Phase I Trial of Methotrexate-Albumin in a Weekly Intravenous Bolus Regimen in Cancer Patients¹

MTX³ (amethopterin) was synthesized by Seeger et al. about 50 years ago (1). The antimetabolite MTX provided the first example of a drug-induced cancer cure by chemotherapy after administration to women suffering from metastatic choriocarcinoma (2). Since then, a broad range of single-agent cytostatic activity has been described, especially when MTX is given as a high-dose treatment counteracted by folinic acid rescue (3). Nowadays, a variety of MTX treatment modalities and promising new antifolates have evolved (4). MTX can be used in low-dose regimens without folinic acid rescue (30 mg/m² once a week or 50–100 mg/m² at 3–4 week intervals). Intermediate- or high-dose MTX treatment (100–1,500 or 8,000–12,000 mg/m² every 1–4 weeks) requires folinic acid rescue to curb toxic side effects such as severe stomatitis, gastrointestinal mucositis, or myelosuppression (5). After i.v. administration, MTX is rapidly removed from circulation by the kidneys (40% after 6 h and 90% after 24 h), causing nephrotoxicity in some instances. The remainder is metabolized to 7-hydroxy-MTX, mostly in the liver. The mean distribution half-lives range from 1.5–3.5 h, whereas the terminal half-life is about 8–15 h (6).

Consequently, the tumor exposition time of MTX is short, if tumor volume doubling times measured in weeks or months for human cancers are considered (7). Not only a short in vivo half-life of MTX impairs an optimal therapeutic effect, but also the high interstitial pressure of solid tumors, caused by a missing lymphatic drainage, results in low tumor accumulation rates for MTX, favoring drug disposition toward healthy tissue (8). To overcome these limitations, we searched for a macromolecular drug targeting system that would offer a longer circulating half-life of MTX and would increase the tumor uptake rate. These requirements are realized by albumin. It carries about 70% of the transportable nitrogen reserves available in blood (9). Recent experimental evidence suggested that albumin preferentially accumulated in tumors (10, 11, 12). Proliferating cells take up albumin by fluid-phase endocytosis. After lysosomal digestion, the derived amino acids serve as a source for nitrogen and energy in the tumor cell. Albumin catabolism by the tumor itself is responsible in part for the nitrogen transfer to the tumor (13). These favorable properties supported the choice of albumin as a drug carrier.

Throughout the last three decades, a variety of attempts have been made to use albumin as a drug carrier for MTX. To our knowledge, the experiments with MTX-HSA conjugates remained in the preclinical state. The discrepant molecular weights of MTX (M_r 454) and albumin (M_r ∼67,000) directed all efforts to improve the therapeutic efficacy of these conjugates by raising the molecular load from an initial value of about 10 mol of MTX conjugated per albumin molecule to MTX-HSA carrying 56 mol of MTX (14, 15, 16, 17, 18). Recent radiopharmacological experiments in tumor-bearing rats demonstrated that MTX-HSA bearing more than 3 mol of MTX was preferentially trapped by the liver. Optimal pharmacokinetic features were found for MTX-HSA with a loading rate of approximately 1 mol of MTX. The distribution pattern of these conjugates in healthy and tumor-bearing rats was indistinguishable from native albumin. High tumor accumulation rates for both albumin and the MTX-albumin conjugate were observed, exceeding 15% of the injected dose after 72 h in rat Walker-256 carcinoma (19, 20). Antitumor activity of this MTX-albumin conjugate was shown in Walker-256 carcinoma in SD rats, in a prostatic adenocarcinoma Dunning (R-3327-Hi) in Copenhagen rats, and in a variety of human xenograft tumors in nude mice models (21, 22, 23). Preclinical toxicology screening in mice showed a favorable profile of MTX-HSA. Mucositis, stomatitis, shaggy fur, and diarrhea were typical side effects that were followed by signs of myelosuppression (predominant leukopenia) if the dose was stepped up. On the basis of these favorable properties, MTX-HSA was selected for further clinical development. We now report on the first Phase I study of MTX-HSA in cancer patients.

The trial was conducted under the guidance of the Phase I Study Group of the Association of Medical Oncology of the German Cancer Society and was sponsored by the German Cancer Research Center, Heidelberg. The objective of this study was to define the MTD of MTX-HSA and to record short-term, delayed, or late toxicity for a weekly administration regimen. Drug administration was monitored to determine steady-state conditions of the drug. All patients were required to have histologically proven cancer that was no longer amenable to standard treatment. Eligibility criteria were as follows: (a) performance status, WHO grades 0–3; (b) age, 18–75 years; (c) life expectancy, at least 3 months; (d) adequate bone marrow, liver, and renal function; (e) no chemotherapy or radiotherapy for at least 4 weeks; and (f) informed consent. Patients with brain metastases or with known contraindications for MTX treatment were excluded. The protocol was approved by the Ethics Committee of the University of Heidelberg, Germany. The study was performed according to the standard operating procedures of the Association of Medical Oncology Phase I/II Study Group (24).

Histories, physical examinations, full laboratory evaluation, and an electrocardiogram were performed at study entry and every 2 weeks during treatment. Blood counts were performed weekly. Extensive laboratory screening included testing for electrolytes, creatinine, urea, cholinesterase, bilirubin, albumin, serum electrophoresis, alanine aminotransferase, aspartate aminotransferase, γ-glutamyltranspeptidase, alkaline phosphatase, glucose, uric acid, blood lipids, Quick, a partial thromboplastin time, fibrinogen, and semiquantitative urine analysis. Radiological investigations comprised a chest X-ray, sonography, and CT or magnetic resonance imaging scans. Patient performance was classified from 0 to 4 according to the WHO scale. Treatment-related toxicity was graded according to the CTC checklist, which was adapted and published by the Phase I/II Study Group (24). Adverse drug reactions were defined as any adverse event whose relation to therapy was certain, probable, or possible. Improvements in disease had to persist for 4 weeks before being considered as a response. Initially, tumor response was checked every 4 weeks; during poststudy treatment, it was checked every 3 months. Duration of response was measured from the beginning of MTX-HSA treatment until the development of progression.

MTX-HSA was provided in 50-ml vials (1 mg of MTX per ml of solution bound to 100 mg of HSA; molar loading rate, 1.3:1) by the German Cancer Research Center, Heidelberg. The drug was administered i.v. once a week. The initial dose was chosen according to the LD₁₀ [lethal dose (for 10% of the mice)] calculation from preclinical toxicology in mice. The MTD of MTX-HSA in mice was 50 mg MTX/kg body weight (LD₁₀) administered weekly for 4 weeks. Converted to body surface, the MTD corresponded to 20 mg MTX/m² bound to albumin (later referred to as 20 mg/m² MTX-HSA; Ref. 25). Dose levels were escalated from 20 mg/m² to 40 mg/m² and 60 mg/m² and then de-escalated to 50 mg/m². Three patients were treated at each dose level. In each group, at least eight injections (one injection/week) were intended. If no DLT was observed during the first three administrations, the next dose level was started. If an episode of DLT occurred, up to six patients were treated at that level. If DLT occurred in more than two of these patients, then accrual stopped. The MTD was defined as one dose level below that at which two of six patients developed DLT. An additional six patients were intended for confirmation of this dose level. After the Phase I study period of 8 weeks, treatment intervals could be prolonged in responding patients. DLT was defined according to CTC. Critical parameters were grade 3 toxicity for hematological parameters; coagulation parameters; stomatitis; diarrhea; gastritis; gastrointestinal tract ulcers; mucositis; diseases of the lung, kidneys, urinary tract, skin, eyes, and ears; and neurological or autoimmune disorders. CTC grade 4 toxicity was chosen for hepatic function, metabolic disorders, vascular diseases, loss of appetite, and alopecia. Toxicity for each patient was assessed every week.

For pharmacokinetic analysis, blood samples were drawn before injections, 1 h after injection of MTX-HSA, 1 day after injection of MTX-HSA, and every 1 or 2 weeks during poststudy treatment. All samples were stored at −20°C until processing. MTX-HSA concentrations were determined using a modified EMIT procedure (20, 26). The EMIT MTX test is a homogeneous enzyme immunoassay. Patient MTX competes with MTX conjugated to a bacterial glucose-6 phosphate dehydrogenase enzyme for common binding sites of a sheep antibody to MTX (27). The antibody reactive to MTX showed cross-reactivity with MTX-HSA. We used serial dilutions of MTX-HSA with known MTX content as standards. This assay was optimized for measuring MTX-HSA with a standard curve ranging from 0.2–2.0 μmol/liter. Coefficients of variation were 2–12% for within-day precision and 3–8% for between-day precision, and accuracy ranged from 88–98%. Samples spiked with radiolabeled MTX-HSA were in good agreement with the immunologically determined MTX-HSA concentrations. Randomly selected MTX-HSA patient plasma samples were analyzed by HPLC. The content of free MTX was below 1%.⁴

Patients’ characteristics are listed in Table 1. Seventeen patients with various solid tumors were eligible. The median performance status was 1 according to the WHO scale. The number of courses administered is shown in Table 2. All 17 patients were evaluable for DLT within the first two consecutive courses; 13 patients received at least three courses, 12 patients received at least seven courses, and 8 patients received more than seven consecutive courses. One patient withdrew informed consent after two courses, and treatment was discontinued in two patients because of tumor-induced health deterioration after the second or third injection.

The data for nonhematological adverse drug reactions are shown in Table 3. Stomatitis was the major DLT. First signs (CTC grade 1) were seen in the 40 mg/m² group after six courses. It occurred with grade 3 intensity in the 50 and 60 mg/m² groups, beginning after four administrations in a patient treated with 50 mg/m² and after two or four courses in two patients treated with 60 mg/m². One female patient showed signs of skin toxicity with slight papulosquamous lesions on both forearms and legs (grade 3) after the first administration of 60 mg/m². This patient had a known history of 5-fluorouracil-induced skin toxicity. All DLT symptoms resolved within 7–14 days after treatment was stopped. Folinic acid rescue was not necessary. Further adverse drug reactions in these patients were intercurrent nausea, diarrhea, constipation, and slight nose bleeding due to mucositis not exceeding CTC grade 1–2. Two patients complained of fatigue possibly related to the treatment. Among other adverse events not considered to be associated with treatment were single events of dyspnea, edema, thrombosis, and fever and one case of supraventricular tachyarrhythmia.

Hematological and laboratory adverse drug reactions are listed in Table 4. After six consecutive weekly administrations of 40 mg/m² MTX-HSA, thrombocytopenia with a nadir (19,000/μl) at day 10 (grade 4) was dose-limiting in one patient. No severe bleeding was seen, and normal values were reached within 15 days. Two patients in the 50 mg/m² dose level showed grade 2 thrombocytopenia after eight consecutive courses. No leukopenia was found, and in a single case, grade 1 granulopenia occurred. Single events of a relative lymphopenia were seen in some patients up to grade 3. Despite ongoing treatment, normal values were re-achieved. In some patients, intercurrent transaminitis (up to grade 3) and slight elevations of alkaline phosphatase occurred. One patient without liver metastases showed CTC grade 3 elevation of transaminases (alanine aminotransferase, 214 units/liter, aspartate aminotransferase, 140 units/liter) once after three administrations (50 mg/m² dose level), which resolved under ongoing treatment. In some patients, grade 2 or grade 3 transaminitis and hyperbilirubinemia were associated with a rapid progression of liver metastases. Further adverse laboratory events (CTC grades 1–3) that were not considered to be associated with treatment included deviations of coagulation parameters, glucose, blood urea nitrogen, creatinine, bilirubin, alkaline phosphatase, transaminases, and electrolytes. Slight intercurrent increases of creatinine were found in two patients. One patient with a known history of choledocholithiasis showed a short episode of a grade 2 increase in amylase. Overall, MTX-HSA could be administered safely to outpatients. Allergic reactions or other acute toxic reactions did not occur, even in patients with tumor responses, each patient receiving more than 35 injections of MTX-HSA.

In 3 of 10 patients treated with 50 or 60 mg/m² MTX-HSA, tumor regressions were observed.

Minor tumor response was observed in a 69- year-old woman who suffered from a life-threatening solitary metastasis of renal carcinoma that had infiltrated the cervical spine, compressing the spinal chord despite radiation therapy. In this patient, therapy began with 60 mg/m² MTX-HSA. After four administrations of MTX-HSA, stomatitis (CTC grade 3) occurred, confined to the regions of the mouth adjacent to the target area of the previous radiotherapy to the cervical spine. Treatment was interrupted until complete recovery 3 weeks later, de-escalated, and continued with weekly injections of 40 mg/m². The neurological symptoms caused by tumor compression of the spinal chord disappeared. After another two injections, slight stomatitis (grade 2) reappeared. MTX-HSA treatment was continued, but intervals were prolonged to injections every second or third week. The patient received 28 injections with a cumulative dose of 2124 mg of MTX conjugated to HSA (152 mg MTX/month). A 30% tumor regression was documented by CT scans and lasted 14 months until tumor progression.

Tumor response was observed in a 61-year- old man who had developed a large cell pleural mesothelioma in March 1996. The tumor infiltrated the right pleura and the mediastinum. Tumor removal by decortication was not feasible. Pleurodesis was unsuccessfully performed three times. The patient suffered from recurrent pleural effusions, and dyspnea worsened until April 1996. The patient was enrolled in the 50 mg/m² MTX-HSA group. At the end of the study treatment period, after eight administrations of MTX-HSA, stomatitis (grade 2) and thrombocytopenia (grade 2) were observed. Treatment was interrupted for 3 weeks until complete restitution and then continued with 50 mg/m² MTX-HSA at 2–4-week intervals. During week 24, slight stomatitis reoccurred, and treatment was paused again for 4 weeks and then resumed. Until September 1998, the patient had received 35 administrations of MTX-HSA (cumulative MTX dose, 3270 mg or 121 mg MTX/month). No significant toxicity was observed. Chest X-rays showed a continuing reduction in tumor burden of about 40%. Dyspnea had vanished during the study period. The patient reported well being and resumed his hobbies as a baritone in a choir and skiing. Poststudy treatment is ongoing, with no signs of a tumor relapse in this patient at the time of completing this article (December 1998; tumor response time, 31 months, ongoing).

Tumor response was also seen in a 63-year-old male patient who had undergone nephrectomy due to renal cell carcinoma in 1989. In 1993, he developed a local solitary lymph node metastasis that was surgically removed; other lymph node metastases reappeared and were no longer surgically amenable. Despite IFN treatment, tumor progression occurred. In June 1996, the patient presented with progressive disease (multiple pulmonary metastases and several intra-abdominal para-aortal metastases of up to 6 cm in diameter). After two injections of 60 mg/m², the treatment had to be paused due to cholangitis (known choledocholithiasis) for 5 weeks. Treatment was resumed with weekly injections of 50 mg/m² MTX-HSA. After five administrations, treatment was paused due to stomatitis (grade 2). Since September 1996, therapy has been continued with 50 mg/m² MTX-HSA injections every 2–4 weeks. No toxicity was observed. CT scans showed a reduction of the pulmonary metastases and the intra-abdominal lymphomas of more than 80%. Until September 1998, 39 courses of MTX-HSA had been administered (cumulative MTX dose, 3430 mg; 123 mg MTX/month). At the time this article was completed, there had been no signs of tumor progression, the patient had reported well-being, and the MTX-HSA treatment was being continued (tumor response time, 30 months, ongoing).

Typical MTX-HSA concentration curves are shown in Figs. 1 and 2 from two patients with tumor responses (see “Case II” and “Case III”). In Fig. 1, the plasma concentrations obtained during 16 courses of MTX-HSA are shown. Peak concentrations of 90 μmol/liter MTX-HSA were measured 1 h after the injection of 50 mg/m² MTX-HSA. Due to the prolonged half-life of MTX-HSA, the drug conjugate accumulated from a baseline value of 15 μmol/liter MTX-HSA to 40 μmol/liter MTX-HSA after eight injections. Due to side effects, treatment was paused for 3 weeks and resumed after the side effects had disappeared and the baseline MTX-HSA concentrations had dropped to 25 μmol/liter MTX-HSA. By injecting MTX-HSA every 2–3 weeks, steady-state conditions were achieved. After an additional six courses, slight side effects (stomatitis) recurred at baseline concentrations of about 30 μmol/liter MTX-HSA. To be on the safe side, the treatment was stopped until baseline concentrations had decreased to approximately 15 μmol/liter MTX-HSA. For further long-term treatment, the MTX-HSA concentrations were kept between 10 and 20 μmol/liter. In Fig. 2, MTX-HSA plasma concentrations of a patient with renal cancer are shown, including the first 15 injections over 36 weeks. Initially, the patient received two administrations of 60 mg/m² MTX-HSA. Due to an episode of a cholangitis, treatment was interrupted and was later resumed with 50 mg/m² MTX-HSA in weekly courses. After five injections, MTX-HSA accumulated to 30 μmol/liter MTX-HSA. Grade 2 stomatitis was observed, and the treatment was paused. On weeks 27 and 28, the patient received injections of MTX-HSA, but only baseline concentrations were measured. After keeping the MTX-HSA concentrations between 10 and 20 μmol/liter, no major side effects of long-term MTX-HSA have been reported in this patient. The declines in plasma baseline concentration during treatment interruptions clearly showed that the half-lives of the MTX-HSA conjugate were close to 3 weeks. If administered in 2–3-week intervals, steady-state conditions were achieved during poststudy treatment.

This Phase I study demonstrated that the prodrug conjugate MTX-HSA could be safely administered to cancer patients. The conjugate comprises MTX, a cytostatic drug, and human albumin, a macromolecular carrier. The loading rate of albumin with MTX was kept low at a molar ratio of approximately 1:1.3. Preclinical studies had revealed that low loading rates were crucial for optimal tumor targeting. Only conjugates with a loading rate below 3:1 (MTX:HSA) enjoyed the same favorable distribution pattern and the long plasma half-lives of native albumin (19).

Seventeen patients received weekly injections over a period of 2–8 weeks (a total of 102 courses) or until DLT. All injections were well tolerated, without signs of allergic reactions or a vascular leakage syndrome. Side effects of MTX-HSA observed in this study did not differ from the side effects commonly known from conventional MTX treatment. The predominant side effect was grade 1–3 stomatitis. If stomatitis grade 2 or grade 3 occurred in a patient, we paused MTX-HSA treatment until restitution (two of four patients in the 50 mg/m² group and three of six patients in the 60 mg/m² group). We did not prematurely resume therapy, taking into account the long biological half-life of MTX-HSA and the risk of further accumulation. Apart from one patient with a thrombocytopenia (grade 4) that resumed without bleeding disorder within 15 days after treatment cessation, no significant hematological toxicity was seen. The short-term toxicity profile of MTX-HSA was very favorable, and all patients retained outpatient status. Four patients (including hree responders) entered poststudy treatment. One patient was treated with MTX-HSA for 14 months until tumor progression and received 28 courses of MTX-HSA. Until September 1998, two other responding patients had received 35 or 39 injections of MTX-HSA. The observation periods are 30 and 31 months without signs of toxicity. MTX-HSA treatment will be continued in these patients. The evaluation of the plasma concentrations indicated that MTX-HSA administered on a weekly schedule will accumulate; depending on the initial dose of MTX-HSA, it took six injections of 40 mg/m² weekly, four to six injections of 50 mg/m², or two to four injections of 60 mg/m² to induce stomatitis. MTX-HSA concentrations of >30 μmol/liter (EMIT MTX Assay) were associated with the incidence of stomatitis.

The MTD was established for four courses of 50 mg/m² MTX-HSA administered at weekly intervals for short-term treatment. Due to drug accumulation, a weekly regimen is not suitable for long-term treatment. The administration of 50 mg/m² MTX-HSA every 2–4 weeks, achieving plasma concentrations between 10 and 20 μmol/liter, proved to be safe and effective, based on observation of the three responding patients.

Three tumor responses were seen, two (a minor response and a partial response) in patients with renal carcinoma and one (a minor response) in a patient suffering from pleural mesothelioma. Both tumor entities are known for their resistance to conventional chemotherapy (28, 29). There is currently no regimen available for renal cancer (28). Only one study by Solheim et al. (30) found that a conventional high-dose MTX regimen (four to eight courses of 3 g of MTX each, followed by folinic acid rescue) increased the median survival of patients with pleural mesothelioma by 11 months. Of 60 patients, 37% responded to the high-dose MTX treatment (30). Renal carcinoma and malignant mesothelioma of the pleura might be interesting target tumors for further Phase II studies with MTX-HSA. The underlying mechanisms for the possible effectiveness of MTX-HSA in these tumors are not yet clear. A variety of hypotheses might apply, based on the information available on tumor pathophysiology and the pharmacology of conventional MTX, the MTX-albumin conjugate, and the carrier protein albumin. Tumor volume doubling times in man range from weeks to months, although the replication process at the cellular level is completed within a time interval of a few tens of hours in the majority of individual tumor cells. This discrepancy is brought about by tumor cell loss due to necrosis or apoptosis (7). In this setting, successful antimetabolite-based chemotherapy, which is only effective during the comparatively short S-phase of the cell cycle, will require a continuous presence to exert an optimal therapeutic impact on the tumor. In this respect, the pharmacokinetics of the prodrug MTX-HSA differ decisively from conventional MTX. The conjugate takes advantage of the long biological half-life of albumin in man (19 days), the role of albumin as a major source for tumor nitrogen, and, consequently, the accumulation of albumin in tumors. We were able to maintain MTX-HSA levels in two cases over more than 30 months, continuously providing the cytostatic drug conjugate in vivo over several tumor volume doubling times. On the basis of the injected dose of MTX-HSA (50 mg/m²), it can be estimated that an injection of 100 mg of MTX-HSA will include approximately 10 g of the carrier protein. The albumin pool in man was estimated to be 350 g (9). One injection of MTX-HSA will cover about 3% of the albumin pool; after repeated injections of MTX-HSA, one can expect a 5% share. Thus, about 1 of 20 albumin molecules might be tagged by MTX.

MTX-HSA is a novel chemotherapeutic prodrug conjugate. Phase I testing revealed an excellent toxicological profile, allowing outpatient treatment and maintaining a high quality of life status for all cancer patients. Altered pharmacological properties of this conjugate in terms of plasma half-life, tumor targeting, and intracellular metabolism might contribute to the tumor responses observed in two patients with renal cancer and in one patient with pleural mesothelioma. MTX-HSA will be evaluated in additional Phase I and Phase II trials.

We thank Dr. A. Moeller (Department of Pharmacokinetics, Klinge Pharma, Fujisawa Group, Munich, Germany) for assistance in evaluating the data from drug level monitoring. We thank Prof. Dr. R. Hehlmann (III. Medical University Clinic Mannheim) and Prof. Dr. h. c. mult. H. zur Hausen (German Cancer Research Center, Heidelberg) for continuous support.