A Phase I Trial of Radioimmunotherapy with ¹³¹I-A5B7 Anti-CEA Antibody in Combination with Combretastatin-A4-Phosphate in Advanced Gastrointestinal Carcinomas

Patients and Methods

Patient eligibility. The study was approved by the Research Ethics Committees at both participating institutions and all patients provided written informed consent. Patients with histologically confirmed gastrointestinal adenocarcinoma not amenable to standard therapy or refractory to conventional therapy were eligible for this study. Other eligibility criteria included a plasma CEA level of 10 to 1000 μg/L; radiologically measurable disease by Response Evaluation in Solid Tumors (RECIST) criteria (8); electrocardiogram (ECG) QTc interval ≤450 ms and no evidence of cardiac arrhythmia or ischaemia; disease assessable by DCE-MRI; life expectancy >3 mo; WHO performance status of 0 or 1; adequate bone marrow function with neutrophils ≥1.5 × 10⁹/L, hemoglobin ≥ 9 g/dL, platelets ≥ 100 × 10⁹/L; adequate clotting with international normalized ratio (INR) ≤1.3, levels and EDTA creatinine clearance >50 mL/min; adequate hepatic function with plasma bilirubin ≤2× upper limit of normal (ULN); and alanine aminotransferase or aspartate aminotransferase ≤2× ULN (or 5× ULN in the presence of liver metastases).

Patients were excluded if they had a history of ischaemic heart disease, congestive heart failure, or significant cardiac arrhythmia; uncontrolled hypertension (defined as blood pressure (BP) consistently >150/100 despite medication); ECG evidence of clinically significant pathology; taking any drug known to prolong the QTc interval; conditions associated with QTc prolongation; major thoracic or abdominal surgery in the preceding 4 wk; autoimmune disorders, inflammatory bowel disease or ≥grade 2 peripheral neuropathy; previous radical radiotherapy; hepatitis B or C, or HIV; positive human antimouse antibody (HAMA); previous anticancer therapy within 4 wk.

Study design. For the first cohort the starting dose for CA4P was 45 mg/m² and that for ¹³¹I-A5B7 was 1,800 MBq/m². Subsequent cohorts were planned with independent dose escalation as follows: cohort 2, CA4P 54 mg/m² and ¹³¹I-A5B7 1,800 MBq/m²; cohort 3, CA4P 54 mg/m² and ¹³¹I-A5B7 2,100 MBq/m²; and cohort 4, CA4P 54 mg/m² and ¹³¹I-A5B7 2,400 MBq/m². After the first patient was entered at each cohort a 7-day interval was required following the administration of ¹³¹I-A5B7 before entry of the subsequent patients. A time interval of 4 wk was required between treating the first patient on a previous dose level and the entry of a new patient at the next dose level. Escalation to the next cohort was permitted if no DLT had occurred in the first three patients. If a DLT occurred in one the first three patients, the cohort was expanded to a maximum of six. The maximum tolerated regimen was defined as the dose level below which two DLTs occurred in up to six patients. DLT was defined as any of the following events, considered to be probably or definitely related to combination study therapy: QTc prolongation ≥500 ms, >grade 2 ventricular or supraventricular arrhythmia, grade 3 or 4 nonhematologic toxicity (except fatigue/asthenia, nausea or vomiting, grade 3 tumor pain controlled with analgesics, and grade 3 hypertension controlled by antihypertensive medication), toxicity resulting in treatment delay >14 d, absolute neutrophil count <500 cells/mm³ for >5 consecutive days or febrile neutropenia with granulocyte count <1,000 cells/mm³, thrombocytopenia <10,000 cells/mm³ for >5 consecutive days, or a bleeding episode requiring platelet transfusion.

Drug preparation and administration. Unlabeled CA4P was supplied by Oxigene Inc to the Cancer Research UK Formulations Unit where it was labeled and released. The drug was reconstituted with 10 mL of sterile water for injection and further diluted with 100 to 150 mL of normal saline and was administered i.v. over 10 min. The drug was protected from light to prevent conversion to the less active trans-isomer. A5B7 was produced by the Biotherapeutics Development Unit (Cancer Research UK, Clare Hall Laboratories) and was formulated in isotonic PBS and stored at 4°C. A fixed dose of 10 mg of A5B7 was radiolabeled in the department of medical physics at the Royal Free Hospital using the N-bromosuccinamide method and administered within 4 h. The N-bromosuccinamide method has been shown to result in a labeling efficiency of 88% to 94% as in a previous study (9). The percent iodine incorporation of the labeled product was measured by thin-layer chromatography and the CEA binding by a CEA column. Labeled ¹³¹I-A5B7 was administered i.v. via an indwelling central venous catheter in 50 mL 0.9% saline over 30 to 40 min using a syringe driver.

Routine premedication was not mandatory prior to CA4P but if significant toxicity occurred with the first dose, premedication with dexamethasone 8 mg i.v. and metoclopramide 10 mg i.v. was recommended prior to subsequent cycles. If systolic blood pressure was >180 mmHg on two readings following CA4P, glyceryl trinitrate 0.3 mg was given sublingually or amlodipine 10 mg orally. No routine allergic prophylaxis was offered prior to ¹³¹I-A5B7; however, hydrocortisone 100 mg and chlorpheniramine 10 mg i.v. were administered if clinically necessary. Potassium iodide 60 mg three times daily for 8 d was commenced 48 h before ¹³¹I-A5B7 to block thyroid ¹³¹I uptake.

An initial dose of CA4P was given within 14 d of the ¹³¹I-A5B7 in order to assess the vascular response by DCE-MRI and enable correlation with RECIST responses. Doing such analysis during the combined therapy was not feasible because patients could not undergo DCE-MRI with high levels of radiation. For similar reasons CA4P pharmacokinetics were done during the initial dose of CA4P. Patients were admitted to hospital on day 1 for administration of ¹³¹I-A5B7. CA4P was given on days 3 and 4, and then weekly, as an outpatient, for up to 7 wk.

Patient assessment and response criteria. Written informed consent was obtained within 4 wk of entering the study. Pretreatment evaluation included history, physical examination and assessment of performance status, ECG, blood count, clotting, urea, electrolytes, creatinine, liver function, thyroid function, HAMA, CEA, and CA19-9, CT, or MRI. Following each administration of CA4P blood was taken for hematology, and vital signs monitored to up to 6 h after the infusion. ECGs were done preinfusion and hourly after the infusion for 4 h. Following administration of ¹³¹I-A5B7 vital signs were recorded hourly for 4 h. Symptoms, signs, performance status, hematology, and biochemistry were assessed weekly, and thyroid function, tumor markers, and HAMA assays were done on days 32 and 53. Radiologic assessment was done every 28 d following the ¹³¹I-A5B7 infusion. Response was classified according to RECIST criteria.

In selected consenting patients, response was also assessed by fluorodeoxygluconse positron emission tomography (FDG-PET). FDG-PET scans were done within 3 wk prestudy, then approximately at days 29 and 57 after ¹³¹I-A5B7, using an ADAC Vertex Plus Gamma Camera. Analysis was visual and semiquantitative. PET response was recorded according to previously published criteria (10).

Toxicities were recorded according to the National Cancer Institute (NCI) Common Toxicity Criteria (CTC) Version 2.0.

Pharmacokinetic studies. CA4P pharmacokinetics was done 1 to 2 wk prior to ¹³¹I-A5B7 administration. Blood was taken into EDTA tubes immediately before the administration of CA4P and then at 5, 30, and 45 min, and 1, 2, and 4 h after completion of the infusion. Blood was placed at 4°C and subsequently centrifuged at 4°C at 1,500 g for 10 min. Plasma was removed immediately after centrifugation and stored at −70°C. CA4P and CA4 levels were quantified by high-pressure liquid chromatography on a system from Waters comprising a 2695 Separations Module, a postcolumn photolysis lamp/coil, and a W474 fluorescence detector (11, 12). The plasma concentration/time data were analyzed using a noncompartmental method. The pharmacokinetic parameters determined for CA4P included the maximum observed plasma concentration (C_max), area under the plasma concentration time curve (AUC), initial and terminal elimination half-life (T_½), clearance, and steady-state volume of distribution. For CA4, C_max, AUC, and T_½ were determined, and in addition, the time to reach C_max.

For ¹³¹I-A5B7, 1 mL of blood was collected into an EDTA tube preinfusion and at 2 min, and 4, 24, 48, 72, and 96 h. The samples were stored in an approved radioactive storage area until they had decayed to appropriate levels for analysis. Radioactivity was measured using a Packard Cobra 11th Series Auto gamma counter. The time of measurement was recorded and a decay correction applied as appropriate to calculate the percent injected dose per kilogram of blood at each time point.

Dosimetry. Antibody distribution and dosimetry was determined using single-photon emission computed tomography (SPECT) imaging done at 4, 24, 48, 72, and 96 h (optional) after administration of ¹³¹I-A5B7. Imaging was done on an ADAC Vertex Plus Dual Headed Gamma Camera. Planar images were acquired at all time points. SPECT images were acquired with scatter windows and with a transmission map for attenuation correction. Reconstruction of SPECT scans was iterative with scatter and attenuation correction applied. Region of interest–based analysis was done on both tumor and normal tissues (liver, lung, spleen, blood/heart, kidney). Olinda/XEM software was used to calculate radiation doses to organs, red marrow, and tumor from the radioactivity uptake within the regions of interest, in blood and from the whole body dose-rate data (13).

HAMA analysis. Serum was collected prestudy (within 12 wk) and on study (days 32 and 53), for the assessment of HAMA response to A5B7 using a standard validated ELISA technique. Briefly, 96-well microtiter plates were coated with 100 ìL of (5 ug/mL) the A5B7 antibody. The wells were blocked with a blocking buffer (PBS/Marvel/Tween) and washed. Serum samples (1/100 dilution, 100 ìL per well) were incubated together with positive and negative serum controls and binding detected with peroxidase-conjugated goat antihuman IgG (Sigma). Samples were termed positive if the absorbance at 1:100 dilution was above the cutoff point optical density (OD) of 0.2 at 490 nm. The result was recorded as positive or negative for HAMA.

DCE-MRI studies. The primary end points for DCE-MRI studies were (a) K^trans, which is the transfer constant reflecting contrast delivery (perfusion) and transport across the vascular endothelium (permeability), with the dominant factor depending on whether delivery is flow- or permeability-limited; and (b) IAUGC₆₀, which is the initial area under the gadolinium concentration time curve for 60 s (14).

MRI was done on a 1.5 Tesla scanner (Siemens Medical Systems), using local coils appropriate for body part. Scans were done 48 h and 24 h before the first infusion of CA4P to obtain repeatability data, and a third scan 4 h following CA4P to assess the antivascular effect of the drug.

Dynamic MRI analysis. We used Magnetic Resonance Imaging Workbench software (MRIW) for analysis (15). Regions of interest (ROI) were drawn around tumor edges by a radiologist who excluded areas of artefact, necrosis, blood vessels, and normal tissues. Metastases >2 cm diameter with no partial volume averaging were chosen and up to five metastases per organ were analyzed. MRIW software converts signal intensities of T1-weighted DCE-MRI datasets into T1-relaxation rates and then into Gd-DTPA concentrations pixel-by-pixel at each of the 40 time points (16).

IAUGC₆₀ was calculated over the first 60 s from the onset of signal increase following injection of contrast (units: mmol/s). The changing tissue Gd-DTPA curve was fitted to a standard compartmental model (17) to characterize the influx of Gd-DTPA into the tumor extracellular extravascular space - transfer constant (K^trans - units/min). A literature-based pooled arterial input function (combined Fritz-Hansen/Weinmann) was used for the modeling (18). Pixel data from all slices per lesion were combined and median values were taken as representative of central tendency.

Statistical analysis. The DCE-MRI data consisted of two kinetic parameters (K^trans and IAUGC₆₀) in multiple lesions (1–5) per patient taken at three time points. Bland and Altman's variance-components methods (19–21) were extended to take into account the clustering of tumors within patients according to methods described by Rabe-Hesketh and Skrondal (22). All statistical tests were two-tailed and P ≤ 0.05 was considered significant.

The results of the reproducibility analysis were used to assess whether there had been a statistically significant change in kinetic parameters due to CA4P, for individual lesions and for all lesions. As there were two pretreatment measurements (days 1 and 2), their mean was taken as the pretreatment value for each parameter. For individual lesions, the repeatability statistic r, expressed as a percentage of the group mean pretreatment value for each parameter, gives a range within which the difference between pretreatment and posttreatment values would be expected to lie for 95% of observations, assuming that no change has occurred. If the difference falls outside this range for a particular kinetic parameter, then a significant change was deemed to have occurred. Similarly, to assess mean response in dose cohorts, the 95% confidence interval for change, expressed as a percentage of the cohort mean pretreatment value, gives the range required.

Results

Patient characteristics

From January 2004 to April 2007, 88 patients were screened for the study and 12 were enrolled. The reasons for screening failure included patient unsuitable for DCE-MRI (19), patient refusal (11), CEA negative or >1,000 (9), WHO performance status >1 (9), no RECIST assessable disease (5), cardiac ischaemia or arrhythmia (5), previous radical radiotherapy (4), abnormal biochemistry (4), and other reasons (10). Baseline characteristics are shown in Table 1 .

Toxicity

Drug-related adverse events are presented in Table 2 . Within the first dose cohort, patients 1 and 6 experienced grade 4 neutropenia meeting the criteria for DLT. Patient 1 had grade 4 neutropenia at day 51 which resolved to grade 2 by day 71, having received one dose of 1,800 MBq/m^{2 131}I-A5B7 and four infusions of 45 mg/m² CA4P, whereas in patient 6, grade 4 neutropenia persisted between days 31 and 38 at the same dose of ¹³¹I-A5B7 and with 10 infusions of CA4P. Neither patient was septic during these episodes. The bone marrow toxicity was attributed to the radioimmunotherapy component of the combination, and the protocol was amended so that the second cohort received ¹³¹I-A5B7 at 1,600 MBq/m². There was no evidence from the first cohort that known toxicities of CA4P had been exacerbated by the combination therapy, nor data suggesting that CA4P contributes to myelosuppression, so the dose of CA4P was escalated to 54 mg/m², which was above the previously shown threshold for reducing tumor blood flow according to DCE-MRI (23). Patient 7 experienced grade 3 ataxia within 24 hours of the first CA4P treatment, which persisted for three days. The patient then received 8 mg dexamethasone prior to subsequent CA4P on days 3 and 4 with no recurrent episodes of ataxia. Patients 8 and 12 both experienced grade 4 neutropenia lasting for between 6 and 8 days and the trial was discontinued at that point. The treatment received for all patients is shown in Table 3 .

Tumor pain was reported in four patients and considered to be related to CA4P, occurring with a median time to onset of 72 minutes (range, 45 minutes to 2 hours 57 minutes) and persisting for a median duration of 65 minutes (range, 5 minutes to 4 hours).

Immunogenicity

Nine patients developed HAMA after the first treatment with ¹³¹I-A5B7 and patient 4 became HAMA-positive after the second dose. Patient 12 did not have a posttreatment measurement and patient 10 did not receive ¹³¹I-A5B7.

Pharmacokinetics

CA4P.Table 4A summarizes the pharmacokinetic parameters for CA4P following the initial test dose, prior to the administration of ¹³¹I-A5B7. CA4P was rapidly converted to the active CA4 with an initial half-life of <10 minutes whereas CA4 was cleared from the plasma with a terminal half-life of 2 hours. Over the narrow dose range studied, clearance, half-life, and volume of distribution were independent of CA4P dose, whereas the peak concentration and AUC of CA4P and CA4 both increased with dose escalation. The pharmacokinetic parameters are in line with previously reported studies (4–6).

¹³¹I-A5B7. For ¹³¹I-A5B7, blood clearance was modeled with a biexponential model. Data for patient 11 could not be fitted by a biexponential model and too few data points were available for patient 12. Patient 10 withdrew from study before receiving ¹³¹I-A5B7 and there are no data for this patient. For the remaining patients, the data are summarized in Table 4B. There is a rapid initial clearance of ¹³¹I followed by a slower second phase, but because the parameters of the model vary widely among patients we used the biexponential model to calculate the 50% and 90% clearance, which were 0.8 and 46 hours, respectively. This model shows a faster clearance than was reported by Lane et al. (9), but in that study a simple monoexponential model was used which would not have captured the early clearance effectively. The range of phase 2 half-lives (9.4-28 hours) was broadly similar to the reported monoexponential half-life (mean 28.6 hours).

Pharmacodynamic studies

Dynamic MRI. Thirty lesions were analyzed, all of which were assessable for IAUGC₆₀ and 29 for K^trans. Figures 1A and B show the percentage change in IAUGC₆₀ and K^trans for each lesion for each patient. The lower control line indicates the 95% repeatability for a single lesion. Three of the twelve patients (patients 6, 7, and 10) had significant or borderline reductions in both IAUGC₆₀ and K^trans at 4 h after CA4P. For individual lesions, 3 of 30 had a significant or borderline reduction in IAUGC₆₀, whereas 5 of 29 had such reductions in K^trans. When all lesions were assessed as a cohort, there was a group reduction of 20.5% (P = 0.015) in IAUGC₆₀ and a 27% reduction in K^trans (P = 0.011) with a repeatability range of ± 10.92% and ± 14.46%, respectively. The graphs show considerable variability in the degree of responsiveness among lesions but the degree of variability among lesions in the same patient seems to be less. There was a trend for greater reductions in all kinetic parameters at higher doses of CA4P (54 mg/m²).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Change in IAUGC (A) and K^trans (B) for 30 lesions in 12 patients at two doses of CA4P (45 and 54 mg/m²).

¹³¹I-A5B7 dosimetry. SPECT imaging data were available for 9 of the 12 patients. Patient 3 had no 4-hour scan, patient 10 did not receive ¹³¹I-A5B7, and patient 12 only had one scan done. Patient 4 received ¹³¹I-A5B7 on two occasions and data are included for both treatments. Figure 2A shows the percentage of injected dose of radioactivity at four time points and shows retention of radiation in the tumor over time compared with normal organs. The absorbed dose per injected activity for normal organs and for the tumor is shown in Fig. 2B and C, respectively. Patients 5, 6, 8, and 9 received a high tumor dose compared with all normal organs with the exception of the lungs.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

The median percentage of the injected dose of radioactivity at 4, 24, 44, and 68 h for normal organs and tumor (A), the median absorbed dose to target organ per injected activity as calculated using Olinda/EXM software (B), and the absorbed dose to tumor per injected activity assuming a spherical lesion with no contribution from the rest of the body (C). It was not possible to calculate absorbed doses to tumor for patients 3 and 10.

Response

Ten of twelve patients were evaluable according to RECIST (8). Two patients were not valuable for response because there was no posttherapy imaging in one (patient 7) and the second withdrew from the trial before receiving A5B7 (patient 10). There were no complete or partial responses. Seven patients had progressed by day 29 and two had stable disease but progressed by day 57. Patient 4 had stable disease (SD) at day 29 and 57 and was retreated with ¹³¹I-A5B7. Patient 6 had a minor response in an iliac lymph node and a decrease in his CA19-9 from 10,507 IU/mL on day 1, to 2,033 IU/mL on day 44.

Four patients (2, 3, 9, and 11) had FDG-PET assessment of response done at days 29 and 57. Using the criteria previously described (10), all four patients had progressive disease (PD) consistent with the standard radiologic assessment.

Discussion

The success in treating hematologic malignancies with radioimmunotherapy has not been matched by similar success in treating solid tumors, and a variety of strategies have been pursued to improve delivery and efficacy (24). Our approach has been to combine radioimmunotherapy with the VDA CA4P. In preclinical models the two agents have been shown to target different compartments of the tumor, with radioimmunotherapy targeting the well-perfused vascular rim and CA4P inducing vascular collapse and necrosis in the poorly perfused center (25). In addition, the delivery of CA4P 48 hours after radioimmunotherapy seems to trap radiation within the tumor, resulting in increased retention of radiation for up to 96 hours (26). Moreover the combination has been shown to be more effective than either drug alone, resulting in the long-term cure of mice with colorectal xenografts (7).

We have confirmed the tumor-specific uptake of ¹³¹I-A5B7 in this cohort of patients. At the early time point of four hours the percent activity in the tumor was low relative to blood and normal organs, but this relationship was reversed at 68 hours consistent with poor initial penetration of the whole antibody but later retention within the tumor. Considering the dosimetry, all normal tissues had an absorbed dose of <0.3 mGy/MBq with the exception of the lung. Direct comparison of absorbed doses to target organs (Fig. 2B) and to the tumor (Fig. 2C) is difficult because the target organ dose includes dose from adjacent organs as modeled by the MIRD (medical internal radiation dose) schema (Olinda/EXM software), but for the tumor, only self-dose is calculated. Furthermore, the distribution of radiation within the tumor is heterogeneous, with greater accumulation in well-perfused viable areas; this is ignored in the analysis that looks at whole tumor volume and assumes uniform distribution. Both of these considerations will tend to underestimate the dose to viable tumor, but in spite of this the dose to tumors in patients 5, 6, 8, and 9 was above that for all normal organs with the exception of the lung.

In our study we selected a starting dose of each drug that was below the maximum tolerated dose reported in the single agent trials (4–6, 9) yet experienced dose-limiting myelosuppression within the first cohort requiring deescalation of ¹³¹I-A5B7. Despite reducing the dose of ¹³¹I-A5B7 to 1,600 MBq/m², well below the single-agent dose of 2,400 MBq/m², we again found dose-limiting myelosuppression. This cannot be explained by greater marrow dose because the red marrow dose in the two studies was very similar with a mean of 0.028 cGy/MBq in the Lane study compared with 0.01 cGy/MBq in the current study. One explanation is that CA4P altered the pharmacokinetics of ¹³¹I-A5B7 resulting in increased marrow toxicity. However, the t_1/2 of the slow clearance phase in this study was similar to the t_1/2 reported for ¹³¹I-A5B7 in the original single agent phase I study (9). Another consideration is that CA4P, in common with other microtubule-binding agents, has independent direct cytotoxic activity (27–29) that may be additive or synergistic with radiation in inducing marrow toxicity. In a phase I trial in which CA4P was given 60 minutes after carboplatin, dose-limiting thrombocytopenia was reported for carboplatin AUC 5 mg min/mL with CA4P at a relatively low dose of 36 mg/m² (30). This was probably due to reduced renal clearance of the carboplatin as CA4P has been shown to cause a temporary reduction in renal blood flow (31), although a pharmacodynamic affect on the bone marrow could not be excluded. It is unlikely that the initial dose of CA4P increased myelosuppression because the previous single agent phase I trials of CA4P did not report myelosuppression nor was myelosuppression increased in two other studies in which CA4P was given the day before paclitaxel and carboplatin (32, 33). Preclinical studies suggest that the efficacy of VDA and chemotherapy combinations are schedule-dependent (34), and toxicity may be similarly affected. Finally, there is a significant difference in previous treatments received by patients in the initial ¹³¹I-A5B7 trial, who had received either no prior treatment or 5-Fluorouracil alone, compared with those in the current study in which all had received at least two different cytotoxic agents and 9 of the 12 patients had two or more lines of chemotherapy. The bone marrow reserve in the current cohort would have been compromised to a greater extent, perhaps accounting for the reduced tolerance of radioimmunotherapy.

The initial study of ¹³¹I-A5B7 reported one partial response out of a cohort of ten patients. In this more heavily pretreated cohort, no partial or complete responses occurred according to RECIST criteria in 11 patients who received both drugs. It is notable, however, that the one minor response, with a decrease in CA19-9 levels from 10,507 to 2,033 IU/mL, occurred in patient 6 whose tumor received the highest radiation dose and who also had a substantial decrease in tumor IAUGC. This is probably the only case in which conditions were achieved which could be expected to produce a tumor response. Patient 10 had the largest decrease in K^trans but experienced tumor pain after the first dose of CA4P and declined further treatment including ¹³¹I-A5B7, so it was not possible to assess the effect of combination therapy in this case. It might be that the benefit of the combination is only be seen in those who have an objective reduction in K^trans and, in this exploratory study, the dose of CA4P was too low to achieve a significant reduction in the majority of patients. The tumor dose of radiotherapy could be also increased by the use of different immune formats such as polyvalent scFvs, diabodies, or minibodies which can improve tumor penetration (35) and also by the use of humanized forms to allow repeat dosing.

In conclusion we believe this to be the first reported trial exploring the combination of a VDA and radioimmunotherapy in advanced solid tumors. We have shown that these agents can be safely combined but that myelosuppression is dose-limiting in heavily pretreated patients. The use of dynamic imaging and dosimetry may help define the group of patients most likely to benefit in further studies, and the data recorded indicate that higher response rates may be achieved by higher levels of radionuclide targeting and of vascular disrupting effect.

Disclosure of Potential Conflicts of Interest

G.J. Rustin has served as a consultant for Oxigene.