Main

High-dose methotrexate (HD-MTX) of more than 1 g m−2 body surface area, administered by a prolonged intravenous infusion with leucovorin (LV, D,L-5-formyltetrahydrofolic acid) rescue, has demonstrated clinical activity in malignancies and diseases, such as acute lymphoblastic leukaemia, lymphoma, osteosarcoma, and head and neck cancer.

Methotrexate is primarily excreted by the renal route. It may cause acute nephrotoxicity, which is presumed to result from its precipitation or its relatively insoluble metabolites in acidic urine. This nephrotoxicity leads to delayed MTX elimination, ineffective rescue and a marked increase of other nonhaematological and haematological toxicities associated with MTX, such as myelosuppression, oro-gastrointestinal mucositis and dermatitis. In early studies using HD-MTX, severe toxicity occurred in approximately 10% of patients and there was a 6% mortality rate (Ahmad et al, 1978). The introduction of pretreatment hydration, alkalinisation of urine, routine monitoring of serum MTX concentrations and postinfusion guided LV rescue has decreased the incidence of severe and life-threatening toxicity after HD-MTX from 10% to less than 1% (Jüergens et al, 1983; Allegra, 1990). However, severe nephrotoxicity of HD-MTX leading to potentially lethal toxicity still occurs.

For patients with decreased MTX clearance due to renal dysfunction, therapeutic options are few and of limited efficacy. When serum MTX concentrations are persistently greater than 10–100 μ M, high doses of LV are not likely to completely reverse the toxicity of MTX (Goldman and Levy, 1968; Pinedo et al, 1976; Frei et al, 1980; Grem et al, 1991).

Thymidine may be used in case of MTX toxicity, but due to its rapid clearance it must be administered in high doses by continuous infusion (Ensminger and Frei, 1977; Howell et al, 1978; Howell et al, 1980; van den Bongard et al, 2001). Neither LV nor thymidine has any effect on the underlying problem of delayed MTX elimination. Peritoneal dialysis is ineffective, most likely due to a combination of factors including the protein binding, the high degree of ionisation and the low lipid solubility of MTX (Ahmad et al, 1978). Haemodialysis and haemoperfusion have variable efficacy and usually produce only transient decreases of plasma MTX concentrations (Gibson et al, 1978; Relling et al, 1988; Saland et al, 2002). Alternative methods to treat persistently elevated MTX levels are therefore needed.

The carboxypeptidase G class of enzymes hydrolyse the C-terminal glutamate residue from folic acid and classical antifolates, such as MTX, rendering them inactive (Levy and Goldman, 1967; Chabner et al, 1972, Goldman, 1975).

Carboxypeptidase G1 (CPDG1) was originally extracted from Pseudomonas stutzeri. During the 1970s, CPDG1 was administered successfully to a small number of patients with brain tumours and to one patient with pre-existing renal failure after MTX administration. However, the bacterial source for CPDG1 is no longer available (Bertino et al, 1974; Abelson et al, 1978; Minton et al, 1983).

A recombinant form of CPDG2, cloned from Pseudomonas strain RS-16, is now available as an investigational drug (Minton et al, 1983; Sherwood et al, 1985). Carboxypeptidase G2 hydrolyses MTX rapidly to the inactive metabolites DAMPA (4-[[2,4-diamino-6-(pteridinyl)methyl]-methylamino]-benzoic acid) and glutamate. In contrast to CPDG1, CPDG2 has a higher affinity for MTX (Km 8 μ M) than LV (Km 120 μ M) and 5-methyltetrahydrofolate (Km 35 μ M), the primary circulating metabolite of LV. In studies performed in rhesus monkeys, a single CPDG2 dose of 50 units (U) kg−1 lowered serum MTX steady-state concentration from 10 μ M to nontoxic levels of <0.05 μ M within 30 min (Adamson et al, 1992). In addition, the bolus injection of CPDG2 (a dose of 50 U kg−1 in combination with LV and partial thymidine rescue) resulted in a rapid 95–99% reduction of plasma levels of MTX in patients with renal dysfunction after HD-MTX (Widemann et al, 1995, 1997; Zoubek et al, 1995).

This study will demonstrate the feasibility, effectiveness and toxicity of an emergency therapy with CPDG2 in a large European patient collective with acute MTX intoxication and renal failure.

Patients and methods

Patients

Study design

The study was a prospective, open, nonrandomised, multicentre trial. The primary study centres were in Bonn, Berlin and Vienna. Their task consisted of consultation and emergency dispatch of CPDG2 for patients with MTX intoxication. The substance was requested by 50 different hospitals in 13 different countries (Austria, Belgium, Czechia, France, Germany, Greece, Italy, Israel, Norway, Slovakia, Spain, Sweden and Switzerland).

Patient eligibility

Patients of any age were eligible if they had (1) serum MTX concentration of >10 μ M at 36 h or of >5 μ M at 42 h or of >3 μ M at 48 h after start of MTX infusion, and (2) decreased diuresis (less than 50% excretion of the input hydration) or (3) a serum creatinine (CreaS) >1.5 times the upper limit of normal and documented increase of CreaS during the infusion period.

Treatment

Carboxypeptidase G2 administration

Recombinant CPDG2 was manufactured by the Centre for Applied Microbiology and Research (Salisbury, UK) and supplied in lyophilised form with vials containing 1000 U of enzyme activity. The vials were each resuspended in 1 ml of sterile isotonic saline; further dilution with sterile isotonic saline (1 : 5 or 1 : 10) was recommended. Carboxypeptidase G2 was administered at a dose of 50 U kg−1 over 5 min intravenously by an infusion pump or by bolus injection. Following each dose of CPDG2, serum MTX concentrations were determined. Patients who experienced a decrease in serum MTX concentration greater than one logarithm following CPDG2 administration but still had serum MTX concentrations >1 μ M might receive additional doses of CPDG2 with the approval of the principal investigator of the study.

Leucovorin

Leucovorin administration had to be stopped 4 h prior to CPDG2 and readministered 1 h following enzyme injection. The recommended LV dose during the first 24 h after the CPDG2 injection was 100 mg m−2 every 6 h followed by an increased rescue for 5 days according to the scheme recommended in the ALL BFM 95 protocol (Schrappe et al, 1995).

Patient monitoring

Patients were evaluated daily for signs and symptoms of MTX toxicity. Complete blood counts with differential, bilirubin and ALT, AST were determined at least twice weekly. Recovery of renal function was documented daily by serum creatinine. The nonhaematological (i.e. renal, liver, neurological, gastrointestinal toxicity and infectious complications) and haematological toxicity of MTX and side effects of CPDG2 (especially allergic reactions) were determined according to National Cancer Institute Common Toxicity Criteria (CTC) version 2.0.

Sampling for determination of MTX concentration

EDTA samples for MTX determination were obtained immediately before and 15, 30, 60 and 120 min following the dose of CPDG2, and then once daily until recovery of kidney function and decrease of MTX serum levels to less than 0.1 μ M. Most of the local hospitals determined the MTX level immediately with the methods such as fluorescence polarisation immunoassay (TDX) or enzyme-multiplied immunotechnique (EMIT). In a few centres, additional blood samples for analysis by high-pressure liquid chromatography (HPLC) were placed on ice, and serum was rapidly separated by centrifugation. To inactivate CPDG2, the serum samples were heated to greater than 80°C for 5 min in a water bath or treated with 1 N HCl to give a final concentration of 0.1 N hydrochloric acid. If the enzyme activity in the samples was not fully stopped, MTX in the sample might continue to be degraded after collection. Serum was stored at −20°C until time of analysis. All serum samples were sent on dry ice to a central laboratory where they were stored at −20°C until analysis by HPLC.

Methods

DAMPA, the catabolic product of CPDG2 action on MTX, is known to crossreact with MTX in most commercial immunological MTX assays (Allegra, 1990; Albertioni et al, 1996; Widemann et al, 1997). Consequently, MTX concentrations determined by commercial laboratories are unreliable following treatment with CPDG2. Methotrexate must be assayed by a specific HPLC or enzyme inhibition method.

High-pressure liquid chromatography method

Serum MTX was measured by HPLC using a modification of a previously reported method (Brandsteterova et al, 1990). After solid phase extraction using C18 cartridges (WatersOasis™ HLB 3 cm3, 60 mg (Waters, Eschborn, Germany)), samples were injected onto a 3 μm, Supelcosil™ LC-18-DB 15 cm × 4.6 mm (Sulpelco, München, Germany) radial compression analytical column with a Supelquard™ LC-18-DB 5 μm (Sulpelco, München, Germany) guard column and eluted isocratically with 80 : 20 (v v−1) 0.1 mol l−1 sodium phosphate (pH 6.8) : methanol (Lichrosolv®) at a flow rate of 1.2 ml min−1. Eluent was monitored using a Waters 490 UV adjustable wavelength detector at a wavelength of 310 nm (ThermoQuest, Egelsbach, Germany).

Under these conditions, the retention times for MTX, DAMPA and the internal standard 4-[[2,4-diamino-6-(pteridinyl)methyl]amino]-benzoic acid (Dm-APA) were approximately 5.6, 9.5 and 3.5 min, respectively. The peak areas were evaluated using the Chromquest software (2.5.1, 1998, ThermoQuest, Egelsbach, Germany). Methotrexate and DAMPA were corrected by means of an internal standard in combination with a calibration curve.

Statistical analysis

Patient data were collected consecutively, recorded on data sheets prepared centrally beforehand, checked by the principal investigator and recorded in a protected database at the University of Bonn. The biometric analysis was performed for all eligible patients who were evaluable (1) for response to the study medication, (2) for the toxicity of CPDG2 and (3) for the toxicity of MTX due to complete data documentation. Data were evaluated using descriptive statistical methods (median, ranges, frequencies and percentages).

Ethics

The study was conducted in accordance with the updated declaration of Helsinki (1996, Somerset West, Republic of South Africa) and approved by the local ethics committees in Bonn, Berlin and Vienna. Prior to enrolment in the study, the patient's parents/legal representatives and/or the patients themselves were informed of the investigational character of the study and the emergency use of CPDG2 and had given their written informed consent.

Results

Patient characteristics

A total of 82 patients from institutions in Europe were enrolled in this compassionate and emergency use protocol between May 1997 and March 2002. Complete or nearly complete data were available from 65 patients. For 17 patients, no data are available despite repeated telephone and written efforts. The evaluable 65 patients aged 0.9–71.8 years (median 15.4 years) suffered from acute lymphoblastic leukaemia (n=26), non-Hodgkin's lymphoma (NHL; n=21), osteosarcoma (n=12), brain tumours (n=3), Hodgkin's lymphoma (n=2) and pleural mesothelioma (n=1). The patient's characteristics and the different MTX regimens are listed in Table 1. Except for one patient who had clearly increased serum creatinine of 156.6 μmol l−1, all patients had an age-dependent normal value of CreaS before treatment (range 17.7–120.4 μmol l−1, median 60.2 μmol l−1). All patients received intravenous hydration and alkalinisation before, during and after the MTX infusion.

Table 1 Patient characteristics prior to CPDG2 administration
Full size table

Contact

Contact with the study centres was established by telephone at a median of 47.5 h (range 19–142 h) after start of the MTX infusion. The MTX levels at the time of contact were between 0.52 and 1082 μ M (median 18.0 μ M) (n=65). Two patients were included in the study with MTX levels of 0.52 μ M at 94 h (case 1) and 1 μ M at 142 h (case 2) after start of MTX and a serum creatinine of six-fold of the initial value. A total of 12 patients had no documented increase of serum creatinine at the time of contact in contrast to the inclusion criteria. Despite the absent renal failure, the principal investigator (UB) decided to give CPDG2 due to very high MTX levels at a late time of contact.

Methotrexate levels and serum creatinine at the first contact and before CPDG2 administration are documented in Table 1.

Carboxypeptidase G2 administration

Immediately before CPDG2 administration, a median MTX serum level of 11.93 μ M (range 0.52–901 μ M) was documented (n=58).

A single dose of 50 U CPDG2 per kg body weight was intended. A vial CPDG2 contains 1000 U dry frozen. In all, 57 of 58 evaluable patients received a dose of between 33 and 60 U CPDG2 per kg body weight (median 50 U kg−1). One patient received only a dose of 17 U CPDG2 kg−1, although the correct amount was available. Carboxypeptidase G2 was administered at a median of 52 h (range 25–178 h) following the start of an MTX infusion of 1.5–12 g m−2 over 4–36 h and was well tolerated. At 15 min after CPDG2 administration, the MTX serum level of these patients was reduced by around 87%.

Nine patients received a second dose of CPDG2. One patient received an additional third dose of CPDG2. Six of these nine patients suffered from an osteosarcoma and received an MTX therapy of 12 g m−2 over 4 h. An additional three patients suffered from ALL and NHL and received MTX infusion of 5 g m−2 24 h−1 (n=2) or 8 g m−2 6 h−1 (n=1), respectively. The median time point of the first CPDG2 dose in these patients was 37.5 h after the start of MTX infusions. The second dose was given at a median of 6 h (range 1.5–90 h) after the first dose. One patient received a second dose of CPDG2 1.5 h after the first dose of CPDG2 due to logistics.

Side effects of CPDG2

After the administration of CPDG2, side effects described by two patients included flushing (n=2) and shaking (n=1). Both had stable vital signs and the symptoms resolved completely without any intervention.

Methotrexate levels

Carboxypeptidase G2 resulted in a rapid reduction of serum MTX concentrations in all patients. The MTX concentration determined with immunological MTX detection such as TDX or EMIT showed a median decrease of the MTX level of about 87% (range 70–99%) 15 min after CPDG2 administration. Subsequent samples also showed an ongoing decrease of MTX concentrations. Additionally, the MTX concentrations after CPDG2 administration were determined by the HPLC method. The MTX levels estimated by HPLC 15 min after the first CPDG2 injection decreased by a median of 97% (range 73–99%) (Figure 1). Almost immediately after CPDG2 administration, DAMPA was detectable at a concentration comparable to the serum concentration of MTX before CPDG2.

Figure 1
figure 1

Number and time of MTX samples determined by TDX/EMIT or HPLC and DAMPA samples determined by HPLC after CPDG2 injection.

Full size image

Nine patients received an additional dose of CPDG2, because their serum MTX concentrations determined by immunological methods remained higher than 1 μ M after the first dose of CPDG2. The first dose of CPDG2 resulted in a median decrease of 88% (Table 2). The second dose of CPDG2 did not result in a significant further decline in the serum MTX concentration. Five patients with an MTX level at a median of 8.9 μ M (range 3.1–77.6 μ M) before the second CPDG2 injection showed a second decrease at a median of 26% (range 5–70%). An additional two patients showed an increase after second CPDG2 injection of 2 and 26%. In one patient, the MTX level after the second dose of CPDG2 is not known and in one case the time between the first and second injections was only 1.5 h and there was no MTX level within the documented level. All patients who received a second dose showed a clearly limited kidney function with median CreaS of 274.4 μmol l−1 (range 168.2–654.9 μmol l−1) before the second administration.

Table 2 Characteristics of patients who got a second CPDG2 injection
Full size table

Methotrexate-related toxicity (CTC)

In all, 50% of the patients had a reduced Karnowsky index (⩽60%) due to the MTX intoxication. Gastrointestinal complaints with nausea and diarrhoea were frequent. Severe stomatitis (CTC scale ⩾2) occurred in 33% of the patients. A total of 66% of the patients showed an increase of serum creatinine of more than 1.5 times of the initial value before MTX therapy. Increased liver values based on increased GPT/GOT were documented in 58% of the patients with CTC scale ⩾2. Only a small number of patients suffered high-grade toxicity of lung and heart/circulation. In five and six patients, respectively, a mild peripheral and/or central neurotoxicity was observed. Two patients suffered from severe central neurotoxic side effects, a female patient additionally from severe peripheral neurotoxic side effects (CTC scale 4). Further data are listed in Table 3.

Table 3 Methotrexate-related toxicity
Full size table

Methotrexate-related mortality

Four patients died due to severe myelosuppression and septic complications 7–22 days following the MTX infusion. Three patients (Table 4, cases 1, 2 and 4) had very high MTX concentrations several hours after the start of MTX, which were lowered by the administration by CPDG2 insufficiently. Thus, progressive renal failure and fatal septic complications were observed. In an additional patient (case 3), the MTX concentration was lowered by CPDG2 adequately without progressive renal failure. This patient died due to prolonged myelosuppression and its complications (pulmonary haemorrhage due to pulmonary aspergillosis), which were caused by the MTX intoxication and the progressive underlying disease (T-cell ALL) too.

Table 4 Characteristics of patients who died related to MTX intoxication
Full size table

The patient characteristics are shown in Table 4. The course of MTX serum levels is documented in Figure 2.

Figure 2
figure 2

Methotrexate levels after first and second CPDG2 injection in patients who died related to MTX intoxication.

Full size image

Discussion

Methotrexate-induced toxicity is frequently associated with delayed MTX excretion due to renal dysfunction. Delayed MTX excretion leads to prolonged drug exposure and the potential for severe life-threatening toxicity. Patients with impaired renal function prior to the MTX infusion, of advanced age or receiving concomitant nonsteroidal anti-inflammatory drugs or other nephrotoxic substances are at an increased risk of developing renal dysfunction during MTX infusion. In this study, a pre-existing reduced kidney function was known before initiating the MTX therapy in only one patient. All other patients showed increased MTX concentrations unexpectedly without risk factors. Rescue with LV can prevent toxicity but has no effect on delayed MTX excretion.

Carboxypeptidase G2 rescue offers the distinct advantage of providing an alternative, rapid route of elimination of MTX by enzymatically catabolising the drug to the inactive metabolite DAMPA.

DAMPA is a minimal inhibitor of dihydrofolate reductase and less soluble than MTX in acidic pH. The faster elimination of DAMPA by comparison with MTX in patients with renal failure, who received CPDG2, is due to the extrarenal elimination of DAMPA (Widemann et al, 2000). After systemic exposure to CPDG2, DAMPA serum concentrations are equivalent to MTX concentrations prior to CPDG2 (compare Figure 1). Because of the risk of additional nephrotoxicity of DAMPA, patients must continue to receive fluid hydration and alkalisation.

After the initial dose of CPDG2, all patients experienced a rapid (<1 h) and marked (>97%) decrease in serum MTX concentration (determined by HPLC).

In the first years of this study, nine patients received a second dose of CPDG2 not resulting in a further marked decrease of the MTX concentration. This could be the result of high DAMPA concentrations inhibiting the hydrolysis of MTX by CPDG2 (Widemann et al, 1997).

In addition, substantial crossreactivity of the MTX antibodies with DAMPA used in immunoassays results in overestimation of serum MTX concentrations. Therefore, the MTX and DAMPA serum concentration after CPDG2 must be monitored by another technique, such as HPLC. The MTX levels estimated by HPLC decreased in median to less than 3% (range 1–27%) of the initial values 15 min after CPDG2 injection.

Due to its molecular size, CPDG2 is restricted to the extracellular compartment, and the intracellular MTX concentration is initially unaffected by its use. In time, the changed equilibrium between intracellular and extracellular MTX will result in the efflux of intracellular MTX back into the serum, resulting in a rise of serum MTX levels some hours after CPDG2 administration, but this is a relatively slow process. For this reason, rescue with LV must be continued following the CPDG2 administration.

In this study, six of 65 patients were treated additionally by haemofiltration and/or haemodialysis. These methods must be repeated continuously or daily, in order to reduce the MTX level to nontoxic concentrations. A further risk of these methods results from the necessary vascular entrance, as well as the risk of bleeding secondary to heparinisation and thrombocytopenia. The effectiveness of these procedures cannot be evaluated from our data. Three patients deceased despite these approaches.

In total, four patients died despite CPDG2 injection due to severe myelosuppression and septic complications 7–22 days following the MTX infusion. The probable reasons for the lethal complications during a prolonged myelosuppression were (1) very high MTX level with an inadequate decrease of MTX levels after the second CPDG2 injection, (2) delayed CPDG2 injection (143.5 h after the start of MTX infusion), (3) a progressive underlying disease (T-cell ALL) and (4) inadequate treatment of severe anaemia in a patient who was a Jehovah witness and refused treatment.

In summary, systemic administration of CPDG2 followed by long-term LV is highly effective in rescuing patients at high risk of life-threatening MTX toxicity. Carboxypeptidase G2 is safe and well tolerated. No severe side effects were observed in this study. Administration of CPDG2 is most beneficial for patients with MTX-induced renal failure, especially if it is administered within 48–72 h after the start of MTX infusion.

For the future, it might be helpful to differentiate between patients with extremly high MTX concentrations and severel renal failure and patients with moderately high MTX concentration and renal failure. Since additional injections at CPDG2 did not result in a significant further decrease of MTX levels, one initial high dose of 100 U kg−1 CPDG2 for patients with extremely high MTX concentrations (for example, >100 μmol l−1 at 48 h after MTX start and renal failure) may be more effective. For the patients with moderately high MTX levels and renal failure, the application of 50 U kg−1 seems to be sufficient. Due to the high costs of the drug, it may be discussed whether an application of 25 U kg−1 CPDG2 would not be sufficient to lower the MTX levels to nontoxic concentrations in patients with marginal high MTX concentrations and renal failure.

The therapy with CPDG2 should always be accompanied by a sufficient hydration, alkalisation and a long-term LV therapy, in order to counterbalance the intracellular MTX. Haemofiltration or haemodialysis might be helpful in oliguric or anuric renal failure. In individual cases with renal failure, severe and lethal courses may occur despite a CPDG2 application, in particular if delayed CPDG2 administration did not prevent overt toxicity.