, red cell fragility may be more pronounced in some patients, resulting in a higher likelihood of in vitro hemolysis.Table 5.2 lists the various causes of in vitrohemolysis.
In vitro hemolysis can cause a positive or negative bias in an analyte. The mechanisms causing bias with in vitro hemolysis include proteolysis of analytes due to release of intracellular compounds, release of throm- boplastic substances, dilution of the analyte due to release of cytoplasmic contents, release of the analyte from erythrocytes, and analytical interference due to hemoglobin and other intracellular substances [15].
Analytes such as potassium, LDH, and AST that are
TABLE 5.2 Common Causes ofin VitroHemolysis and the Various Stages in which They Occur
Phlebotomy
Collection from catheter
Collection of capillary blood (finger stick or heel stick) Needle gauge
Length of time that tourniquet is used Fist clenching by patient
Tube underfilled
Vigorous mixing of sample following collection Lack of mixing following collection
Transfer from syringe into Vacutainer tube Specimen transport
Transport via pneumatic tube
Significant time delay in specimen transport to laboratory Transport of specimen at inappropriate temperature Specimen processing
Significant time delay between receipt of specimen and centrifugation
Centrifuge temperature extremes Speed of centrifugation
Re-centrifugation of previous centrifuged specimen Poor barrier separation if gel barrier tubes used Specimen storage
Improper storage temperature following analysis Length of time that specimen is stored following analysis
1Adapted from Blanket al.[13].
56 5. HEMOLYSIS, LIPEMIA, AND HIGH BILIRUBIN
present within erythrocytes at concentrations greater than approximately 10 times those seen in extracellular fluids will cause an increase following hemolysis.
Also, the greater concentration for some analytes, such as potassium, neuron-specific enolase, and acid phos- phatase, observed in serum compared with plasma is due to release of these compounds from platelets dur- ing fibrin clot formation[16]. Interference from hemo- globin may be due to the spectral properties of this compound, which has an absorbance peak at 420 nm and shows significant absorbance between 340 and 440 nm and between 540 and 580 nm. In addition to the spectral interference effects of hemoglobin, this compound may interfere due to the reactivity of its iron atoms, which can participate in oxidation- reduction reactions or in reactions that involve hydro- gen peroxide (see Chapter 8 for examples of tests using hydrogen peroxide and peroxidase assays). Although many analytes are subject to interference effects from in vitro hemolysis, the influence of in vitro hemolysis on measured potassium concentrations is probably the most widely recognized. Based on the frequency at which potassium is measured in the clinical laboratory, along with the serious consequences of misdiagnosis of hypokalemia or hyperkalemia, this analyte is widely recognized as being affected byin vitrohemolysis.
The prevalence of in vitro hemolysis can vary widely depending on the patient population being tested; whether trained phlebotomists or inexperienced individuals are collecting the sample; and whether the sample is processed on-site or is sent to a remote site for processing, with significant time delays between collections and processing. With respect to the type of patient population that is being tested, the prevalence of in vitrohemolysis in outpatients has been found to be approximately 90 times less than that of samples collected from patients in the emergency department, in which studies often show in vitro hemolysis to be present in approximately 10% of samples [17]. Other hospital locations associated with a high prevalence of in vitro hemolysis include pediatric and neonatal wards, in which finger stick and heel stick samples are often collected and which are associated with high rates ofin vitrohemolysis.
Poor blood collection practices contribute toin vitro hemolysis. A study that evaluated the root causes of in vitrohemolysis found that approximately 80% of the samples with in vitro hemolysis were attributed to aspirating blood too vigorously through a needle into a syringe (31%), collecting blood from a butterfly nee- dle into a syringe (20%), collecting from an intrave- nous catheter into a syringe (17%), or collecting from an infusion port into a syringe (12%) [14]. Of interest, errors in handling, including freezing of specimens, accounted for 1% of the causes ofin vitrohemolysis.
The use of automated analyzers for measuring plasma hemoglobin in patient blood samples has proven to be a reliable means of identifying in vitro hemolysis. Some instruments report in vitro hemolysis using a semiquantitative scale, whereas others report the actual plasma hemoglobin concentration. However, once identified, laboratories utilize a variety of differ- ent mechanisms for dealing with specimens that are hemolyzed, including outright rejection of hemolyzed samples, analysis of hemolyzed samples and reporting of results with a disclaimer stating that those analytes affected by in vitro hemolysis may be incorrect, or correction or adjustment of the measured analyte con- centration using a correction factor based on the mag- nitude of hemolysis. The vast majority of laboratories either reject samples within vitrohemolysis or analyze the samples and report the results with a comment stating that the results may not be accurate.
Reporting a result that has been corrected or adjusted based on the magnitude of in vitro hemolysis is not a recommended approach[18]. The wide range of correc- tion factors that have been proposed for correcting mea- sured potassium for in vitro hemolysis demonstrates how problematic the use of correction factors can be. For example, correction factors that have been proposed for adjusting potassium in samples with in vitro hemolysis range from an increase of 0.20 mmol/L of potassium per 100 mg/dL of plasma hemoglobin to an increase of 0.51 mmol/L per 100 mg/dL of plasma hemoglobin [18,19]. The wide range of correction factors that have been proposed highlights the fact that factors influenc- ing hemolysis are not as simplistic as they seem. Factors such as interindividual variability in erythrocyte hemo- globin concentrations or in the concentration of erythro- cytes, the effect of erythrocyte age on intracellular concentrations of various analytes, and differences in red cell membrane fragility likely contribute to the dif- ferences that are seen in recommended correction factors [19 21]. For example, decreased erythrocyte concentra- tions may lead to faster flow of cells through the needle during phlebotomy, resulting in increased shear forces and cell membrane rupture [22]. Also, variability in erythrocyte membrane permeability as a result of dis- ease can impact the effects of in vitrohemolysis. Older erythrocytes have greater permeability to cations com- pared to younger cells, and they contain approximately half the potassium content of younger cells [23,24].
Hemolysis that occurs as a result of mechanical trauma to cells during blood collection is likely to cause lysis of older cells containing different concentrations of certain analytes, whereas hemolysis induced by lysis of all cells—due, for example, to freezing and thawing of blood—will induce lysis of all cells. In addition, condi- tions such as chronic lymphocytic leukemia have been associated with increased fragility of leukocyte
membranes, with release of intracellular contents from these cells during the collection of blood[25].
The mechanisms causing in vitro hemolysis during collection of blood from a patient may be very different from the mechanisms that have been employed in stud- ies designed to investigate the effect ofin vitrohemoly- sis. Mechanisms used to simulate in vitro hemolysis include osmotic lysis of cells, freeze thaw cycles, phys- ical disruption of cells by forcing through a small-bore needle, physical disruption of clotted blood with a wooden applicator stick, and homogenization of whole blood in a blender [19,22,24,26 29]. In addition, some studies remove platelets and leukocytes prior to inducing hemolysis, whereas other studies do not.
The use of these different methods to simulatein vitro hemolysis probably accounts for the wide variability in results that is sometimes reported regarding the effects of in vitro hemolysis on measured analyte concentrations.
Evaluation of the effects of hemolysis on laboratory test results using paired blood samples collected dur- ing the same phlebotomy draw, in which one sample is hemolyzed and the other sample is free of hemoly- sis, shows that the effects of in vitro hemolysis can be very different from those in studies in which hemoly- sis is introduced by artificial means [18]. Methods used to mimicin vitro hemolysis, such as the addition of a whole blood lysate prepared by the freezing and thawing of whole blood or osmotic lysis of cells fol- lowing the addition of distilled water to packed cells, typically result in the lysis of all cells, including reticu- locytes, mature erythrocytes, platelets, and leukocytes.
However,in vitrohemolysis due to mechanical disrup- tion of cells during blood collection typically does not result in lysis of all cells. Older erythrocytes are more prone to shear stresses that occur during sample col- lection compared with younger cells [30]. Shear stress that occurs during phlebotomy can result in the forma- tion of erythrocyte membrane pores that allow the leak of small ions such as potassium but block passage of larger molecules such as LDH, AST, and hemoglobin [31]. Thus, the use of plasma hemoglobin as a marker of in vitro hemolysis may not show a direct relation- ship with the loss of small ions, such as potassium, from cells due to shear stress forces.
CASE REPORT A 74-year-old male was seen by his physician for a routine physical. A general chemistry screen consisting of electrolytes and liver and renal function tests was ordered, and a blood sample was collected into a plain evacuated tube. The phlebotomist mentioned to the physician that she had a difficult time collecting blood from the patient due to the lack of “good veins,” and she needed to use a 23-gauge needle to collect the sample because she was out of
21-gauge needles. After collecting the sample, the phle- botomist placed the tube of blood in a rack and took her lunch break. She returned 1 hr later, and she cen- trifuged the sample and removed the serum from the cells. The technologist who analyzed the sample observed that the serum had a pink to red color but failed to note this observation on the report that was sent to the physician. The following are laboratory results obtained on serum:
Test (Reference Interval) Result
Glucose (#99 mg/dL) 55
Potassium (3.2 4.6 mmol/L) 5.8 Lactate dehydrogenase (133 248 U/L) 322 Aspartate transaminase (#48 U/L) 62 Alanine aminotransferase (#55 U/L) 18 Creatinine kinase (#200 U/L) 93
The results obtained from this patient were consis- tent for a sample that had been delayed in processing to remove the serum from cells and that was also hemolyzed. The relatively low glucose suggested some delay in processing, with metabolism of glucose by cells. The rate of disappearance of glucose in the pres- ence of blood cells has been reported to be approxi- mately 10 mg/dL per hour, but the rate increases with glucose concentration, temperature, and white blood cell count. The increased AST and LDH could be indic- ative of liver or skeletal muscle injury, but the normal alanine aminotransferase and normal creatinine kinase values suggested that liver or skeletal muscle injury was not present in this patient and that AST and LDH were likely increased due to in vitro hemolysis. The increased potassium was also consistent with in vitro hemolysis. In addition, the lengthy delay in processing of the sample could have led to leakage of potassium out of cells into the serum. The use of small-bore nee- dles like the one used for collection of blood from this patient is discouraged because needles larger than 21 gauge have been found to be associated with increased rates of in vitro hemolysis. Small-bore needles can cause a larger vacuum force being applied to the blood, causing increased shear stress on cells and thus causing them to rupture.