Investigational New Drugs 10: 17-22, 1992. 9 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Activity of a recombinant transforming growth factor-a-Pseudomonas exotoxin hybrid protein against primary human tumor colony-forming units Daniel D. Von H o f f 1, Martha H. Marshall 1, David C. Heimbrook 2, Steven M. Stirdivant 2, Janet D. Ahern 2, Wayne K. Herbert 2, Robert Z. Maigetter 2 and Allen Oliff 2 1Cancer Therapy and Research Center of South Texas, 4450 Medical Drive, San Antonio, TX 78229, USA; 2Merck Sharp & Dohme Research Laboratories, West Point, PA 19486, USA

Key words: transforming growth factor-a, cloning assay, colony-forming units, epidermal growth factor receptor

Summary

Transforming growth factor-a-Pseudomonas exotoxin-40 (TP40) is a recombinant fusion protein. TP40 consists of the entire human transforming growth factor-a (TGF~) protein fused to a 40,000 Da. segment of the Pseudomonas exotoxin A protein. TP40 is a bifunctional molecule that possesses the epidermal growth factor (EGF) receptor binding properties o f T G F a and the cell killing properties of Pseudomonas exotoxin A. These properties make TP40 a selective cytotoxic agent that kills EGF receptor bearing cells. TP40 has been shown to effectively kill human tumor cell lines that possess EGF receptors in vitro and in nude mice. In the present study, TP40 was tested against tumors taken directly from patients and grown in a soft agar human tumor cloning system. A total of 107 patients' tumors (taken from patients with tumors refractory to chemotherapy) were tested with a continuous exposure to 0 . 5 - 5 0 nM concentrations of the agent. TP40 exhibited a clear dose response effect against a wide variety of human solid tumor colony-forming units with _> 84% of evaluable tumors responding at a drug concentration > 24 nM. When used as a continuous exposure, concentrations of TP40 as low as 5 nM demonstrated substantial in vitro activity. This activity included cytotoxicity against breast, colorectal, endometrial, head and neck, non small-cell lung, gastric, sarcoma, and pancreatic cancer tumor colony-forming units. Additional in vivo testing of this compound is warranted.

Abbreviations: TP40 - recombinant transforming growth factor-a-PE40 toxin hybrid protein; H T C A - human tumor cloning assay; T F G - a - transforming growth factor-a; E G F - epidermal growth factor; PBS - phosphate buffered saline

Introduction

Transforming growth factor alpha (TGFc0 is a member of the epidermal growth factor (EGF) family of polypeptide hormones that stimulate the growth of several types of mammalian cells [1-4]. The cell surface receptor for EGF binds both EGF and T G F a with almost equal affinities [5]. Some

human tumors such as glioblastomas and squamous cell carcinomas have been found to overexpress the EGF receptor [6-8]. Other cancer cells may employ T G F a as an autocrine growth factor to stimulate their own growth [9,10]. We have found that TGFc~ can stimulate colony formation from cells obtained from primary human tumors [11]. This enhancement in tumor colony formation

18 presumably occurs through an EGF-receptorT G F a interaction. We have also found that T G F a activity in malignant effusions is a reliable marker of tumor burden and patient prognosis [12,13]. Since many human tumors contain EGF receptors which bind TGFc~, it may be possible to exploit the presence of these receptors on tumor cells for therapeutic purposes. To this end, Pastan and his collegues engineered a hybrid fusion protein consisting of TGF~ and a 40,000 Da. segment of the Pseudomonas exotoxin A protein [14,15]. We have also constructed a modified TGFa-Pseudomonas exotoxin A fusion protein (TP40) [16]. TP40 is a bifunctional molecule that possesses both "celltargeting" and cell-killing activities. TGFo~ functions as the "cell-targeting" domain of TP40 by binding specifically to the surface of cells that possess EGF receptors. The segment o f Pseudomonas exotoxin A functions as the cell-killing domain. Pseudomonas exotoxin A kills mammalian cells by ADP-ribosylating elongation factor 2 which results in the inhibition of protein synthesis [17]. This combination of "cell-targeting" and cell-killing properties makes TP40 potentially useful as an anticancer agent capable of killing EGF receptor bearing tumor cells. TP40 was recently shown to kill a variety of human cancer cell lines in vitro [15,16], and to prolong the survival of nude mice bearing human tumor cell line xenografts [18]. However, until now no information was available concerning the sensitivity and histologic spectrum of primary human tumor cells that are susceptible of TP40. The human tumor cloning assay (HTCA) is an in vitro soft agar technique developed to grow colonyforming units from primary human tumor specimens [19,20]. The assay has been used to attempt to predict whether or not a patient's tumor would respond to a particular antineoplastic agent and to screen for new antineoplastic agents [21-24]. Recent improvements in the H T C A have made predictive assays and drug screening assays more reproducible, reliable, and accurate [25-28]. In the present study we have utilized the H T C A to test the in vitro activity of TP40 against primary human tumors growing in soft agar. As will be seen below, TP40 has remarkably good activity in vitro against a wide spectrum of human tumors (e.g.,

colon, non small-cell lung, endometrial carcinoma, ovarian cancer, and others). These impressive in vitro results suggest that TP40 merits a clinical evaluation to assess its utility in cancer patients.

Material and methods

Collection of specimens Utilizing routine clinical methods, malignant pleural fluid or ascites were collected in sterile vacuum bottles or plastic bags (without preservatives) containing 10 units of preservative-free heparin for each ml of malignant fluid collected. Under a pathologists supervision, solid tumor specimens were minced into 2 m m - 5 mm fragments in the operating room and were placed immediately in a 50 cc plastic centrifuge tube containing McCoy's 5A medium plus 10% newborn calf serum, 1% penicillin and streptomycin, 15 mM H E P E S buffer and 2 mM pyruvate (all from GIBCO Laboratories, Grand Island, NY). Upon arrival at the cloning laboratory, single cell suspensions were prepared from both the solid tumors and effusions using mechanical or mechanical and enzymatic methods as previously described [24,29]. After single-cell suspension preparation, the cells were washed in McCoy's plus 10% heatinactivated fetal calf serum (GIBCO).

In vitro exposure of tumor cells to TP40 and other antineoplastic agents Both a 1 hr and a continuous exposure to TP40 were explored. TP40 was kept frozen in 100/zl aliquots until it was ready for use. The vials were thawed (but kept on ice) and diluted with phosphate buffered saline (PBS). No thawed vials were kept (on ice) for more than 24 h. Concentrations of TP40 tested included 0.5, 5.0, and 50 nM as a 1 h exposure and 5, 25, 50, and 400 nM as a continuous exposure. For purposes of comparison, a number of standard agents were also tested against the same patients' tumors. Stock solutions of I.V. formulations

19 of the standard agents were prepared in sterile buffered 0.9~ NaC1 solution, water, or dimethyI sulfoxide and stored at - 7 0 ~ in aliquots sufficient for individual assays. Subsequent dilutions for incubation were made using a 0.9% NaC1 and water solution. The tumor cells were exposed to standard anticancer agents at the following concentrations (mcg/ml): actinomycin D, 0.01; bis(chlorethyl) nitrosourea, 3.0; doxorubicin, 0.04, etoposide, 3.0; 5-fluorouracil, 6.00; methotrexate, 0.30; melphalan, 0.30; mitomycin C, 0.10; cytosine arabinoside, 100; streptozotocin, 0.5; vinblastine, 0.05; and vincristine, 0.01. H u m a n leukocyte interferon was used at concentrations of 100 units per ml. The concentration o f each drug corresponded to approximately one-tenth of the peak plasma concentration of each drug in humans [30]. T u m o r cell suspensions were transferred to tubes and adjusted to a final concentration of 5 • 105 cells/ml in the presence of the appropriate TP40 or drug dilution of control medium. Cells were incubated with and without the TP40 or drug for 1 h at 37 ~ in McCoy's plus 10% heat-inactivated fetal calf serum. These were then centrifuged at 150 • g for 10 minutes, washed with McCoy's solution and placed in the assay system described below. For the continuous exposure study with the TP40 the cells were not washed and the TP40 was incorporated with the celts into the top layer of agar.

Assay for tumor colony-forming units The culture system used in this study has been extensively described elsewhere [24,29,31,32]. Colonies (_> 50 cells) usually appeared by day 14 of culture, and the number of colonies on the plates was determined by counting the colonies on an inverted stage microscope at 30 • magnification. To be considered evaluable for colony growth, the plates had to have excellent single cell suspensions (which are defined as a _< 30% survival of colonies on the chromomycin A3 positive control plates) as well as _> 30 colonies on control plates.

Positive control To assure the presence of an excellent single cell suspension for each specimen, a positive control consisting of chromomycin A3 (Sigma Chemical Co.) at a concentration of 100 mcg/ml was utilized. The chromomycin does not allow growth of human tumor colony-forming units and assures the investigator that a single cell suspension is present on the day of culture. In order for an experiment to be considered evaluable, the chromomycin had to produce a < 30% survival of colony-forming units. The use of a positive control has been shown to greatly increase the reproducibility of the human tumor cloning system [25].

Data analysis and statistical consideration The results of the in vitro cloning assay were expressed as percentage of survival of tumor colonyforming units for a particular drug relative to its control. This quantity was calculated as the ratio between the mean number of colonies surviving in the 3 drug-treated plates versus the number of colonies growing in three control plates. The definition of an in vitro response was a < 50~ survival of tumor colony-forming units. This definition was based on prior retrospective and prospective trials of the human tumor cloning system [23,24]. Overall, with the definition of a response in vitro the percent true positives for the assay (e.g., the in vitro assay predicted the patient would respond and they did) was 68% while the percent true negatives for the assay (the assay predicted the patient would not respond and they did not) was 96% [28,33,34]. Recombinant TP40 was produced in bacteria and purified to homogeneity as previously described [16,18]. The purity and protein concentration of TP40 was assessed by gel electrophoresis with silver staining, amino acid composition analysis, and Nterminal amino acid sequence determinations.

Results

A total of 29 patients' tumors were placed in the soft agar cloning system with a 1-h exposure to

20 Table 2. Tumor specific activity of TP40 (continuous exposure)

Table 1. Summary of in vitro activity of TP40

Concentration (nM)

Exposure time

# Responses*/ # evaluable (%)

0.5 5.0 50.0 5.0 25.0 50.0 400.0

1 hour i hour 1 hour Continuous Continuous Continuous Continuous

0/15 1/15 2/15 17/31 21/26 27/31 2/2

(0%) (7%) (13%) (55%) (81%) (87%) (100%)

* Response is defined as _< 50% survival of tumor colonyforming units.

T P 4 0 a n d 78 p a t i e n t s ' t u m o r s were p l a c e d in the system with a c o n t i n u o u s e x p o s u r e to TP40. A l l specimens were f r o m p a t i e n t s whose t u m o r s were r e f r a c t o r y to at least one p r i o r c h e m o t h e r a p e u t i c regimen. F i f t e e n o f the 29 1-hr e x p o s u r e specimens a n d 31 o f t h e 78 c o n t i n u o u s e x p o s u r e specimens were e v a l u a b l e for d e t e r m i n a t i o n o f sensitivity to T P 4 0 (e.g., sufficient c o l o n y f o r m a t i o n a n d an acc e p t a b l e positive c o n t r o l result). T a b l e 1 s u m m a r i z e s the n u m b e r o f in vitro responses ( d e f i n e d as ___ 50% survival o f t u m o r c o l o n y - f o r m i n g units) n o t e d for each c o n c e n t r a t i o n a n d length o f exposure. As can be seen in t h a t table the c o m p o u n d does exhibit a d o s e - r e s p o n s e effect. In a d d i t i o n , the activity o f the c o m p o u n d a p p e a r s s c h e d u l e - d e p e n d e n t as the c o m p o u n d has m o r e in vitro c y t o t o x i c i t y when used as a c o n t i n u o u s exp o s u r e t h a n when used as a 1-h e x p o s u r e . T a b l e 2 details the t u m o r specific activity o f T P 4 0 (when used as a c o n t i n u o u s exposure). T h e c o m p o u n d a p p e a r s to have a b r o a d s p e c t r u m o f in vitro c y t o t o x i c i t y against b r e a s t , c o l o n , e n d o m e t r i al, h e a d a n d neck, n o n small-cell lung, o v a r i a n , pancreas, sarcoma, and stomach tumors. I n a d d i t i o n to testing for the in vitro activity o f TP40, o t h e r c o n v e n t i o n a l a n t i n e o p l a s t i c agents were also tested against the same p a t i e n t ' s t u m o r s . T a b l e 3 details in vitro activity o f t h e c o n v e n t i o n a l c y t o t o x i c agents against the same p a t i e n t ' s t u m o r s . O f n o t e is t h a t these t u m o r s a p p e a r quite resistant to the s t a n d a r d agents (in p a r t no d o u b t due to the p r i o r in vivo t r e a t m e n t status o f these p a t i e n t ' s malignancies). The s u b s t a n t i a l in vitro activity o f

# of responses/# of evaluable Tumor type

5

25

50 nM

Breast Colon Endometrial Head and Neck* Lung - non-small cell Ovary Pancreas Sarcoma Stomach Thyroid

3/4 2/6 3/3 1/3 2/5 3/5 1/ 1 1/1 0.2 1/1

3/3 4/4 2/2 1/4 3/4 4/5 1/1 2/2 1/1

4/4 5/6 3/3 2/3 4/5 4/5 1/ 1

1/1 2/2 1/1

* squamous histology Table 3. In vitro response rate for other conventional agents

Agent

TP40 5 Fluorouracil Vinblastine Doxorubicin Mitomycin C Cis-platinum Melphalan 4 hydroperoxycyclophosphamide Bleomycin VP16 Methotrexate BCNU Vincristine Cytosine arabinoside Actinomycin D Interferon alpha Streptozotocin

Concentrations (#g/ml)* 5.0 6.0 0.05 0.04 0.10 0.20 0.10 3.0 0.2 3.0 0.3 0.I 0.01 100.0 0.1 100.0 0.5

# responses**/# attempted (%) 14/26 2/18 1/14 0/14 0/14 0/11 1/10

(54%) (11%) (7~ (0%) (0%) (0%) (10%)

1/9 0/6 1/4 0/6 0/4 0/2 0/2 0/1 0/1 0/1

(11%) (0%) (25O7o) (0%) (0%) (0%) (0%) (0o70) (0%) (0%)

* Except for TP40 which is nM. ** Response is defined as < 50% survival of tumor colonyforming units.

the T P 4 0 versus t h a t o f t h e s t a n d a r d agents is o f g r e a t interest.

Discussion

This is the first r e p o r t o f the in vitro activity o f a genetically engineered h y b r i d T F G a - b a c t e r i a l t o x i n

21 fusion protein against human tumor colonyforming units. The compound demonstrated substantial broad spectrum in vitro activity particularly when compared to the in vitro activity of conventional cytotoxic agents used at clinically achievable concentrations against the same tumors (Table 3). The activity of TP40 was maximal when the compound was used as a continuous exposure. This result is consistent with earlier cell culture studies that showed TP40 was ten fold more cytotoxic when used to treat human tumor cells for three hours versus one hour. Unfortunately, it is uncertain what concentration or exposure time to TP40 can be achieved in patients. The current study also clearly demonstrated a dose-response effect. Fifty-five percent of the evaluable tumor specimens responded to TP40 at 5 nM while > 81% of evaluable specimens responded at drug concentrations > 25 nM (Table 1). These results indicate TP40 could best be used clinically in a regimen which could achieve a high concentration x time profile (perhaps, via an intracavitary route). TP40 theoretically works by binding of the agent to the EGF receptor with subsequent internalization of the ligand-toxin conjugate. TP40 induced killing of A431 cells in vitro has been shown to proceed via attachment of TP40 to unoccupied EGF receptors [16]. TP40 is only eytotoxic against cells expressing EGF receptors [15,16]. Similarly, TP40 treatment of A431 cell tumors in animals has been shown to require EGF receptor binding activity for suppression of tumor growth [18]. Therefore, the in vitro activity of TP40 noted in this study also was presumably mediated by EGF receptors on the tumor cells. Unfortunately, the E G F receptor status was not determined for the patients' tumors examined in this study. No tumor cells are left from these specimens. A future study correlating the in vitro response of human tumors and their EGF receptor status would be of great interest for this compound. There may be a minimum receptors concentration required on a patient's tumor cells versus on normal cells to achieve the necessary selectivity with TP40. Based on the present in vitro responses noted against a wide variety of patients' tumors and

TP40's reported activity in nude mouse xenograft studies [18], this compound should be advanced to clinical trial. Concerns in clinical trials will include selectivity o f TP40 as well as the host's immune response to the conjugate. It is well known that normal human tissues such as liver contain EGF receptors. A major concern in clinical trials is whether the TP40 will cause damage to these normal organs. However, the work by Heimbrook and colleagues [18] in nude mouse xenograft studies indicates there may be an acceptable therapeutic index for the compound. The tumors in those mice responded without the mice suffering adverse effects.

Acknowledgement The authors wish to thank Drs. Ira Pastan and David FitzGerald for helpful discussions, and Ms. Rose Alvizo and Julia Perkins for their expert assistance in the preparation of this manuscript.

References 1. Todaro G J, Fryling C, De Larco JE: Transforming growth factors produced by certain human tumor cells: Polypeptides that interact with epidermal growth factor receptors. Proc Natl Acad Sci USA 77:5258-5262, 1980 2. Marquardt H, Hunkapiller MW, Hood LE, Twardzik DR, De Larco JE, Stephenson JR, Todaro G J: Transforming growth factors produced by retrovirus-transformed rodent fibroblasts and human melanoma cells: Amino acid sequence with epidermal growth factor. Proc Natl Acad Sci USA 80:4684-4688, 1983 3. Roberts AB, Frolik CA, Anzano MA, Sporn MB: Transforming growth factors from neoplastic and non-neoplastic tissues. Fed Proc 42:2621-2626, 1983 4. Roberts AB, Sporn MB: Transforming growth factors. Cancer Sruv 4:683-705, 1985 5. Marquardt H, Hunkapiller MW, Hood LE, Todaro G J: Rat transforming growth factor type 1: Structure and relation to epidermal growth factor. Science 223:1079-1082, 1984 6. Ro J, North SM, Gallick GE, Hortobagyi GN, Gutterman JU, Blick M: Amplified and overexpressed epidermal growth factor receptor gene in uncultured primary human breast carcinoma. Cancer Res 48:161-164, 1988 7. Liberman TA, Rason N, Bartal AD, Yarden Y, Schlessing J, Soreq H: Expression of epidermal growth factor receptors in human brain tumors. Cancer Res 44:753-760, 1984 8. Thompson DM, Gill GN: The EGF receptor: Structure, regulation and potential role in malignancy. Cancer Surv

22 4:767-788, 1985 9. Sporn MB, Todaro G J: Autocrine secretion and malignant transformation of cells. N Engl J Med 303:878-880, 1980 10. Derynck R, Goeddel DV, Ullrich A, Gutterman JU, Williams RD, Bringman TS, Berger SH: Synthesis of messenger RNA species for transforming growth factors alpha and beta and the epidermal growth factor receptor by human tumors. Cancer Res 47:707-712, 1987 ll. Hanauske AR, Hanauske U, Buchok J, Von Hoff DD: Recombinant human transforming growth factor-c~ stimulates in vitro colony formation of fresh human tumor specimens. Int J Cell Cloning 6:221-229, 1988 12. Hanauske AR, Arteaga CL, Clark GM, Buchok J, Marshall M, Hazarike P, Pardue RL, Von Hoff DD: Determination of transforming growth factor activity in effusions from cancer patients. Cancer 61:1832-1837, 1988 13. Arteaga CL, Hanauske AR, Clark GM, Buchok J, Marshall M, Hazarika P, Pardue RL, Von Hoff DD: Immunoreactive of transforming growth factor activity in effusions from cancer patients as a marker of tumor burden and patient prognosis. Cancer 48:5023-5028, 1988 14. Chaudhary VK, FitzGerald D J, Adhya S, Pastan I: Activity of a recombinant fusion protein between transforming growth factor type alpha and Pseudomonas toxin. Proc Natl Acad Sci USA 84:4538-4542, 1987 15. Siegall CB, Xu Y-H, Chaudhary VK, Adhya S, FitzGerald D, Pastan I: Cytotoxic activities of a fusion protein comprised of TGF~ and Pseudomonas exotoxin. FASEB 3:2647-2652, 1989 16. Edwards GM, Defeo-Jones D, Tai JY, Vuocolo GA, Patrick DR, Heimbrook DC, Oliff A: Epidermal growth factor receptor binding is affected by structural determinants in the toxin domain of transforming growth factor-alphaPseudomonas exotoxin fusion proteins. Mol Cell Biol 9:2860-2867, 1989 17. Iglewski BH, Kabat D: NAD-dependent inhibition of protein synthesis by Pseudomonas aeruginosa toxin. Proc Natl Acad Sci USA 72:2284-2288, 1975 18. Heimbrook DC, Stirdivant SM, Ahern JD, Bolishen NL, Patrick DR, Edwards GM, Jones DD, FitzGerald DJ, Pastan I, Oliff A: Transforming growth factor-aPseudomonas exotoxin fusion protein prolongs survival of nude mice bearing tumor xenografts. Proc Natl Acad Sci USA 87:4697-4701, 1990 19. Hamburger AW, Salmon SE: Primary bioassay of human tumor stem cells. Science 197:461-463, 1977 20. Hamburger A, Salmon SE: Primary bioassay of human myeloma stem cells. J Clin Invest 60:846-854, 1977 21. Salmon SE, Hamburger AW, Soehnlen B, Durie BG, A1berts DS, Moon TE: Quantitation of differential sensitivity of human tumor stem ceils to anticancer drugs. N Engl J

Med 298:1321-1327, 1978 22. Salmon SE, Alberts DS, Durie BGM, Meyskens FL, Jones SE, Soehnlen B, Chen H-SG, Moon T: Clinical correlations of drug sensitivity in the human tumor stem cell assay. Recent Results Cancer Res 74:300-305, 1980 23. Von Hoff DD, Casper J, Bradley E, Sundbach J, Jones DG, Makuch R: Association between human tumor colonyforming assay results and response of an individual patient's tumor to chemotherapy. Am J Med 70:1027-1041, 1981 24. Von Hoff DD, Clark GM, Stogdill B J, Sorosdy MF, O'Brien MT, Casper JT, Mottox DE, Page CP, Cruz AB, Sandbach JF: Prospective clinical trial of a human tumor cloning system. Cancer Res 43:1926-1931, 1983 25. Shoemaker RH, Wolpert-DeFilippes MK, Kern DH, Lieber MM, Makuch RW, Melnick NR, Miller WT, Salmon SE, Simon RM, Venditte JM, Von Hoff DD: Application of a human tumor colony-forming assay to new drug screening. Cancer Res 45:2145-2153, 1985 26. Von Hoff DD, Foreseth BJ, Huong M, Buchok JB, Lathan B: Improved plating efficiencies for human tumors cloned in capillary tubes versus petri dishes. Cancer Res 46:40124017, 1986 27. Von Hoff DD, Sandbach JF, Clark GM, Turner J, Forseth B, Piccart M J, Colombo N, Muggia FM: Selection of cancer chemotherapy for a patient by an in vitro assay versus a clinician. J Natl Cancer Inst 82:110-116, 1990 28. Von Hoff DD: He's not going to talk about in vitro predictive assays again, is he? J Natl Cancer Inst 82:96-101, 1990 29. Hamburger AW, Salmon SE, Kim MB, Trent J, Soehnlen BJ, Alberts DS, Schmidt H J: Direct cloning of human ovarian carcinoma cells in agar. Cancer Res 38:3438-3444, 1978 30. Alberts DS, Chen HSB: Tabular summary of pharmacokinetic parameters relevant to in vitro drug assay. In: J.S. Salmon (ed.), Human Tumor Stem Cells, Alan R. Liss, Inc., New York, 1980, pp 351-359 31. Alberts DS, Salmon SE, Chen HS, Surwit EA, Soehnlen B, Young L, Moon TE: In vitro clonogenic assay for predicting response of ovarian cancer to chemotherapy. Lancet 2:340-342, 1980 32. Pike BL, Robinson WA: Human bone marrow colony growth in agar-gel. J Cell Physiol 76:77-84, 1970 33. Von Hoff DD: In vitro predictive testing: The sulfonamide era. Int J Cell Cloning 5:179-190, 1987 34. Von Hoff DD: Human tumor cloning assays: Applications in clinical oncology and new antineoplastic agent development. Cancer and Metastasis Review 7:357-371, 1988

Address for offprints: D.D. Von Hoff, Cancer Therapy and Research Center of South Texas, 4450 Medical Drive, San Antonio, TX 78229, USA