| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Laboratory of Endocrinology/Bone Metabolism, Laboratory of Clinical Chemistry and Supportive Care Clinic, Service de Médecine et Laboratoire d’Investigation Clinique H.J. Tagnon, Institut Jules Bordet, Université Libre de Bruxelles, 1000 Brussels, Belgium;
2
DiaSorin Corp., Stillwater, MN 55082;
a address
correspondence to this author at: Institut Jules Bordet, Rue Héger-Bordet 1, 1000 Brussels, Belgium
Parathyroid hormone-related protein (PTHrP) plays a key role in hypercalcemia of malignancy (HM), particularly in humoral HM (HHM) (1)(2). Serum PTHrP is increased in 47–100% of HM patients (3)(4)(5)(6) and in almost all patients with HHM. In up to two-thirds of hypercalcemic patients with bone metastases, serum PTHrP is increased (5)(7). PTHrP assays may aid in the diagnosis of HHM (8).
IRMAs that measure large N-terminal regions appear to be optimal for HM evaluation (9). With the first commercially available IRMA (Nichols Institute), PTHrP concentrations were found to be increased in 40 of 81 (49%) HM patients. Moreover, healthy subjects had detectable values up to 2.6 pmol/L (10), contrary to some reports (5)(8)(11). With an IRMA from Mitsubishi Petrochemical Co (12), PTHrP was increased in all 46 studied patients with hypercalcemia and solid tumors, and most healthy subjects had values <1 pmol/L. However, almost all patients had HHM, and the specificity toward normocalcemic cancer (C.NoCa) patients was not thoroughly investigated.
We have evaluated a commercially available PTHrP IRMA. We particularly studied the assay specificity for HM patients compared with normocalcemic patients presenting an active cancer, with patients in complete remission from their cancer, and with patients suffering from osteopenia, osteoporosis, or chronic renal failure (CRF).
The first control group included 94 patients (88 women and 6 men), median age 61 years (range, 26–80 years), referred to our bone clinic for a low bone mass or established osteoporosis. The patients were classified (13) as follows: 39 osteopenic (T score at the lumbar spine of -1.78 ± 0.09, mean ± SE); 29 osteoporotic (T score -3.16 ± 0.10); 26 established osteoporosis (low bone mass and presence of one or more fragility fractures). The second control group included 31 patients in end-stage CRF. The third control group included 62 patients (56 women and 6 men), median age 57 years (range, 36–80 years), in complete remission from malignancies for a median of 5 years (range, 2–22 years). The tumors included breast (n = 47), cervix (n = 3), Hodgkin lymphomas (n = 3), head and neck (n = 2), prostate (n = 2), thyroid (n = 2), and 3 miscellaneous (1 endometrium, 1 ovary, and 1 pituitary). The fourth control group included 85 normocalcemic cancer patients (74 women and 9 men), median age 62 years (range, 33–89 years; C.NoCa group). The tumors included breast (n = 65), lung (n = 5), prostate (n = 5), cervix (n = 3), endometrium (n = 3), and 4 miscellaneous (1 head and neck, 1 stomach, 1 bladder, and 1 soft tissue sarcoma). Bone metastases were definite in 74 (extensive in 66), absent in 8, and doubtful in 3 patients (14).
The HM group included 38 patients (18 women and 20 men), median age 57 years (range, 26–85 years), recruited consecutively. They were studied before any specific hypocalcemic therapy; 5 were treated by radiotherapy and 15 by chemo- or hormonotherapy at the time of evaluation. The tumors included breast (n = 12), head and neck (n = 8), lung (n = 4), and 14 miscellaneous (3 of unknown origin, 2 bladder, 2 vulva, 1 melanoma, 1 rectum, 1 cervix, 1 esophagus, 1 gallbladder, 1 teratocarcinoma, and 1 sarcoma). Twenty patients had epidermoid tumors, 15 had adenocarcinomas, and 3 had other histologies. Bone metastases were definite in 19 (extensive in 12), absent in 18, and doubtful in 1. Mean (± SE) serum calcium (corrected for protein) was 3.37 ± 0.07 mmol/L (13.5 ± 0.3 mg/dL); the median was 3.24 mmol/L (range, 2.62–5.24 mmol/L). PTH was adequately suppressed at 6.3 ± 0.6 ng/L; the median was 4.5 ng/L (range 4.0–17.2 ng/L) (15).
PTHrP was measured (IRMA of DiaSorin Corp., Stillwater, MN) with human recombinant PTHrP 1-84 for calibrators and controls. Calibrators were supplied lyophilized and were reconstituted in PTHrP-free human serum (zero calibrator). The method for the IRMA is as follows. Anti-PTHrP 1-40 antibody is bound to polystyrene beads, and an anti-PTHrP 57-80 antibody is labeled with 125I. Samples are incubated sequentially with the antibodies, and the polystyrene beads are then washed. The remaining radioactivity is measured.
Blood was collected into EDTA-coated tubes containing aprotinin at 1000 kallikrein inhibitor units/mL blood (16), put on ice immediately after collection, and separated within 60 min; the plasma was stored at -20 °C for <6 months.
As described previously (9)(14)(15), blood markers included total serum calcium (reference interval or "normal values", 2.12–2.57 mmol/L), calcium corrected for protein concentrations (17) (2.12–2.62 mmol/L), ionized Ca (Ca2+, measured by the Ciba-Corning electrode; 1.05–1.27 mmol/L), Pi (0.71–1.45 mmol/L), magnesium (0.65–1.05 mmol/L), alkaline phosphatase (<110 U/L), intact PTH (10–55 ng/L), osteocalcin (0.7–5.6 µg/L), and 1,25(OH)2vitamin D3 (15–42 ng/L); the last three assays were from DiaSorin. In 2-h fasting morning samples, we determined phosphorus, calcium, creatinine (calcium:creatinine ratio, <0.59 mol/mol in healthy postmenopausal women), hydroxyproline (<40.4 µmol/mmol creatinine), the renal threshold for phosphorus [TmP/glomerular filtration rate (GFR); 0.81–1.36 mmol/L] (18), and renal reabsorption of calcium [algorithm provided by J.P. Bonjour (TRCaI; 2.36–2.86 mmol/L)] (19).
Data are expressed as the mean ± SE and/or by the median (range)
when indicated. We performed classical tests (ANOVA,
2, and nonparametric correlations) with
StatviewTM II (Abacus Concepts) and the
log-rank test for survival analysis with
StatisticaTM 4 (Statsoft).
The detection limit (mean + 3 SD of 20 determinations of the zero calibrator) was 0.3 pmol/L. Results were considered reportable only when they were greater than the first calibrator (1 pmol/L). The within-run imprecision (CV) was 4.0%, 3.3%, and 3.4% at PTHrP concentrations of 3.3, 12.5, and 92 pmol/L (n = 20). Interassay CVs (n = 11) were 8.8%, 5.2%, and 3.8% at 3.4, 12.8 and 96 pmol/L. Dilution curves from plasma samples containing high endogenous PTHrP concentrations (four patients with HM) were parallel to the calibration curve. Recovery of 5–100 pmol/L PTHrP 1-84 from three PTHrP-free serum samples was 83–100% (93% ± 3%). Although all of our clinical samples were collected into aprotinin-containing tubes, we have started to determine the need for protease inhibitors. EDTA-whole blood samples from 10 healthy donors were supplemented with 30 pmol/L PTHrP 1-84 and stored for up to 4 h at room temperature with or without aprotinin (400 kIU/mL). The recovery was 98%, suggesting that protease inhibitors may not be necessary for samples stabilization.
Values were unreportable in 124 healthy subjects (<1 pmol/L). PTHrP
was also unreportable in all 94 patients of the osteoporotic group and
in all 31 patients with CRF, was reportable in only 1 of 62 (2%)
patients in complete remission from cancer, in 4 of 85 (5%) patients
in the C.NoCa group, and was increased (reportable) in 31 of 38 (82%)
cancer patients with HM (Fig. 1
). In patients with HM, mean PTHrP concentrations were 7.4
± 1.1 pmol/L; the median value was 6.0 pmol/L (range, <1–26.5
pmol/L). PTHrP was increased in all 20 patients with epidermoid tumors
and in 10 of 15 (67%) patients with adenocarcinomas. PTHrP was
increased in all 18 patients without bone metastases and in 12 of 19
(63%) patients with bone metastases. Mean PTHrP thus was higher in
patients without bone metastases, 9.6 ± 1.3 pmol/L, than in
patients with bone metastases, 5.2 ± 1.6 pmol/L (P
<0.05).
|
Baseline values of the biochemical markers in patients with HM are
summarized in Table 1
. Patients were divided according to the presence (n = 31)
or the absence (n = 7) of increased PTHrP. Patients with increased
PTHrP had lower osteocalcin, Pi, and TmP/GFR.
Their survival was also shorter than patients with PTHrP concentrations
within the reference interval (log-rank test, P <0.05), but
this was not confirmed when only patients with breast cancer were
considered.
|
In patients with HM, correlations were significant between PTHrP and Pi (rs = -0.52; P <0.01), TmP/GFR (rs = -0.51; P <0.01), and TRCaI (rs = 0.38; P <0.05). Correlations between PTHrP and Ca2+ or calcium corrected for protein concentrations were significant in the group of patients without bone metastases (rs = 0.53–0.54; P <0.05).
The technical validation demonstrates that this assay is robust and reproducible, and that it satisfies the criteria of recovery and linearity. From a clinical point of view, the data indicate that it is highly sensitive (82% of increased values in patients with HM) and specific for HM (specificity of 95% in patients with an active cancer, 99% in patients in remission from cancer, and 100% in patients with benign diseases or healthy subjects). It appears to be at least as sensitive and specific as in-house assays (2)(5)(11)(20) and more so than other commercial IRMAs (10)(12). Moreover, this IRMA is the first commercial assay able to demonstrate the pathological relevance of circulating PTHrP concentrations: we observed expected correlations between PTHrP and Pi, TmP/GFR and TRCaI, corresponding to the well-known effects of PTHrP on phosphate and calcium tubular reabsorption (21)(22)(23).
In patients with HM, PTHrP was increased in 100% of the patients with epidermoid tumors, whether or not they had bone metastases. PTHrP was increased in 67% of patients with adenocarcinomas and in 58% of patients with breast cancer and bone metastases. These findings confirm and expand the original work of Grill et al. (5) and indicate that, in addition to tumor-derived locally produced bone-resorbing factors, hypercalcemia of breast cancer can be attributable to paraneoplastic secretion of PTHrP, just as is classically observed in HHM.
Acknowledgments
This study was supported in part by grants from Fondation Medic, Fonds National de la Recherche Scientifique (Belgium, Contract FRSM 3.4577.96 and Télévie Contract 7.4509.97), and Fondation Lambeau-Marteau. We thank Dr. MacFarlane (DiaSorin) for assistance in providing the most recent data, and DiaSorin for the gift of the PTHrP assay kits.
Footnotes
fax 32-2-541-3310, e-mail jj.body{at}bordet.be
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
