CASE PRESENTATION
A 45-year-old male with recurrent metastatic Ewing sarcoma presented to Memorial Sloan Kettering Cancer Center for potential enrollment in a clinical trial. Initial laboratory results were notable for increased alkaline phosphatase levels [(ALP2, 495 U/L; (reference interval: 33–97 U/L); Abbott Architect C8000] and γ-glutamyl transferase levels [GGT, 393 U/L (reference interval: 0–73 U/L); Abbott Architect C8000] but were otherwise unremarkable, including normal levels of aspartate aminotransferase (AST) and alanine transaminase (ALT). Elevations in ALP and GGT levels suggested liver disease, although the patient had no previous history of liver abnormalities. Given the patient’s history of metastatic bone disease, it was suspected that the observed elevation in ALP was due to increased bone ALP isoenzyme. However, it was important to document the origin of the increased ALP as the patient was to be enrolled in an investigational trial that would define drug dosage and side effects. Elevations in ALP over the course of the trial that were attributable to liver-specific isoenzymes would be reported as a dose-limiting toxicity and could delay treatment or potentially exclude the patient from the clinical trial. Conversely, elevations in ALP originating from bone and secondary to disease would not be reported as toxicity.
The laboratory was consulted regarding testing options to identify the specific isoenzyme(s) responsible for this patient’s increased ALP. To this end, a bone-specific ALP chemiluminescent immunoassay (Beckman Access Ostase) was performed. The result of this assay was 68.5 μg/L (reference interval: 0.0–20.1 μg/L), which represents a significant elevation of bone ALP. To investigate the potential contribution of other ALP isoenzymes, electrophoresis was preformed at a reference laboratory. This method uses electrophoresis to separate ALP isoenzymes followed by densitometry for quantification. Similar to results obtained at MSKCC, the patient’s total ALP level was increased at 524 U/L (reference interval: 45–115 U/L). Interestingly, the bone ALP level was abnormally low at 7.8% (reference interval: 19.1–67.7%) of total ALP level, whereas liver ALP level was increased and accounted for 92.2% [reference interval: 27.8–82.1%] of total ALP levels. Neither placental isoenzymes nor intestinal isoenzymes were detected. In light of these results, the clinical care team started a work-up for potential hepatobiliary conditions while the laboratory further investigated the discrepant results.
An alternative method using a combination of enzyme activity and heat inactivation was performed at a reference laboratory. The results of this assay were also increased for both liver (359 U/L; reference interval: 0–94 U/L) and bone (101 U/L; reference interval: 0–55 U/L) isoenzymes. Imaging studies were performed and revealed the presence of congestive hepatopathy resulting from a perfusion abnormality secondary to hepatic venous congestion. This condition typically presents with increased ALP and GGT levels with normal AST and ALT levels. An MRI confirmed progression of disease in the right humerus and new metastatic lesions in the scapular body. The imaging studies confirmed the presence of liver dysfunction and progressive bone disease, both of which could result in increased liver ALP and bone ALP levels. Owing to these complications, the patient was excluded from the clinical trial.
DISCUSSION
ALP is a hydrolase enzyme responsible for the de-phosphorylation of many types of molecules, including proteins and nucleotides. There are 4 primary isoenzymes of ALP, including bone, liver, intestinal, and placental. Approximately 95% of total ALP activity in human serum is derived from bone and liver sources that occur in a ratio of approximately 1:1 in healthy adults (1). In adults, elevations in ALP can be observed in numerous conditions, including pregnancy, congestive heart failure, ulcerative colitis, and bacterial infections. Increased liver ALP level is most frequently observed in hepatobiliary conditions, particularly cholestasis, whereas elevations in bone ALP levels are encountered in pathologies of increased osteoblast activity such as Paget disease or certain cancers that either originate from bone or have spread to bone (2). Because ALP can arise from several sources and can be increased as a result of a variety of conditions, it is often necessary to test for ALP isoenzymes to determine the origin.
There are 3 primary methods available to evaluate ALP isoenzymes. Electrophoresis is 1 method, where ALP isoenzymes are separated on the basis of differences in charge. However, because liver and bone isoenzyme mobility is virtually equivalent, an additional step is required. Wheat germ lectin (wheat germ agglutin: WGA) is added to exploit the differences in sialation between liver and bone isoenzymes. The bone isoenzyme is typically rich in sialic acid and will react with WGA in the agarose gel and precipitate. Once separated, isoenzymes are read on a densitometer for quantification and expressed as percentages of total ALP (3). The second method takes advantage of the differences in heat stability between isoenzymes. In this method, total ALP activity is measured from the sample directly and after heating. Bone ALP is heat-labile; therefore, measurements taken after heating will be absent of bone ALP activity. Liver ALP is heat stable and will be active in both measurements. Bone ALP can then be determined by subtracting the 2 measurements (4). The third method is an immunoassay used to quantify bone ALP levels exclusively (5). There are several commercially available bone ALP immunoassay kits, although the Beckman Access Ostase assay was used in this study. This method is a 1-step immunoenzymatic assay. In short, a mouse monoclonal antibody specific to bone ALP is added to paramagnetic particles coated with goat antimouse polyclonal antibody. Patient serum is then added to the coated particles, and any bone ALP present will bind to the antibone ALP monoclonal antibody. A chemiluminescent substrate is added and light generated by the reaction is directly proportional to the concentration of bone ALP in the sample.
In this case, given the patient’s history of Ewing sarcoma and lack of symptoms or history of liver dysfunction, elevations in bone ALP levels were investigated first. Bone ALP was increased; however, due to a difference in units between the bone ALP immunoassay and the total ALP enzymatic assay (immunoassay measuring quantity in microgram per liter vs enzymatic assay measuring activity in units per liter), the results cannot be used interchangeably and must be interpreted appropriately. The observed increase in bone ALP provided evidence that the total ALP elevation was at least partially from bone origin, but it was not possible to definitively identify bone ALP as the sole contributor. Subsequent tests using electrophoresis and heat stability to identify elevations in other ALP isoenzymes did reveal an elevation in liver ALP levels, but these also demonstrated a discrepancy in the bone ALP level. Imaging studies revealed an underlying pathology for both the increased liver and bone isoenzymes. It was therefore concluded that the discordant results were the result of decreased bone ALP determined by the electrophoresis method.
The ability of the electrophoretic method to distinguish bone ALP from liver ALP is based on the premise that bone ALP present in serum is rich in sialic acid. In serum, there are 3 primary bone ALP isoforms, namely, B1, B2, and B/I (6). Interestingly, it has been demonstrated that these isoforms differ in the number of sialic acid residues present. A previous study estimated the number of sialic acid residues to be 29 and 45 for each B1 and B2 homodimer, respectively. B/I consists of 70% bone and 30% intestinal isoenzyme and has very few sialic acid residues (7). The different isoforms precipitate to varying degrees in the presence of WGA based upon the number of sialic acid residues present, with B2 showing the highest level of precipitation followed by B1 and B/I. Similarly, a study by Farley et al. demonstrated that heat inactivation and immunoassay were superior to WGA precipitation assay for detecting bone ALP in the serum of post-menopausal osteoporotic subjects (8). It is plausible that the isoform present in this patient had fewer sialic acid residues that could account for the discrepant normal bone ALP result using the electrophoretic method. It is interesting to note that the 3 ALP isoforms exhibited similar heat inactivation kinetics (7).
In this case, 3 different methods were used to determine the source of increased ALP in a patient with metastatic bone disease and no history of liver dysfunction. The correct interpretation of the initial bone ALP result by immunoassay was instrumental in the subsequent diagnosis and treatment of a previously undiagnosed liver condition. This investigation highlights the importance of understanding the specific methodologies and limitations behind different ALP assays.
Table 1.
Summary of laboratory results.
| Method | Results | |
|---|---|---|
| Total ALP (performed at MSKCC) | 529 U/L [33–97] | |
| Electrophoresis/Densitometry | Total ALP | 524 U/L [45–115] |
| Bone ALP | 92.2% [27.8–82.1] | |
| Liver ALP | 7.8% [19.1–67.7] | |
| Heat inactivation (56 °C for 10-min)/activity | Total ALP | 460 U/L [40–120] |
| Bone ALP | 101 U/L [0–55] | |
| Liver ALP | 359 U/L [0–94] | |
| Bone-specific (ELISA) | 68.5 μg/L [0.0–20.1] | |
| GGT | 393 U/L [0–73] | |
| AST | 27 U/L [10–37] | |
| ALT | 35 U/L [5–37] | |
TAKEAWAYS.
Total ALP level can be increased because of many reasons, and >1 isoenzyme may contribute to the elevation.
Investigating the source of alkaline phosphatase elevation may require testing by several different methods.
Understanding the methods and limitations of each assay used to measure ALP isoenzymes is important for correct result interpretation.
Footnotes
Nonstandard abbreviations: ALP, alkaline phosphatase; AST, aspartate aminotransferase; ALT, alanine transaminase; GGT, γ-glutamyl transferase; WGA, wheat germ agglutin.
Authors’ Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
REFERENCES
- 1.Magnusson P, Degerblad M, Sääf M, Larsson L, Thorén M. Different responses of bone alkaline phosphatase isoforms during recombinant insulin-like growth factor-I (IGF-I) and during growth hormone therapy in adults with growth hormone deficiency. J Bone Miner Res 1997;12: 210–20. [DOI] [PubMed] [Google Scholar]
- 2.Siddique A, Kowdley KV. Approach to a patient with elevated serum alkaline phosphatase. Clin Liver Dis 2012;16:199–229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schreiber WE, Whitta L. Alkaline phosphatase isoenzymes resolved by electrophoresis on lectin-containing agarose gel. Clin Chem 1986;32:1570–3. [PubMed] [Google Scholar]
- 4.PetitClerc C Quantitative fractionation of alkaline phosphatase isoenzymes according to their thermostability. Clin Chem 1976;22:42–8. [PubMed] [Google Scholar]
- 5.Broyles DL, Nielsen RG, Bussett EM, Lu WD, Mizrahi IA, Nunnelly PA, et al. Analytical and clinical performance characteristics of Tandem-MP Ostase, a new immunoassay for serum bone alkaline phosphatase. Clin Chem 1998;44:2139–47. [PubMed] [Google Scholar]
- 6.Magnusson P, Häger A, Larsson L. Serum osteocalcin and bone and liver alkaline phosphatase isoforms in healthy children and adolescents. Pediatr Res 1995;38: 955–61. [DOI] [PubMed] [Google Scholar]
- 7.Magnusson P, Farley JR. Differences in sialic acid residues among bone alkaline phosphatase isoforms: a physical, biochemical, and immunological characterization. Calcif Tissue Int 2002;71:508–18. [DOI] [PubMed] [Google Scholar]
- 8.Farley JR, Hall SL, Ilacas D, Orcutt C, Miller BE, Hill CS, Baylink DJ. Quantification of skeletal alkaline phosphatase in osteoporotic serum by wheat germ agglutinin precipitation, heat inactivation, and a two-site immunoradiometric assay. Clin Chem 1994;40: 1749–56. [PubMed] [Google Scholar]