Inhibition of Angiogenesis

Cortisone acetate 0.9 mg in 0.9 ml saline was flooded over the chorioallantoic membrane of 8-day chick embryos through a window in the shell made previously. On day 9, tumor extract (100 .mu.g) from hepatoma cells (as described in Zetter, Nature Vol. 285, 41-43, 1980) in 5 .mu.l H.sub.2 O was placed on the center of a round 15 mm diameter plastic coverslip and allowed to dry. To the center of each coverslip was then added a 5 .mu.l aliquot containing either heparin (6 .mu.g, i.e., 1 unit), or water, at least twenty embryos being subjected to each. After drying, the coverslip was placed on the chorioallantoic membrane: Additional control embryos received tumor extract and/or heparin, but were not pre-treated with cortisone acetate. The membranes were viewed on day 11 with a .times.12 stereoscope. Angiogenesis was present if new capillaries were seen to converge on the spot in the center of the coverslip. All of the embryos treated with water or heparin but no cortisone exhibited angiogenesis, as well as 80% of those treated with cortisone acetate alone. Less than 2% of those treated with both heparin and cortisone acetate exhibited angiogenesis.

Porcine mucosal heparin was exhaustively degraded using heparinase by the procedure of Langer et al , Science Vol. 217, 261-3 (1982) and the products were fractionated using Sephadex columns equilibrated with 1M NH.sub.4 OAc. The degraded heparin had no anticoagulant activity as determined by activated partial thromboplastic time or whole blood recalcification time. The product or product mixture (250 mg) was dissolved in 1 cc of 1M NH.sub.4 OAc, loaded onto a 75.times.2.5 cm G-15 column, and eluted at 0.5 cc/min. This resulted in several incompletely resolved peaks corresponding to tetra-, hexa-, and higher oligosaccharides and a separate peak corresponding to disaccharide product. The disaccharide peak was freeze-dried, dissolved in 0.2 cc of 1M NH.sub.4 OAc and rechromatographed on G-15 resulting in the same sharp peak which was freeze-dried. The mixture of tetra-, hexa-, and oligosaccharides was freeze-dried, redissolved in 1 cc of 1M NH OAc, and eluted from a 50.times.1.25 cm G-50 at 2 cc/min, resulting in an unresolved double peak corresponding to tetra- and hexasaccharide fragments and an additional peak corresponding to oligosaccharides which was freeze-dried. The tetra- and hexasaccharide fragments were combined, freeze-dried, redissolved in 1 cc of 1M NH.sub.4 OAc, and loaded onto a G-15 column. The tetrasaccharide was eluted from a G-15 column in a single peak, the center cut of which was freeze-dried. The hexasaccharide fraction was freeze-dried, redissolved in 0.3 cc 1M NH.sub.4 OAc, eluted from a G-15 column in a single peak, the center cut of which was freeze-dried.

Fragment size was determined by dissolving a weighed amount of each fraction into 0.03M hydrochloric acid and measuring the absorbance of this solution at 232 nm. The molecular weight of each fragment was calculated using a molar absorptivity, for the .alpha., .beta. unsaturated carboxylate end group present in each of these products, of .xi.=5500. The di- and tetrasaccharides were further characterized by comparing their K avg-values on G-15 with mono-, di-, and trisaccharide standards. Measured molecular weights were 530, 1210, 1600, and 1870 for the di-, tetra-, hexa-, and oligosaccharide fractions respectively.

The various heparin fragments were dissolved into methylcellulose discs either alone or with cortisone acetate. The discs were then applied to the 4-day yolk sac membrane of chick embryos cultured in Petri dishes as described by Taylor and Folkman, Nature, Vol. 297, 307-312 (1982). In the presence of cortisone acetate (100 .mu.g), as shown in the following table, the hexasaccharide fragment demonstrated the highest antiangiogenesis activity.

Tetra- and disaccharides were inactive. Oligosaccharides were less active and were toxic at higher concentrations. Therefore, the hexasaccharide fragment was used in subsequent experiments. In the growing 6-day chorioallantoic membrane, discs containing hexasaccharide (12 .mu.g) and cortisone acetate (100 .mu.g) produced large avascular zones up to 12.6.+-.0.1 mm diameter by 48 hours. As in the case of heparin (with cortisone acetate), capillaries in the mesodermal layer were absent while the other two tissue layers of the membrane were intact and viable. Hexasaccharide alone did not promote tumor angiogenesis as heparin did.

All discs contained a combination of cortisone acetate (100 .mu.g) and a heparin fragment. No avascular zones developed in the presence of any heparin fragment alone, or with cortisone or methylcellulose alone. Ten embryos were used for each group. With hexasaccharide (plus cortisone), the area of the avascular zone was 17% of the vascular membrane at 12 .mu.g and 15% at 0.1 .mu.g. For the oligosaccharides, the maximum avascular area was 10%.

Fertilized chick embryos were removed from their shell on day 3 (or 4) and incubated in a Petri dish in high humidity and 5% CO.sub.2 as previously described by Auerbach et al., J. Devel. Biol., Vol. 41, 391-4 (1974), except that an outer dish and antibiotics were not used. On day 6, a methylcellulose disc (10 .mu.l) containing either heparin (6 .mu.g), or hexasaccharide heparin fragment (12 .mu.g), or cortisone acetate (Sigma, powder free of preservatives and suspending agents), or a combination of cortisone acetate+heparin or cortisone acetate+hexasaccharide was implanted on the chorioallantoic membrane. The embryos were examined 48 hours later, and if a clear avascular zone appeared around the methylcellulose disc, the diameter of the zone was measured with a Nikon Profile projector at .times.20. Thirty embryos were used in each group. India ink was injected into the heart of some embryos just before formalin fixation so that vessels could be followed to the edge of the avascular zone in histological sections.

Hexasaccharide+cortisone acetate produced avascular zones of 12.6.+-.0.1 mm diameter in all embryos. Heparin+cortisone acetate produced avascular zones of 8.9.+-.0.7 mm diameter. There were no avascular zones in the presence of any compounds alone, or with methylcellulose alone.

Histologic cross-sections of the chorioallantoic membranes, revealed that capillaries developed normally in the presence of any compound alone. In contrast, capillaries were completely absent from the mesodermal layer in the face of either hexasaccharide+cortisone acetate, or heparin+cortisone acetate, while the ectodermal and endodermal cell layers remained unaffected. In the mature chorioallantoic membrane where vessels are no longer growing, the cortisone acetate-heparin or -hexasaccharide fragment combinations were without effect.

Avascular zones were also observed using essentially the same procedure with a combination of heparin (Panheprin) (10 units) with 25-200 .mu.g of each of the following steroids:

11.beta., 17.alpha., 21-trihydroxypregn-4-ene-3,20-dione

11.alpha., 17.alpha., 21-trihydroxypregn-4-ene-3,20-dione

17.alpha.,21-dihydroxypregn-4-ene-3,20-dione

17.alpha.,21-dihydroxypregna-4,9(ll)-diene-3,20-dione

9.alpha.-fluoro-11.alpha.,16.alpha.,17.alpha.,21-tetrahydroxypregna-1,4-die ne-3,20-dione (or triamcinolone)

9.alpha.,11.beta.-dichloro-17.alpha.,21-dihydroxypregn-4-ene-3,20-dione-21- acetate

17.alpha.,21-dihydroxy-16.beta.-methylpregna-4,9(11)-diene-3,20-dione-21-be nzoate

17.alpha.,21-dihydroxy-16.beta.-methylpregna-1,4,9(11)-triene-3,20-dione-21 -succinate sodium monohydrate

6.alpha.-fluoro-17.alpha.,21-dihydroxy-16.beta.-methylpregna-4,9(11)-diene- 3,20-dione

6.alpha.-fluoro-17.alpha.,21-dihydroxy-16.beta.-methylpregna-4,9(11)-diene- 3,20-dione-17,21diacetate

15.alpha.,17.alpha.,21-trihydroxy-4-pregnene-3,20-dione

16.alpha.,17.alpha.,21-trihydroxy-6.alpha.-methylpregn-4-ene-3,20-dione-21- acetate-16,17-cyclic ketal of acetone

Of the foregoing, the last four displayed the greatest anti-angiogenic activity by this test.

All of the effective combinations of active agents, for the purposes of the present invention, exhibit an avascular zone when implanted on an immature chick chorioallantoic membrane as described in this Example 3.

Polymer pellets of ethylene vinyl acetate copolymer (EVA) of approximately 1 mm diameter were impregnated, using the procedure of Langer et al., Nature, Vol. 263, 797-800 (1976), with heparin 180 .mu.g (Sigma), or hexasaccharide fragment 300 .mu.g, or cortisone acetate 1.5 mg (Sigma), or a combination of cortisone and heparin. The pellets were implanted in the cornea of a rabbit eye 1 mm from the limbus and a 1 mm.sup.3 piece of V2 carcinoma was implanted distal to the polymer, 2 mm from the limbus. In the opposite eye of each rabbit, control pellets that were empty were similarly implanted in juxtaposition to the tumor.

Release rates averaged 15 .mu.g/day for heparin; 21 .mu.g/day for hexasaccharide fragment; and 5 .mu.g/day for cortisone. When the compounds were mixed, they released at the same rates. By spectrophotometry, the pellets released heparin for 14 days, hexasaccharide for 11 days, and cortisone for more than 30 days.

As capillary blood vessels grew towards the tumor implant, maximum vessel length was measured every 3 days with a stereoscopic slit lamp at .times.10 (.+-.0.1 mm). On day 14 the rabbits were killed, and India ink was injected into each carotid artery. The corneas were removed and examined with a stereoscope.

New capillary blood vessels were observed growing towards the tumor and passing over an empty pellet or a pellet containing heparin alone at a mean rate of 0.44 mm/day; and over a pellet containing cortisone alone at 0.22 mm/day. The tumors behind these pellets were vascularized by 6-8 days. When the pellets contained both cortisone and heparin, there was no capillary growth for 13 days. When the heparin-cortisone pellets were removed or when the pellets were depleted of heparin, capillary growth resumed. Histologic sections showed that tumor cells remained viable and capable of replication even when they were adjacent to the heparin-cortisone pellet.

In the presence of implanted pellets in which the hexasaccharide fragment of Example 2 replaced the heparin, new capillaries grew toward the tumors at a mean rate of 0.30 mm/day in the presence of the hexasaccharide pellets; 0.14 mm/day when the pellets contained cortisone; and 0.32 mm/day when the pellets were empty. In the presence of the hexasaccharide-cortisone combination, there was no capillary growth throughout the 13-day observation period in 4 rabbits, and in one rabbit a few capillaries grew at 0.07 mm/day.

(a) Reticulum cell Sarcoma: 1 mm.sup.3 pieces of tumor were implanted with a trocar subcutaneously in the backs of 30 mice; 5 per group. Treatment was begun 10 days later, when mean tumor volume was 1.5.times.10.sup.2 mm.sup.3. Oral heparin was 200 U/ml in drinking water, average daily consumption was 5-10 ml water per 22 g mouse. Cortisone acetate was administered subcutaneously once daily in a dose of 250 mg/kg for six days, 125 mg/kg for one day, 75 mg/kg for one day, and thereafter a daily maintenance dose of 37 mg/kg (a "tapered" dosage). Control animals received either saline injections, or heparin alone or cortisone acetate alone. All controls were dead by day 34 with large primary tumors and lung metastases. All mice treated with oral heparin+cortisone tapered dosage became tumor-free by day 15 and remained so after treatment was discontinued.

An additional group of mice was treated similarly except that heparin was administered twice daily subcutaneously in a dose of 627 units, and cortisone acetate was administered subcutaneously once daily in a uniform dose of 75 mg/kg. Response of the mice was the same as in the first set except that tumors recurred after cessation of treatment; these mice became permanently tumor-free when subjected to the regimen of oral heparin and cortisone acetate tapered dosage described above. One of these mice died on day 31 with no gross primary tumor and no metastases.

(b) Lewis Lung Carcinoma: Treatment began 7 days after implantation of a 1 mm.sup.3 piece of tumor in 42 mice: 7 per group. Treatment was with oral heparin and subcutaneous cortisone acetate tapered dosages described above. All controls died by day 33 with large tumors and numerous lung metastases. In the heparin+cortisone acetate groups treatment was discontinued for each mouse after tumor had been invisible for approximately 7 days. In the oral heparin+cortisone acetate all mice were off treatment by day 33, and remained tumor-free. In an additional group treated with subcutaneous heparin and subcutaneous cortisone acetate (75 mg/kg), 5 mice were off treatment at day 37 and remained tumor-free. Two mice died of pneumonia on days 30 and 33 respectively with small primary tumors (<75 mm.sup.3). One metastasis was found in one mouse.

To determine if other steroids could substitute for cortisone, heparin was administered with hydrocortisone, dexamethasone, or medroxyprogesterone. Only hydrocortisone was as effective as cortisone acetate in causing tumor regression when administered with heparin. At the highest tolerable doses neither dexamethasone (3.2 mg/Kg), nor medroxyprogesterone (112 mg/Kg), caused regression of Lewis lung tumors with or without heparin.

(c) B-16 Melanoma: 7.4.times.10.sup.6 melanoma cells were injected subcutaneously into 40 mice; 5 per group. Treatment of one group was with oral heparin as described above and oral hydrocortisone, 0.45 mg/ml in drinking water Another group received oral heparin and subcutaneous cortisone acetate tapered dosage, and a third group received subcutaneous heparin and cortisone acetate 75 mg/kg. Controls received either water, or heparin alone, or hydrocortisone or cortisone acetate alone. All control animals died by day 31 with large tumors and lung metastases. In the heparin+cortisone acetate groups, treatment was discontinued after tumor had become invisible for approximately 7 days. In the oral heparin+cortisone acetate tapered dosage group, 1 mouse died on day 24 with a partially regressed tumor and 2 lung metastases that were avascular and measured less than 0.1 mm. All other mice in the group became tumor-free and remained so after their treatment was discontinued by day 32. In the subcutaneous heparin+cortisone acetate group one mouse died on day 18 and another on day 21; neither had lung metastases. Treatment was discontinued for the other 3 mice on day 32; tumors 3 weeks later were successfully re-treated with oral heparin+cortisone acetate tapered dosage, and these mice have remained tumor-free In the oral heparin+oral hydrocortisone group, all mice remained tumor-free after their treatment was discontinued on day 47. The regimen of oral heparin+oral hydrocortisone seemed to be more effective for melanoma than it was for ovarian sarcoma or Lewis lung carcinoma.

(d) Bladder Carcinoma: 70 mice: 7 per group received a 1 mm.sup.3 implant of tumor subcutaneously. All control animals died by day 31, with large primary tumors. No mice bearing bladder carcinoma developed lung metastases. One group was treated with oral heparin and subcutaneous cortisone acetate tapered dosage, starting on day 9 when mean tumor volume was 140 mm.sup.3. Tumors stopped growing, but regressed only partially, and then reached a steady state where tumor volume remained at approximately 70 mm.sup.3 for as long as the treatment was continued (i.e., 61 more days). Only one mouse died of pneumonia on day 19. The "dormant" tumors were viable, as evidenced by resumption of tumor growth whenever treatment was discontinued for one mouse at a time, beginning at day 70. For a second group treated with subcutaneous heparin and subcutaneous cortisone acetate 75 mg/kg as described above, the result was similar; i.e., long-term tumor "dormancy". One mouse died at day 21.

Because of the inability the standard regimen of oral heparin+cortisone acetate to bring about complete regression, higher concentrations of oral heparin were used with other groups. With oral heparin (600 U/ml)-cortisone acetate tapered dosage, there was more significant tumor regression and a steady state ("dormancy") was reached at a smaller tumor volume of approximately 45 mm.sup.3. One mouse died. However, with 1000 U/ml heparin, there was complete regression; mice remained tumor-free after discontinuation of treatment on ay 39. No mice died in this group.

In summary, all tumors either stopped growing or regressed when the heparin-cortisone acetate combination was administered. In contrast, when either compound was used alone, tumor growth continued at the same rate as in animals receiving only saline injections; all such control animals died with a large tumor burden.

In the majority of animals treated with heparin+cortisone acetate, it was possible to achieve "complete regression", i.e., tumors did not recur after treatment was discontinued. Thus, with the most effective regimen, oral heparin (200 U/ml)+cortisone (s.c. 250-tapered dosage), it was possible to obtain "complete regression" in 100% of ovarian sarcomas, 100% of Lewis lung carcinomas and 80% of B-16 melanomas. However, when bladder carcinomas were treated with this regimen, there were no "complete regressions" until heparin was increased to 1000 U/ml, following which 100% of tumors regressed without recurrence. With the less effective regimen, heparin (s.c.)+cortisone (75 mgm), complete regression rate was ovarian, 80%; Lewis lung, 71%; B-16 melanoma 60%; and bladder 0%.

The hexasaccharide heparin fragment of Example 2 was dissolved in saline, 1.5 mg/ml. Three mice bearing implanted ovarian sarcoma were treated with subcutaneous injection of the hexasaccharide fragment twice daily at a dosage of 7 mg/kg and subcutaneous injection of cortisone acetate- tapered dosage. Control mice received either cortisone acetate alone or saline. While control tumors grew progressively, the hexasaccharide+cortisone acetate treated tumors regressed rapidly and were barely visible 4 days later. Their treatment with hexasaccharide was then discontinued, and the tumors reappeared 3-5 days later.

To directly observe avascular tumors during systemic therapy, Lewis lung tumors were implanted in the mouse cornea by the procedure of Muthukkaruppan et al., Science, Vol. 205, 1416-18 (1979) and treatment was begun 24 hours later. Heparin (oral)-cortisone acetate tapered dosage significantly inhibited capillary growth (0.02 mm/day) compared to cortisone acetate alone (0.24 mm/day), heparin alone (0.32 mm/day) or saline (0.23 mm/day). In the presence of heparin-cortisone acetate, a thin plate of tumor remained avascular. Three-dimensional tumor-growth did not occur. In contrast, in the saline controls or when either heparin or cortisone acetate were administered alone, tumors became vascularized and grew as a three-dimensional mass until they eventually perforated the cornea. These large tumors could be regressed to the flat, thin intracorneal phase by the resumption of the heparin-cortisone acetate combination. However, the intracorneal tumor cells could not be eradicated; discontinuation of the heparin-cortisone acetate led to recurrence of a vascularized tumor.

Lung metastases were counted in all animals that died. A .times.6 stereoscope was used. In all control animals, the lungs were heavily studded with metastases from the three types of metastasizing tumors. In contrast, when any combination of heparin+cortisone acetate was used, no metastases were found in mice bearing ovarian sarcoma; only 1 metastasis was found in a mouse bearing Lewis lung carcinoma; and 2 avascular metastases less than 0.1 mm diameter were found in one mouse bearing B-16 melanoma. The nearly complete absence of metastases in heparin+cortisone treated mice was so striking, that the effect can be more readily appreciated by the following expression of data

Furthermore, no lung metastases appeared in any surviving animals that were off treatment.

To exclude the possibility that tumor regression might be caused by direct cytotoxicity, all 4 types of tumor cells were cultured in the presence of 10% serum obtained from mice receiving either heparin, cortisone acetate, heparin-cortisone acetate, or no drug. Heparin-cortisone acetate did not inhibit cell growth, but in fact stimulated it. Furthermore, histological sections shows no evidence of a cytotoxic effect on bone marrow or intestinal mucosa in animals receiving heparin-cortisone acetate.

To exclude the possibility that heparin-cortisone acetate might induce tumor regression by promoting an immune reaction, mice were inoculated with fresh tumor cells at various intervals after they were off treatment. These tumors grew at the same rate as the original implants. Furthermore, if tumor regression was nearly complete and heparin-cortisone acetate was discontinued prematurely, the original tumor resumed its growth. Finally, by stopping and starting treatment, or by using sub-optimal doses of heparin-cortisone, bladder tumors could be maintained at nearly a constant small size (i.e., 45 to 70 mm.sup.3) for periods of more than 8 weeks.

Inflammatory angiogenesis induced by implantation of silica particles into the rabbit cornea, immune angiogenesis induced by implantation of lymph node from a different rabbit, were also completely prevented by the cortisone-heparin pellets. Cortisone by itself temporarily delayed the onset of both types of angiogenesis (compared to an empty pellet), and heparin by itself delayed the onset of immune angiogenesis, but neither alone prevented angiogenesis for an extended period of time as did the cortisone-heparin combination.

A human colon carcinoma was inoculated subcutaneously into nude [athymic] mice, and allowed to grow to a volume of 0.5 cm.sup.3. Controls and treated animals received the same compounds, except that heparin in the drinking water was 1000 U/ml. also an additional treatment group included oral hydrocortisone (0.45 mg/ml) and oral heparin. Animals were housed in cages protected by millipore filter. Tumors grew progressively in all control animals, but regressed in animals treated with heparin and cortisone or heparin and hydrocortisone. The treated tumors were barely palpable after 6 weeks of therapy.

CD-1 "Swiss" mice were used because this strain breeds easily and the fertilized females almost always conceive and subsequently deliver a full litter.

One male was left alone in the cage for 24 hours. His bedding was then changed and two females were added at 5 p.m. The females were then checked at 8 a.m. the next morning for the presence of a cervical plug, which indicates insemination. Sufficient male-female cages were set up to obtain at least 20 pregnant females.

Treatment of the inseminated females was started on the day after insemination by offering drinking water to each of four groups, as follows:

The treatments were continued for only four days, after which female mice were placed one per cage, and followed closely for the presence of offspring. The cages were checked daily for any sign of abortion (fur, fetal remains, etc.).

Groups I, II and IV all produced healthy litters. In Group III there were no mice born and no evidence of abortion. This supports the conclusion that anti-angiogenesis by heparin-cortisone inhibits implantation, presumably by inhibiting capillary growth from the uterus.

17.alpha.,2l-Dihydroxypregna-4,9(11)-diene-3,20-dione (or delta-9(11)-deoxycortisol) (90 mg/ml) and Sigma heparin (1.8 mg/ml) were dissolved in 2 cc of 10% ethylene vinyl acetate (EVA) copolymer. A 1-2 mm pellet was cut from the polymer and implanted into the pocket of a rabbit cornea 1 mm from the limbus. The polymer was positioned between the limbal vessels and a 1 mm.sup.3 piece of V2 Carcinoma implanted distal to the pellet. The opposite eye of each of the 3 rabbits tested contained empty polymers. Mean vessel growth or 0.45 mm/day over a period of 16 days was seen with the control corneas and 0.07 mm/day was seen in the eyes containing steroid and heparin Measurements are expressed as the mean.+-.SD. Similar results have been demonstrated with cortexolone and heparin implanted in the same manner.