In 1965, Frederickson, Levy, and Lees published an article that caught the attention of the medical world. They divided lipoprotein disorders into five basic “phenotypes,” based primarily on electrophoresis of serum obtained after 10 hours of overnight fasting. The sixth phenotype was added later when type II was split into IIa and IIb. Lipoprotein phenotyping was originally proposed as a means of classifying congenital disorders of lipid metabolism according to the lipoprotein abnormality involved to provide more specific therapy. In this way abnormal serum levels of specific lipids such as cholesterol could be traced to abnormalities in specific lipoprotein groups, which, in turn, could suggest certain congenital or acquired etiologies. In time, however, the original intent and limitations of the system were sometimes forgotten, and treatment was sometimes begun—based on the phenotype suggested by the patient’s lipid profile or lipoprotein electrophoretic pattern—as though the patient had a congenital lipoprotein disease. Congenital disease cannot be treated directly, so therapy is directed against the abnormal lipids. Since congenital disease is present in only a small percentage of persons with abnormal serum lipid values, and since lipid disorders due to acquired conditions are best treated by therapy directed at the condition responsible for the lipid abnormality, some of these persons were not being managed appropriately. In other words, the symptoms (abnormal levels of lipids or lipoproteins) were being treated rather than the underlying etiology.

Laboratory tests in lipoprotein disorders

Screening tests for lipoprotein abnormality include determination of serum cholesterol and TG levels plus visual inspection of serum (or plasma) for presence of chylomicrons after the specimen has been kept overnight at refrigerator temperature. After this incubation, chylomicrons will rise to the surface as a creamlike surface layer. If the serum or plasma remains cloudy without formation of a definite surface layer, this represents VLDLs. The specimen should be obtained after a 10- to 12-hour fast. This triad of tests serves not only as a screening procedure but in the majority of cases is sufficient to establish the phenotype. Normal results on all three tests are reasonable evidence against serious lipoprotein disease. However, occasional patients with disease can be missed, due sometimes to laboratory variation, borderline abnormality, or the overlap of normal and abnormal persons in statistically established reference ranges referred to earlier. Lipoprotein electrophoresis and ultracentrifugation are useful in a minority of cases as confirmatory or diagnostic tests. Electrophoresis is helpful in differentiating Frederickson type I from type V disease, some cases of type IIa from IIb, and some cases of type II from type IV. Ultracentrifugation is useful mainly in diagnosis of type III disease. Electrophoresis is not needed in the majority of patients. Lipoprotein phenotyping is done preferably on outpatients rather than hospitalized patients due to the short-term effects of serious illness on lipid metabolism as well as the other factors mentioned in the previous discussion of serum cholesterol measurement. However, screening of hospital inpatients can detect possible abnormality that can be verified later under more basal conditions.

A short summary of lipoprotein phenotype patterns shows that the presence of a creamlike layer of chylomicrons after overnight incubation of fasting serum in the refrigerator indicates type I or type V. If the serum below the chylomicron layer is clear, this suggests type I; if the serum below the chylomicron layer is cloudy or turbid, this suggests type V. Fasting serum obtained from patients with the other phenotypes does not contain chylomicrons. If fasting serum does not contain chylomicrons, elevation of serum cholesterol levels with normal TG levels suggests type II disease, whereas the reverse suggests type IV disease. If both cholesterol and TG levels are significantly abnormal, the disease may be type II or type III. Type II disease has recently been subdivided into IIa and IIb. Type IIa has increased cholesterol but normal TG. Type IIb has elevated cholesterol and TG. Type III is uncommon. It is similar to IIb in that both cholesterol and TG levels are elevated, but it frequently has a slightly different electrophoretic pattern (broad beta) and always has a peculiar “floating beta” component (a beta-mobility protein that floats at 1.006 density instead of 1.013), which can be demonstrated only by ultracentrifugation. Type IV phenotype has elevated TG and normal cholesterol levels.

Although the majority of patients can be phenotyped using the triad of biochemical tests, some overlap occurs among the phenotypes because cholesterol is present to some extent in all of the major lipoprotein fractions and because triglyceride is found in both chylomicrons and the VLDL fractions. Also, the laboratory reference range may inadvertently influence a phenotype decision in some cases, depending on whether the upper limit of the reference range was derived from a sample of the local population, obtained from data published in the literature, or structured according to findings in populations with low risk of atherosclerosis.

Phenotypes I, III, and V are uncommon. Phenotypes II and IV constitute the majority of hyperlipoproteinemias. Type IV is probably more common than type II. Most type IV patients have the acquired form. The majority of type II patients also have the acquired form (the most common etiology being a high-cholesterol diet), but type II is more frequently congenital than type IV.

Some serum specimens from patients with types IIb, III, or IV disease may have a somewhat cloudy or faintly milky appearance. This must be differentiated from the thicker, creamlike precipitate characteristic of increased chylomicrons.

Certain considerations affect interpretation of these laboratory results. Patients should be on a normal diet for several days before testing and must be fasting for at least 10 hours before a specimen is drawn. If a test cannot be done the same day, the serum must be refrigerated but not frozen; freezing alters prebeta and chylomicron fractions, although cholesterol and triglyceride determinations can be done. Various diseases may produce certain phenotype patterns or may falsely change one pattern into another. Changes in diet, medications, activity levels, stress, and other factors may alter a mild or borderline abnormality or produce mild abnormality. In addition, the laboratory results have ± 5%-10% built-in variability for technical reasons.

On electrophoresis, occasional persons display increased prebeta but normal TG levels. If laboratory error is ruled out, these patients may have a congenital variant called “Lp system” or “sinking prebeta.” This is found in up to 10% of the U.S. population and is the same as the lipoprotein group now collectively called Lp(a). Lp(a) was discussed earlier as an independent risk factor for CHD.

Plasma or serum may be used for lipoprotein analysis. Plasma collected with EDTA is preferred if the specimen cannot be tested the same day. Plasma values for TG are about 2%-4% lower than for serum.

Two rare diseases display characteristic lipoprotein patterns on electrophoresis. Tangier disease has no alpha peak. Bassen-Kornzweig syndrome (associated with “pincushion” RBCs called acanthocytes, and neurologic abnormalities) lacks a beta peak.