Validation of 125I-hCG as a marker for elimination of hCG and stability of 125I-hCG after in vivo injection in humans

We have recently introduced 125I-hCG as an elimination marker in patients with human chorionic gonadotrophin (hCG) producing testicular cancer. 125I-hCG is a well-known reagent in clinical biochemistry and is used extensively in hCG assays. Previous studies have shown that the iodination process leaves the hCG molecule mainly intact. The iodination, purification and stability of 125I-hCG tracer are described. The aim of the present study was to determine whether or not 125I is associated with hCG after the injection of 125I-hCG intravenously (i.v.) in humans. Three different methods were used. Following injection of 125I-hCG, the plasma disappearance of radioactivity and hCG were followed for a period of 28 days in 13 normal subjects. Serum from a normal healthy male following injection of 125I-hCG was analysed using a double antibody direct binding radioimmunoassay specific for holo-hCG and high performance liquid chromatography (HPLC). Following injection of 125I-hCG in eight normal healthy males and five normal healthy females, the disappearance of radioactivity and hCG showed identical paths in the 28 days follow-up period. The bindable radioactive fraction of immunologically active hCG in serum of a normal healthy male following injection of 125I-hCG was between 57.0% and 72.1%, and was constant over time. HPLC showed similar elution pattern of serum from a normal healthy male injected i.v. with 125I-hCG and 125I-hCG. Using three different methods, we were able concurrently to demonstrate the association of 125I with hCG in humans up to 28 days after injection of radiolabelled hCG i.v. Thus, information about the expected elimination of hCG can be obtained by following the elimination of activity in plasma after injection of 125I-hCG. © 1999 Cancer Research Campaign

the slow disappearance as caused by a slow elimination rather than a production of hCG from vital tumour cells. 125 I-hCG is a well-known reagent in clinical biochemistry and is used extensively in hCG assays. Previous studies have shown that the iodination process leaves the hCG molecule mainly intact (Catt et al, 1972). During iodination, two epitopes on the α chain are destroyed (Berger et al, 1988). The structural integrity of hCG and its ability to bind to specific antibodies and LH/hCG receptors is well retained after iodination. After bolus injection of 125 I-hCG and hCG into rats, no difference in disappearance was seen (Markkanen et al, 1980). The present investigations were initiated to answer the following questions of importance for the utilization of 125 I-hCG as an elimination marker: Does iodination of hCG affect its elimination in humans? Does the drop of radioactivity in serum reflect elimination of 125 I-hCG, metabolites, or dissociation of 125 I from injected 125 I-hCG i.v.?

MATERIALS AND METHODS
The study was approved by the local ethics committee and all volunteers gave their informed written consent before entering the study.

Iodination of hCG
The hCG used for labelling was supplied by LEO Pharmaceutical Co., Denmark. The crude hCG was obtained from the urine of pregnant women and delivered without additives. LEO used the crude hCG as raw material for their human ovary luteinizing hormone (LH)-stimulating pharmaceutical Physex ® . The hCG was purified to 3000 IU mg -1 , sterilized, tested for pyrogens, oestrogens and abnormal toxicity, and conformed according to (Council of Europe, 1986). The labelling of the hCG was carried out by Isopharma A/S, Kjeller, Norway. The following procedure was used: for labelling 1.5 ml hCG (1 mg ml -1 ) were dissolved in 0.01 M phosphate-buffered saline (PBS), pH 7.4, mixed with 90 MBq Na 125 I in a tube containing 60 µg iodogen®. The mixture was allowed to incubate for 10 min before it was transferred to an empty tube. For purification, the reaction mixture was then applied to a PD-10 column (Pharmacia, Sephadex G-25, medium coarse), equilibrated with 1 mg human serum albumin (HSA) ml -1 PBS and eluted with 1 mg HSA ml -1 PBS. Following the first 2.5 ml, eight fractions of 0.5 ml each were collected. Fractions with less than 2% iodide were collected, mixed and diluted to 2 MBq ml -1 with 1 mg HSA ml -1 PBS. The purification procedure was carried out twice. The product was sterile filtered using a 0.22 µm filter (Durapore), tested for radiochemical purity, using high-voltageelectrophoresis on the day of production, and for pH, pyogens and bacterial growth 14 days after production. The 125 I-hCG used had a labelling per cent of > 88%, radiochemical purity with < 1.7% iodide, no detectable contamination with 125 I-albumin, specific activity 50-60 MBq mg -1 , pH 7.0-7.5, and was sterile and free of pyogens (LAL < 175 EU 20 ml -1 ).
High-voltage-electrophoresis was used to test the radiochemical stability of 125 I-hCG following incubation at 4°C for 5, 7, 14, 20 and 29 days respectively. The iodide content was < 5%. To investigate the influence of the iodination process on the integrity of hCG, one sample of hCG was taken through the same procedure for labelling except for the iodogen and iodination step. The product was then analysed for intact hCG and free β-hCG following 18 days of storage using immunofluorometric assays. In the labelled and unlabelled product, intact hCG concentrations of 42 000 IU l -1 and 56 000 IU l -1 and free β-hCG concentrations of 270 IU l -1 and 550 IU l -1 , respectively, were found. Differences in concentrations are caused by difference in the dilution step.

Dosimetry
After injection of 2.66 MBq of 125 I-hCG in a healthy normal male, serum samples were obtained once a day for 21 days. The plasma disappearance curve of activity from 125 I counted in the spectrum of 15-75 KeV was constructed. The data were fitted a triexponential model minimizing least squares using non-linear regression according to the Marquardt-Levenberg method. Area under curve was calculated. Assuming the total activity was deposited in the patient, a volume of distribution of 7.3l, the total accumulated radiation dose was calculated to be < 0.05 mSv following injection of 3 MBq 125 I-hCG.
Elimination of 125 I-hCG in normal healthy subjects Giving 400 mg KI orally at least 1 h before injection of 125 I-hCG prevented thyroid uptake of 125 I. A well-functioning i.v. catheter was placed in the cubital vein of the subject. Serum samples were drawn prior to injection of 125 I-hCG to determine background activity. Then, 3 MBq 125 I-hCG (Isopharma A/S, Kjeller, Norway) were given as a single i.v. bolus injection. The specific activity of the labelled hCG was 50 MBq mg -1 and the concentration was 3000 IU mg -1 . Thus, the quantity of hCG given was 180 IU. The elimination of hCG was studied in two ways: (a.) Immunologically by measuring the plasma disappearance of the total hCG, using a modified Wallac immunofluorometric assay. The assay was modified by doubling the sample volume and changing detecting antibody to INN-hFSH-158 specific for the α-chain epitope α 5 . The capturing antibody used was INN-hCG-2 specific for the β-chain epitope β 5 (Dirnhofer et al, 1994). The sensitivity was 0.1 IU l -1 (3.IS) and LH cross-reactivity 0.4%. (b.) By measuring the rate of plasma disappearance of activity of 125 I. Radioactivity was measured in a well counter (COBRA auto gamma, Packard, USA) set to detect within the energy range 15-75 KeV. The background activity has been subtracted from all the presented data. All plasma samples were counted at the same time to avoid calculated correction for the decay of 125 I. The relative counting error was less than 5%. Serum samples were taken twice a week for the first 2 weeks and then once a week for the following weeks. The 125 I concentrations are presented as counts min -1 for a 3 ml aliquot of plasma. To determine non-specific binding (NSB), 0.4 ml of serum were incubated with mAb and excess (100 IU) of non-labelled hCG (Pregnyl ® , Organon, The Netherlands). After addition of 0.2 ml immunoadsorbent consisting of goat Ig-antimouse Ig covalently linked to sepharose 6B (Pharmacia, Uppsala, Sweden) and incubation for 3 h at room temperature with vigorous agitation, activity was counted in a γ-counter (Wallac, Turku, Finland) for 15 min to determine the total activity (T a ) for each sample. The tubes were then washed three times using PBS-Tween-80 and counted for 15 min to determine total bound activity (B t ) and NSB. Counter background activity (BG) was determined with empty tubes. As more than 90% of radioactivity after iodination of hCG is associated with the α-subunit of hCG, this assay will measure holo-hCG (Markkanen et al, 1980). All samples were run in duplicate. Mean values were used for calculation.

Immunoadsorption
Total added activity corrected for background activity (T acorr ) was calculated as T acorr = T a -BG and specifically bound activity as B s = B t -NSB.
High-performance liquid chromatography Serum samples drawn from a normal healthy volunteer (male, age 40 years, weight 75 kg, height 181 cm) at 2 h, 4 h, 1, 2, 3 and 4 days after injection of 1.75 MBq 125 I-hCG were analysed by highperformance liquid chromatography (HPLC) using a Waters Deltapac C18 column equilibrated with 0.1% TFA (trifluoracetic acid) in water, and flow 1 ml min -1 . From each serum sample, five samples of 100 µl each of serum diluted to 50% by sterile saline (0.9% sodium chloride in water) to prevent clotting of the system, B s *100 Per cent B s of T acorr was calculated as %B s = % T acor were applicated to the column consecutively. After the fifth application, the chromatography was continued as reverse phase (RP) chromatography by increasing the concentration of (acetonitrile + 0.1% TFA) from 0% at 5 min to 100% at 20 min. From the 25th min the column was again eluted with 0.1% TFA.
The outflow from the column was collected in 3-min fractions in order to determine the low radioactivity. This procedure, unfortunately, loses some resolution in detection of elution time but this was necessary due to the low count rate which was undetectable by the HPLC-radioactivity flow detector. The first six fractions were collected during application of the first four samples. The remaining ten fractions were collected during the RP period. The fractions were counted for radioactivity in the energy range 15-75 KeV covering the energy range of 125 I for 10 min each using a γ-well counter (Packard). Samples were analysed in order of increasing activity to avoid carry over between samples in the column. Counter background activity was determined on empty tubes and subtracted from each sample. Analysing unlabelled hCG and 125 I-hCG used for i.v. injection, UV detector (280 nm) and radioactivity detector (15-75 KeV) were used.

The in vivo elimination of 125 I-hCG in normal healthy subjects
Eight healthy males with a median age of 37 years (range 29-45), median weight of 78 kg (range 66.4-83.0) and a median height of 185 cm (range 171-195) were injected with a median dose of 2.92 MBq 125 I-hCG (range 2.48-3.10). Plasma disappearance of hCG and radioactivity are shown in Figure 1.
Five healthy females with a median age of 41 years (range 37-42), median weight of 60.0 kg (range 49.2-62.8) and a median height of 166 cm (range 163-168) were injected with a median dose of 2.89 MBq 125 I-hCG (range 2.63-3.02). Plasma disappearance of hCG and radioactivity are shown in Figure 2.
The plasma disappearance of radioactivity and hCG showed identical paths from injection of 125 I-hCG to day 28 in both males and females. The plasma disappearance showed an initial fast elimination phase followed by a phase with a decreasing rate of disappearance over time. In order to describe the covariation between 125 I-hCG and hCG, measurements were logarithmically transformed and a linear least squares regression analysis was applied followed by a t-test procedure for significance. An intercept α = 5.53 (P < 0.0001), slope β = 1.045 (P < 0.0001) and correlation coefficient between 125 I-hCG and hCG, r 2 = 0.94 (P < 0.0001) were found.

Immunoadsorption
The radioactivity counted on 3 ml of proband serum before injection of 125 I-hCG was identical to background activity. The percentage of immunoreactive tracer in the serum of a healthy proband injected with a bolus of 125 I-hCG remained constant over time between 67.1 and 72.1% of total radioactivity (Table 1). Only after 4 days did it apparently drop to 57%, but this might be due to the vulnerability of the test system at extremely low count rates.

HPLC
The radioactivity counted on 3 ml of proband serum before injection of 125 I-hCG was identical to background activity. No radioactivity was detected in the first eight fractions, indicating no free iodine. The elution pattern of radioactivity in serum samples drawn from 2 h to 3 days after injection of 125 I-hCG is identical to the elution pattern of 125 I-hCG (Figure 3). The radioactivity of 125 I-hCG eluted in fractions 13 and 14. The radioactivity in serum after injection of 125 I-hCG eluated in fraction 13 and 14 constant over time. In Figure 4, the percentage of radioactivity applied to the column from each sample is shown. The recovery was 101% at 2 h, 92% at 4 h, 91% at day 1, 118% at day 2 and 69% at day 3.
HPLC of 125 I-hCG used for injection showed a UV 280 nm and radioactivity detector signal at the same elution time. HPLC of the  unlabelled hCG used as raw material for 125 I-hCG showed a UV 280 nm detector signal at the same time as radioactivity detector signal during HPLC of 125 I-hCG.

DISCUSSION
We have recently introduced 125 I-hCG as an elimination marker in patients with hCG producing testicular cancer (Christensen et al, 1999). For this purpose, it is important that 125 I is associated with hCG in vivo. Following injection of 125 I-hCG in eight healthy males and five healthy females, the plasma disappearance of radioactivity ( 125 I-hCG) and immunologically measured hCG showed identical paths from the day of injection to 28 days after injection (Figures 1 and 2). The correlation between 125 I-hCG and hCG measurements was found to be statistically significant (r 2 = 0.9443, P < 0.0001). The multiphasic shape of the plasma disappearance of 125 I-hCG and hCG clearly demonstrates the inappropriateness of calculating one half-life for the elimination of hCG, as the half-life steadily increases over the 28-day observation period in this study. Serum hCG values in the eight healthy males were less than 0.25 IU l -1 and in the five healthy females less than 0.5 IU l -1 before injection of 125 I-hCG. Thus, the main part of immunologically measured hCG originates from the BtT acorr (%) as a function of T acorr* Linear regression y = 1.66*ln(x) +51.7. SE on slope is ± 1.77. The value 1.66 found is not significantly different from 0 (t-test) injected 125 I-hCG. No side-effects were seen following the injection of 125 I-hCG. The crude hCG was delivered without additives. Following iodination we were not able, with the present methods, to find any difference in immunological reactivity or elution pattern of labelled, unlabelled, or crude hCG material, suggesting that the hormone remains intact following iodination using the described iodogen and purification method. Furthermore, we analysed LH concentrations in serum samples from some normal healthy males and females after injection of 125 I-hCG and were able to detect a small transient decrease in LH concentrations (data not presented) suggesting that the labelled hormone was still hormonally active. After iodination of hCG, immunoreactivity of 125 I-hCG is maximum 70-80% of the total radioactivity. These percentages are only found when highly purified hCG (14 000 IU mg -1 ) is used for labelling purposes. We used material purified from urine of pregnant women with an immunological activity of 3000 IU mg -1 . Using double-antibody direct-binding radioimmunassay, we were able to bind a constant fraction of immunologically active hCG in serum following injection of 125 I-hCG. The bindable fractions were between 57 and 72% (Table 1, Figure 5). This is comparable to the known immunoreactivity achievable after iodination of hCG. At day 4, we were only able to bind 57%, but this could be due to the vulnerability of the test system at low count rates. The bindable fraction of 125 I-hCG was 49%. This is probably due to a content of damaged hCG molecules following iodination and storage (radiation damage). Iodination of hCG destroys two out of 22 epitopes on the hCG molecule as described by Berger et al (1988). Clearance of radioactive iodide, following ingestion of KI, damaged hCG molecules, β-hCG and α-hCG produced from urine of pregnant women, is very rapid from the circulation of man (Wehmann and Nisula, 1980;Liu et al, 1989). Thus, shortly after injection of 125 I-hCG, the main part of radioactivity in serum will be associated with hCG, as damaged fractions of protein will be cleared from the circulation.
HPLC on serum from a healthy normal male following i.v. injection of 125 I-hCG showed no radioactivity in the first eight fractions, indicating no free iodine. The radioactivity constantly eluted in fractions 13 and 14 (Figures 3 and 4). HPLC of 125 I-hCG also showed the radioactivity to elute in fractions 13 and 14. Thus, 125 I-hCG and radioactivity in serum after injection of 125 I-hCG showed an identical elution pattern. As HPLC of crude hCG and 125 I-hCG showed the same elution times, 125 I was associated with hCG in the 125 I-hCG used for injection.
In summary, we were able to demonstrate concurrently in each method the association of 125 I with hCG in humans for up to 28 days after i.v. injection of radiolabelled hCG by use of three different methods. Thus, 125 I-hCG is stable in vivo and information of the expected elimination of hCG can be obtained by following the elimination of activity in plasma after injection of 125 I-hCG.