Download CCK Receptor Targeting: Stability & Affinity of CCK & Gastrin Peptides and more Exercises Medicine in PDF only on Docsity!
new class of receptor-bindingpeptidesfor diagnosisand therapy
ofCCK-Breceptor-expressingtumors,suchasMTCorsmallcell
lungcancer.Nonsulfatedgastnnderivativesmaybe preferable
because of their CCK-B receptor selectivity,hence lower accre tion in normal CCK-A receptor-expressing organs. Key Words: gastrin;cholecystokinin;cholecystokinin-B/gastnn receptors;radiolabeledpeptides;medullarythyroidcancer;small cell lung cancer J NucI Med 1999;40:1029-
he development of regulatory peptides as tools to visualize and, more recently, to treat malignant tumors has been an important focus of interest over the past years (1—3). The successful development of sufficiently stable radiola beled somatostatin analogs, such as ‘23I-1@'r3-or II1In diethylenetriamine pentaacetic acid (DTPA)-D-Ph&-octreo tide, for diagnostic puposes (1,2) and the introduction of ‘61ThD@fpAor @°Y-DOTAconjugates for therapeutic appli cations have opened new horizons in nuclear oncology (3—5).Other regulatory peptides, such as vasoactive intesti nal polypeptide (6), substance P (7) and gastrin-releasing peptide/bombesin derivatives (8), have emerged as poten tially useful candidates for in vivo scintigraphy and radiopep tide therapy. Although somatostatin receptor scintigraphy has proven to be a valuable tool for staging gastroenteropancreatic tumors (e.g., carcinoids), its sensitivity and accuracy in other neoplasms, such as medullary thyroid cancer (MTC) and small cell lung cancer, are limited (9—11).Because of the outstanding diagnostic accuracy of the pentagastrin test in detecting the presence, persistence or recurrence of malig nant C cells (even when they are smaller than a size detectable by conventional morphological imaging meth ods), we postulated the expression of the corresponding receptor type in human MTC (12). Indeed, receptor autora diographic studies have demonstrated the expression and presence of cholecystokinin (CCK)-B/gastrin receptors not only in more than 90% of MTCs (13) but also in a high percentage of other tumor types, such as small cell lung cancer, stromal ovarian cancers and astrocytomas (14).
The high sensitivity of pentagastrin stimulation in detecting pnmary or metastatic medullarythyroid cancer (MTC) suggests widespread expression of the corresponding receptor type on human MTC. Indeed, autoradiographicstudieshave demon strated cholecystokinin (CCK)-B/gastrin receptors not only in more than 90% of MTCs but also in a high percentageof small cell lung cancers, some ovarian cancers, astrocytomas and potentially a variety of adenocarcinomas. The aim of this study was to systematicallyscreen and optimize,in a preclinicalmodel and a pilot clinical study, suitable radioligands for targeting CCK-B receptors in vivo. Methods: A variety of CCK/gastrin related peptides, all bearing the C-terminal CCK receptor-binding tetrapeptide sequence Trp-Met-Asp-PheNH2 or derivatives thereof, were studied.They were radioiodinatedby the lodogen or Bolton-Hunter procedures. The peptides were members of the gastrin or CCK families, which differ by the intramolecular positionof a tyrosylmoiety.Their stabilityand affinitywere studied in vitro and in vivo; their biodistnbution and therapeutic efficacywere tested in nude mice bearing subcutaneoushuman MTC xenografts. Diethylenetriamine pentaacetic acid (DTPA) derivativesof suitablepeptideswere synthesizedsuccessfully, and their preclinical and initial clinical evaluations were per formed, labeled with 111ln.Results: All members of the CCK or gastrinfamilieswere stablein serum(withhalf-livesof several hours at 37°C);nevertheless, the stability of those peptides bearing N-terminal pGlu residues or D-amino acids was signifi cantlyhigher.In accordancewiththeircomparablylow affinity, nonsulfatedmembersoftheCCKfamilyshowedfairlylowuptake in the tumor and other CCK-B receptor-expressingtissues. SulfatedCCK derivativesperformedsignificantlybetterbutalso displayed a comparably high uptake in normal CCK-A receptor expressing tissues. This effect was probably due to their similar affinityforbothCCK-AandCCK-Breceptors.Besttumoruptake and tumor-to-nontumor ratios were obtained with members of the gastrin family because of their selectivity and affinity for the CCK-B receptor subtype. Pilot therapy experiments in MTC bearinganimals showed significantantitumorefficacycompared with untreated controls. DTPA derivatives of minigastrin were successfullydeveloped.In a pilotclinicalstudy,radioiodinated and 1111n-labeledderivatives showed excellent targeting of physi ological CCK-B receptor-expressing organs, as well as all known tumorsites.Conclusion:CCK/gastnnanalogsmaybe a useful
Received Jul. 7, 1998; revision accepted Feb. 12, 1999. Forcorrespondenceor repnntscontact:ThomasM. Behr,MD,Department of Nuclear Medicine, Georg-August-University, Robert-Koch-Str. 40, D- Gottingen, Germany.
CH0LECYsT0KININ-B RECEPTOR-EXPRESSINGTUMORS •Behr et al. 1029
Radiolabeled Peptides for Targeting
Cholecy stokinin-B/Gastrin
Receptor-Expres sing Tumors
Thomas M. Behr, Niels Jenner, Martin Béhé,Christa Angerstein, Stefan Gratz, Friedhelm Raue and Wolfgang Becker
Department ofNuclear Medicine, Georg-August- University, Gottingen, Germany; and Department ofEndocrinology, Ruprecht-Karl University, Heidelberg, Germany
lC@Gastrin PeptideChain length Aminoacidsequence Label derivatIves
Previous studies suggested an even more widespread occur rence of gastrin receptors in colon, pancreatic and stomach cancers (15). Therefore, we undertook pilot experiments in nude mice bearing human MTC xenografts and were able to demonstrate the feasibility of radiolabeled gastrin-I to target
CCK-B receptor-expressing tumors, as well as CCK receptor
expressing normal organs such as the stomach, in vivo in animals and patients (12). However, a variety of problems remained to be solved,
such as the molecular characteristics that render the peptide
an optimal candidate for in vivo targeting of CCK-B receptor-expressing tumors, considering in vivo stability;
affinity to and selectivity for the CCK-B receptor; or the potentially unfavorable accretion in normal organs, such as the liver, bowel, or kidney. The aim of this study, therefore, was to develop and optimize, in a preclinical model as well as in a pilot clinical study, suitable radioligands for targeting
CCK receptors in vivo.
MATERIALS AND METhODS Peptides and Radlolodination We investigated various peptides in this study (Table 1). They either belong to the gastrin or CCK superfamilies or possess characteristics of both (16). Both families differ by the location of
TABLE
Structures and Estimates of the lC@Values of the Radiolabeled Peptides Tested
Cholecystoldnln(CCK)derivatives
Big-gastrin
Gastnn-l
[Leu15]-gastnn-I
[Gln@1°]-gastrin-l
Minigastrin
(D-Leu1]-minigastrln
[Gln@]-minigastnn
Gastrin-Ifragment11— Gastnn-lfragment1—
34 pGlu-Leu-GIy-Pro-GIn-GIy-Pro-Pro-His-Leu-Val-Ala-Asp Pro@Ser@Lys@Lys@GIn-GIy-Pro-Trp-Leu-GIu-Glu-Glu**
rtTrp-p-
17 pGIu-GIy-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyrt Gly-Trp-Met-Asp-PheNH 17 pGlu-Gly-Pro-Trp-Leu-Glu-Glu-Glu-Glu-Glu-Ala-Tyrt GIy-Trp.Leu-Asp-PheNH 17 pGlu-Gly-Pro-Trp-Leu-GIn-GIn-Gln-GIn-Gln-AIa-Tyrt
Gly-Trp-Met-Asp-PheNH
13 Leu-Glu-Glu-GIu-Glu-GIu-Ala-Tyrt-Gly-Trp-Met-Asp PheNH @ 13 PhNH 13 Leu-GIn-Gln-GIn-GIn-Gln-Ala-Tyrt-Gly-Trp-Met-Asp PheNH@ 7 AIa-Tyrt-Gly-Trp-Met-Asp-PheNH 14 pGlu-Gly-Pro-Trp-Leu-GIu-Glu-Glu-Glu-Glu-Ala-Tyrt-Gly
T@
I, BH —0.5x 108M
I —109M
I —109M
I —108M
l,BH —109M
I —109M
I >104M
BH —0.5x 109M
BH —0.5x 109M I, BH —108M
BH >104M
Caerulein
[Thr28,NIe31]-sCCK-25-
sCCK- nsCCK-
[Tyr(SO3H)@]CCK26-
10 pGIu-GIn-Asp-Tyrt-Thr-Gly-Trp-Mat-Asp-PheNH **SO3- 9 ArgM@TyrmrGlyT@NleAsp..pMNHa SO3- 8 Asp-Tyr-Met-Gly-Trp-Met-Asp-PhONH @ 8 SO3- 4 MpTyrMetG@yNH Peptides with common features of both gastrin and CCK SO3-S03 - 8 *MnTyrTyrGlyT@Mp..p@NHa 5 *@AlaT@@Mpp@NH 4 @Trp-Met-Asp-PheNH
13 DTPA-Leu-Glu-GIu-GIu-GIu-Glu-Ala-Tyr-Gly-Trp-Met Asp-PheNH
Cionin des-BOC-pentagastrin CCKfragment30-33(CCK-4) Radiometal-chelate derivative DTPA-minigastnnorits[D-Leu']analog
BH —0.5x 109M BH —0.5x 107M BH
111ln3@ (^) —109M
The potential attachment sites ofthe radioiodinatedBolton-Huntermoieties. tTyrosylmoietiesthatmayberadiolodinatedbythelodogenprocedure. I = lodogen;BH = Bolton-Hunterradlolodinatlon. Aminoacidsequencesin boldfaceindicatethe receptorbindingsequence.
1030 THE Joui@.r'@iOF Nuci@i@ MEDIcINE •Vol. 40 •No. 6 •June 1999
with Ham's F12K medium, supplemented with 20% fetal-calf serum, to yield a 20% cell suspension. Two hundred fifty microli ters of this suspension were injected subcutaneously. After approxi mately 3—4wk, tumors approximately 100—200mg in size (the size used in this study) had developed in more than 90% of the animals.
Blodistribution Studies Thmor-bearing animals were injected intravenously into the tail with the ‘@‘I-,l25I@.or ‘DIn-labeledpeptide (740—1480kBq [20- pCiJ,0.5—1.0pg).Themicewerekilledat10mmor1,3,5or24h. They were bled by retro-orbital puncture, and, after cervical dislocation, the animals were dissected. For detennination of the whole-body retention, whole mice were measured in a well counter. The amount of activity in the tumor and tissues (lung, liver, spleen, kidney, stomach, pancreas, gallbladder, intestine, brain, adrenals, muscle, bone and blood) was determined by gamma scintillation counting using an injection standard to account for physical decay (23,24). The number of animals used for each study was typically 5 per group at each time point. In several studies, animals were coinjected with a mixture of ‘@‘I-and ‘@I-labeled peptides. In these instances, windows were set for each radionu @ clide, and the backscatter of 131!into the window was corrected.
To assess specificityof receptorbinding,parallelgroupsof mice
were given either 1 pig of radiolabeled peptide alone or 1 pg of radiolabeled peptide supplemented with 1 mg ofunlabeled peptide.
Experimental Radiopeptide Therapy Thmor sizes were determined by caliper measurement in three dimensions immediately before therapy and at weekly intervals thereafter. Tumor volumes were estimated by multiplying the product of the three perpendicular diameters by 0.5, assuming an elliptical geometry (23). Tumors were left untreated (controls), given unlabeled gastrin-I or injected with a single dose of ‘31I-labeledgastrin-I or gastrin-I fragment 1—14.Ten to 20 animals were studied in each group. Body weight was recorded weekly, and survival was monitored (23,24). The maximum tolerated dose was defined as the highest possible activity under the respective conditions that did not result in any animal deaths, with the next lO%—20%higher dose level resulting in atleast 10% ofthe animals dying (23,24). Total and differential white blood cell and platelet counts were determined on the day of therapy and at weekly intervals thereafter. Seventy-five-microliter heparinized specimens were collected by retro-orbital bleeding. The samples were counted on a Technicon H3 Auto-Analyzer (Bayer-Diagnostik, Munich, Germany). Blood urea nitrogen and creatinine as indicators of renal function, as well as glutamate oxaloactetate transaminase and alkaline phosphatase as liver function parameters, were determined at the same time points as the blood counts. Means ±SDs were calculated for each group (23,24).
Initial Clinical Evaluation of 131I-L.abeledGastrln-l and
‘11ln@DTPA-Minigastrin
After premedication with omeprazole (20 mg orally, given 2 h before the radiolabeled peptide) and, in the case of radioiodinated peptides, with potassium iodide (200 mg/d for 3 d), 4 patients with metastatic MTC were injected over a 10-mm period with 370 MBq (10 mCi) ‘311-labeledgastrin-I (n = 1) or with 130—170MBq (3.5-4.6 mCi) @In-DTPA-minigastrin(n = 3) (approximately0. pg/kg body weight each). Whole-body scans and SPECT images of the thorax and abdomen were obtained over a 48-h period, using a
double-head gamma camera (Picker Prism 2000; Picker Interna tional) (11,12).
StatisticalAnalysis
Differences in the tissue uptake values and the biodistribution of the radioligands in various tissues were statistically analyzed with the Student t test for unpaired data. Pairwise comparisons were performed with the Wilcoxon rank-sum test. Differences in the therapeutic efficacy between the treatment modalities were ana lyzed by assuming an exponential tumor growth pattern; nonlinear regression analysis based on asymptotic approximation was used.
RESULTS
Stabilfty and Affinity of the Various Peptides In Vitro The peptides (Table 1) were radioiodinated by the Iodo gen or Bolton-Hunter procedures as indicated, depending on the availability of a nonsubstituted tyrosyl moiety (lodogen) or a free amino group (Bolton-Hunter). HPLC analysis of
the radioiodinated peptides revealed a single peak (95% of
the total activity) for Bolton-Hunter-labeled peptides in
contrast to several (up to four) closely related peaks for their Iodogen-labeled analogs. Previous studies had shown that the methionine moieties of CCK or gastrmn derivatives are readily susceptible to oxidation, even under rather mildly oxidizing conditions such as those occurring during Iodogen radioiodination. Accordingly, iodination of [Leu15}-gasttin I, lackinganyMetresidues,led to a single(95% of the total activity) radioiodinated peptide.
To further study the role of methionine oxidation, redox
experiments were undertaken by incubating the labeled peptides for 1 h at 37°Cin 1-mollL H202 or for 48 h in
l-mol/L dithiothreitol (21). HPLC analysis showed nearly
complete interconversion of the various peptide peaks to a homogeneous product under each of these respective condi tions. However, pilot receptor-binding studies failed to show any major influence of the methiomne oxidation on the receptor-binding affinities (data not shown), which is in
accordance with previous reports (25). Furthermore, pilot
studies comparing the biodistribution ofbig-gastrmn, minigas
tin and nonsulfated CCK-8 labeled by the lodogen (1311)
versus the Bolton-Hunter (‘@I)procedure intraindividually did not show any significant difference between the biologi cal behavior of both labels (data not shown). Therefore, the radiolabeled peptides were used for the subsequent studies without further consideration of partially oxidized Met residues.
Table 1 shows the IC50 values of the radioiodinated
peptides obtained from cell-binding studies. Sulfated CCK
and cionin analogs had the highest affinities (nmollL
range), whereas desulfation or the complete removal of the
N terminally located tyrosine-containing parts of the peptide (as is the case in des-BOC-pentagastrmn or CCK-4) led to a loss of affinity by several orders of magnitude. Furthermore,
the presence of an intact C-terminal Trp-Met-Asp-PheNH
sequence is crucial for receptor binding, whereas the methio
nine can be replaced by leucine or norleucine. Similarly high
affinities as those seen for the sulfated CCK analogs were
1032 THEJoui@.r@OFNUCLEARMEDICINE•Vol.40 •No. 6 •June 1999
@ F I‘ Ipoororgans-@-@-.-.-@--n
musde@-@--bone-0-
@
I0 I ‘I • 10 20
observed for some nonsulfated members of the gastrmn family. Stability testing of the various peptides was performed in human serum. At 37°C, all peptides had in vitro serum half-lives of several hours, but those peptides bearing N-terminal pyroglutamic acid (pGlu) residues exhibited significantly longer half-lives (24 h at 37°C)than those with a free N-terminal amino group (data not shown).
On HPLC analysis, a small fraction of the peptides
(@10%) rapidly shifted to a higher molecular weight fraction on only a short serum incubation at 2°C,suggesting binding to serum proteins, whereas the majority of the peptides remained unbound. Incubation at 37°Cled to small molecular weight metabolites, coeluting with di-, tri- or tetrapeptides or monoiodotyrosine (data not shown). A much more rapid metabolic breakdown of minigastrin was ob served compared with the pGlu-protected gastrmn-I. In addition, the replacement of the L-Leu1-moiety in minigas trin by its D-isomer led to an improved serum stability (tl,@ 8 h versus 45 mm at 37°C). Biodistribution Studies in Human Medullary Thyroid
CancerXenograft-BearingNudeMice
Biodistribution studies of the various peptides used were performed in U xenograft-bearing nude mice. Figures 1 and 2 showthe biodistributionpatternsand the tissueuptake values, respectively, of those peptides essentially confined to the known receptor recognition region Trp-Met-Asp PheNH2 (e.g., CCK-4 and des-BOC-pentagastrmn) and of the nonsulfated CCK analogs (nonsulfated CCK-8 and caer ulein). Because similar biodistribution kinetics were ob served for most members within each subfamily (e.g., nonsulfated versus sulfated CCK versus gastrin derivatives), Figures 1, 3 and 4 show the biodistribution of a typical member of the various families over a 24-h period (differen tiating those tissues known or expected to express CCK-B receptors [top panels], organs involved in the excretion of the radiopeptide [middle panels] and pure “blood-pool― organs [bottom panels]), respectively. Furthermore, Figures 2, 5 and 6 quantitativelycomparethe tissueuptakeof the most important tested peptides belonging to these families at 1-h postinjection. This time point was chosen because it was typically the time point with the highest target-specific uptake (Figs. 1, 3 and 4). Accordingly, Figure 1 shows the biodistribution pattern of nonsulfated CCK derivatives. With nonsulfated CCK-8, over the first hour, increasing uptake was seen in the stomach (as the organ with the highest physiological CCK-B receptor density) (26) and the gallbladder as well as, in comparison with other tissues, some prolonged retention of activity in the tumor (Fig. 1, top panel). The major route of excretion was renal, with an initially intense uptake in the kidney (Fig. 1, middle panel). In addition to a potentially specific CCK receptor binding in the gallbladder, a biliary excretion pathway was suggested by an activity transfer into the bowel (Fig. 1, middle panel). All other organs studied, such as the lung, liver, spleen, muscle and bone, behaved as
CCK-Breceptorexpressingtissues
— @mor
--‘-. brain
— s@math -.•—pancreas
0•
excretoryorgans
— @dney
---@-. gaH@adder -@•--b@
I
FIGURE1. Biodistnbutionof nonsulfatedCCK-8in CCK-B
receptor-expressingtissues(top),organsinvolvedinitsexcretion
(middle)andblood-poolorgans(bottom)intimecourse.
typical blood-pool organs with a rapid, exponential clear ance pattern that was parallel to the blood clearance (Fig. 1, bottom panel).
The comparison of the 1-h tissue uptake values of
nonsulfated CCK-8 and nonsulfated caerulein with those of peptides essentially confined to the receptor-binding region, i.e., des-BOC-pentagastrin and CCK-4, shows hardly any stomach or tumor uptake with the latter, which is in contrast
to the, although still rather low, significantly (P < 0.05)
higher uptake of nonsulfated CCK-8 and caerulein in these tissues. This is in accordance with the external scintigraphy (Fig. 7, first scan), which shows hardly any tumor and only rather weak stomach uptake for nonsulfated CCK-8 as the organs with the highest CCK receptor expression (mainly
CHOLECYSTOKININ-BRECEPTOR-EXPRESSINGTUMORS •Behr et al. 1033
timi p.i. (h)
.@
CCK-Breceptorexpressingtissues
@ —tumor @ @•io-@
CCK-Breceptorexpressingtissues 40-i 301 —.—-tumor @
and 9). The uptake in the tumor, stomach and pancreas was significantly lower at the 1-mg saturation dose. Interestingly, despite receptor blocking, the gallbladder uptake was still quite substantial, indicating that this uptake may be due to nonspecific biliary excretion. Surprisingly, the renal accre tion was lowered under receptor saturation conditions as well (P < 0.05), whereas the blood clearance seemed to be slightly slower under high-dose peptide challenge condi tions. No significant differences were found in lung, liver, spleen, bowel, brain, adrenal, muscle or bone uptakes. In addition, the biodistribution of fragment 1—14,lacking an intact C-terminal receptor-binding tetrapeptide, was studied
@--@----
FIGURE3. BiodistributionofsulfatedCCK-8inCCK-Breceptor expressingtissues(top),organsinvolvedinitsexcretion(middle) and blood-pool organs (bottom) in time course.
-@- s@math --•- pancreas --*-. brain
excretoryorgans
— kidney — ..-..-- gallbladder -4-- stomath
- ..- pancreas --‘-. brain
“bloodpoor organs
excretoryorgans^ 5. —.- lung -@ l@
.-@‘4-- muscle bone
Sblood
—kidney
2
-@ .--.. gallbladder --•--b@
(^0) 20
“bloodpoor org@
5. (^) FIGURE 4. Biodistributionof gastrin-l in CCK-B receptor expressing tissues (top), organs involved in its excretion (middle) and blood-pool organs (bottom) in time course.
(Fig. 8). Again, significantly lower uptake of this peptide occurred in the tumor, stomach and pancreas (P < 0.01). whereas there was no difference in the renal accretion of gastrmn-I. ObservatIons on the Influence of Molecular Characteristics on Renal Accretion Figure 10 shows the kidney accretion dynamics for nonsulfated (top panel) and sulfated (middle panel) CCK analogs and members of the gastrmn family (bottom panel). Whereas the initial renal uptake and retention ofthe sulfated, thus highly anionic, CCK analogs (middle panel) were very low, the initial uptake of their nonsulfated analogs was
- lung
---.-.muscle bone
-0-- blood
.@
a
(^0 )
CHOLECYSTOKININ-B RECEPTOR-EXPRESSING TUMORS •Behr et al. 1035
120-i 100-I @. 80-I r •@60@
3 rn
0
@50 I \
tlms p.1. (h)
time p.1. (h)
tu lu H apI ki Sb pa ga bw tissues
ad mu bo
tu lu H
FIGURE 5. Comparisonof 1 h postinjec tion tissue uptake values of radioiodinated sulfated CCK-8 (5CCK-8), [Thr28,Nle31]- sCCK-25—33,cionin and [Tyr(SO3H)27]-
CCK-26—29.tu = tumor;lu = lung;Ii = liver;
spl = spleen;ki = kidney;sto = stomach;
pa = pancreas; ga = gallbladder; bw = bowel; br = brain; ad = adrenal gland; mu = muscle; bo = bone; bI = blood.
spl ki sto pa ga bw br ad mu bo bi tissues
approximately twofold higher (top and middle panels), which held true for members of the gastrmn family (bottom panel) as well. The retention of peptides bearing an N- terminal pGlu moiety was significantly longer (dashed lines in Fig. 10) than with “open-ended―peptides (various dotted lines in Fig. 10). Finally, big-gastrmn, which is the only peptide with two positively charged lysine groups and an N-terminal pGlu moiety, had the highest renal uptake and longest retention of all peptides studied.
InItial Therapeutic Experiments with 131I-Labeled
Gastrln Analogs Because ofthe encouraging “diagnostic―results of CCK-B receptor targeting, we undertook initial experiments to assess whether radiolabeled gastrin may be therapeutically useful. To determine the maximum tolerated dose and dose-limiting organs, varying amounts of gastrin-I or its
fragment 1—14were injected into groups of 10 animals each, starting at 74 MBq (2 mCi) and increasing in lO%—20% steps. Myelotoxicity was found to be dose-limiting (blood count nadirs at 1- to 2-wk postinjection, complete recovery within 2—3wk) with both ‘31I-labeledpeptides, and the maximum tolerated activity was reached at 170 MBq (4. mCi), corresponding to a blood dose of approximately 3. Gy (12). No sign of second-organ toxicity was observed; values for blood urea nitrogen, creatinine and the liver enzymes remained within normal limits. This is in contrast to big-gastrin, in which we observed signs of radiation nephrotoxicity (rising blood urea nitrogen and creatinine levels; data not shown), which is in accordance with its higher renal accretion and retention. The tumor growth was significantly retarded with ‘31I-gastrin-I(P 0.03) corn pared with untreated controls, with animals given the same
—biggastrin T@ 4 — minigastrin E: [GIn@2@-minigastrin
FiGURE 6. Comparisonof 1 h postinjec
tion tissueuptakevaluesof radioiodinated
big-gastnn, gastnn-l, [Leu151-gastnn-l,mini gastrin and [GIn@-@]-minigasthn.tu = tu mor; lu = lung; Ii = liver; spl = spleen; ki = kidney; sto = stomach; pa = pancreas; ga = gallbladder; bw = bowel; br = brain;
ad = adrenalgland;mu = muscle;bo =
bone; bI = blood.
1036 T@ Joui@.r'iiu.OF NUCLEAR MEDICINE •Vol. 40 •No. 6 •June 1999
I 150.@ .1 sCCK- so.:1. flS@3l}sCCK@ —cionin EJ [Tyr(SO3H)2@]-CCK-26-
250, @ 1501
50.1.
.@
1iig@ Img gastñn@ 1pgg@thn-lfr@m.1-
FIGURE8. SpecificityofCCK-Breceptor
binding.Comparisonof biodistributionof 1 pg ‘31l-labeledgastnn-l versus 1 pg radiola beled gastrin, supplemented with 1 mg unla
beled peptideor 1 pg fragment(fragm.)
1—14(lacking receptor-binding tetrapeptide *@ sequence). @ < 0.05; < 0.01; tu = tumor; lu = lung; Ii = liver; spl = spleen; ki = kidney;sto= stomach;pa = pancreas; ga = gallbladder; bw = bowel; br = brain; ad = adrenal gland; mu = muscle; ho = bone; bI = blood.
“In-labeled derivative. No significant difference was ob served between “In-DTPA-minigastrin and its [D-Leu'] analog (data not shown). Pilot Clinical Evaluation of 131l-Labeled Gastrin-I and 111ln-DTPA-Minigastrln Four patients with known or occult metastatic MTC were investigated with either ‘3'I-labeledgastrin-I (one patient) or “In-DTPA-minigastrmn (three patients) (Table 3). Three patients had known lesions, and one patient had occult
lpg
FIGURE 9. Externalscintigramcomparing1 pg ‘311-labeled gastnn-l(left)with1 mgblockingdose(right).
metastatic MTC, as indicated by pathologically elevated carcinoembryonic antigen (CEA) and calcitonin levels with out radiological evidence of disease. All patients had undergone somatostatin receptor and anti-CEA immunoscin tigraphy within 3 mo. A peptide dose of 0.5 j.tg/kg body weight was slowly injected intravenously, which was toler ated by two patients without any side effects, whereas the other two patients had a short episode of pentagastrin stimulation-test-like symptoms (nausea, “strangefeeling― in the abdomen), lasting for only a few minutes and resolving spontaneously. In accordance with the preclinical studies, receptor targeting was seen in all four patients in physiologically CCK-B receptor-expressing tissues (mainly in the stomach), as well as in all known metastatic lesions (Table 3 and Figs. 14 and 15). In the second patient with occult metastatic MTC (without evidence of disease on somatostatin or anti-CEA scintigraphy), “In-DTPA-niinigas tin detected a previously occult liver lesion that was confirmed by CT 4 wk later, whereas splenic involvement that was suspected by “In-DTPA-minigastrin scanning remained unconfirmed positive.
DISCUSSION
The outstanding sensitivity of pentagastrin in detecting the presence of malignant C cells suggests the expression of @ a corresponding receptor type in human MTC (27). Indeed, receptor autoradiographic studies by Reubi and Waser (13) and Reubi et al. (14) have demonstrated the presence of CCK-B/gastrin in more than 90% of MTCs and in a high percentage of other important tumor types, such as small cell lung cancer, stromal ovarian cancers and astrocytomas. Even though the expression of relevant amounts of CCK-B receptors in typical adenocarcinomas is a matter of debate, some investigators were able to demonstrate and character ize this receptor type in colon, gastric and pancreatic cancers
1038 Twi Jou@r'miOFNUCLEARMEDICm@•Vol. 40 •No. 6 •June 1999
120 lh I @ 100 80 60
1.
@ ma - @I@asI@a
tIssues
1mg
gastrin - I
30@
tb.
--0@-@ nsCCK- -..- caerulein -,@.-. des-BOC-.pentagastiin --,.-. CCK-
I • I bO
t
--@-- sCCK- --@-. ‘@31]@CCK --=-- cionin @.1@@. [Tyr(SO3H)21I-CCK-26-
FIGURE10. ComparIsonof biokineticsof various peptides In kidneys for nonsulfated (top) and sulfated (middle) CCK analogs and members of gastnn family (bottom). Solid lines represent peptides containing cationic amino acids (lysine), dashed lines represent peptides with N-terminal pGlu moiety and various dotted lines represent peptides lacking both of these charactens tics.
@ I • I I
with previous results (25), they also seem to confirm that (at least partial) oxidation of the Met residues (e.g., to methio nine sulfoxide) does not affect the receptor-binding capabili ties ofthese radioiodinated peptides. Accordingly, no signifi cant differences in biodistribution were noticed, whether using peptides labeled according to the Bolton-Hunter or the Iodogen procedure. Because standard methods for determining the binding affinity of peptides use either quantitative receptor autoradi ography in frozen tissue sections or cortex membrane preparations, our data obtained by a different methodology
were intended to provide only rough estimates. That is why
only the approximate orders of magnitude are given in Table
1. Nevertheless,the valuesobtainedare in accordancewith published values (28,29). Our affinity/IC@ results show that
as well (15). Thus, there are ample reasons to regard CCK-B receptors as an attractive target for radiolabeled peptides for diagnostic and therapeutic purposes. Although we were previously able to demonstrate the feasibility ofradiolabeled gastrin-I to target CCK-B receptor expressing tissues in animals and patients (12), a variety of problems remained unsolved to identify an optimum candi date for in vivo targeting of CCK-B receptor-expressing tumors in larger clinical studies. The aim of this study, therefore, was to develop and optimize suitable radioligands for targeting CCK receptors in vivo. Previous Studies had provided evidence for an unusual susceptibility of the methionine moieties of CCK and gastrin derivatives toward oxidation, e.g., during radioiodination. Our data seem to confirm this sensitivity, but, in accordance
CHOLECYSTOKININ-B RECEFrOR-EXPRESSING TUMORS •Behr et al. 1039
sulfatedCCKanalogs
@-big-gastrin -.4- gastrin-l -.,- [Leu'@]-gastrin-l .--.-- minigastñn .-..-- [Gn52'J-minigasthn
@nderivadves
@b
.‘- gastrin-l fragm. 1-
tlm•p.1.(h)
1111n-DTPA-mlnIgastrln
I0minp.i. —I hp.i. 3hp.i. 5hp.i. 24hp.i.
I
.@ I
FIGURE 12. Quantitativebiodistnbution of 1111n-DTPA-minigastrin.tu = tumor; lu = lung;Ii = liver;spl = spleen;ki = kidney; sto = stomach; pa = pancreas; ga = gallbladder; bw = bowel; br = brain; ad = adrenal gland; mu = muscle; bo = bone; bI = blood.
tissues
hand, are known to contribute to the renal tubular reabsorp
tion of peptides and low-molecular-weight proteins (33,34).
In accordance with their high affinity to the CCK-A receptor, which is almost ubiquitously present throughout the gastrointestinal tract (especially in the pancreatic acini
and islets, the gallbladder, neurons and smooth muscle cells
throughout the bowel, as well as gastric mucosal cells) (35), sulfated CCK analogs showed high uptake not only in the stomach but also in the bowel, pancreas, liver and gallblad der. Therefore, we postulate that for high affinity to and selectivity for the CCK-B receptor, nonsulfated gastrmn derivatives may be superior to CCK analogs, the latter
FIGURE 13. Externalscintigramof subcutaneouslyTT-xeno graftednude mouseat 1 h postinjectionof 1111n-DTPA-minigastnn (Picker Prism 2000 gamma camera, medium-energy, parallel hole collimators; 100 kcts). Arrow designates tumor (ventral view).
having been proposed for CCK-B receptor visualization in vivo by Reubi et al. (36). The main excretion pathways of gastrin and CCK-related peptides seem to be renal filtration and, to a lesser extent, biliary excretion. Interestingly, the molecular structure of the various peptides was also reflected by their renal handling. The highest renal accretion occurred with big-gastrmn, which is the only peptide studied that bears positively charged lysine residues. The latter are known to be involved in renal tubular reabsorption (33,34). In contrast, the sulfated, thus strongly anionic, CCK derivatives showed the lowest renal uptake and retention of all peptides studied. In con trast, N-terminal pGlu moieties seem to protect not only from serum peptidase digestion but also from lysosomal degradation in the renal tubuli, because peptides bearing such N-terminal pGlu moieties had longer activity retention in the kidneys than their “open-chained―counterparts (Fig. 10). Because these data identify the amino acid sequence ... -(Glu)@-X-Tyr-(Y),2-Trp-Met-Asp-PheNH2 (where X and Y may be any neutralaminoacid, such as Ala, Gly or Thr; Met may be replaced by Leu or Nle; and n = 5, but other chain lengths remain to be tested) as the basic structure combining selectivity for and high affinity to the CCK-B receptor, we embarked on developing CCK-B receptor targeting peptides, which are suitable for labeling with @@ radiometals, such as ‘ Such a peptide appears to be much more convenient for clinical purposes because of the more favorable physical imaging characteristics of indium com pared with iodine and because of an anticipated easier labeling formulation. In addition, because indication for an internalization of CCK-B receptor ligands exists (37), residualizing radiolabels, such as radiometals, may be expected to improve tumor-to-nontumor ratios over rapidly
CHOLECYSTOKININ-BRECEPTOR-EXPRESSINGTUMORS •Behr et al. 1041
PatientHistoryKnown
lesions attime of presentationTumor
marker levelsScintigraphic
diagnosis analogs71-y-old withgastrin manSpontaneous MTC(pT4 pNl Ml) 3 moearlierBilateral
hilarlymphnodeand lungmetastasesCEA
95.9ng/mL Calcitonin 4,417ng/L
MBq(10mCi)131I-gastrin I: bilateralhilarlymphnode metastases50-y-old and lung manSpontaneous MTC(pT3pNl MO)7 y earlierNone
(occultmetastaticMTC)CEA (^) 18.3ng/mL Calcitonin 6,582 ng/L
MBq(3.5mCi)111ln DTPA-minigastrin: liver metastasis,questionable involvement43-y-old splenic manMEN 2AsyndromeMTC (TNM?) 14 y earlierLeft
hilarlymphnodeinvolve mentCEA
65.6ng/mL Calcitonin 11,502 ng/L
33 MBq(3.6mCi)111ln DTPA-minigastrin:left hilar metastasis33-y-old lymph node manMEN 2B syndromeMTC (TNM?J7 y earlierCervical
lymphnode,lung metastasesCEA
296.8ng/mL Calcitonin 34,508ngIL
MBq(4.6mCi)1111n- DTPA-minigastrin:cervical and mediastinal lymph node,lung,liverandbone metastasesMTC
= meduIlarythyroidcancer;CEA= carcinoembryonicantigen;MEN=multiple (^) endocrineneoplasia.
TABLE 3
Patient Characteristics of Pilot Clinical Evaluation of 1311-LabeledGastrin-l and 111In-DTPA-Minigastrin
released labels, such as conventional iodination (38). The DTPAderivativeof minigastrintestedin this studyseemsto fulfill all these requirements. The animal biodistribution data show good targeting of CCK-B receptor-expressing tissues (mainly tumor and stomach) and more prolonged retention of the activity in these tissues (Figs. 4 and 12) than with iodine. Interestingly, in contrast to our observations with radioiodinated gastrin analogs, “‘In-DTPA-minigastrinseems to exhibit some increased adrenal uptake. Future studies will have to address whether this is due to receptor-specific binding (as is the case for somatostatin analogs in rodent adrenals) or merely to nonspecific processes. In accordance with these preclinical data on the favorable targeting properties of ‘‘‘In-DTPA-minigastrin, our pilot
FIGURE 14. Targetingof CCK-B com pared with somatostatin receptors in 43-y- old man with MTC in connection with mul tiple endocrineneoplasiatype 2A(patient 3, Table 3). 111ln-DTPA-minigastrinwhole body scan (left) clearly shows uptake in stomach, kidneys, and left mediastinal me tastasis known from CT (Fig. 15). Right scan shows 111ln-pentetreotide biodistnbu tion in same patient3 mo earlier (bothscans at 24 h postinjection). Difference in size of metastaticlesionis due to tumorgrowthin 3-mo time intervalbetweenboth scans.
clinical data confirm that this peptide is capable of targeting physiological (e.g., in the stomach) and pathological CCK-B receptor expressing tissues (e.g., in tumors). This is in accordance with previous data showing that phylogenetic differences concerning the receptor affinity toward human gastrmn-I, at least between mammalian species, are minimal (28,29). Interestingly, the liver uptake of ‘‘‘In-DTPA minigastrmn seems to be considerably lower than that ob served with ‘‘‘In-DTPA-pentetreotide (Figs. 14 and 15), which may favor the detection of liver metastases. Accord @ ingly, in one patient with occult metastatic disease, ‘‘In DTPA-minigastrin was able to demonstrate a liver metasta sis that only became detectable by CT 1 mo later. Further clinical studies are warranted to show whether radiolabeled
CCK-B Somatostatin
1042 THEJOURNALOFNUCLEARMEDICINE•Vol. 40 •No. 6 •June 1999
15. Smith JP, Stock EA, Wotring MG, McLaughlin PJ, Zagon IS. Characterization of the CCK-B/gastrin-like receptor in human colon cancer. Am J PhysioL 1996;271: R797—R805. 16. Mutt V. Historical perspectives on cholecystokinin research. Ann N Y Acad Sci. 1994;713:1—10. 17. Bolton AE, Hunter WM. The labelling of proteins to high specific radioactivities by conjugation to a ‘@I-containingacylating agent. Biochem J. 1973;133:529— 539. 18. Hnatowich DJ. Layne WW. Childs RL, et al. Radioactive labeling of antibody: a simple and efficient method. Science. l983;220:613—615.
- Behr1, FischerW, Peter-KatalinicJ, EggeH. The stnictureof pneumococcal lipoteichoic acid: improved preparation, chemical and mass spectrometric studies. EurJ Biochem.1992;207:1063—1075. 20. Behr T. Becker W, Hannappel E, Goldenberg DM. Wolf F. Targeting of liver metastases of colorectal cancer with lgG, F(ab')@, and Fab' anti-CEA antibodies **labeled with w@@1'c:the role ofmetabolism and kinetics. CancerRes. 1995;55:5777—
- Bacarese-Hamilton AJ, Adrian it, Bloom SR. Prevention of cholecystokinin** oxidation during tissue extraction. Ann N YAcadSci. 1985:448:571—572.
- Zabel M, Grzeszkoviek F. Characterization of thyroid medullary carcinoma 17 celiline. HistolHistopathoL 1997;12:283—289. 23. Behr TM, Wulst E, Radetzky 5, Ct al. Improved treatment of medullaiy thyroid cancer in a nude mouse model by combined radioimmunochemotherapy: doxoru bicin potentiates the therapeutic efficacy ofradiolabcled antibodies in a radioresis tant tumor type. Cancer Res. 1997;57:5309—5319. 24. Behr TM. Memtsoudis 5, Sharkey RM, Ctal. Experimental studies on the role of antibody fragments in cancer radioimmunotherapy: influence of radiation dose and dose rate on toxicity and anti-tumor efficacy. a'm J Cancer 1998;77:787—795. 25. Rehfeld iF, Hansen HF, Marley PD, Stengaard-Pedersen K. Molecular forms of cholecystokinin in the brain and the relationship to neuronal gastrins. Ann N Y AcadSci. 1985;448:1 1—23. 26. Reubi JC. Waser B, LäderachU, Ct al. Localization of cholecystokinin A and cholecystokinin B-gastrin receptors in the human stomach. Gastroenterology. l997;l12:1197—1205.
- Ball DW, Baylin SB, de Bustros, AC. Medullary thyroid carcinoma. In: Braverman LE, Utiger RD. ads. Werner and Ingbar's the Thyroid: A Fundamental
and Clinical Text. 7th ed. Philadelphia, PA: LippincottlRaven Press; 1996:946— 960.
- Vanderhagen ii, Crawley JN, ads. Neumnal Cholecystokinin. Vol. 448. New York, **NY: New York Academy of Sciences; 1985.
- Reeve Jr. JR. Eysselein V. Solomon TE, Go VLW, ads. Cholecystokinin. Vol. 713.** **New York, NY: New York Academy of Sciences; 1994.
- Deschodt-Lanckman M. Enzymatic degradation of cholecystokinin in the central** nervous system. Ann N YAcad Sd. 1985:448:87—98. 31. Jakubke HD. Peptide: Chemie urn! Biologie. Heidelberg, Germany: Spekirum Akademischer Verlag; 1996.
- Reeve JR. Eysselein YE, Ho Fi, et al. Natural and synthetic CCK-58: novel reagents for studying cholecystokinin physiology. Ann N YAcadSci. 1994;713:1 I— **21.
- Behr TM, Sharkey RM, Juweid ME, et al. Reduction of the renal uptake of** radiolabeled monoclonal antibody fragments by cationic amino acids and their derivatives. Cancer Res. 1995;55:3825—3834. 34. Behr TM, Becker WS, Sharkey RM, et al. Reduction of the renal uptake of monoclonal antibody fragments by amino acid infusion. JNuclMed. l996;37:829— 833.
- Wank SA, Pisegna JR. DeWeerth A. Cholecystokinin receptor family: molecular cloning, structure, and functional expression in rat, guinea pig, and human. Ann N YAcadSci. 1994;713:49—66. 36. Reubi JC, Waser B, Schaer JC, et al Unsulfated DTPA- and DOTA-CCK analogs as specific high-affinity uganda for CCK-B receptor-expressing human and rat tissues in vitro and in vivo. EurJ NuclMed. 1998;25:481—490.
- Nouel D, Faure M, St. Pierre J, Alonso R, Quirion R, Beaudet A. Differential binding profile and internalization process of neurotensin via neuronal and glial receptors. JNeurosci. 1997:17:1795—1803. 38. Duncan JR, Behr TM, DeNardo Si. Intracellular fate of radiometals. J NucI Med. 1997;38:829. 39. Reubj JC, Chayvialle JA, Franc B, Cohen R, Calmettes C, Modigliani E. Somatostatin receptors and somatostatin content in medullary thyroid carcinomas. Lab Invest. 1991;64:567—573. 40. Bohuslavizki KH, Brenner W, GUnther M, et al. Somatostatin receptor scintigra phy in the staging of small cell lung cancer. NucI Med Commun. 1996:17:191—
1044 THE Joum@i OF NUCLEAR MEDICINE •Vol. 40 •No. 6 •June 1999
