| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Department of Medicine, Division of Endocrinology (R.L., M.-F.L.), and Department of Biochemistry (M.M.), Centre Hospitalier de l’Université de Montréal, Montréal, Québec, Canada; and Collaborative Research for Effective Diagnostics (C.E.A.) and Division of Clinical Biochemistry (G.D.F.), Faculté de Médecine et des Sciences de la Santé (R.L., M.-F.L., C.E.A., G.D.F.), Université de Sherbrooke, Sherbrooke, Québec, Canada
Address all correspondence and requests for reprints to: Dr. Guy Fink, Division of Clinical Biochemistry, Faculté de Médecine et des Sciences de la Santé, Université de Sherbrooke, 3001 12th Avenue North, Sherbrooke, Québec, Canada J1H 5N4. E-mail: gfink.chus{at}ssss.gouv.qc.ca.
 |
Abstract |
|---|
 Top Abstract References |
|---|
Results: We describe here the case of a 41-yr-old man with metastatic MTC. Despite extensive disease in the neck as well as metastatic lesions in the liver, his serum calcitonin, measured with a commercial one-step immunoradiometric assay, was only minimally elevated (244 ng/liter). After serial dilutions, a nonlinear relationship became evident, suggesting the presence of a "hook effect." Treatment of the serum with heterophilic blocking reagent revealed no change. Calcitonin was then measured with a different immunoradiometric assay and revealed a much higher level. Similar discrepancies were found in different samples from various patients when analyzed with different calcitonin immunoassays.
Conclusion: To our knowledge, this is the first reported case of a phenomenon such as the hook effect in a calcitonin immunoradiometric assay in patients with MTC. Being aware of this phenomenon is important, because a low calcitonin result could give false reassurance to both the patient and the clinician and could dramatically change the prognosis of the patient.
MEDULLARY THYROID CARCINOMA (MTC) is a relatively rare disease, accounting for 3–5% of all thyroid malignancies, and has a higher mortality rate than well-differentiated thyroid cancers. Serum calcitonin, a well-known tumor marker for MTC, can be quantified by a variety of immunoassays, such as immunoradiometric assays (IRMAs), and its level is usually proportional to the extent of the tumor burden. Because surgery is the only effective treatment for patients with MTC, the postoperative level of serum calcitonin will indicate whether residual disease was left behind and whether reintervention is necessary (1). However, limitations exist in the calcitonin assays, because interferences may cause erroneous results. In fact, the presence of heterophilic antibodies may give erroneously high results of calcitonin by cross-linking the capture and tracer antibodies in the absence of the analyte (2). Although it is well known that the "high-dose hook effect" may give falsely low results for several tumor markers, such as prostate-specific antigen, CA 19-9, CA 125, and others (3), it has never been described for calcitonin. We recently encountered a male patient with metastatic MTC. Despite extensive disease in the neck as well as metastatic lesions in the liver, his serum calcitonin level was only minimally elevated. We therefore hypothesized that a hook effect could be responsible for falsely low results of calcitonin measurements.
The patient, a 41-yr-old man, presented with enlarged lymph nodes, and biopsy was suggestive of MTC. Before the initial surgery, a staging set of imaging was performed, which suggested extensive bilateral lymph node involvement on computed tomography of the neck. Also, a possible liver metastasis was confirmed to be metabolically active by 18F positron emission tomography. His preoperative serum calcitonin level was only 103 ng/liter (reference interval, 0–8 ng/liter), and his serum carcinoembryonic antigen (CEA) level was 127 µg/liter (reference interval, 0–5 µg/liter). A total thyroidectomy with modified radical neck dissection was performed in May 2004, and the pathological report confirmed metastatic MTC in the majority of lymph nodes. In July 2004, the patient underwent right liver lobectomy, which revealed four liver metastases. A computed tomography of the neck, performed a few months later, revealed the presence of suspicious lymph nodes in the cervical area that were hypermetabolic on 18F positron emission tomography and were subsequently removed.
Because we found the level of calcitonin to be low for the extensive load of neoplasia, we wondered whether a hook effect was possible. All serum calcitonin measurements were executed with the one-step DSL-7700 ACTIVE IRMA kit (Diagnostic Systems Laboratories, Inc., Webster, TX), which has a linearity of 1–500 ng/liter and a normal range of less than 8 ng/liter. The assays used were all performed according to the manufacturer’s instructions. After several dilutions, a nonlinear relationship became evident, suggesting the presence of a hook effect. In fact, in September 2004, his serum CEA was 50 µg/liter, and his serum calcitonin levels were as follows: 48 ng/liter undiluted, and 205, 342, 546, 702, 619, and 662 ng/liter after 1:5, 1:10, 1:20, 1:50, 1:100, and 1:200 dilutions, respectively (Table 1
). Repeated dilutions in subsequent specimens showed the same pattern, but the serum CEA level remained stable over time, varying from 47–53 µg/liter. To verify whether the presence of heterophilic antibodies could cause this inappropriate response, we treated the serum specimen with heterophilic blocking reagent (HBR; Scantibodies Laboratory, Santee, CA). The calcitonin ratio before and after HBR treatment was the same as that with control serum (99 ng/liter before and 107 ng/liter after HBR), suggesting no interference by heterophilic antibodies. Also, rheumatoid factor, which can interference with the aggregation of Igs, was negative in our patient. Because it was unusual to find a phenomenon such as the hook effect within the linearity of our assay, we analyzed the calcitonin specimen from our patient with a different IRMA, the IRMA-human calcitonin (hCT) (CIS Biointernational, Gif-sur-Yvette, France), which has a linearity of 0–1500 ng/liter and a normal range of 0–10 ng/liter. The serum calcitonin level analyzed with DSL-7700 was 75 ng/liter undiluted and 974 ng/liter after 1:200 dilution. The same calcitonin specimen was measured with IRMA-hCT and was greater than 1500 ng/liter undiluted and 2728 ng/liter after 1:10 dilution. In contrast to the DSL 7700 assay (Table 1), the IRMA-hCT showed a curve parallel to the standard curve using serial dilutions of a sample from our patient (>1500 ng/liter undiluted). We, therefore, compared the serum calcitonin levels of several patients with MTC using the DSL-7700 ELISA BioSource (BioSource International, Camarillo, CA), DSL-1200 RIA (Diagnostic Systems Laboratories, Inc.), and IRMA-hCT. Because the Nichols-Advantage, an assay frequently used in the United States, was previously shown to be similar to the IRMA-hCT (correlation coefficient, 0.91) (4), we did not use it in this study. The results were consistently lower with the DSL-7700 kit compared with the three other methods (Fig. 1). Furthermore, for one patient, the level of calcitonin obtained with DSL-7700 suggested minimal residual disease, but the calcitonin level obtained from IRMA-hCT was significantly higher, changing the clinical management (calcitonin, 7 ng/liter with DSL-7700 and 42 ng/liter with IRMA-hCT). However, even if IRMAs are more specific than RIAs, variability in results between different assays is observed. This can be explained by the use of different antibodies recognizing different epitopes on the calcitonin molecule or by other poorly understood factors.
|
|
In a one-step assay, if an extremely high antigen concentration is present, the capture and signal antibodies are saturated with the antigen, thus preventing a sandwich formation. When the liquid phase is discarded, most of the antigen is lost with the signal antibodies; thus, the antigen concentration measured is falsely low. The DSL-7700 ACTIVE calcitonin IRMA is a one-step method, using two polyclonal antibodies. Potential errors may occur with this type of assay, explaining the results we obtained. Because the assay is a one-step method, the calcitonin molecule may be sandwiched by two signal antibodies, because both capture and signal antibodies are polyclonal, leaving a lower result than expected when the liquid phase is discarded. However, because the calcitonin IRMA-hCT kit is a two-step method, using two monoclonal antibodies, this problem is less likely to occur.
The first description of the hook effect in the literature was made by Miles et al. (5) in the early 1970s with a two-site IRMA for ferritin. It was then described when hormones were being oversecreted and measured using IRMA. In fact, hook effects in the measurements of prolactin (6, 7, 8, 9, 10, 11, 12), GH (13), TSH (14, 15), gonadotropin (LH and FSH) (16, 17), and human chorionic gonadotropin-ß (18, 19, 20) assays were reported. When tumor markers are measured by two-site IRMA in situations where there is extensive disease, a hook effect may be possible. It is now well described for prostate-specific antigen (21, 22, 23, 24, 25), CA 125 (26), and CA 19–9 (27). Although calcitonin is a well-known tumor marker for MTC and is measured by different immunoassays, the hook effect was never reported previously. However, interference in its dosage was reported by Tommasi et al. (2), causing falsely high calcitonin levels in the presence of heterophilic antibodies in a patient with a multinodular goiter without MTC. Other substances have been reported to produce erroneously high measurements of calcitonin. Morimoto et al. (28) demonstrated that high levels of vitamin C, urea, and creatinine were able to cause falsely high levels of calcitonin when measured by RIA. Even if our results are lower than the classical hook effect, we definitively observed a phenomenon resembling a hook effect with the DSL-7700 assay compared with other immunoassays.
Another explanation for a low serum calcitonin level in patients with metastatic MTC could be tumor dedifferentiation. In this situation, a decreased or stable level of calcitonin is usually accompanied by an elevated or rising CEA level (29). This does not seem to be the explanation in our case, because the CEA level remained stable. Therefore, when an inverse relationship appears in the levels of calcitonin and CEA (low calcitonin, but high CEA) in a patient with MTC, a problem with the calcitonin assay should be suspected, and either dilutions should be performed or the results confirmed with another assay in addition to considering the possibility of tumor dedifferentiation.
In conclusion, clinicians following patients with cancer and using tumor markers need to be aware of phenomena such as the hook effect, because it could give false reassurance to both the patient and the clinician. In patients with MTC, it is important to know that the calcitonin level could be falsely low, because in the presence of metastases, surgery is the only effective treatment for this cancer (1). Furthermore, it is essential to use a reliable calcitonin assay, because some assays may give consistently erroneous results, as we have found with the DSL-7700 ACTIVE IRMA, rendering the follow-up very difficult. We therefore suggest that clinicians use the DSL-7700 ACTIVE IRMA in the care of patients with MTC to verify whether similar problems could occur in their laboratory. In addition, before using any commercial method in clinical practice, it is important to perform dilution curves and spiking experiments. To our knowledge, this is the first reported case of a phenomenon such as the hook effect for calcitonin immunoassay in patients with metastatic MTC. Being aware of this phenomenon will help clinicians to manage patients with this type of carcinoma.
|
Footnotes |
|---|
M.-F.L. is a Junior 2 Clinician Researcher with the Fonds de Recherche en Santé du Québec.
First Published Online November 8, 2005
Abbreviations: CEA, Carcinoembryonic antigen; HBR, heterophilic blocking reagent; hCT, human calcitonin; IRMA, immunoradiometric assay; MTC, medullary thyroid carcinoma.
Received June 28, 2005.
Accepted October 27, 2005.
|
References |
|---|
|
TopAbstractReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Tommasi and S. Raspanti Comparison of Calcitonin Determinations by Polyclonal and Monoclonal IRMAs Clin. Chem., April 1, 2007; 53(4): 798 - 799. [Full Text] [PDF]
|
|
Read all eLetters
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
