A phase II study of bryostatin 1 in metastatic malignant melanoma.

Bryostatin 1 is a protein kinase C partial agonist which has both antineoplastic and immune-stimulatory properties, including the induction of cytokine release and expansion of tumour-specific lymphocyte populations. In phase I studies, tumour responses have been observed in patients with malignant melanoma, lymphoma and ovarian carcinoma. The dose-limiting toxicity is myalgia. Sixteen patients (age 35-76 years, median 57 years) with malignant melanoma were treated. All had received prior chemotherapy. In each cycle of treatment, patients received bryostatin 25 degrees g m(-2) weekly for three courses followed by a rest week. The drug was given in PET diluent (10 microg bryostatin ml(-1) of 60% polyethylene glycol, 30% ethanol, 10% Tween 80) and infused in normal saline over 1 h. The principal toxicities were myalgia (grade 2, eight patients and grade 3, six patients) and grade 2 phlebitis (four patients), fatigue (three patients) and vomiting (one patient). Of 15 patients evaluable for tumour response, 14 developed progressive disease. One patient developed stable disease for 9 months after bryostatin treatment. In conclusion, single-agent bryostatin appears ineffective in the treatment of metastatic melanoma in patients previously treated with chemotherapy. It should, however, be investigated further in previously untreated patients.

w ith several biochemical forms that are membrane associated and phosphory late other downstream proteins (Nishizuka. 1986). PKC isoenzymes are important components of signal transduction in response to growth factors. hormones and tumour-promoting phorbol esters. and are a common pathw ay in the regulation of cell growth by oncoproteins. PKC lex els may be altered in tumour cells (Guillem et al. 1987: O'Brian and Ward. 1989: Barr et al. 1991: Couldwell et al. 1991. and cultured fibroblasts induced to o-erexpress PKC by transfection with PKC cDNA exhibit a transformed phenotype (Housey et al. 1988). Thus. PKC represents a rational target for anti-cancer drug development to block a common pathwax of oncogene activation.
Brvostatin I is a novel anti-cancer dmua derixed from the marine inxvertebrate Bugula neriina (Pettit et al. 1981 ). It is the prototype of a novel class of structural1v related macrocvclic lactones w-hich interact with PKC to affect cellular growth and differentiation. cytokine secretion and stimulation of imnmunocompetent and haemopoietic cells (Berkow and Kraft 1985: Fields et al. 1988: Tuttle et al. 1992: Steube and Drexler. 1995). Brvostatins interact with PKC through the phorbol ester bindincg site. binding with high affinitx and a slow rate of release. Brvostatins induce some of the responses of phorbol esters and antagonize those responses to phorbol esters that they themselves do not induce (Dell'Aquila et al. 1988: Gschwendt et al. 1988: Kennedy et al. 1992: Lewin et al. 1992). The nature of the response to brvostatin is probably a function of the target cell population and PKC isoenzymes. Some isoforms are affected similarly by brvostatin and phorbol esters. and some Received 30 January 1998Revised 20 Apnl 1998Accepted 29 April 1998 Correspondence to: AL Hams differentially (Hocevar and Fields. 1991: Hocexar et al. 1992: Kennedy et al. 1992: Lewin et al. 1992: Szallasi et al. 1994a. 1994b In Xivo. brx-ostatin has anti-tumour activity against B 16 melanoma. M5076 oxarian reticulum cell sarcoma. P388 acute leukaemia and L1OA B-cell lymphoma (Pettit et al. 1982: Schuchter et al. 1991: Homung et al. 1992). In addition. brx ostatin has immunostimulatorproperties that may contnrbute to its in vivo anti-tumour activitx-. It stimulates cytokine release. enhances T-and B-cell actixation and lxmphokine-acti-ated killer (LAK) cell activitv as w ell as neutrophil phagocytic actix ity and degranulation (Berkow and Kraft. 1985: May et al. 1987: Mohr et al. 1987: Drexler et al. 1990: Esa et al. 1990: Tuttle et al. 1992: Scheid et al. 1994: Steube and Drexler. 1995. Three phase I clinical trials of bryostatin hax-e been conducted (Philip et al. 1993: Prendix-ille et al. 1993: Jayson et al. 1995. In the first. brvostatin was administered as a 1 infusion in 60% ethanol evenr 2 weeks for three cycles (Prendiville et al. 1993). Nineteen patients received brvostatin at doses ranging between 5 and 65 .cg m-'. The dose-limiting toxicity w-as myalgia. and the maximum tolerated dose (MTD) w as 35 jg m-'. At 65 jgc m-'. but not at lower doses. there was sinnificant haematological toxicitv. The second trial. undertaken by the same centre. used brn ostatin infused in PET diluent (10 go brvostatin mlrof 60% polyethylene glycol. 30%ethanol. 10% Tween 80) and infused with nonnal sahine oxer 24 h everv week for 8 weeks (Jayson et al. 1995). The MTD w-as 25 jg m-' and arain the dose-limiting toxicity was myalcia. Of 12 patients treated. there w ere three responses: two minor responses in patients with low-grade lvmphoma and a partial response in a patient heavily pre treated for ovarian carcinoma.
In our prexvious phase I study. 35 patients were treated wxith brvostatin. Initially. the druo was dissolx-ed in ethanol and subsequently. in order to reduce the incidence of phlebitis. with PET diluent. Bryostatin was infused oxer 1 h on days 1. 8 and 15 of a 28-day cycle. The MTD u-as 25 jgc m' (Philip et al. 1993). The dose-limiting toxicitvxwas also myalgia. There were two objective responses. both in patients wxith metastatic melanoma: one had a partial response in lung metastasis and the other a partial response in skin metastases.
Immunological mechanisms are implicated in melanoma regression. In viexx of the responses observed in two patients with melanoma in our phase I trial (Philip et al. 1993) and the known immunostimulatorx properties of bryostatin. we undertook a phase II study of the effect of bryostatin 25 jgc m-' given over 1 h on days 1. 8 and 15 of a 28-day cycle in patients x-ith metastatic malignant melanoma.

PATIENTS AND METHODS Eligibility criteria
Eligibilitv criteria for entrv included: histologically or cytologically proven metastatic malignant melanoma w-ith objective exidence of progressive disease. Eastern Cooperative Oncology Group (ECOG) performance status of 0-2. white cell count greater than 3.0 x 10 1-'. platelet count greater than 100 x 109 L-'. normal renal and hepatic function. negative historof cardiac disease. absence of active infection. life expectancy of at least 3 months. presence of measurable or evaluable disease. and informed consent. Patients had not receixed radiotherapy or chemotherapy in the 4 weeks (6 weeks for nitrosoureas or mitomycin C) before commencing, the study. The study was approved by the Central Oxford Research Ethical Committee (COREC). and conducted

Assessment of toxicity
Baseline investigations included full blood count A ith a differential white cell count. serum biochemistry. urinalvsis. chest radiograph and electrocardiogram. Patients were reviewed by a physician weekly. and new signs and symptoms and performance status (ECOG) were documented. At each visit, a full blood count with a differential white cell count. serum biochemistrx and urinals sis were performed. Additional tests w ere performed as appropriate.
World Health Organization (WHO) toxicity criteria were used to grade the toxicity of bryostatin except for mvalgia. w-hich was graded according to the follow'ing scale: grade 0no pain: grade 1 mild short-lived pain not requiring analgesics: grade 2 moderately sexvere pain but patients remained ambulators w ith irreaular analgesic intake: grade 3 -moderate to sexvere pain which significantlv affected ambulation and required regular analgesia (nonopiate): and grade 4 -verv severe incapacitating pain necessitatine constant bed rest and regular opiates.   Assessment of tumour response Ev-aluable and measurable disease sites w-ere assessed before entering the study by phy sical examination. plain radiography and computerized tomography where appropriate. and repeated evenr t-wo cycles. Physical examination was repeated w eekly and imaging investigations for the purposes of tumour measurement were repeated after two cycles of treatment or at the time of suspected disease progression. Standard W'HO criteria for objectix e response assessment vvere employed. Partial response was defined as a 50%7 or greater reduction in the sum of the products of the largest perpendicular diameters of all measurable disease sites. Progressive disease wxas indicated by a greater than 25% increase in the size of at least one measurable lesion. or the appearance of a new lesion. Stable disease was defined as an increase in disease measurements of less than 25% or a decrease by less than 50%. Patients with progressixe disease w ere withdrawn from the study.

Patients
Sixteen patients (eight men. eicht w-omen: age range 35-76 y ears) were recruited to the study. All patients had prexiously receixed chemotherapy w-ith dacarbazine (DTIC) given as I g m-oxver I h once even 3 w-eeks. and four had in addition receixed radiotherapy. One of the patients A as treated A ith DTIC in combination u-ith BCNU (carmustine). cisplatin and tamoxifen after dex eloping progressix e disease on DTIC. This regimen is further detailed elsewhere (Del Prete et al. 1984). In 14 patients. there had been disease progression in response to chemotherapy. one patient achieved stable disease. and one v.-as not exaluable because the chemotherapy x-as given in the adjuxvant setting. Their characteristics are show-n in Table 1. STATISTICS To ensure a low probability of erroneously rejecting, a treatment that is active in 20% of patients. at least 14 patients were treated. according to previously described principles (Gehan. 1961).
The toxicities associated wxith treatment are shown in Tables 3 and   4. Eight patients developed grade 2 myralgia and six had grade 3 myralaia. In aeneral. the myalgia orsened ith each course of bryostatin (Table 4). Studies in xixo haxe suaaested that this mvalgia may be caused by impairment of oxidative metabolism. possibly as a result of x-asoconstriction. In an attempt to reverse xvasoconstriction. six patients were treated wxith nifedipine but this was ineffectixe. as prexviously reported (Thompson et al. 1996). Apart from my algia. the incidence of sexere toxicitw as lowx. Of note. there w as no significant biochemical or haematolocical toxicirx.

DISCUSSION
In this studx of 15 exaluable patients treated xxith brvostatin. only one patient. whose disease stabilized. obtained significant benefit from the drug. Apart from this patient. the remainder were xithdrax n from treatment because of disease progression and. in sexven patients. this occurred ithin 1 month of starting bn ostatin. Hence. single-acent bry-ostatin. gixven by this formulation at a dose of 25 go m-' administered oxver I h weekly for 3 wxeeks of a 4wxeek cycle. is not an effectixe therapy for metastatic melanoma in patients prexiously treated ith chemotherapy. In our prexious phase I study of bryostatin. disease responses wxere observed in two patients ith melanoma (Phihip et al. 1993). and there are theoretical reasons hy bn ostatin might be construed as a potentially effectixe therapy for melanoma. It has anti-tumour effects in murine models of melanoma and on melanoma cell lines (Schuchter et al. 1991: Szallasi et al. 1996. Immunolocgical mechanisms are implicated in the regression of melanoma. and br ostatin stimulates cvtokine release and augrments specific anti-tumour immunitx (Mohr et al. 1987: Tuttle et al. 1992: Steube and Drexler. 1995. There are sex eral reasons w-hich might explain the drug's lack of clinical effect in this phase H study. Animal data suggest that the  (Berkowet al. 1993: Zhang et al. 1996. and its anti-tumour effects are potentiated w-hen given over a prolonged penrod (Homung et al. 1992). Indeed. in a previous phase I trial (Jayson et al. 1995): when the druc was given as a 24-h infusion. partial tumour responses were observed using the same drug dose as aiven in the current study. but. when the same or higher doses were Diven over 1 h. no tumour responses were observed (Prendiville et al. 1993). In this latter studv. however. the drug was administered on a 2 weekly. rather than a wveekly schedule. Nevertheless. we have observed partial disease responses in tWo patients w-ith melanoma who were treated with brvostatin bv 1 h bolus infusion at the same dose and by the same weekly schedule as used in the current study (Philip et al. 1993).
Expression of PKC isotypes in tumours varies (Guillem et al. 1987: O'Brian and Ward. 1989: Barr et al. 1991: Couldwell et al. 1991. and not all are equally down-regulated by bryostatin (Szallasi et al. 1994b). PKC isoenzvmes are involved in both oncogene and tumour-suppressor gene acti-ation and could have opposing effects on tumour growth. dependent on tumour type.
Hence the clinical effects of brvostatin are likely to be complex. Toxicity in this study. apart from mvalaia. was low. Phlebitis. a prevalent feature w-hen the drug w-as dissolved in ethanol (Philip et al. 1993). wAas minimized by the PET diluent. Myalgia was a prominent feature in the patients studied here. The myalgia occurred within 2-3 days of drug administration and lasted 3-5 days. As in previous studies. it worsened incrementallyA with further courses of brvostatin. Its aetiolo5v is unknown. and appears to be a direct drug effect on muscle (Hickman et al. 1995: Thompson et al. 1996. Other studies have shown that the myalgia is not reversed by nifedipine. a drug that does reverse brvostatininduced vasoconstriction (Thompson et al. 1996 . Corticosteroids. non-steroidal anti-inflammatory drugs and a variety of analgresics. including morphine. have not proven effective in the treatment of this toxicitv. Further assessment of the aetiologv of the myalaia and development of methods of reversing it could allow higher doses than used here to be administered which would perhaps attain therapeutic levels. Some brvostatin analogues have been shown. in murine models. to have equal anti-tumour effects but less toxicity than brvostatin 1 . Furthermore. there is evidence that the toxicity. but not the anti-tumour effects of some bryostatin analogues. is mediated by direct interaction with PKC (Szallasi et al. 1996).
Hence. clinical testing of brvostatin analogues may be of value.
All patients in this study had previously received chemotherapy. and four had received radiotherapy. Therefore. it is likely that there was significant suppression in lymphocyte function. which could have reduced the possibility of brvostatin actinc by immunemediated mechanisms. Further studies are indicated in chemotherapy naive patients.
In vitro. bryostatin potentiates cvtotoxic agent activity (Basu and Lazo. 1992: Mohammad et al. 1995). Also. although there is conflicting evidence. bryostatin may act as a multidruc, resistance (MDR) modulator (Kamanda et al. 1994: Scala et al. 1995. The lack of significant myelotoxcity in this and previous phase I studies indicates that the drugr could be nriven in combination with cytotoxic agrents. In peripheral blood lymphocytes (PBLs) obtained from patients receiving brvostatin. LAK cell generation and proliferation were enhanced following in vitro stimulation with interleukin 2 (IL-2) w-hen compared Awith PBLs obtained from healthy control subjects. In conjunction w-ith 1L-2. brxrostatin up-regulated IL-' receptor expression and augmented cvtotoxic T-lymphocyte (CTL) numbers in vitro (Scheid et al. 1994). Hence. brvostatin and cytokine therapy may be synergistic.
The findincs of this study showv that brvostatin is not effective as a single agent in metastatic malicnant melanoma. Experimental evidence suggests that it may warrant further study in combination with cvtotoxic or biological agents.