α-Particles are suitable to treat cancer micrometastases because of their short range and very high linear energy transfer. α-Particle emitter 213Bi-based radioimmunotherapy has shown efficacy in a variety of metastatic animal cancer models, such as breast, ovarian, and prostate cancers. Its clinical implementation, however, is challenging due to the limited supply of 225Ac, high technical requirement to prepare radioimmunoconjugate with very short half-life (T1/2 = 45.6 min) on site, and prohibitive cost. In this study, we investigated the efficacy of the α-particle emitter 225Ac, parent of 213Bi, in a mouse model of breast cancer metastases. A single administration of 225Ac (400 nCi)–labeled anti-rat HER-2/neu monoclonal antibody (7.16.4) completely eradicated breast cancer lung micrometastases in ∼67% of HER-2/neu transgenic mice and led to long-term survival of these mice for up to 1 year. Treatment with 225Ac-7.16.4 is significantly more effective than 213Bi-7.16.4 (120 μCi; median survival, 61 days; P = 0.001) and 90Y-7.16.4 (120 μCi; median survival, 50 days; P < 0.001) as well as untreated control (median survival, 41 days; P < 0.0001). Dosimetric analysis showed that 225Ac-treated metastases received a total dose of 9.6 Gy, significantly higher than 2.0 Gy from 213Bi and 2.4 Gy from 90Y. Biodistribution studies revealed that 225Ac daughters, 221Fr and 213Bi, accumulated in kidneys and probably contributed to the long-term renal toxicity observed in surviving mice. These data suggest 225Ac-labeled anti–HER-2/neu monoclonal antibody could significantly prolong survival in HER-2/neu–positive metastatic breast cancer patients. [Cancer Res 2009;69(23):8941–8]
- rat HER-2/neu
Radioimmunotherapy of metastatic cancer using α-particle emitter-labeled monoclonal antibodies (mAb) is promising because α-particles can deliver highly focused energy along their short path length (1). The high-energy α-radiation typically causes complex DNA double-strand breaks that are difficult to repair, thus leading to effective tumor cell kill. Preclinical studies using 213Bi-labeled mAbs have shown substantial efficacy in various tumor models including leukemia (2), prostate cancer (3), ovarian cancer (4), and colon cancer (5). We have also shown that 213Bi-labeled anti–HER-2/neu mAb (7.16.4) is effective in prolonging the survival of HER-2/neu transgenic mice that if left untreated develop widespread metastases, including bone and liver metastases (6). Clinical trials using 213Bi-labeled anti-CD33 mAb to treat myeloid leukemia have shown feasibility and safety (7, 8). However, the short half-life of 213Bi complicates the preparation of the radioimmunoconjugates for clinical use and demands a large amount of 213Bi activity that is limited by worldwide availability of its parent, 225Ac. As a result, the maximum tolerated dose was not reached in the clinical trial at the largest administered activity of 37 MBq/kg (8).
To overcome the short half-life of 213Bi, McDevitt and colleagues (9) proposed the concept of an in vivo 225Ac (T1/2 = 10 days) generator that would deliver four α-particles to the target site per decay of 225Ac. This compares to one α from 213Bi, making 225Ac much more potent. Indeed, 225Ac-labeled mAbs have greatly improved the survival in lymphoma and ovarian cancer models (9, 10). More recently, 225Ac-labeled antivascular endothelial cadherin mAb targeting tumor neovasculature has been shown to inhibit tumor growth in a prostate cancer (LNCaP) model especially when it is combined with sequential chemotherapy (11). The in vivo generator concept has also been investigated in other α-particle emitters. 212Pb (T1/2 = 10.6 h, α-particle-emitting daughter 212Bi) in vivo generator (212Pb-trastuzumab) has prolonged survival in a colon cancer xenograft model (12). Dahle and colleagues showed that a single injection of 227Th-rituximab (T1/2 = 18.7 days, α-particle-emitting daughters 223Ra, 213Rn, 215Po, and 211Bi) is able to completely eradicate 60% of B-cell lymphoma xenografts (13). Because 223Ra, the first daughter of 227Th, has a 10-day half-life and rapidly localizes to bone, α-particles are mainly delivered from 227Th itself and the studies have shown that subsequent daughters do not lead to toxicity.
Several studies have been published to compare the efficacy of α- and β-radiation directly in metastatic tumor models. Behr and colleagues (14) found that 213Bi is therapeutically more effective than the β-emitter, 90Y, in a metastatic colon cancer model. Few studies, however, directly compared the efficacy of in vivo α-particle generators with that of conventional α- and β-particle emitters. In this work, we compare the efficacy of targeted therapy using the 225Ac in vivo generator with targeted 213Bi and 90Y in a syngeneic HER-2/neu metastatic breast cancer model.
HER-2/neu is a tumor cell surface tyrosine kinase associated with aggressive phenotype and poor prognosis (15). Targeting HER-2/neu with trastuzumab has shown significant clinical benefit in patients with metastatic breast cancer (16). In this study, we show the efficacy of 225Ac-7.16.4 in targeting rat HER-2/neu–positive pulmonary metastases. Rat HER-2/neu is also expressed on normal lung tissue in this mouse model as determined by Western blot (17). This allows for evaluating efficacy and toxicity of 225Ac-7.16.4 in a model that closely mimics clinical cases where cross-reactivity of tumor antigen expressed on normal organs is common.
Materials and Methods
Mice, cell line, and mAbs
neu-N transgenic mice, ages 6 to 8 weeks, expressing rat HER-2/neu under the mouse mammary tumor virus promoter were obtained from Harlan. All experiments involving the use of mice were conducted with the approval of the Animal Care and Use Committee of The Johns Hopkins University School of Medicine. NT2.5, a rat HER-2/neu–expressing mouse mammary tumor cell line, was established from spontaneous mammary tumors (18). The NT2.5 cells are maintained in RPMI containing 20% fetal bovine serum, 0.5% penicillin/streptomycin (Invitrogen), 1% l-glutamine, 1% nonessential amino acids, 1% sodium pyruvate, 0.02% gentamicin, and 0.2% insulin (Sigma) at 37°C in 5% CO2. 7.16.4, a mouse anti-rat HER-2/neu mAb, was purified from the ascites of athymic mice. The hybridoma cell line was kindly provided by Dr. Mark Greene (University of Pennsylvania). Rituximab (IDEC Pharmaceuticals), an anti-human CD20 mAb, was used as a negative control.
Radiolabeling of antibody with 213Bi, 90Y, and 225Ac
7.16.4 was conjugated to SCN-Chx-A′′-DTPA following published protocol (19). 90Y was purchased from Perkin-Elmer and labeled to 7.16.4-Chx-A′′-DTPA (10 mCi/mg) at 37°C for 30 min in acetate buffer (pH 4.5). 225Ac/213Bi (Institute for Transuranium Elements) generator was constructed and 213Bi was labeled to 7.16.4-Chx-A′′-DTPA (10 mCi/mg) as described previously (6). Both 90Y- and 213Bi-labeled 7.16.4 radioimmunoconjugates were purified by MicroSpin G-25 column (GE BioSciences).
225Ac was purchased from Curative Technologies. 225Ac was labeled to mAb in a two-step reaction following McDevitt and colleagues (20). First, 225Ac (0.15-0.2 mCi in 20-80 μL) was chelated to 1 μL (10 mg/mL) p-SCN-Bn-DOTA (Macrocyclics) at 56°C for 1 h. Ascorbic acid (1 μL, 150 mg/mL) was added as a radioprotectant and 2 mol/L sodium acetate (40-60 μL) was added to raise the pH to 6.5. The efficiency of 225Ac chelation to DOTA was determined by Sephadex C-25 column (GE Bioscience). Second, 100 μg mAb (∼20 μL, 5 mg/mL) was incubated with p-SCN-Bn-DOTA-225Ac at 37°C for 45 min (pH 8.5). 225Ac-labeled mAb was purified with a Centricon centrifuge filter unit (YM-10; Millipore).
The reaction efficiency and purity of the radioimmunoconjugates were determined with instant TLC using silica gel impregnated paper (Gelman Science). Instant TLC paper strips were counted the next day with a γ-counter (LKB Wallac; Perkin-Elmer) to allow 225Ac to reach equilibrium. 225Ac-7.16.4 immunoreactivity was determined by incubating 5 ng 225Ac-7.16.4 with excess antigen binding sites (1 × 107 NT2.5 cells) twice on ice for 30 min each time. Immunoreactivity was calculated as the percentage of 225Ac-7.16.4 bound to the cells. Stability of 225Ac-7.16.4 was measured by incubating 225Ac-7.16.4 in cell culture medium containing 20% fetal bovine serum for 30 days and the fraction of 225Ac chelated to DOTA was measured with Sephadex C-25 column and instant TLC. To determine internalization of 7.16.4, NT2.5 cells (2 × 106/mL) were incubated with 111In-labeled 7.16.4 (1 μg/mL) at 37°C. After 0, 10, 20, 30, 60, 120, 240, 360, and 480 min (two samples for each time point) of incubation, reaction was stopped and NT2.5 cells were washed three times with cold PBS. Cell pellet was then incubated with 1 mL NaCl 150 mmol/L/50 mmol/L glycine (pH 2.0) for 10 min at room temperature and washed twice with PBS. Both pellet and supernatant were collected and counted with γ-counter. The activity fraction of the cell pellet was the fraction of internalized 7.16.4 (10).
Specific cell kill in vitro by 225Ac-7.16.4
Specific cell kill in vitro was determined by colony formation assay. NT2.5 cells were seeded into 96-well plates at 2.0 × 103 per well. NT2.5 cells were then treated with serially diluted 225Ac-7.16.4 or 225Ac-rituximab (0.2-20 nCi/mL at a specific activity of 0.10 μCi/μg). After incubation for either 3 or 5 days, NT2.5 cells were trypsinized and replated on cell culture Petri dishes for colony growth. Blocking of cell kill by 225Ac-7.16.4 using unconjugated 7.16.4 (50 μg/mL) was also tested.
Biodistribution of 225Ac-, 213Bi-, and 111In-7.16.4
neu-N mice (3 per group) bearing s.c. tumors in the mammary fat pad were injected i.v. with 400 nCi 225Ac-7.16.4. At 1, 6, 24, 72, 144, and 288 h after injection, mice were sacrificed and major organs, including blood, heart, lung, liver, spleen, kidney, stomach, intestine, femur, and tumor, were collected. 225Ac in each organ was counted the next day using an energy window of 150 to 500 keV. To determine the distribution of free 221Fr (T1/2 = 4.9 min) and 213Bi that was released after 225Ac decay, the organs were counted right away repeatedly for 221Fr or 213Bi using the 190 to 250 and 400 to 480 keV energy window, respectively. An exponential expression was fitted to the decay curves thus obtained. The fitted activity coefficient at time 0 is the activity concentration at the time of sacrifice. Differences in 221Fr and 213Bi activity at the time of sacrifice and the levels from equilibrium of 225Ac reflect clearance or accumulation of free 221Fr and 213Bi. The time activity curves of free 213Bi and 221Fr in each organ were then constructed with free 221Fr and 213Bi activities at multiple sacrifice time points. Injectates with equivalent injected activity were counted as standards for decay correction. In a separate series of animals, the biodistribution of 225Ac-, 213Bi-, and 111In-7.16.4 to lung metastases was also measured and compared. neu-N mice (3 per group) were injected with either 225Ac-, 213Bi-, or 111In-labeled 7.16.4 and sacrificed at 0.5, 1.0, 3.0, and 6.0 h after injection; 111In radionuclide was used as a surrogate for 90Y (21). Lung metastases were collected and counted for 225Ac, free 221Fr, and 213Bi.
Efficacy of 225Ac-, 213Bi-, and 90Y-labeled anti-rat HER-2/neu mAb to treat lung metastases
The maximum tolerated dose was determined as the highest administered activity that allows 100% survival with no significant body weight loss (>15%). Healthy neu-N mice were injected (5 per group) with 100, 200, 400, 500, 600, 700, or 1,000 nCi 225Ac-7.16.4. Mice were weighed twice per week for 90 days. To evaluate the efficacy of radiolabeled mAbs to treat early-stage micrometastases, 3 days after neu-N mice were injected with 1 × 105 NT2.5 cells (i.v.), mice were treated i.v. with (a) 400 nCi 225Ac-7.16.4 (n = 12); (b) 120 μCi 213Bi-7.16.4 (n = 10); (c) 120 μCi 90Y-7.16.4 (n = 5); (d) 200 nCi 225Ac-7.16.4 (n = 5); (e) 200 + 200 nCi 225Ac-7.16.4 (injected 1 week apart; n = 5); (f) 100 μg 7.16.4, unlabeled control (n = 5); (g) 400 nCi 225Ac-rituximab, nonspecific control (n = 5); and (h) untreated control (n = 10). To evaluate the efficacy of treating late-stage lung metastases, at 18 days after tumor cell inoculation, neu-N mice were treated with (a) 400 nCi 225Ac-7.16.4 (n = 5), (b) 120 μCi 213Bi-7.16.4 (n = 5), and (c) 120 μCi 90Y-7.16.4 (n = 5). Mice were monitored and weighed three times a week and were euthanized when significant body weight loss (>15%) or breathing difficulties developed. In survival and long-term toxicity studies, animals were followed for up to 1 year.
At time of sacrifice, all major organs were collected for histopathologic examination. The number of visible lung metastases was counted. Lung metastases were snap-frozen in liquid nitrogen and sectioned with a cryotome into 8-μm-thick slices. The section was fixed in acetone and immunostained with 7.16.4 and a biotinylated rat anti-mouse IgG2a antibody (BD Pharmingen). The stain was developed with a Vectastain ABC kit (Vector Lab) according to the manufacturer's instruction.
Organ and tumor absorbed doses for 225Ac were calculated based on measured biodistribution data (22). Time activity curves in each organ and in s.c. tumors and small lung metastases were measured and integrated over time to obtain the total disintegration of 225Ac, free 221Fr, and 213Bi. 213Po and 217At, with half-lives of 4.2 μs and 32.3 ms, were assumed to decay at the same position as 213Bi or 221Fr. Absorbed doses for three tumor geometries were calculated: s.c. tumor (∼1 cm diameter), small lung metastases (300 μm diameter; day 18), and single cells (day 3). In the s.c. tumor calculation, the α-particle and electron energies of 225Ac were assumed completely absorbed within the tumor. Photon doses were not included because they typically account for <1% of the total absorbed dose. Absorbed doses for normal organs and the s.c. tumor were thus calculated as Dα = Ã * Δα / M; De = Ã * Δe / M; D = Dα + De, where Dα, De, and D are the α-particle, electron, and summed α and electron absorbed dose, respectively; Ã is the total number of disintegrations in an organ/tumor; Δα and Δe are the mean energy emitted per nuclear transition for α-particles and electrons; and M is the weight of the organ/tumor. The total absorbed doses were calculated as the sum of doses from 225Ac at equilibrium and its free daughters. If an organ is accumulating or depleting free daughters, the dose from free daughters is added or subtracted from the 225Ac doses at equilibrium. Absorbed doses for single tumor cells (day 3) and small lung metastases (day 18) were also calculated and compared for 225Ac-, 213Bi-, and 90Y-7.16.4. 213Bi and In-111 (surrogate for 90Y) biodistribution data for small lung metastases were used for 213Bi and 90Y dosimetry. The absorbed fractions of electron and α-particle emissions in small lung metastases were calculated using Monte Carlo; these calculations accounted for the range of all of the βs and αs (23). Absorbed doses to single cells were obtained using MIRD cellular S values (24, 25). In the single-cell calculations, daughter radionuclides resulting from the decay of 7.16.4-conjuagated 225Ac that was tumor-cell bound but not internalized were assumed to diffuse away and not contribute to the target cell absorbed dose. The fraction of internalized 225Ac-7.16.4 was obtained from the internalization assay.
Kaplan-Meier survival analysis or statistical comparisons between groups (Student's t test) were done with MedCalc (MedCalc Software). For all studies, the level of statistical significance is set at P < 0.05.
Radiolabeling, immunoreactivity, and antibody internalization
Both 90Y and 213Bi labeling efficiencies were ∼90% and purities reached ∼98% after size-exclusion purification. The two-step 225Ac labeling efficiency was 12.0 ± 3.8% (n = 12). After purification, the purity of 225Ac-7.16.4 was 97.0 ± 1.8% (n = 12). The specific activity of 225Ac-7.16.4 was between 0.06 and 0.10 μCi/μg. The immunoreactivity of 225Ac-7.16.4 was 62.4 ± 5.9% (n = 3), worse than that of 111In- or 213Bi-7.16.4 (∼80%). After incubation in 20% bovine serum albumin at 37°C for 30 days, purity of 225Ac-7.16.4 became 80.9 ± 8.2% and 65.9 ± 3.9% as measured by Sephadex C-25 column and instant TLC (n = 3), respectively. 7.16.4 internalizes after binding to NT2.5 cells. The internalized 111In-7.16.4 increased from 1.55 × 10−2 μg/106 cells (28.2%) at 5 min after binding started to 3.05 × 10−2 μg/106 cells (55.5%) at 120 min and plateaued afterwards to 3.25 × 10−2 μg/106 cells (59.2%) at 1,440 min.
In vitro cell kill
225Ac-7.16.4 was very effective in killing rat HER-2/neu–expressing NT2.5 cells (Fig. 1A). The activity concentration of 225Ac-7.16.4 that can kill 50% of the NT2.5 cells (ED50) was 2.3 and 6.4 nCi/mL for 5- or 3-day incubation, respectively. When 50 μg/mL unlabeled 7.16.4 was used to block 225Ac-7.16.4, the ED50 increased to 9.4 nCi/mL (5-day incubation). The ED50 for 225Ac-rituximab was 19.0 and 18.8 nCi/mL when it was incubated with NT2.5 cells for 3 or 5 days. In comparison, the ED50 for 213Bi-7.16.4 was 940 nCi/mL.
Biodistribution of 225Ac-, 213Bi-, and 111In-7.16.4
The biodistribution of 225Ac-7.16.4 at equilibrium in (s.c. mammary fat pad) tumor-bearing neu-N mice is shown in Fig. 1B. The effective half-life of antibody clearance from blood in the first 72 h was 26.8 h, similar to that of 111In-labeled 7.16.4 (26.3 h). 225Ac-7.16.4 targeting to tumors increased continuously and accumulation peaked at 20.6%/g 6 days after injection. Antibody localization in the tumor was 17.3%ID/g 12 days post-injection. The peak tumor uptake of 225Ac-7.16.4, however, was lower than that of 111In-7.16.4 (∼38%ID/g), probably reflecting impaired immunoreactivity of 225Ac-7.16.4.
The biodistributions of free 213Bi and 221Fr at 1 h after 225Ac-7.16.4 injection are shown in Fig. 1C and D. For 213Bi (Fig. 1C), the activity concentration in blood increased from 2,780 Bq/g at time of sacrifice to 3,500 Bq/g at equilibrium (P = 0.03), whereas the kidneys activity decreased from 2,450 Bq/g at time of sacrifice to 887.0 Bq/g at equilibrium (P = 0.005). Similar pattern was observed for 221Fr (Fig. 1D). These data strongly suggest that both free 213Bi and 221Fr cleared from blood through kidneys. The fractions of free 213Bi and 221Fr in blood that ended up in kidneys were 41.1% and 18.7%, suggesting a higher kidney retention rate for 213Bi, which has been attributed to the higher positive charge of 213Bi (Bi3+) that interacts with the negatively charged basement membrane of the glomerulus. In biodistribution study of lung metastases, 225Ac-7.16.4 was found to be 13.7 ± 1.5%ID/g at 144 h, lower than 30.8 ± 1.5%ID/g of 213Bi-7.16.4 at 6 h.
Treating rat HER-2/neu–expressing lung metastases with 225Ac-, 213Bi-, and 90Y-7.16.4
The maximum tolerated dose of 225Ac-7.16.4 in neu-N mice was found to be 400 nCi in a single injection. Maximum tolerated dose of 213Bi-7.16.4 (120 μCi) and 90Y-7.16.4 (120 μCi) were obtained from literature and confirmed in the neu-N model (6, 26). Kaplan-Meier survival curves of neu-N mice bearing lung metastases after treatment by radiolabeled anti-rat HER-2/neu mAb are shown in Fig. 2A. All untreated mice (100%) developed pulmonary metastases, showing signs of labored breath and stress at later stages. Median survival improved to 50 days in 90Y-7.16.4 group (P = 0.013) and 61 days in 213Bi-7.16.4 group (P = 0.0001) compared with untreated mice (41 days). Eight of 12 mice in 225Ac-7.16.4–treated group achieved long-term (1-year) survival (P < 0.0001). Both 213Bi-labeled (P = 0.025) and 225Ac-labeled (P < 0.001) 7.16.4-treated mice lived significantly longer than the mice treated by 90Y-7.16.4. 225Ac-7.16.4 also improved survival of neu-N mice compared with α-emitter 213Bi-7.16.4 (P = 0.001). 225Ac-rituximab did not show efficacy in these models with median survivals of only 33 days (P = 0.99). Unlabeled 7.16.4 (single i.v. dose 100 μg/mouse) slightly improved median survival to 42 days (P = 0.014) and the median survival of neu-N mice treated with 200 nCi 225Ac-7.16.4 improved to 47 days (P = 0.029). Notably, when a second dose of 200 nCi 225Ac-7.16.4 was administered 1 week after the first 200 nCi injection, median survival improved to 84 days (P = 0.004), with 2 of 5 mice achieved long-term survival for up to 1 year (Fig. 2B).
To examine the efficacy of 225Ac-7.16.4 on later-stage metastases, neu-N mice were treated 18 days after tumor cell injection (Fig. 2C), where the average diameter of lung metastases was 296 ± 94 μm (n = 29). 213Bi-7.16.4–treated mice have a median survival of 51 days, not a statistically significant improvement over 90Y-7.16.4 with a median survival of 50 days (P = 0.76). Although mice treated with 213Bi-7.16.4 at 18 days had a significantly shorter survival compared with those treated at 3 days (P = 0.0001), mice treated with 90Y-7.16.4 at 18 and 3 days have about the same median survival of 50 days (P = 0.96), suggesting that targeting with the 90Y is less sensitive to the size of the metastases. The median survival decreased to 66 days for mice treated with 225Ac-7.16.4 at 18 days compared with 3 days (P = 0.0005). Nevertheless, 225Ac-7.16.4–treated mice still survived significantly longer compared with 213Bi-7.16.4 (P = 0.004).
Number of lung metastases
The numbers of visible metastases on the lungs of neu-N mice and representative images are shown in Fig. 3. All treated mice have reduced number of lung metastases compared with untreated mice. When treatment was initiated at 3 days, mice treated with 213Bi-7.16.4 and 225Ac-7.16.4 had 8.5 ± 1.8 (P < 0.0001) and 2.9 ± 6.1 (P < 0.0001) lung metastases, significantly less than 90Y-7.16.4–treated mice with 36.7 ± 8.1 metastases. No lung metastases were found in neu-N mice surviving 225Ac-7.16.4 treatment (Fig. 3A, bottom right). It is also evident that the sizes of the metastases were much larger in 213Bi-7.16.4–treated mice (Fig. 3A, bottom left) than those from 90Y-7.16.4–treated mice (Fig. 3A, top right), which suggests that 213Bi-7.16.4 is able to eliminate more metastases-initiating cells compared with 90Y-7.16.4 and it takes longer for the surviving tumor cells to grow to a lethal tumor burden. When treatment was initiated at 18 days, the number of metastases found on 90Y-7.16.4–treated mice (32.8 ± 13.2) was similar to that obtained for mice treated with 90Y-7.16.4 at 3 days. More metastases (34.0 ± 13.5) were found on mice treated with 213Bi-7.16.4 at 18 days compared with 3 days (P < 0.0001), indicating reduced efficacy of 213Bi to treat later-stage metastases. 225Ac-7.16.4–treated mice again presented with the fewest number of lung metastases (11.5 ± 3.5) compared with the other two radioisotopes (Fig. 3B).
Histopathology and immunohistostaining of rat HER-2/neu
Immunohistostaining showed that the HER-2/neu expression level decreased on the metastases surviving the treatment of 90Y-7.16.4 (Fig. 4A, middle) and 213Bi-7.16.4 (Fig. 4A, right) compared with untreated tumors (Fig. 4A, left). At 1 year after treatment, no residual tumor cells were found on the lungs of mice treated with 400 nCi 225Ac-7.16.4 (Fig. 4B, left) and a single metastatic tumor nodule was found on one of the lungs from a mouse treated with 200 + 200 nCi 225Ac-7.16.4 (Fig. 4B, right). No damage to the normal lung tissue was observed in groups treated with 225Ac-7.16.4. Examination of kidneys from these mice, however, revealed shrunken and pale kidneys (Fig. 4C, left). H&E staining showed widespread loss of tubular epithelium in the kidney cortex (Fig. 4C, middle). In the medulla, there is extensive loss of tubules and replacement with marked fibrosis and a mixed inflammatory infiltrate composed mainly of mononuclear cells and lesser numbers of neutrophils. Scattered dilated cystic structures are visible throughout the cortex and medulla. These observations are also present but less evident in kidneys treated with 200 + 200 nCi 225Ac-7.16.4 (Fig. 4C, right).
The absorbed doses to major organs and tumors are shown in Table 1. Reductions or increases in the absorbed dose from 225Ac at equilibrium with its free daughters are indicated by negative or positive dose contributions. The dose calculations show that free 221Fr and 213Bi are cleared from the blood, lungs, liver, and spleen, whereas kidney, stomach, and intestine are accumulating these daughters. A 400 nCi administration of 225Ac-7.16.4 gives a tumor absorbed dose of 10.49 Gy; no significant depletion or accumulation of free daughters is observed. Absorbed doses of 225Ac-, 213Bi-, and 90Y-7.16.4 to single tumor cells and small lung metastases (∼300 μm in diameter) are shown in Table 2. The 225Ac absorbed dose to small metastases is 9.6 Gy, much higher than that of 2.0 Gy from 213Bi and 2.4 Gy from 90Y. Similarly, the 25.3 Gy 225Ac dose to single tumor cells is also higher than that of 213Bi at 2.6 Gy and 90Y at 0.6 Gy. When no antibody-receptor internalization or 100% internalization was considered, 225Ac dose to single cells became 12.0 or 38.6 Gy, respectively.
We showed that α-particle emitter 225Ac-labeled anti-rat HER-2/neu mAb is effective in eliminating breast cancer lung metastases, leading to long-term animal survival but also renal toxicity most likely caused by the free daughters of 225Ac (27).
The improved efficacy of α-emitter 225Ac over 213Bi and 90Y can partially be attributed to the higher radiation doses that micrometastases receive from 225Ac. 225Ac emits four α-particles along its decay chain and deposits a total energy of 4.50 × 10−12 J/Bq s, 3.2 and 30.0 times higher than that by 213Bi and 90Y. Furthermore, the majority of α-radiation will be absorbed locally, whereas most of the β-particle energy from 90Y will be deposited outside of micrometastases. An important finding from the biodistribution studies is that only a small fraction of 221Fr and 213Bi was released from major organs/tumors, suggesting that these daughters are either retained in the cells due to antibody internalization or there is simply not enough time for a substantial fraction of the daughters to diffuse out of the organs/tumors. The long half-life of 225Ac also helps to deliver more doses to the lung metastases than 213Bi. The biodistribution studies showed that antibody localization to s.c. tumors peaked at least 24 h after injection. For lung metastases, antibody localization is much faster, reaching peak at ∼6 h after injection. Nevertheless, for 213Bi, a substantial fraction (>90%) of the decays will have already occurred, greatly reducing the number of α-particles that can be delivered.
The higher potency and longer half-life of 225Ac overcame the ∼35% lower immunoreactivity of 225Ac-7.16.4 compared with 213Bi-7.16.4. The immunoreactivity of 225Ac-labeled mAbs seems to vary significantly among different antibodies (20) independent of the labeling procedure. When the incubation time in the second step of the reaction was reduced to 30 or 15 min, the immunoreactivity of 225Ac-7.16.4 did not improve in the biodistribution study, with a 16.2%ID/g and 15.4%ID/g tumor localization at 72 h after injection (3 mice per group). Furthermore, stability assay suggested slow but significant release of 225Ac from the antibody, which could also contribute to the lower tumor localization. More studies are needed to understand and improve the immunoreactivity and stability of 225Ac-7.16.4.
The main concern of radioimmunotherapy with α-emitter 225Ac-7.16.4 is the long-term renal toxicity caused by free 221Fr and 213Bi (27). Furosemide or chlorothiazide and competitive metal blockade were found to be effective in reducing renal uptake of the free daughters (22). The extent to which such interventions are needed will likely depend on the residence time of the antibody in circulation and the internalization properties of the antibody-antigen complex. In a recent update of results from a clinical trial of 225Ac-HuM195 antibody in leukemia patients, no acute toxicity was seen and no evidence of radiation nephritis has been seen, with follow-up to 10 months in patients receiving up to 4 μCi/kg (28). Liposomal encapsulation of 225Ac (29) has also been proposed to trap free 225Ac daughters. Extracorporeal metal chelation of blood 221Fr and 213Bi could also be used to reduce renal toxicity (30). Interestingly, 225Ac-labeled anti-thrombomodulin mAb (201B, targeting lung vasculature) caused lethal toxicity to the lungs, whereas it was able to inhibit metastatic growth (31). No evident damage to lung tissue was observed when rat HER-2/neu was targeted in this study, showing the importance of selecting tumor antigen for targeting to reduce toxicity.
The reduced efficacy when treating micrometastases at 18 day after cancer cell inoculation was expected and may be partially explained by increased tumor burden and increased tumor heterogeneity leading to nonuniform dose distributions (32). Monte Carlo modeling has shown that, when treating a micrometastases containing 1,000 cancer cells, the tumor control probability could drop from 93% to 0% when a 50% variance of antigen expression was introduced to a uniform expression (33). A recent clinical trial has shown that consolidation of chemotherapy regimen with 213Bi-HuM195 in acute myeloid leukemia patients is highly effective (34). The efficacy data from days 18 and 3 metastatic models suggest that 225Ac-based radioimmunotherapy should be tested ideally in patients undergoing remission after first line chemotherapy treatments.
In conclusion, we have shown that α-particle emitter 225Ac-labeled mAb is very effective, better than the α-emitter, 213Bi, and the β-emitter, 90Y, in prolonging the survival of neu-N transgenic mice bearing lung metastases. Long-term renal toxicity was observed and approaches designed to mitigate such toxicity (22) may need to be implemented in clinical studies.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Grant support: National Cancer Institute grant R01 CA 113797 and Department of Defense Fellowship BC0444176 to H. Song.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Received May 20, 2009.
- Revision received September 3, 2009.
- Accepted September 17, 2009.
- ©2009 American Association for Cancer Research.