Farnesylprotein transferase (FT), an enzyme that catalyzes the first step in the posttranslational modification of ras and a number of other polypeptides, has emerged as an important target for the development of anticancer agents. SCH66336 is one of the first FT inhibitors to undergo clinical testing. We report a Phase I trial to assess the maximum tolerated dose, toxicities, and biological effectiveness of SCH66336 in inhibiting FT in vivo. Twenty patients with solid tumors received 92 courses of escalating SCH66336 doses given orally twice a day (b.i.d.) for 7 days out of every 3 weeks. Gastrointestinal toxicity (nausea, vomiting, and diarrhea) and fatigue were dose-limiting at 400 mg of SCH66336 b.i.d. Moderate reversible renal insufficiency, secondary to dehydration from gastrointestinal toxicity, was also seen. Inhibition of prelamin A farnesylation in buccal mucosa cells of patients treated with SCH66336 was demonstrated,confirming that SCH66336 inhibits protein farnesylation in vivo. One partial response was observed in a patient with previously treated metastatic non-small cell lung cancer, who remained on study for 14 months. This study not only establishes the dose for future testing on this schedule (350 mg b.i.d.) but also provides the first evidence of successful inhibition of FT in the clinical setting and the first hint of clinical activity for this class of agents.
Several small, hydrophilic eukaryotic G-proteins involved in cell signaling, including ras and rho, as well as several polypeptides of diverse function such as lamins A and B, rhodopsin kinase, transducin,and PxF are synthesized as cytoplasmic precursors that require a series of posttranslational modifications for conversion to mature membrane-bound forms (1, 2, 3, 4). The first and obligatory step in this process is the FT3-catalyzed transfer of a farnesyl moiety (15-carbon isoprene lipid) from farnesyl pyrophosphate to the cysteine residue located in a tetrapeptide CAAX (A,aliphatic; X, methionine or serine) sequence at the COOH terminus of all these proteins.
This pathway initially attracted attention because of its role in the processing of ras proteins, membrane-localized guanine nucleotide-binding polypeptides that function as molecular switches linking receptor and nonreceptor tyrosine kinase activation to downstream cytoplasmic and nuclear events. Three mammalian ras proto-oncogenes encode four related and highly conserved proteins, H-ras, N-ras, and the splice variants K-ras 4A and K-ras 4B(5). Activating mutations in these Ras polypeptides result in constitutive signaling, thereby stimulating cell proliferation and inhibiting apoptosis (6). Oncogenic ras mutations have been identified in approximately 30% of human cancers(7). K-ras mutations are frequent in non-small cell lung,colorectal, and pancreatic carcinomas; H-ras mutations are found in bladder, kidney, and thyroid carcinomas; and N-ras mutations are found in melanoma, hepatocellular carcinoma, and hematological malignancies(7). Because of the role of farnesylation in ras maturation (8), FT inhibition was envisioned as a strategy for interfering with ras-dependent transformation.
Whereas FTIs clearly inhibit ras farnesylation, it is unclear whether the antiproliferative effects of these compounds result exclusively from effects on ras. Geranylgeranylated forms of K-ras and N-ras, which are themselves capable of transforming cells, are observed in cells treated with FTIs (9). Despite this alternative prenylation pathway, FTIs inhibit proliferation of many tumor cells expressing activated K-ras in vitro and in vivo. In addition, several cell types that lack ras mutations are also sensitive to FTIs in vivo and in vitro(10). Collectively, these findings argue that inhibition of the farnesylation of other proteins might also contribute to the observed antitumor properties of FTIs (11, 12).
SCH66336, a novel nonpeptide tricyclic FTI, competes with the protein substrate for binding to FT (13). Farnesylation of H-ras and K-ras-4B in vitro by purified human FT is inhibited by SCH66336 with IC50 values of 1.9 and 5.2 nm,respectively. No inhibition of the related geranylgeranyl protein transferase GGT-1 occurs at SCH66336 concentrations of up to 50μ m, confirming the selectivity of this agent for FT. In tissue culture, SCH66336 blocks the growth of human neoplastic cell lines and fibroblasts expressing activated K-ras proteins. SCH66336 induces significant tumor regressions in WAP-ras transgenic animals(14). In addition, SCH66336 demonstrates potent antineoplastic activity in nude mice bearing human lung, prostate,pancreas, colon, and bladder cancer xenografts (14). On the basis of these findings, a Phase I study of cyclic oral administration of SCH66336 was undertaken in patients with advanced solid tumors. The goals of this study were as follows: (a)to determine the MTD of SCH66336; (b) to determine the dose-limiting toxicity of SCH66336; (c) to determine the recommended dose for subsequent studies; (d) to seek preliminary evidence of antitumor activity; and (e) to demonstrate FT inhibition at doses achieved in vivo.
MATERIALS AND METHODS
Patients with histological or cytological evidence of metastatic or locally advanced cancer for which there was no established curative or life-prolonging therapy were eligible for this study. Eligibility criteria also included age ≥18 years, ECOG performance status≤2 (moderate physical impairment from the effects of the disease),prior radiation to ≤30% of bone marrow completed at least 3 weeks before study enrollment, life expectancy of at least 12 weeks, and adequate bone marrow (platelets ≥100 × 109/liter, ANC > 1.5 × 109/liter, and hemoglobin ≥10 g/dl), hepatic (total bilirubin ≤1.5 × the upper limit of normal and aspartate transaminase ≤2.5 × normal), and renal(serum creatinine <1.5 × the upper limit of normal)functions.
SCH66336 was supplied by Schering-Plough Research Institute,(Kenilworth, NJ) in solid gelatin capsules containing 25, 100, or 200 mg of SCH66336. This agent was administered orally twice daily for 7 days out of a 21-day cycle. Antiemetic prophylaxis was not given.
One new patient was entered at each dose level until moderate organ toxicity (NCI CTC grade 2; Ref. 15) was observed. The schedule was then changed to three patients at each dose level. Dose escalation was not allowed in individual patients. The MTD was defined as one dose level below the dose that induced dose-limiting toxicity in more than one-third of patients (at least two of a maximum of six patients). Severe or life-threatening (NCI CTC grades 3 or 4) nonhematological toxicity (with the exception of nausea and vomiting) were considered dose limiting. NCI CTC grade 3 or 4 nausea and vomiting in patients who had received prophylactic treatment with an optimal antiemetic regimen were considered dose limiting. ANC <1.0 × 109/liter, platelet count <50 × 109/liter, or hemoglobin of 6.5 g/dl was also considered dose limiting.
Clinical Care of Patients.
Complete patient histories, physical examinations, complete blood cell counts, serum electrolytes, chemistries, urinalysis, and electrocardiograms were performed at baseline and before each course of treatment. Laboratory studies were repeated weekly while patients were on study. Ophthalmologic examination, including retinal photography,was performed at baseline and before every second cycle. Radiological studies (roentgenograms, computed axial tomographic scans, and magnetic resonance imaging) were performed at baseline and after every two courses of therapy to assess tumor response.
Standard response criteria were utilized. A PR was defined as a≥50% reduction in the sum of the products of the largest perpendicular diameters of indicator lesion(s), single or multiple,chosen before therapy. A CR was the total disappearance of all evidence of tumor. For a patient to qualify for CR or PR, none of the factors constituting progression could be present. Tumor progression was defined as the appearance of new lesion(s) or a 25% increase in size of indicator lesion(s). Stable disease was documented when there was a failure to meet the criteria for CR, PR, or progression. All objective responses were required to last for at least 4 weeks.
Monoclonal antihuman lamin A (16) was a kind gift from Frank McKeon (Harvard Medical School, Boston, MA). A high titer polyclonal serum that recognizes the COOH-terminal domain of human prelamin A was raised by immunizing rabbits with the peptide CLLGNSSPRTQSPQN coupled to keyhole limpit hemocyanin as described by Sinensky et al. (17). Generation and characterization of the monoclonal anti-PxF antibody will be described in greater detail elsewhere.4 Affinity-purified peroxidase- and fluorochrome-coupled secondary antibodies were from Kirkegaard & Perry (Gaithersburg, MD).
Replicate 100-mm dishes containing 30–40% confluent A549 human non-small cell lung cancer cells (American Type Culture Collection,Manassas, VA) grown in RPMI 1640 containing 5% (v/v) heat-inactivated fetal bovine serum, 100 units/ml penicillin G, 100 μg/ml streptomycin, and 2 mm glutamine were treated with the indicated concentrations of SCH66336 (added from 1000× concentrated stocks prepared in DMSO) or the corresponding volume of diluent. After a 24-h incubation, cells were washed three times with ice-cold RPMI 1640 supplemented with 10 mm HEPES (pH 7.4) and solubilized in buffer consisting of 6 m guanidine hydrochloride, 250 mm Tris-HCl (pH 8.5 at 21°C), 10 mm EDTA, 1%(v/v) β-mercaptoethanol, and 1 mmα-phenylmethylsulfonyl fluoride (freshly added from a 100 mm stock in anhydrous isopropanol). Samples were prepared for electrophoresis as described previously (18). Aliquots containing 50 μg of protein [assayed by the method of Smith et al. (19)] were subjected to electrophoresis on SDS-polyacrylamide gels containing 8% (w/v) acrylamide (for lamins) or 12% acrylamide (for Pxf), transferred to polyvinylidene difluoride membrane, and probed with the antibodies or sera described above. Antigen-antibody complexes were detected using peroxidase-coupled secondary antibodies and an enhanced chemiluminescence kit (Amersham Pharmacia Biotech, Piscataway, NJ).
A549 cells grown on 20-mm glass coverslips were treated with≥200 nm SCH66336 or diluent for 24 h as described above. At the completion of the incubation, coverslips were washed twice with ice-cold PBS and fixed in acetone for 15 min at −20°C. Samples were then rehydrated with two to three changes of PBS and blocked for a minimum of 1 h at 4°C in buffer A, which consisted of 10% (w/v) powdered milk in 150 mm NaCl, 10 mm Tris-HCl (pH 7.4 at 21°C), 100 units/ml penicillin G,100 μg/ml streptomycin, and 1 mm sodium azide. Cells were treated with a mixture of murine monoclonal anti-lamin A (1:3000) and rabbit anti-prelamin A (1:750) in buffer A at 4°C for 12–16 h. After removal of the primary antibodies, samples were washed six times with PBS over a 20-min period, incubated for 30 min with a mixture of affinity-purified rhodamine-conjugated antimouse IgG and fluorescein-conjugated antirabbit IgG (20 μg/ml each), washed six times with PBS over a 20-min period, mounted in Vectashield (Vector Laboratories, Burlingame, CA), sealed with clear nail polish, and examined on a Zeiss LSM 310 confocal microscope (Carl Zeiss, Inc., New York, NY). Control experiments indicated that the epitope recognized by the prelamin A serum was attenuated when samples were exposed to air for 24–48 h before fixation but was stable for at least 3 months when samples were fixed and then stored in buffer A.
Buccal smears obtained before therapy and again on day 8 (12 h after the last dose of SCH66336) were air-dried and fixed in acetone within 3 h of harvest. Samples were stored in buffer A and subjected to the immunohistochemical assay in batches. With each batch, A549 cells treated with SCH66336 or diluent were included as positive and negative controls, respectively. Sensitivity of the photomultiplier tubes on the confocal microscope was adjusted so that the signal for prelamin A in diluent-treated A549 cells was just below the limit of detection, a result consistent with the appearance of the specimens by conventional fluorescence microscopy. With the sensitivity of the confocal microscope fixed at this level, all other specimens were then examined in the conventional and laser scanning modes. When images were subsequently imported into Adobe Photoshop 3.0, any adjustments to brightness or contrast were applied identically to paired samples harvested before and after therapy with SCH66336.
Based on promising preclinical results obtained with SCH66336(14), a Phase I clinical trial was performed. Twenty patients (Table 1), 8 females and 12 males, received 92 assessable courses of therapy through seven dosage levels (Table 2). The median number of courses administered per patient was 2.5 (range,1–21 courses). The median age of study participants was 58.5 years(range, 43–72 years). Patients were minimally symptomatic from their cancer, with only two patients being significantly impaired in their daily activities. Four patients had received no prior chemotherapy, and 16 patients had received up to two prior chemotherapy regimens. Six patients had received prior radiation therapy to <30% of bone marrow. The tumor types treated on this study are shown in Table 1.
The hematological effects of SCH66336 and the number of patients experiencing various grades of toxicity are shown in Fig. 1. Brief and mild noncumulative leukopenia (CTC grade 1) was the most common hematological toxicity. This toxicity was seen in 5 of 92 treatment courses. The leukopenia was associated with neutropenia in four of the five episodes, with ANCs dropping to ∼60% of baseline values during the first cycle. Brief and mild thrombocytopenia was the next most common hematological toxicity, occurring in 4 of 92 treatment courses (one grade 2 and three grade 1). Anemia was not observed in this study. The incidence and severity of hematological toxicity did not appear to be related to SCH66336 dose.
Diarrhea was the principal nonhematological toxicity of SCH66336. It was severe and dose limiting at 400 mg b.i.d. Secretory diarrhea, which occurred without steatorrhea, tenesmus, or bloody stools, typically started on day 3 of treatment. Abdominal examination during the diarrhea episodes was benign. Diarrhea promptly resolved with loperamide therapy and/or drug discontinuation. Of the 52 episodes of diarrhea, 38 were grade 1, 11 were grade 2, and 3 were grade 3.
In addition to diarrhea, nausea and vomiting were observed. Nausea was reported by 11 of the 20 patients, and vomiting was reported by 6 patients. Nausea and vomiting were sporadic and mild at doses of SCH66336 up to 350 mg b.i.d. Above this dose, nausea and vomiting were frequent and severe. In all instances, nausea and vomiting resolved quickly and completely with antiemetic therapy.
Fatigue precluded dose escalation above 350 mg b.i.d. Both patients treated at 400 mg b.i.d. experienced severe fatigue, necessitating bed rest for most of the day; neither patient could complete the 7-day course of treatment. The fatigue, which was constitutional in nature,led to a two-grade decline of performance status on the ECOG scale. This fatigue was worsened by dehydration from GI toxic effects but slowly resolved after discontinuation of SCH66336. Twenty-five episodes of less severe fatigue (23 grade 1 and 2 grade 2 episodes) were reported at lower doses.
Moderate reversible renal insufficiency (serum creatinine = 2–4 mg/dl, representing a 2–4-fold elevation above baseline)occurred in both patients receiving 400 mg b.i.d. In both instances,serum creatinine normalized in 24 h with hydration, suggesting that the renal effects reflected dehydration-associated prerenal azotemia secondary to nausea, vomiting, and diarrhea. One patient at the 400 mg b.i.d. dose level (subsequently reduced to 200 mg b.i.d.)developed hypokalemia with each treatment and required potassium supplementation. No other possible etiological factors for the hypokalemia could be identified.
Because the visual proteins rhodopsin kinase and transducin are farnesylated, ophthalmologic examinations were performed at regular intervals. No evidence of ocular toxicity was found.
Based on the results of this study, the MTD and recommended dose of SCH66336 for subsequent clinical testing on this schedule is 350 mg orally twice a day.
One patient with metastatic NSCLC who had previously received combined modality treatment with radiation and chemotherapy achieved a PR after two courses of treatment. This patient remained on study for 14 months. Eight patients had stable disease for 5–10 treatment cycles, and 10 patients progressed on therapy after 1–3 treatment cycles.
FT Inhibition: Development of an Assay Suitable for Clinical Material.
As indicated in the “Introduction,” a number of intracellular polypeptides are farnesylated. In principle, any one of these could serve as an indicator for FT inhibition. Our studies focused on PxF(Hk33), a peroxisomal protein of unknown function that is widely expressed (20), and lamin A, a major structural protein of the nuclear envelope. Processing of both of these polypeptides involves farnesylation followed by further posttranslational modifications(20, 21, 22, 23).
When A549 lung cancer cells [which contain mutated K-ras(24)] were treated with increasing concentrations of SCH66336, evidence for inhibited farnesylation of both polypeptides was observed. Twenty-four h after the addition of 400 nmSCH66336 to the cells, levels of the Mr∼33,000 PxF polypeptide (Fig. 2,A,top) decreased by ∼50%, accompanied by appearance of a new Mr ∼35,000 species (see Lanes 1and 2) that has previously been reported to represent unprocessed PxF. This species was observed at SCH66336 concentrations as low as 12.5 nm (Lane 7). Likewise, 24 h after the addition of 400 nm SCH66336, a slower migrating species of lamin A (Fig. 2,A, middle) was detected(see Lanes 1 and 2). This species, which was the sole antigen that reacted with antiserum raised against the lamin A prepeptide (Fig. 2,A, bottom and B),was detected after exposure of A549 cells to SCH66336 concentrations as low as 6 nm (bottom, Lane 8). In additional studies, this serum was used to develop a histochemistry-based assay for prelamin A. As indicated in Fig. 3 (right), prelamin A was detected at the periphery of the nucleus in SCH66336-treated cells (middle) but not in control cells (top). Binding of the antibody was inhibited by the immunizing peptide (lower panels). Binding of an antibody that recognizes mature lamin A polypeptide was unaffected by these experimental manipulations (Fig. 3, left). These observations form the basis for the histochemical assay used to detect prelamin A in clinical samples.
Inhibition of FT in a Dose-dependent Manner in Patients Receiving SCH66336.
To determine whether FT was inhibited in patients receiving SCH66336,the preceding assays were applied to samples of normal human tissues. Peripheral blood mononuclear cells were subjected to immunoblotting with the anti-PxF antibody. When samples harvested on day 8 of therapy were compared with samples harvested before therapy, alterations in the mobility of PxF were not detectable (Fig. 4). The lack of unprocessed PxF in these specimens might reflect slow cell turnover with a concomitant low synthesis rate of PxF or altered prenylation and processing of PxF in these cells (Pxf, like K- and N-ras, can be alternatively prenylated by geranylgeranyl protein transferase in FTI-treated cells).5
In additional studies, the presence of prelamin A was assessed in buccal mucosa cells using the histochemical assay described above. Four considerations led to the examination of this cell type: (a)the presence of lamin A in these cells but not in circulating lymphocytes(25);6(b) the relatively rapid turnover of this cell population;(c) the ready accessibility of this tissue source; and(d) the absolute dependence of prelamin A proteolytic processing on farnesylation (21, 23). Before treatment,none of the buccal smears contained detectable prelamin A. After treatment, however, prelamin A was readily detectable in a subset of the specimens (e.g., patients B, C, and D in Fig. 5). Examination of the relationship between dose and inhibition of lamin processing revealed that prelamin A was detected in 60% of buccal mucosa samples from patients treated with 200 mg of SCH66336, in 67% of patients treated with 300 mg of SCH66336, in 75% of patients treated with 350 mg of SCH66336, and in 100% of patients treated with 400 mg of SCH66336, indicating the possibility of a dose-dependent trend. These results provide evidence that FT is inhibited at clinically achievable doses and also suggest that the degree of FT inhibition at any particular dose level might vary from patient to patient. Whereas the sample sizes are too small to be tested statistically with any degree of accuracy, these results are worthy of further study in a larger group of patients.
To our knowledge, the present study is the first full-length report of a human trial of a FTI. The historical development and structural diversity of FTIs have been reviewed recently(26). These agents represent the first class of inhibitors to be introduced into clinical trials based on their effects on proteins that are genetically altered in cancer cells. Because of their unique mechanisms of action, at least four FTIs are in clinical trials in the United States and Europe, with several more at different levels of preclinical development. Several of these agents competitively block the CAAX binding site on FT. L731735 and L744832 were designed to mimic the action of the tetrapeptide CVFM, the earliest identified inhibitor of FT. SCH66336 and R115777, two other FTIs in clinical trials, were identified by random screening and then determined to be CAAX mimetics. In contrast, other FTIs under development, including PD 169451 and RPR 130401, compete for the isoprenoid binding site of FT (27, 28, 29, 30, 31).
Because of the large number of farnesylated mammalian proteins, FTIs were expected to have serious toxicities. The present Phase I study,however, demonstrates that SCH66336 is well tolerated, with reversible and manageable GI toxicity. Rigorous physical examination and laboratory tests did not identify any other significant toxicities. Several factors might account, in part, for the relative lack of toxicity of SCH66336. First, SCH66336 is highly selective for FT as compared with GGT-1. As a result, several farnesylated proteins might be geranylgeranylated if farnesylation is inhibited, providing an alternative pathway for generating active forms of the proteins. Second, it is possible that survival pathways in normal cells might be less dependent on ras-containing signal transduction pathways than those in ras-transformed cells. Additional studies in preclinical models and clinical specimens are required to determine the basis for the selectivity of FTIs.
Despite the growing literature on FTIs, there is only limited evidence that FT is inhibited in animals at therapeutic doses. Accordingly, we set out to determine whether clinically achievable doses of SCH66336 inhibited FT in patients. Mammalian cells contain at least 20 FT substrates, several of which require prenylation for further processing(1, 2, 3, 4). These include prelamin A, which is farnesylated before proteolytic removal of the COOH-terminal peptide (Fig. 2 B; Refs. 21, 23, and 32), and the ras proteins. Recent work has shown that N-ras and K-ras can undergo geranylgeranylation when farnesylation is inhibited by SCH66336 (9). This alternative prenylation makes these ras isoforms unsuitable as biochemical markers for FT inhibition. Although H-ras is not alternatively prenylated, this isoform is difficult to detect in normal tissue such as peripheral blood mononuclear cells. Moreover, as indicated in the “Introduction,” there is evidence that antiproliferative effects of FTIs might result from effects on farnesylated substrates other than ras.
In this study, we examined processing of prelamin A in buccal mucosa cells as a potential marker of in vivo activity of SCH66336. Removal of the COOH-terminal peptide occurs only if prelamin A is farnesylated (21, 23, 32). We observed a dose-dependent increase in the frequency of unprocessed prelamin A when accessible lamin A-expressing tissue was examined. These data provide the first evidence of successful FT inhibition in humans. Because prelamin A is more resistant than ras to the effects of FT inhibition(33), the ability of clinically achievable levels of SCH66336 to inhibit prelamin A farnesylation suggests that the modification of several FT substrates has very likely been inhibited in vivo.
A >50% shrinkage of a metastatic NSCLC lesion in the adrenal gland was noted in a patient previously treated with radiation and chemotherapy. This patient received 21 treatment cycles (14 months). Seven additional patients with refractory tumors had stable disease for 5–10 treatment cycles. Thus, the present Phase I study in a pretreated population provides a hint that this class of agents might have anti-neoplastic activity in humans. Phase II studies of this agent are currently in progress.
In summary, the present study establishes that the FTI SCH66336 has a toxicity profile different from that of most conventional anticancer agents and identifies a suitable dose (350 mg orally twice a day) for subsequent clinical trials on this 7-day schedule. In addition, this study provides the first demonstration that protein farnesylation can be safely inhibited in vivo in humans and the first evidence of clinical activity of this class of agents. These findings not only encourage the future clinical development of FTIs but also establish a paradigm for the study of other small molecule inhibitors of signal transduction.
We gratefully thank Dr. Michael Sinensky for an aliquot of anti-prelamin A and advice on the prelamin A assay, Dr. Robert Patton for an aliquot of anti-prelamin A raised by the Schering Plough Research Institute, Kim Jensen for data management, Michelle Daiss for protocol development, and Gail Prechel and Deb Strauss for secretarial assistance.
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.
Supported in part by NIH Grants CA77112 and RR00585.
The abbreviations used are: FT, farnesyl protein transferase; ANC, absolute neutrophil count; CR, complete response;CTC, Common Toxicity Criteria; FTI, farnesyl transferase inhibitor;MTD, maximum tolerated dose; PR, partial response; GI,gastrointestinal; NCI, National Cancer Institute; ECOG, Eastern Cooperative Oncology Group; b.i.d., twice a day; NSCLC, non-small cell lung cancer.
P. Kirschmeier et al., manuscript in preparation.
P. Kirschmeier and W. Robert Bishop, unpublished observations.
P. A. Svingen and S. H. Kaufmann,unpublished observations.
|No. of courses (fully evaluable)||92 (92)|
|Median no. of courses/patient (range)||2.5 (1–21)|
|Median age, years (range)||58.5 (43–81)|
|ECOG performance status|
|Prior chemotherapy regimens|
|No. of courses (fully evaluable)||92 (92)|
|Median no. of courses/patient (range)||2.5 (1–21)|
|Median age, years (range)||58.5 (43–81)|
|ECOG performance status|
|Prior chemotherapy regimens|
|Level .||SCH66336 (mg) .||No. of patients .||No. of courses .|
|Level .||SCH66336 (mg) .||No. of patients .||No. of courses .|