BIOPHARMACEUTICALS

BioDrugs 1998 Apr; 9 (4): 325-335 1173-8804/98/0004-0325/$05.50/0 © Adis International Limited. All rights reserved.

Matrix Metalloproteinase Inhibitors A Review Susan A. Watson and Gill Tierney Cancer Studies Unit, Department of Surgery, Queens Medical Centre, University of Nottingham, Nottingham, England

Contents Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. What Are Matrix Metalloproteinase Inhibitors (MMPIs)? . . . . 1.1 Naturally Occurring versus Synthetic MMPIs . . . . . . . . 1.2 How Do They Function? . . . . . . . . . . . . . . . . . . . 1.3 Synthetic MMPIs . . . . . . . . . . . . . . . . . . . . . . . . 2. Preclinical Evaluation . . . . . . . . . . . . . . . . . . . . . . . 2.1 Hydroxamate-based MMPIs . . . . . . . . . . . . . . . . . 2.2 TIMP-Based Inhibition . . . . . . . . . . . . . . . . . . . . . 2.3 Additional Classes of MMPIs . . . . . . . . . . . . . . . . . 2.4 Combination Therapy with MMPIs and Cytotoxic Agents 3. Design of Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . 4. Results of Preliminary Clinical Trials . . . . . . . . . . . . . . . . 5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Summary

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

325 326 326 326 327 327 327 330 331 332 332 333 333

The matrix metalloproteinases (MMPs) are a family of closely related, zincdependent proteolytic enzymes. Collectively, they are capable of degrading all the components of the extracellular matrix and as such are involved in a number of physiological and pathological processes. The extracellular matrix is the principal barrier to tumour growth and spread, and there is evidence that MMPs play a role in the processes of tumour growth and metastasis. Therefore, inhibitors of MMPs may be of value in the treatment of malignant disease. There exist naturally occurring inhibitors of these enzymes known as ‘tissue inhibitors of MMPs’, or TIMPs. Although there have been considerable preclinical studies on these inhibitors, they are as yet unavailable for use as therapeutic drugs. Research in this field has focused largely on the development of low molecular weight (<500D) synthetic inhibitors of MMPs. In this review we focus on the various subgroups of MMP inhibitors now available, their preclinical evaluation and the limited information available from preliminary clinical trials. We comment on the suitability of the preclinical models used and the difficulty in designing clinical trials of these drugs. We focus on future developments which may involve the use of these drugs in combination with existing chemotherapeutic regimens to achieve a synergistic effect.

326

The matrix metalloproteinases (MMPs) are a family of closely related zinc-dependent proteolytic enzymes. Collectively, these enzymes are capable of degrading all the components of the extracellular matrix (ECM) [table I]. The ECM is the principal barrier to tumour growth and spread. Inhibitors of MMPs may be of therapeutic value in the treatment of malignant disease.[1] 1. What Are Matrix Metalloproteinase Inhibitors (MMPIs)? There are 2 groups of MMPIs: naturally occurring ‘tissue inhibitors of matrix metalloproteinases’ (TIMPs) and synthetic, broad spectrum inhibitors. 1.1 Naturally Occurring versus Synthetic MMPIs

The 4 naturally occurring TIMPs are potent native inhibitors of cysteine proteinases[2] and metalloproteinases.[3] TIMP-1 is a 28.5kD glycoprotein which forms a 1 : 1 complex with activated MMP1 and -3. It also forms a complex with the latent and active forms of MMP-9. TIMP-2 is a 21kD nonglycosylated protein which has a high affinity for inactive MMP-2. It forms a 1 : 1 complex with the latent and active forms of this enzyme, but also blocks the hydrolytic activity of all the activated MMPs. TIMP-3 is localised to the ECM and preferentially binds to components of the ECM. TIMP-2 can also bind to MMP-2/MT-MMP-1 to form a complex. TIMPs are not orally bioavailable, making them unsuitable for use as therapeutic drugs.[1] Research has focused on the development of low molecular weight (<500D) synthetic inhibitors, several of which have been developed. The majority of them are substituted peptide derivatives, composed of analogues of the cleavage site in the collagen molecule with a metal-binding group in the position of the cleaved peptide bond.[4] Their potencies are in the low nanomolar range. 1.2 How Do They Function?

X-ray crystallography data on mammalian © Adis International Limited. All rights reserved.

Watson & Tierney

MMPs has identified the active site of the enzymes and has allowed detailed structural studies to be performed.[5] The active site of the enzyme is composed of a cleft in which resides a zinc atom. The substrate of the enzyme fits tightly into the cleft, allowing exposure of a principal cleavage site which is then acted upon. Enzyme inhibitors have been designed to mimic the substrate such that they lie within the cleft and bind to the zinc atom within the active site, rendering the metalloproteinase inactive. Chemical groups which are able to interact with the zinc atom include both hydroxamate and carboxylate groups. Compounds have been derived which are composed of a peptide backbone which has a structure resembling a principal cleavage site in the collagen molecule. This allows a tight fit within the active site, with the hydroxamate/carboxylate group positioned so that it binds the zinc atom, resulting in reversible but potent inhibition. Table I. Biochemical characteristics of matrix metalloproteinases Enzyme name

MMP number

Molecular mass (kD)

Proteins degraded

Collagenase

1

55

Fibrillar collagens I, II, III, VIII, X

8

75

Proteoglycan

13

65

Fibrillar collagen

Gelatinase A

2

72

Denatured collagens (gelatins)

Gelatinase B (type IV collagenase)

9

92

Type IV collagens, elastin

Stromelysin-1

3

57

Proteoglycan core protein, non-helical regions of type IV collagen, laminin, fibronectin

Stromelysin-2

10

57

Procollagens I, II, III, gelatinase B, collagenase

Stromelysin-3

11

51

Like stromelysin

Matrilysin

7

28

Like stromelysin

Metalloelastase

12

54

Like stromelysin

MT-MMP-1

14

63

Progelatinase A

MT-MMP-2

?

Progelatinase A

MT-MMP-3

?

Progelatinase A

MT-MMP-4

?

? Abbreviations and symbol: MMP = matrix metalloproteinase; MTMMP = membrane-traversing MMP; ? = unknown.

BioDrugs 1998 Apr; 9 (4)

Matrix Metalloproteinase Inhibitors

1.3 Synthetic MMPIs

Tetracycline, doxycycline and minocycline are weak inhibitors of MMPs as well as having antimicrobial activity.[6] Tetracyclines can inhibit MMP-1 in vivo and in vitro, as well as MMP-2 and MMP-12 in vitro.[7] Synthetic tetracyclines are more efficient than the parent compound [IC50 (concentration for 50% inhibition) of 15 mmol/L for doxycycline compared with 350 mmol/L for tetracycline].[8] The antineoplastic anthracycline drugs daunorubicin, doxorubicin and epirubicin inhibit basement membrane collagen degrading activity in a reversible, non-competitive manner.[9] These compounds are likely to function like tetracycline, by chelating the metal ion. Batimastat (BB-94) is a low molecular weight (477D) broad-spectrum MMPI that is active against MMP-1 (IC50 3 nmol/L), MMP-3 (120 nmol/L), MMP-2 (4 nmol/L), MMP-9 (4 nmol/L) and MMP-7 (6 nmol/L). It is a selective inhibitor for the family of MMPs and shows little activity against other metalloproteinases such as angiotensin-converting enzyme.[1] Ilomastat (GM-6001) has also been tested in clinical trials. It is not orally bioavailable, but is in development as a topical agent for the treatment of corneal ulceration.[10] Second generation orally available MMPIs have begun to be tested in clinical trials. RO319790 was being developed as an anti-arthritic agent but development has been discontinued because of toxicity. Trials with the analogue RO323555 are continuing. Marimastat (BB-2516), an orally bioavailable MMPI, is currently undergoing phase II/III trials in patients with advanced malignancy. 2. Preclinical Evaluation The characteristics necessary for successful preclinical evaluation of MMPIs as potential anticancer agents are: • ability to prevent tumour invasion and spread • ability to inhibit neovascularisation. © Adis International Limited. All rights reserved.

327

The ability of MMPIs to inhibit tumour invasion may be evaluated by in vitro invasion assays involving ECM components and evaluating tumour invasion microscopically. In vivo models exist which follow experimental metastasis, such as the ability to lodge within the lungs following intravenous administration or the liver after intrasplenic administration. Such models require tumour cells to lodge within the first capillary bed encountered and their subsequent outgrowth may potentially involve MMP expression. Spontaneous metastasis models involve the spread of tumour cells from a primary site of implantation; as MMP expression is likely to be recruited at the primary and secondary sites of implantation, assessment of MMPIs may be more appropriate in the latter models. The ability of MMPIs to inhibit neovascularisation is important in preventing the establishment of a cancer at both the primary and secondary sites. In the following sections we discuss the preclinical screening of new MMPIs and comment both on the effects of the drugs and on the relevance of the cancer models used. 2.1 Hydroxamate-based MMPIs 2.1.1 Activity In Vitro

Although hydroxamate-based MMPIs have been shown to have either weak or no direct effects on tumour cell proliferation,[11-16] batimastat has been shown to inhibit invasion of human endothelial cells through ECM components.[14] 2.1.2 Activity In Vivo in Tumour Models Effect on subcutaneous growth

Assessment of this parameter may reflect the action of MMPIs on neovascularisation necessary to maintain the growth of subcutaneously grafted tumour tissue, which is generally not invasive. Such action on neovascularisation may potentially be responsible for necrosis and reduced growth in liver-invading xenografts.[15] Batimastat 30 mg/kg injected directly into the peritoneal cavity at day 11 after tumour inoculation inhibited the subcutaneous growth of the human melanoma line B16BioDrugs 1998 Apr; 9 (4)

328

BL6 by 33% (p < 0.05). The level of inhibition was increased to 58% when batimastat was administered from the day of tumour implantation.[11] Effect on Ascitic Cell Growth

The end-point of tumour progression in certain malignancies, such as ovarian, gastric and colorectal cancer, is the formation of ascites within the peritoneal cavity. Because batimastat is administered directly into the peritoneal cavity, it has been evaluated in a number of models in which ascites was induced with different tumour types. In an initial study by Davies et al.,[12] an ovarian ascites model was initiated with the HU xenograft. When cells were introduced into the peritoneal cavity, the tumour cells grew as a mucinous ascites, resulting in morbidity within 2 to 3 weeks. Batimastat was administered intraperitoneally from day 7 after tumour injection until day 20, which resulted in resolution of ascites and an improvement in survival from 18 days in the control animals to 105 days in the batimastat-treated group. Additionally, batimastat treatment resolved the ascites into small nodules which, when assessed histologically, were found to be avascular and composed of nests of tumour cells surrounded by a dense fibrous capsule.[12] A second ascites model has been initiated by injecting the human colorectal tumour C170HM2 into the peritoneal cavity of SCID mice.[17] When batimastat 40 mg/kg was administered from day 1, the volume of ascites in the treated mice was reduced by 79%, but ascites volume was not significantly affected when batimastat was administered from day 10. However, reduction in the size of peritoneal tumour deposits by 23 to 31% was seen with both treatment regimens.[17] Both ascites models again indicate the necessity for early treatment to maximise any subsequent therapeutic effects. Full or partial resolution of ascites has been achieved with batimastat, which may be due to the reduction in leakiness of endothelium in the peritoneal cavity, which is believed to be mediated by MMP, and/or a fibrous response to the tumour cells. © Adis International Limited. All rights reserved.

Watson & Tierney

The final model discussed in this section involves intraperitoneal administration of nonHodgkin’s lymphoma cells into SCID mice. This resulted in secondary spread to spleen, liver and lung.[18] Mice treated with batimastat showed no reduction in tissue colonisation by the lymphoma cells. The authors concluded that MMPIs do not affect non-Hodgkin’s lymphoma to the same extent as other solid tumour types. Effect on Angiogenesis

The effect of batimastat was evaluated in an experimental model of haemangioma formed by murine endothelioma cells previously transformed by the polyoma middle T oncogene (denoted eEnd.1).[14] eEND.1 endothelioma cells were injected subcutaneously into nude mice, which resulted in the formation of haemorrhaging haemangiomas. Daily treatment with batimastat 0.3 to 30 mg/kg administered at the site of cell injection resulted in inhibition of proliferation. When the treated tumours were analysed histologically, a reduction in the degree of haemorrhage was found. It was concluded that batimastat may have blocked endothelial cell recruitment by the transformed eEND.1 cells or disrupted vascular cell organisation.[14] Effect on Experimental Metastasis

Intravenous injection of the melanoma line B16-BL6 into mice results in the establishment of lung nodules. Treatment of mice with batimastat 30 mg/kg, administered into the peritoneal cavity at the time of cell injection, resulted in a 68% inhibition of lung colony formation (p < 0.001).[11] By radiolabelling the injected melanoma cells, the same authors found that batimastat treatment did not affect the retention of the cells within the lung but possibly reduced extravasation. Experimental lung metastasis of the rat mammary carcinoma HOSP.1P was inhibited by 80% following intraperitoneal administration of batimastat 30 mg/kg (6 injections). Both the number and size of the lung nodules were reduced in this particular model, showing that seeding efficiency and subsequent proliferation and outgrowth were both affected.[16] BioDrugs 1998 Apr; 9 (4)

Matrix Metalloproteinase Inhibitors

A slightly more discriminating experimental metastasis model involving specific liver invasion by a human colorectal tumour cell line was used to further evaluate batimastat, as MMPs have been shown to be expressed at the invading edge of the tumour.[15] Nude mice bearing C170HM2 liver tumours were treated with batimastat 40 mg/kg from day 10 to day 39 following cell injection, which resulted in a significant reduction in both the number and the size of the resultant liver tumours (10 of 20 batimastat-treated mice had no liver tumours compared with 2 of 20 in the vehicle-treated controls).[15] Histological examination of the liver tumours showed that treatment with batimastat resulted in larger areas of advanced necrosis when compared with the control tumours, which was indicative of poor vascularisation (fig. 1). SC-44463 is also a hydroxamate-based inhibitor and has been shown to inhibit experimental metastasis of the B16-F10 melanoma and the M2 clone of the K-1735 cell line.[19] SC-44463 1mg was administered at the time of cell injection (mixed with the cells) and an additional 2.5 to 5mg was injected directly into the peritoneal cavity 2 hours later. This regimen significantly reduced lung lesions from a mean of 133/mouse down to 5 to 15/mouse. Effect on Spontaneous Metastasis

Models of spontaneous metastasis allow the greatest evaluation of the role of MMPIs in preventing tumour spread, as they reflect every aspect of true metastasis such as growth of primary tumour, extravasation of cells into the circulation with subsequent extravasation at the secondary site, and metastatic growth. Some of these models allow the effect of drugs to be evaluated on both the primary and secondary sites simultaneously. Additional models involve removal of the primary tumour to allow subsequent outgrowth of the metastases. However, it must be clarified that if growth of the primary tumour is reduced indirectly by the action of MMPIs, then this may affect subsequent metastasis purely in terms of cell number and not by inhibition of enzymes required for invasion. © Adis International Limited. All rights reserved.

329

Fig. 1. Effect of batimastat on histology of liver tumours induced in nude mice by intraperitoneal administration of liver-invasive human colorectal tumour cells. Batimastat 40 mg/kg/day was administered intraperitoneally from day 10 to termination of therapy on day 39. (Top) Vehicle-treated animal: this nest of tumour cells comprises a peripheral zone of viable cells and a necrotic centre. (Bottom) Batimastat-treated animal: necrosis is advanced in this tumour nest and the peripheral zone of viable cells is thinner than in vehicle-treated animals (from Watson et al.,[15] with permission).

A spontaneous metastasis model was created using the B16-BL6 melanoma cell line.[11] The line was grown as solid subcutaneous syngeneic mouse tumours; after surgical removal of the primary tumour, the animals developed secondary spread to the lungs. Treatment with batimastat 30 mg/kg/day for 18 days resulted in a 76% reduction in the median weight of lung colonies with only a small reduction in the number of lung colonies.[11] The effect of batimastat in an orthotopic model of colon cancer metastasis was described by Wang BioDrugs 1998 Apr; 9 (4)

330

© Adis International Limited. All rights reserved.

Control rats Batimastat-treated rats

100

80

Survival (%)

et al.[20] Fragments of human colon carcinoma were sutured to the colon wall of nude mice. This allowed the effect of batimastat to be determined on local invasion within the abdominal wall, peritoneum and local lymph nodes, as well as on metastasis to more distant sites such as liver and lung. Batimastat 30 mg/kg/day was administered to the tumour-bearing animals 7 days after tumour implantation for the first 60 days. Treatment with 30 mg/kg thrice weekly was then continued to therapy termination. Batimastat treatment resulted in a reduction in the median weight of the primary tumour from 294 to 144mg and also reduced local and regional invasion from 12 of 18 to 7 of 20 mice. In addition, 6 of 18 control mice had metastatic disease (liver, lung, peritoneum, abdominal wall, lymph nodes) compared with 2 of 20 in the batimastat group, in which metastases were confined to the abdominal wall. Survival was followed in the 2 groups of mice; control survival was 110 days which was significantly increased to 140 days in the batimastattreated group. Thus, in a model which reflects the complexity and sites of recurrence existing in colorectal cancer, batimastat had significant effects on limiting growth and dissemination. Two mammary carcinoma models have been used to compare the effects of batimastat on outgrowth of secondary tumours following resection of the primary tumour.[13,16] The models were complimentary in that the first evaluated the effect of batimastat on a human breast cancer model in immunocompromised nude mice, and the second evaluated the effect of batimastat on a syngeneic rat mammary carcinoma in immunocompetent rats. The former model has the advantage that the tumour is of human origin; however, the immune system of nude mice can only make a limited contribution to therapeutic outcome, whereas the latter model allows the effect of batimastat to be evaluated with the assistance of a fully functional immune system. The nude mouse model involved transplantation of the human breast cell line MDA MB 435 into mammary fat pads. After 9 weeks, the mammary

Watson & Tierney

60

40

20

Tumour excision

0 0

30

60

90

120

150

Days after tumour inoculation

Fig. 2. Effect of batimastat on metastasis-free survival of rats

following excision of HOSP.1P mammary carcinomas. Primary tumours, initiated from the subcutaneous inoculation of 1 million HOSP.1P cells, were excised at day 14 (from Eccles et al.,[16] with permission).

fat pads were removed and mice were treated with either vehicle or batimastat 30 mg/kg/day. Mice were killed at week 16 and treatment with batimastat was shown to have reduced the incidence, number and total volume of lung metastases.[13] In the rat syngeneic model, HOSP.1P mammary tumour cells were inoculated into mammary fat pads of immunocompetent rats and tumours were excised on day 16. Animals were killed when metastatic disease manifested. Batimastat was administered intraperitoneally, either in a limited treatment regimen (5 or 14 days after removal of the primary tumour) or an extended regimen (2 days prior to surgery with continuation for 70 days). In the limited treatment regimen, lung metastasis was inhibited but there was no effect on lymphatic metastasis. However, survival was significantly increased. With the extended peri-operative treatment regimen, 100% of animals survived with no evidence of metastatic disease (fig. 2).[16] 2.2 TIMP-Based Inhibition

The activity of MMPs in the extracellular milieu is regulated by inhibitors, including TIMPs. The balance between the levels of MMPs and TIMPs is BioDrugs 1998 Apr; 9 (4)

Matrix Metalloproteinase Inhibitors

thought to be critical in the homeostasis of the ECM. Disruption of this balance in favour of the enzymes may be a critical determinant in tumour cell invasion. Four TIMPs have been identified and characterised: TIMP-1, TIMP-2, TIMP-3 and TIMP-4 are potent, broad spectrum MMPIs. Their potential use as therapeutic agents has been discussed in section 1.1. Although recombinant TIMPs have been used to clarify the role of these proteins in in vitro invasion assays[21,22] and human amnion invasion and lung colonisation by B16-F10 melanoma cells,[23] the potential drawbacks of using recombinant TIMP in an in vivo setting make it likely that the in vivo potential of TIMP-based MMP inhibition will lie in the ability to target TIMP genes to tumour cells that over-express MMP. The mouse melanoma line B16-BL6 was transfected with TIMP-1, which the cells subsequently overproduced, and this was shown to reduce the ability of the cells to invade Matrigel in an in vitro transwell invasion assay.[24] Experimental metastasis was also shown to be inhibited in vivo when assessed in the chick embryo[24] and in an experimental metastasis model in mice.[25] Extravasation of parental B16-F10 cells was compared with that of B16-F10 cells transfected to overproduce TIMP-1 by using intravital videomicroscopy of chick embryo chorioallantoic membrane.[26] No effect was seen on extravasation, which occurred at 36 hours for both cell lines. However, at day 7 there was a reduced number and size of tumours in the cells that over-expressed TIMP-1. An orthotopic model was used to evaluate the effect of transfection of gastric cancer cells with TIMP-1.[27] The human gastric carcinoma line KKLS was transfected with TIMP-1 and transplanted orthotopically into nude mice. A solid tumour, grown previously at a subcutaneous site, was implanted into the gastric wall, resulting in metastatic spread to the liver. Two transfectants were compared to the parental line: KTCL-1 with low TIMP-1 expression and KTCL-14 with high TIMP-1 expression. The parental line metastasised to the liver at week 1 and grew exponentially, as © Adis International Limited. All rights reserved.

331

shown by polymerase chain reaction amplification of the human β-globin gene fragment from murine liver samples. KTCL-1 and KTCL-14 cells metastasised to the liver but did not grow.[27] Cells have also been manipulated to overexpress TIMP-2. Invasive H-Ras-transfected rat embryo fibroblasts, when injected subcutaneously, grow as a primary tumour which subsequently metastasises spontaneously. When the cells were manipulated to over-express TIMP-2, both primary growth and metastases were inhibited. At the primary site, tumours were associated with a capsule of connective tissue, resulting in inhibition of local invasion.[28] TIMP-2 over-expression, again induced by transfection of the gene, inhibited the subcutaneous growth and spontaneous metastasis of a human melanoma in SCID mice.[29] As efficient methods of gene delivery are still being developed, drugs which can upregulate TIMP production may have potential therapeutic value. Oncostatin M, a lymphocyte and monocytederived cytokine, has been shown to upregulate expression of mRNA for TIMP-1 in human lung fibroblast cultures.[30] Use of such a cytokine in the tumour scenario may modulate metastatic potential. Cinnamic acid, which is a naturally occurring fatty acid, reduces tumour cell proliferation and invasion through Matrigel. This reduction is associated with downregulation in expression of collagenase type IV mRNA at the same time as upregulation of expression of TIMP-2 mRNA.[31] This product is now being more fully evaluated in both preclinical and clinical studies. 2.3 Additional Classes of MMPIs

CDP-845 is an orally active carboxylate-based gelatinase inhibitor which has been shown to have in vivo activity in a mouse metastasis model.[32] However, its development has now been suspended. BE-16627B has been evaluated in preclinical studies on 2 subcutaneously grafted xenografts: HT1080, a fibrosarcoma with high MMP expression, and HCT116, a human colon carcinoma with BioDrugs 1998 Apr; 9 (4)

332

low expression of MMPs.[33] When administered to mice by an osmotic minipump (2 mg/day/animal for 3 weeks), BE-16627B inhibited the subcutaneous growth of HT1080 by 71% but did not significantly inhibit the growth of HCT116. BE-16627B was also evaluated in an experimental metastasis model using the HT1080 fibrosarcoma which was injected intravenously, resulting in lung nodules.[33] BE-16627B at 2 mg/day/animal for 3 weeks, administered by an osmotic minipump implanted 1 week after tumour cell injection, reduced the number (to 24.3%) and size (to 46.4%) of lung nodules compared with vehicle-treated controls. Lung weight increase due to tumour growth was also inhibited by 85.5%. Ilomastat is a specific inhibitor of MMPs involved in angiogenesis[34] and was shown to reduce the number and area of new blood vessels in rat corneas implanted with the Walker 256 carcinosarcoma. Its potential therapeutic application in tumour progression has not yet been documented. This is also the same for minocycline, which has been reported to have anticollagenase activity.[34] Platelet factor 4 has been shown to inhibit collagenase activity[35] in addition to its multiple other properties, including an effect on endothelial migration and proliferation.[36,37] Systemic administration of recombinant platelet factor 4 was shown to inhibit the experimental metastasis of B16F10 melanoma cells. This resulted in an approximate 50% reduction in both the number of lung metastases and lung weight.[38] However, due to the multiple effects of platelet factor 4, it is difficult to know the extent to which MMP inhibition contributed to the overall antitumour effect. 2.4 Combination Therapy with MMPIs and Cytotoxic Agents

As MMPIs and cytotoxic agents target different aspects controlling malignant cell growth, combination treatment may result in reduced invasion, neovascularisation and direct cell proliferation. The most investigated MMPIs, batimastat and marimastat, have been shown to have no direct cytotoxic effects and thus a small number of pilot © Adis International Limited. All rights reserved.

Watson & Tierney

studies have been initiated to evaluate these MMPIs in combination with cytotoxics. Batimastat has been shown to potentiate cisplatin activity in the human ovarian xenograft HOC22.[39] Administration of batimastat 60 mg/kg increased the survival of the mice, but it was not active on late stage tumours. However, when given in combination with cisplatin at 4 mg/kg, all mice with early stage tumours were cured and mice with late stage tumours had increased survival. Marimastat has been given in combination with cisplatin for the treatment of small cell lung cancer xenografts.[40] Marimastat 50 mg/kg orally twice daily reduced the growth of the 841 xenografts from 265 mm2 in the vehicle control to 190 mm2, resulting in an survival increase from day 31 to day 38. When marimastat was combined with intravenous cisplatin 4 mg/kg, the tumour size was further reduced to 58 mm2 with a survival increase to day 70. The therapeutic effect of the orally active gelatinase inhibitor CDP-845 together with cyclophosphamide was evaluated in the Lewis lung carcinoma model.[41] Cells were injected subcutaneously into the hind legs of C57BL/6 mice; primary tumours grew by day 7 and metastasised to the lung, inducing morbidity by days 21 to 25. CDP845 plus cyclophosphamide was more effective at delaying local tumour growth (by 31 days) compared with CDP-845 alone (2.6 days) and cyclophosphamide alone (19.5 days). In addition, the number of lung metastases was reduced to 5 in mice receiving the combination regimen compared with 15 in the CDP-845 group, 11 in the cyclophosphamide group and 24 in the no-treatment group. 3. Design of Clinical Trials Defining the end-points of any trial evaluating a novel anticancer therapy for use in advanced disease is contentious. An attractive concept is to measure the incidence of metastases, which should in theory be decreased. Clinical evaluation of proteinase inhibitors is complicated by the fact that these drugs are designed to limit disease progression only and are unlikely to cause measurable tumour BioDrugs 1998 Apr; 9 (4)

Matrix Metalloproteinase Inhibitors

shrinkage. Conventional methods of measuring therapeutic efficacy, such as serial computed tomography scans or cancer antigen levels, are not applicable, but other measures such as quality of life and survival are measurable end-points. [1] These clinical methods of measuring drug efficacy are more relevant to drugs of this type than laboratory methods such as the assessment of net proteolytic activity. Cancer metastasis is a biological system with many variables, and the end result of prolonged survival with improved quality of life is the most important measurable quantity. 4. Results of Preliminary Clinical Trials As the role of MMPs in tumour growth and angiogenesis has been elucidated, MMPIs have been studied as antimetastatic agents. The first clinical trials with batimastat started in 1991. The drug taken orally as a capsule showed poor bioavailability, and results from preclinical work indicated that the drug might be more effective when administered via the intraperitoneal route.[1] Batimastat has been administered to patients via the intraperitoneal and intrapleural routes for the treatment of malignant ascites and malignant pleural effusion, respectively. A phase I study of intrapleural batimastat in patients with malignant pleural effusion reported that significantly fewer pleural aspirations were required after treatment for 1 month compared with before treatment.[42] Peak serum batimastat concentrations were seen 4 to 24 hours after administration, and concentrations greater than the IC50 were detectable for up to 4 weeks. A phase I/II study of batimastat in patients with symptomatic malignant ascites was started in 1993. Patients received a single intraperitoneal dose of batimastat in 500ml of 5% dextrose (150 to 1350 mg/m2) There was reduced discomfort and a decreased requirement for paracentesis in a proportion of patients. The drug was well tolerated. High and sustained plasma concentrations of drug were achieved, with batimastat 100 to 200 ng/ml detectable 28 days after administration.[43] © Adis International Limited. All rights reserved.

333

Oral batimastat has been used in a phase I study in patients with metastatic bone disease. The drug was well tolerated and drug concentrations were measurable in the serum of patients after treatment.[44] Marimastat is a second-generation MMPI with good oral bioavailability. It has progressed to phase II clinical trials in patients with advanced malignancy, including pancreas, ovarian, gastric, breast and colorectal cancer. Preliminary results, including cancer antigen studies and time to disease progression, have been encouraging. Randomised placebo-controlled trials in patients with advanced malignancy are currently ongoing. 5. Discussion The cancer models described in this article indicate the range of antitumour effects of MMPIs. These drugs are capable of inhibiting invasive tumour growth, organ colonisation and metastatic spread. Inhibition of tumour growth is inconsistent, but this may reflect the organ associated with metastatic spread, the site of cell administration and the tumour type. MMPIs also inhibit tumourinduced angiogenesis, but there is no evidence of a direct cytotoxic effect. Overall the effects of these drugs would appear to be ‘tumourostatic’. The question of which clinical end-points are appropriate for studying such drugs has been discussed. The conventional definitions of biological response seen for cytotoxic drugs would be inappropriate. Current clinical studies have used survival data and time to disease progression as measurements of efficacy. No drug is without adverse effects, and little is known about the effects of long term inhibition of a family of enzymes which have a normal physiological role in addition to their pathological role. However, as MMPs are so widely over-expressed in malignancy compared with TIMPs, administration of MMPIs may be looked upon as a way of redressing this balance. Future applications for MMPIs may include combining them with cytotoxic agents. Animal models have shown the potentiation of cytotoxics BioDrugs 1998 Apr; 9 (4)

334

Watson & Tierney

by anti-angiogenesis agents. The possibility exists that, in combination with conventional chemotherapy, MMPIs may improve the prognosis for patients with advanced malignancy. References 1. Brown PD, Giavazzi R. Matrix metalloproteinase inhibition: a review of anti-tumour activity. Ann Oncol 1995; 6: 967-74 2. Jarvinen M, Rinne A, Hopsu-Havu VK. Human cystatins in normal and diseased tissues – a review. Acta Histochem 1987; 82: 5-18 3. Stetler-Stevenson WG, Krutzsch HC, Liotta LA. Tissue inhibitor of metalloproteinase (TIMP-2). J Biol Chem 1989; 264 (29): 17374-8 4. Johnson WH, Roberts NA, Borkakoti NJ. Collagenase inhibitors: their design and potential therapeutic use. Enzyme Inhibition 1987; 2: 1-22 5. Bode W, Reinemer P, Huber R, et al. The X-ray crystal structure of a catalytic domain of human neutrophil collagenase inhibited by a substrate analogue reveals the essentials for catalysis and specificity. EMBO J 1994; 13: 1263-9 6. Lee W, Aiken K, Kulkarni G, et al. Collagenase activity in recurrent periodontitis: relationship to disease progression and doxycycline therapy. J Periodont Res 1991; 26: 470-85 7. Golub L, Ramamurthy NS, MacNamara TF, et al. Tetracyclines inhibit connective tissue breakdown: new therapeutic implications for an old family of drugs. Crit Rev Oral Biol Med 1991; 2: 297-322 8. Vincenti MP, Clark IM, Brinkerhoff CE. Using inhibitors of metalloproteinases to treat arthritis. Easier said than done? Arthritis Rheum 1994; 37: 1115-26 9. Karakiulakis G, Missirlis E, Maragoudakis ME. Basement membrane collagen-degrading activity from a malignant tumour is inhibited by anthracycline antibiotics. Biochim Biophys Acta 1990; 1035: 218-22 10. Galardy RE, Grobelny D, Foellmer HG, et al. Inhibition of angiogenesis by the matrix metalloproteinase inhibitor N-(2R-2(hydroxamidocarbonymethyl) - 4 - methylpentanoyl - L - tryptophan methylamide. Cancer Res 1994; 54: 4715-8 11. Chirivi RGS, Garofalo A, Crimmin MJ, et al. Inhibition of the metastatic spread and growth of B16-BL6 murine melanoma by a synthetic matrix metalloproteinase inhibitor. Int J Cancer 1994; 58: 460-4 12. Davies B, Brown P, East N, et al. A synthetic matrix metalloproteinase inhibitor decreases tumour burden and prolongs survival of mice bearing human ovarian carcinoma xenografts. Cancer Res 1993; 53; 2087-91 13. Sledge Jr GW, Qulali M, Goulet R, et al. Effect of matrix metalloproteinase inhibitor batimastat on breast cancer regrowth and metastasis in athymic mice. J Natl Cancer Inst 1995; 87: 293-8 14. Taraboletti G, Garofalo A, Belotti D, et al. Inhibition of angiogenesis and murine hemangioma growth by batimastat, a synthetic inhibitor of matrix metalloproteinases. J Natl Cancer Inst 1995; 87: 293-8 15. Watson SA, Morris TM, Robinson G, et al. Inhibition of organ invasion by the matrix metalloproteinase inhibitor batimastat (BB-94) in two human colon carcinoma metastasis models. Cancer Res 1995; 55: 3629-33 16. Eccles SA, Box GM, Court WJ, et al. Control of lymphatic and hematogenous metastasis of a rat mammary carcinoma by the

© Adis International Limited. All rights reserved.

17. 18. 19.

20.

21. 22.

23.

24.

25.

26.

27.

28. 29.

30. 31. 32. 33. 34.

matrix metalloproteinase inhibitor batimastat (BB-94). Cancer Res 1996: 56: 2815-22 Watson SA, Morris TM, Parsons SL, et al. Therapeutic effect of the matrix metalloproteinase inhibitor, batimastat, in a human colorectal cancer ascites model. Br J Cancer 1996; 74: 1354-8 Zubair AC, Ali SA, Rees RC, et al. The effect of BB-94 (batimastat) on the in vivo behaviour of human lymphoma cells in SCID mice [abstract]. Br J Cancer 1996; 73 Suppl. 26: 42 Reich R, Thompson EW, Iwamoto Y, et al. Effects of inhibitors of plasminogen activator, serine proteinases, and collagenase IV on the invasion of basement membranes by metastatic cells. Cancer Res 1988; 48: 3307-12 Wang X, Fu X, Brown PD, et al. Matrix metalloproteinase inhibitor BB-94 (batimastat) inhibits human colon tumour growth and spread in a patient-like orthotopic model in nude mice. Cancer Res 1994; 54: 4726-8 Mignatti P, Robbins E, Rifkin DB. Tumour invasion through the human amniotic membrane: requirement for a proteinase cascade. Cell 1996; 47: 487-98 DeClerck YA, Yean DT, Chan D, et al. Inhibition of tumour invasion of smooth muscle cell layers by recombinant human metalloproteinase inhibitor (rMI/rTIMP-2). Cancer Res 1991; 51: 2151-7 Schultz RM, Silberman S, Persky B, et al. Inhibition by human recombinant tissue inhibitor of metalloproteinases of human amnion invasion and lung colonisation by murine B16-F10 melanoma cells. Cancer Res 1988; 48: 5539-45 Khokha R, Zimmer MJ, Wilson SM, et al. Up-regulation of TIMP-1 expression in B16-F10 melanoma cells suppresses their metastatic ability in chick embryo. Clin Exp Metastasis 1992; 10: 365-70 Khokha R. Suppression of the tumourigenic and metastatic abilities of murine B16-F10 melanoma cells in vivo by overexpression of the tissue inhibitor of metalloproteinases-1. J Natl Cancer Inst 1994; 86: 299-304 Koop S, Khokha R, Schmidt EE, et al. Overexpression of metalloproteinase inhibitor in B16-F10 cells does not affect extravasation but reduces tumour growth. Cancer Res 1994; 54: 4791-7 Watanabe M, Takahashi Y, Ohta T, et al. Inhibition of metastasis in human gastric cancer cells transfected with tissue inhibitor of metalloproteinase-1 gene in nude mice. Cancer Suppl 1996; 77 (8): 1676-80 Declerck YA, Perez N, Shimada H, et al. Inhibition of invasion and metastasis in cells transfected with an inhibitor of metalloproteinases. Cancer Res 1992; 52: 701-8 Montgomery AMP, Mueller BM, Reisfeld RA, et al. Effect of tissue inhibitor of the matrix metalloproteinases-2 expression on the growth and spontaneous metastasis of a human melanoma cell line. Cancer Res 1994; 54: 5467-73 Richards CD, Shoyab M, Brown TJ, et al. Selective regulation of metalloproteinase inhibitor (TIMP-1) by oncostatin M in fibroblasts in culture. J Immunol 1993; 150: 5596-603 Liu L, Hudgins WR, Shack S, et al. A natural product with potential use in cancer intervention. Int J Cancer 1995; 62: 345-50 Chander SK, Antoniw P, Beeley NR, et al. An in vivo model for screening peptidomimetic inhibitors of gelatinase A. J Pharm Sci 1995; 84: 404-9 Naito K, Kanbayashi N, Nakajima S, et al. Inhibition of growth of human tumour cells in nude mice by a metalloproteinase inhibitor. Int J Cancer 1994; 58: 730-5 Tamargo RJ, Bok RA, Brem H. Angiogenesis inhibition by minocycline. Cancer Res 1991; 51: 672-5

BioDrugs 1998 Apr; 9 (4)

Matrix Metalloproteinase Inhibitors

35. Zucker MB, Katz IR. Platelet factor 4: production, structure and physiologic and immunologic action. Proc Soc Exp Biol Med 1991; 198: 693-702 36. Maione TE, Gray GS, Petro J, et al. Inhibition of angiogenesis by recombinant human platelet factor-4 and related peptides. Science 1990; 237: 77-9 37. Sharpe RJ, Randolph HR, Scott CF, et al. Growth inhibition of murine melanoma and human colon carcinoma by recombinant human platelet factor-4. J Natl Cancer Inst 1990; 82: 848-53 38. Kolber DL, Knisely TL, Maione TE. Inhibition of development of murine melanoma lung metastases by systemic administration of recombinant human platelet factor-4. J Natl Cancer Inst 1995; 87: 304-8 39. Nicoletti MI, Garofalo A, Lucchini V, et al. Batimastat potentiates cisplatin activity on human ovarian carcinoma xenografts [abstract]. Eur J Cancer 1996; 32A Suppl. 1: 6 40. Watson SA, Morris TM, Parsons S, et al. Combined therapeutic effect of marimastat and cisplatin on the in vivo growth of a human small cell lung cancer [abstract]. Br J Cancer 1996; 73 Suppl. 26: 29 41. Anderson IC, Shipp MA, Docherty AJP, et al. Combination therapy including a gelatinase inhibitor and cytotoxic agent

© Adis International Limited. All rights reserved.

335

reduces local invasion and metastasis of murine Lewis lung carcinoma. Cancer Res 1996; 56: 715-8 42. Macaulay V, O’Byrne K, Saunders M, et al. Phase I study of the matrix metalloproteinase inhibitor batimastat (BB-94) in patients with pleural effusions [abstract]. Br J Cancer 1995; 71 Suppl. 24: 11 43. Beattie GJ, Young HA, Smyth JF. Phase I study of intraperitoneal metalloproteinase inhibitor BB-94 in patients with malignant ascites [abstract]. 8th NCI-EORTC Symposium of New Drugs in Cancer Therapy. 1994 Mar 15-18; Amsterdam 44. de Takats PG, Young A, Kiernan J, et al. Phase I study of oral batimastat in patients with metastatic bone disease [abstract]. Br J Cancer 1996; 73 Suppl. 26: 51

Correspondence and reprints: Dr Susan A. Watson, Cancer Studies Unit, Department of Surgery, Queens Medical Centre, University of Nottingham, Nottingham NG7 2UH, England.

BioDrugs 1998 Apr; 9 (4)