Paclitaxel Couplets With Cyclophosphamide or Cisplatin in Metastatic Breast Cancer

Anthony W. Tolcher

Determining active combinations containing paclitaxel (Taxol; Bristol-Myers Squibb Company, Princeton, NJ) to treat metastatic breast cancer has been the focus of recent clinical development. Paclitaxel combined with eithe r cyclophosphamide or cisplatin has several poten­tial advantages: cisplatin and cyclophosphamide are ac­tive single agents against previously untreated meta· static breast cancer, colony-stimulating factors can modulate overlapping toxicities like myelosuppression, and no mechanisms of cross-resistance between pacli­taxel and these agents are yet known. Major questions include the optimal schedule of administration and the sequence dependence of toxicities with these combina­t ions. Paclitaxel schedules with cisplatin include either two dose levels using the 24.hour infusion o r a novel biweekly l·hour infusion. The sequence in the three available studies was paclitaxel followed by cisplatin. Hematologic toxicities were dose limiting with the bi­weekly and low-dose 24-hour pacl itaxel/cisplatin com· binations; with granulocyte colony-stimulating factor, neurotoxicity became a prominent cumulative toxicity of the high-dose paclit axel/cisplatin combination. Re· sponse rates in the first-line treatment of metastatic breast cancer ranged from 49% to 85%. In the three completed studies with cyclophosphamide, paclitaxel has been administered over e ithe r 72, 24, or 3 hours. Paclitaxel followed by cyclophosphamide had greater he matologic toxicity than the opposite schedule or con­current administration. Pharmacokinetic factors do not see m to account for this sequence-dependent toxic· ity. As expected, dose-limiting toxicity in all studies has been hematologic. However, granulocyte colony-stim­ulating factor has ameliorated myelosuppression and allowed considerable dose escalation of cyclophospha· mide. This combination has demonstrated activity in previously treated patients with m etastatic breast can­cer, including the anthracycline-refractory subpopula· t ion that will be reviewed. Copyright © 1996 by W.B. Saunders Company

PACUTAXEL (Taxol; Bristol-Myers Squibb Company, Princeton, NJ) is an active single

agent in the treatment of metastatic breast cancer with reported response rates up to 62% in first-line therapy and up to 30% in extensively pretreated patients. H The logical progression for the clinical development of this agent includes combination studies of paclitaxel and other drugs that possess proven activity against breast cancer. Cyclophos­phamide and cisplatin are two such agents: both have proven activity in rhe treatment of breast cancer and neither has any known mechanisms of cross-resistance with paclitaxel.

Seminors in Oncology, Vol 23. No I. Suppl I (February). 1996: pp 37-43

A lthough seldom used in treating breast cancer, cisplatin is active, with single-agent response rates of 47% to 54%5

·6 as first-line therapy for metastatic

breast cancer. A paclitaxcl/cisplatin combination has several potential advantages: patients generally have nor received adjuvanr cisplatin and, with the exception of peripheral neuropathy, overlapping toxicities for the two drugs are uncommon and are not usually hematologic.

Cyclophosphamide is the standard alkylating agent used in both the adjuvant and palliative treat­ment of breast cancer. Considerable cl inical experi­ence has been ga ined from its past 30 years of use against this disease. Single-agent activity, reported from the earliest clinica l trials, ranged from 25% to 45% in patients with advanced disease.7

'8 Myelosup­

pression, particularly neutropenia, is the major over­lapping toxicity of cyclophosphamide and pacli­taxel; this can now be diminished with the use of colony-stimulating factors.

Combining two chemotherapy agems with dis­tinctly different mechanisms of action and charac­teristics into a couplet represents a challenge. The optimal dose and schedule of both agents in combi­nation needs to be determined, which requires bal­ancing the combined, and at times unexpected, clinical toxicities against the need for maximum cytotoxicity from each agent. Similarly, the optimal sequence of administration may need tO be deter­mined. Potential sequence-dependent toxicities and cytotoxic activity have been associated with other chemotherapy combinations in clinical oncol­ogy.9·t0 If sequence dependence is clinically signifi­cant, phase I evaluation must determine whether pharmacokinetic or biologic factors are responsible.

PACLITAXEL/CISPLATIN COUPLET

Preclinical and Clinical Sequence Rationale

Paclitaxel/cisplatin combinations have demon­strated additive and synergistic cytotoxicity in

From rhe British Columbia Cancer Agency, Uni••ersity of British Columbia, Vancouver, BC, Canada.

Addre.1.1 reprint r~quests co Anthony W. Tolcher, MD , FRCPC, Bricish Columbia Cancer Agi'llC)', University of British Columbia, 600 \~ 10th Ave, Vancouver, BC V7C IY4 , Canada.

Copyright© 1996 by W .B. Saunders Company 0093-7754196/2301-0113$05.0010

37

1 of 7 Celltrion, Inc., Exhibit 1013

38

Table I. Paclitaxel/Clsplatin In Breast Cancer:

Dose and Schedules

Investigator Dose (mg/m'}nnfusioo Length

Low dose BU Paclicaxel I 35121 hr

Cfsplatin 75

High dose NYU Paclicaxel 200121 hr

Cisplatin 75

I G·CSF

Biweekly BCCA Paclicaxel 90/3 hr

Cispladn 60

Abbreviations: BU. Brown University; NYU. New York Uni-

versity: BCCA. British Columbia Cancer Agency.

some cell lines in vitro.it.i i Sequence analysis of this combination has been studied in vitro by many investigators. n·14 Paclitaxel followed by cisplatin appears to produce maximal cytotoxicity in many cell lines, whereas the reverse sequence has dem­onstrated antagonism_ l).l 4 Paclitaxel is a cell cycle phase-specific agent, and perturbations of the cell cycle by cisplatin have been hypothesized as a bio­logic explanation for this sequence-dependent an­tagonism.14·1 5

In an elegant phase I and pharmacokinetic study, Rowinsky et al examined the sequence-de­pendent clinical toxicities of the paclitaxel/cis­plarin combination. In this study, cisplatin admin­istered before a 24-hour paclitaxel infusion was associated with significantly increased granu locyte toxicity compared with the reverse sequence of paclitaxel fol lowed by c isplatin .16 Pharmacokinetic analysis of these sequences suggested that adminis­tering cisplatin first significantly delayed clearance of the myelosuppressive agent paclitaxel compared with the alternate sequence and that the delayed clearance was responsible for the increased granu­locyre toxicity. Based on this compelling labora­tory and clinical evidence, paclitaxel followed by cisplatin has become the standard sequence of ad­ministration in clinical investigation.

Clinical Scudies

Interim results from three phase 11 studies are available for analysis. Table 1 lists the doses and schedules, which are illustrated in Fig 1.

New York University Medical Cencer trial. Wasserheit et al studied paclitaxd 200 mg/m2 by 24-hour infusion followed immediately by cisplatin

ANTHONY W. TOLCHER

P200mg/m2 ,

G-CSF 22 NYU - ---------------+- -C 75mg/m2

t t , 22 BU P135mg/m2 - -C7Smgtm2 t t , 15 BCCA P90mg/m2 • • C60mg/m2 t t

Fig I. The schedule and sequence of the three phase II stud­

ies of paclitaxel (P) in combination with cisplatin (C). N YU,

New York University; BU, Brown University; BCCA, British

Columbia Cancer Agency; G·CSF, granulocyt e colony-stimulat­lng factor.

75 mg/m2 every 3 weeks (C. Wasserheit, MD, prin­cipal investigator, unpublished data, July 1995). Based on previous clinical trial experience with paclitaxel in ovarian cancer and, as a single agent, in breast cancer, neutropenia was anticipated to

be dose limiting. Thus, granulocyte colony-stimu­lating facror (G-CSF) was administered to all pa­tients following chemotherapy.

Open to women with metastatic breast cancer the study has treated 44 patients. Prior adjuvan; chemotherapy was permitred and three patients had received prior chemotherapy for metastatic disease. Twenty-three patients had prior adjuvanr chemotherapy and 19 had received anchracyclines. Tahle 2 summarizes patient characteristics.

Although G-CSF modified hematologic toxic­ity, instances of febrile neutropenia and one toxic death were recorded. Neuropathy was a prominent nonhemaLOlogic and dose-limitingcumulacive tox­icity in chis study. Peripheral neuropathy grade 2 o r greater occurred in 88% of patients over a me­dian of five cycles. Periphera l neuropathy was dose limiting in the phase I study of pacliraxel/cisplatin and G-CSF due to rhe G-CSF-relaced ameliora­t ion of the usually dose-limiting neutropenia and

Table 2. PaclitaxeVClsplatln in Breast Cancer:

Patient Characteristics

NYU BU BCCA

No. of patients 44 16 29 Median age (yr) 49 53 47

Median no. of disease siles 2 2 2

No. of patients with prior chemotherapy 23 13 27 Anthracycline containing 19 12 23

Abbreviations- NYU. New York University: BU. Brown Uni-

versity; BCCA, British Columbia Cancer Agency.

I

2 of 7 Celltrion, Inc., Exhibit 1013

PACUTAXEL W ITH CJSPlATIN OR CYCLOPHOSPHAMIDE

the consequent increased dose delivery of these two neurotoxic agents. 17

This schedule of paclitaxel/cisplatin at these doses was active, with five complete responses (CRs) and 15 partial responses (PRs) of 41 evalu­able patients, for an overall response rate of 49%.

Brown University Hospitals study. Browne and colleagues initiated a phase II study of paclitaxel 135 mg/m2 by 24-hour infusion followed immedi­ately by cisplatin 75 mg/m2 in women with meta­static breast cancer (M. Browne, MD, principal investigator, unpublished data, July 1995). The doses and schedule were derived from phase I stud­ies in ovarian cancer. Granulocyte colony-stimu­lating factor was used only as indicated clinically. Only patients with untreated metastatic breast cancer were eligible, although prior adjuvant che­motherapy was allowed. Sixteen patients have been entered to date, 13 of whom had received prior adjuvant chemotherapy.Twelve patients' ad­juvant therapy contained anthracycline. Table 2 summarizes patient characteristics. The largely he­matologic clinical toxicities have been tolerable. Nonhematologic toxicities have included alopecia, rash, nausea and vomiting, and fatigue, a ll which were anticipated in a cisplatin-containing regimen. Peripheral neuropathy has been tolerable and con­sistent in frequency with the incidence reported in ovarian cancer trials.

Two CRs and six PRs have been observed with this pacliraxel/cisplarin dose and schedule in 15 patients evaluable for response, for an interim re­sponse rare of 53%. This level of activity is encour­aging considering that most of the women had received anthracyclines.

British Columbia Cancer Agency trial. This phase I/II study (K. Gelmon, MD, principal inves­tigator, unpublished data, Ju ly 1995) was planned to deliver biweekly pacliraxel by 3-hour infusion followed immediately by cisplatin. The biweekly schedule of paclitaxel combined wirh cisplatin was based on the results of the European and Canadian study of single-agent paclitaxel in refractory ovar­ian carcinoma, which compared the 3- and 24-hour infusion schedules at dose levels of 135 and 175 mg/m2•

18 In this comparative study, the major­ity of patients who received paclitaxel by the 3-hour infusion at both dose levels had neutrophil recovery to greater than 1,500/ µL by day 15, sug­gesting that biweekly administration of a 3-hour paclitaxel infusion was possible. The paclitaxel/

39

cisplatin combination in a biweekly schedule was hypothesized to be tolerable due to the modest myelosuppression associated with cisplatin. The schedules and sequence of the agents are demon­strated in Fig l.

The starting dose level in the phase I portion of the study in metastatic breast cancer cou ld not be escalated, however, because the patients' granu­locyte counts had not recovered to greater than 750/µL by day 14. The phase !I dose level was thus determined to be paclitaxel 90 mg/m2 by 3-hour infusion followed by cisplatin 60 mg/m2 adminis­tered every 2 weeks. The use of G-CSF was not permitted. The study was open to patients with chemotherapy-naive metastatic breast cancer, al­though prior adjuvant chemotherapy was permissi­ble. Twenty-nine patients were entered, 27 of whom were evaluable for response in the phase II portion. Twenty-seven patients had received prior adjuvant chemotherapy, 23 with anthracycline­containing regimens. Table 2 summarizes patient characteristics.

Hematologic toxicity was common, although brief grade 4 neutropenia, median 2 days' duration (range, 0 to 7), was observed in only 34% of pa­tients. One patient was admitted for febrile neutro­penia and bacteremia was documented. During the septic episode, which responded to antibiotic man­agement, this same patient experienced the only case of grade 4 thrombocytopenia.

Mild nonhematologic toxicities were common. Alopecia, nausea and vomiting, and fatigue, al­though frequent with this combination, were gen­erally not dose limiting. Arthralgia and myalgia occurred in 41 % of patients but were dose limiting in only one. Mild sensory changes were observed, but only one patient experienced grade 3/4 periph­eral neuropathy with this schedu le.

This combination and schedule have been asso­ciated with a provocative level of activity. Three patients have achieved CRs and 20 achieved PRs, for an overall response rate of 85% in a patient population exposed to prior adjuvant, mostly an­thracycline-based, chemotherapy. A confirmatory phase II study is presently being performed.

Summary of Paclitaxel/Cisplatin Couplet Trials

The pacliraxel/cisplarin combination has dem­onstrated an encouraging level of antitumor activ­ity in women with metastatic breast cancer and has an acceptable level of toxicity. These response

3 of 7 Celltrion, Inc., Exhibit 1013

40

rates occurred despite previous adjuvant chemo­therapy, in many cases with anchracyclines. The use of G-CSF permits higher doses of the drugs to be delivered, but peripheral neuropathy then becomes a prominent dose-limiting toxicity.

The differenc response rates in these phase II trials should be interpreted with caution. Results with the biweekly schedule generate several ques· tions. Is the antitumor activity schedule depen­dent? A phase I/II study of biweekly paclitaxel alone is presently under way to determine whether frequent dosing of the 3-hour paclitaxel infusion alone is associated with a high level of activity. An alternate explanation for the high level of anti· tumor response includes the different cisplatin dose intensity. Nonetheless, the actual difference in cisplatin dose intensity between the studies is quite small (30 v 25 mg/m2/wk). The biweekly schedule also allowed treatment on day 15 if the granulocyte count was at least 750/µL and the platelets had recovered to at least 75,000/ µL. Given this same level of hematologic recovery, would a similar level of activity be observed with the other regimens if treatment was also initiated earlier than day 21? It is hoped that some of these questions will be answered should results of the confirmatory study support the biweekly paclitaxel/ cisplatin schedule.

PACLIT AXEU CYCLOPHOSPHAMIDE COUPLET

Preclinical RatioTlllle

Combinations of paclitaxel and alkylating agents have demonstrated additive but also antago­nistic cyrotoxicity in vitro. Preclinical pacl itaxel/ cyclophosphamide sequence data were not avail­able before phase l tria ls of this combination be­gan. Recently, however, in vitro results with some cell lines have demonstrated that the sequence of an alkylating agent before paclitaxel is associated with reduced cell kill compared with the reverse sequence, much like the in vi tro results with the cisplatin/paclitaxel combination. 14 This finding was not universal, however; other cell lines did not demonstrate sequence dependence. 14

•19

Phase I Clinical Studies

Results of three phase l stud ies are mature. The dose and schedules used in each of these studies

ANTHONY W. TOLCHER

are summarized in Table 3 and illustrated in Fig 2.

Johns Hopkins Oncology Center trial. Kennedy et al initiated a phase I study of paclitaxel/cyclo­phosphamide with G-CSF in patients with anthra­cycline-refractory metastatic breast carcinoma. 1

'.I

Table 4 briefly summarizes the patient characteris­tics.

Paclitaxel was administered by a 24-hour infu­sion with intravenous bolus cyclophosphamide (Fig 2). The elegant clinical trial design incorpo­rated sequence analysis of toxicity for al l dose lev­els when possible. All patients received both drug administration sequences. The use of G-CSF per­mitted considerable dose escalation of this combi­nation, with nine dose levels (paclitaxel l35 to 200 mg/m2 and cyclophosphamide 750 to 2,000 mg/m2

). Dose-limiting toxicity was neutropenia (one patient) and typhlitis (one patient) in two of four patients. The ma,'<imum tolerated dose was considered to be paclitaxel 200 mg/m2 and cyclo­phosphamide 1,600 mg/m2 while the recom­mended doses for phase II investigation were 200 and 1,250 mg/m2

, respectively. Dose-limiting thrombocytopenia was uncommon, occurring in on ly two patients. Myalgias and neuropachy were uncommon nonhemarologic dose-limiting toxici­ties, occurring in only five patients. The dose-lim­iting toxicity of typhlitis with this combination and schedule had been observed in a phase l study of the paclitaxel/doxorubicin combination and

Table 3. Padit:axel/Cyclophosphamide in Breast Canc,er

Dose (mg/m'}/lnfusion Length

NCI

Pacliaxel 160172 hr

Cyclophosphamide 2,700 (divided dose)

G-CSF

JHOC

Pacliaxel 200124 hr

Cyclophosphamide 1,600

G-CSF

Bell inzona

Pacliaxel 20013 hr

Cyclophosphamide 1,250

Abbreviations: NCI, National Cancer Institute: JHOC, Johns

Hopkins Oncology Center. Bellinzona, Ospedale San Giovanni.

Bellinzona, Switzerland.

4 of 7 Celltrion, Inc., Exhibit 1013

PACLITAXEL WITH CISPLATIN OR CYCLOPHOSPHAMIDE

may represent a unique, albeit an uncommon, tox­icity for paclitaxel combinations.20

Sequence-dependent toxicities were observed. Paclitaxel fo llowed by cyclophosphamide was asso­ciated with greater neutrophil and platelet tox icity compared with the alternate sequence, as con­firmed by the rates of admission for febrile neutro­penia (27% v 11 %, respectively). The pharmacol­ogy of these agents did not differ significantly with either sequence. Therefore, and in contrast to the paclitaxel/cisplatin data, a ltered pharmacokinetics does not explain the observed d ifference in hema­tologic toxicity, and a biologic explanation for se­quence-dependent cytotoxicity may apply.

ln this phase I study of heavily pretreated, doxo­rubicin-refractory patients, the overall response rate of 28% demonstrated encouraging activity. Phase 11 data will clarify the true activity of this schedule and combination .

Currently, the Johns Hopkins group is evaluat­ing the same combination using the 3-hour pacli­taxel infusion schedu le with G-CSF.

Ospedale San Giovanni, Bellinzona, phase I trial. Sessa and Pagani (0. Pagani, MD, principa l inves­tigator, unpublished data, July 1995) and their col­leagues initiated a phase I study of pacliraxel by 3-hour infusion combined with cyclophosphamide. The characteristics of the 43 patients are summa­rized in Table 4. Dose levels of paclitaxel 175 to

200 mg/m2 and cyclophosphamide 750 to 1,250 mg/m2 without G-CSF were explored. Further dose escalation ro 1,750 mg/m2 cyclophosphamide was possible with the use of G-CSF. Grade 4 neutro­penia for greater than 7 days limited further dose escalation. Nonhemarologic toxicities were nor dose limiting, and ryphlit is was not observed. Se­quence-dependent toxicity analysis is planned but nor yet avai lable. However, caution is warranted

1 2 3 NCI P 160 mgfm2 72 hr G-CSF

C 2,700 mg/m2 t ---------.-

t t

P 200 mg/m2 24 hr 1 G-CSF JHOC --------------·

C 1,600 mg/m2

t

Bellinzona P 200 mgtm2 3 hr • G-CSF -------------· C 1,250 mg/m2

t

Fig 2. T he schedule and sequence of the three phase I stud­ies of paclitaxel (P) in combination with cyclophosphamide (C). N CI, N ational Cancer Institute; )HOC, Johns H opkins Oncol­ogy Cente,r.

41

Table 4. Paclitaxel/Cyclophosphamlde In Breast

Cancer: Patient Characterist ics

NCI JHOC Bell

No. of patients 55 37 43

Median age (yr) 45 so 5 1 Median no. of disease sites 2 2 2

No. of patients with prior chemotherapy 53 37 36

Median no. o f prior chemotherapy

regimens 2 2 -Anthracycline-containing 47 37 26

Anthracycline-resistant 19 29 14

in interpreting the existence of sequence effects for the shorter paclitaxel infusion schedule. Sequence dependence may be less apparent with concurrent or near-concurrent administration of paclitaxel and cyclophosphamide. With such schedu les, the actual toxicities observed may reflect not the order of administration but, rather, the relatively differ­ent pharmacokinetic half-lives of the two drugs and the sequence that this engenders.

Preliminary activity data demonstrate objective responses including CRs at multiple dose levels in th is phase I study.

Medicine Branch, National Cancer lnscicute phase I study. This phase I study (A. T olcher, MD, prin­cipal investigator, unpublished data, July 1995) was initiated to determine the tolerability of ad­ministering paclitaxel by 72-hour continuous infu­sion (CIVI) in combination with high-dose cyclo­phosphamide and G-CSF in the ambulatory setting. In vitro data suggest that paclitaxel is a cell cycle phase-specific cytotoxic agent. Cytotox­icity in cell lines in vitro reaches a plateau at a threshold paclitaxel concentration beyond which increased pacliraxel concentrations do nor yield greater cell kill; however, prolonging the duration of drug exposure over multiple drug concentrations to a maximum of 72 hours has increased cell kill further in some assays.2

1.22 In the clinical setting,

prolonged infusions of paclitaxel can theoretica lly increase the number of cancer cells exposed to pacliraxel during the G 2/M phase of the cell cycle, in which cells are most vulnerable to this agent's cytotoxic effects. Alkylating agents like cyclophos­phamide are associated with steep dose-response relationships in vitro and considerable dose escala­tion of this agent can be achieved clinically with the use of G-CSF. Divided-dose cyclophosphamide

5 of 7 Celltrion, Inc., Exhibit 1013

42

theoretically permits ambulatory administration of high doses of this agent, while reducing urotoxic­ity, emesis, and the risk of cardiotoxicity. The schedule of administration for this study appears in Fig 2. This phase I study included 55 patients, 54 of whom are evaluable for toxicity. The patient characteristics are summarized in Table 4. Forty­seven patients had already received doxorubicin and 52 had received cyclophosphamide. The study was open to patients with metastatic breast cancer who had received up to two prior chemotherapy regimens (one adjuvant, one for metastasis) and the median number of prior regimens was two. Dose escalation was completed over 10 levels from a starting dose of paclitaxel 135 mg/m1 C IVl and cyclophosphamide 600 mg/m2 to paclitaxel 160 mg/m2 C IVI and cyclophosphamide 3,300 mg/m2

•

Dose-limiting toxicity was reached at the latter dose level, with two patients experiencing grade 4 neutropenia for more than 5 days and one patient experiencing both this level of grade 4 neutropenia and platelets less than 20,000/ µL. The maximum tolerated dose was therefore paclitaxel 160 mg/m1

C IVI over 72 hours and cyclophosphamide 2,700 mg/m1 in three divided doses. Febrile neutropenia commonly complicated this dose-intense regimen, causing at least one hospitalization in 86% of pa­tients; one patient died of a septic complication.

Dose-limiting nonhematologic toxicities were uncommon. Diarrhea was a cumu lative toxicity, occurring in six patients, and may have been re­lated to the use of broad-spectrum antibiotics for febrile neutropenia; in three patients with Clostrid­ium difficile, toxin was detected. Six patients devel­oped gross hematuria requiring dose reduction and increased hydration. In those patients whose he­maturia recurred, oral mesna prevented further bleeding and permitted ambulatory treatment to be continued. Despite a premedication regimen of dexamethasone, diphenhydramine, and cimeti­dine, five patients experienced grade 1 or 2 acute hypersensitivity reactions to paclitaxel. Four of these patients cou ld be re-treated; the o ther re­fused further therapy. No episodes of typhlitis were observed with this paclitaxel/cyclophosphamide schedule.

A pilot study of the sequence of paclitaxel by 72-hour CIVI followed by a single bolus dose of cyclophosphamide was initiated to determine whether the sequential schedule was tolerable and comparable with the concurrenc schedule (A.

ANTHONY W. TOLCHER

Tolcher, MD, principal investigator, unpublished data, July 1995). Dose-limiting toxicity was ob­served in three patients at the first dose level of paclitaxel 160 mg/m2 C IVJ followed by cyclophos­phamide 1,600 mg/m2 and G-CSF. The sequential schedule was associated with intolerable hemato­logic and gastrointestinal toxicity (grade 4 mucosi­tis), despite a cyclophosphamide dose that was only 60% of the previous maximum tolerated dose. These results suggested that the sequence of pacli­taxel fo llowed by cyclophosphamide was more toxic than the concurrent schedule of paclitaxel and divided-dose cyclophosphamide. Since data from Kennedy at al 19 demonstrated a similar find­ing and no alterations in pharmacology were ob­served with these two agencs, one can hypothesize that a perturbation of paclitaxel's cell cycle-de­pendent cytotoxicity may occur to the hemato­logic and gastrointestinal progenitor cells with the concurrent schedule, which is not seen when pacli­taxel is fo llowed by cyclophosphamide. The pre­clinical data suggest that this reduction in normal tissue toxicity may result from a reduction in the number of progenitor cells in the G 2/M phase, where paclitaxel exerts its greatest cytotoxic ity. 14

The question remains whether the schedu le and sequence of cyclophosphamide also a lter rumor re­sponsiveness to this couplet.

In this phase I study using a spectrum of dose levels, the responses of 44 patients with measurable disease were assessed: 23 patients achieved a PR and one patient a CR, for an overall response rate of 55%. In the doxorubicin-refractory subset, seven (54%) of 13 patients evaluable for response ob­tained an objective response (one CR and six PRs). In this heavily pretreated popu lation, this response rate is encouraging.

Summary of Pacliraxel/Cyclophosphamide Couplet Trials

In these three phase I studies, the paclitaxel/ cyclophosphamide combination has demonstrated both an acceptable level of toxicity and encourag­ing anritumor activity in metastatic breast cancer patients previously treated with doxorubicin and/ or cyclophosphamide. Sequence analysis suggests that paclitaxel followed by cyclophosphamide is associated with greater hematologic toxicity com­pared with either the opposite sequence or with concurrent adm inistration. Based on the Johns Hopkins data, 19 altered pharmacokinetics do not

6 of 7 Celltrion, Inc., Exhibit 1013

PACLITAXEL WITH CISPLATIN OR CYCLOPHOSPHAMIDE

cause this sequence dependence. Whether the ad­ministration sequence of paclitaxel/cyclophospha­mide wi ll affect amitumor activity, as it does nor­mal organ toxicity, has yet to be determined. Phase 11 assessments of the role of this couplet, and the schedule and sequence effects, are necessary to de­termine the place of this combination in the treat­ment of both early and advanced breast cancer.

REFERENCES

1. Holmes F, Walcers RS, Theriault RL, et al: Phase ll trial of Taxol. an active drug in the trcatmenc of metastatic breast cancer. J Natl Cancer Inst 83:1797-1805, 1991

l. Reichman BS, Seidman AD. CrownJPA, er a l: Paclitaxel and recombinant human granulocyte colony-stimulating factor as initial chemotherapy for metastatic breast cancer. J Clin Oncol 11:1943-1951. 1993

3. Vermorken JB. Huizing MT. Liefting AJM, er al: High­dose Taxol (HOT) with G-CSF in patients with advanced breast cancer (ABC) refractory to anthracycline (ANT) ther­apy. Eur J Cancer 29A:8.3, 1993 (suppl 6) (abstr 435)

4. Nabholtz JM. Gelmon K, Bontenbal M. et al: Randomized trial of two doses of Taxol in metastatic breast cancer. Proc Am Soc Clin Oncol 12:60. 1993 (abstr 42)

5. Kolaric K, Roth A: Phase II clinical trial of cisdichlorodi­ammine platinum (cis-DDP) for antitumorigenic activity in previously unrreared patients with metastatic breast cancer. Cancer Chemother Pham1acol 11:108-112, 1983

6. S ledge OW, Loehrer PJ, Roth BJ, et al: Cisplatin as first· line therapy for metastatic breast cancer. J Clin Oncol 6:181 l-1814, 1988

7. Joslin CAF, Kunkler PB. Howell Evans I. et a l: Cyclo­phosphamide in the management of advanced breasc cancer. Proc R Soc Med 63:81-84, 1969

8. Mouridsen J, Palshof T, Brahm M, et al: Evaluation of single-drug versus multiple-drug chemocherapy in the rrear­menr of advanced breast cancer. Cancer Treat Rep 61:47-50, 1977

9. Adel AL, Dorr RT, Liddil JD: The effect of anticancer drug sequence in experimental combination chemotherapy. Cancer Invest ll:IS-24, 1993

IO. Maksymiuk AW, Jen JR, Earle JD, et al: Sequencing

43

and schedule effec!.S of cisplatin plus ecoposide in small cell lung cancer: Results of a Norch Central Cancer Treatment Group randomized clinical trial. J Clin Oneal 12:70-76, 1994

11. Rose WC: Taxol: A review of its preclinical in vivo antitumor activity. Anticancer Drugs 3:311-321, 1992

12. Jekunen A, Christen R, Shalinsky D, et al: Synergistic intemction between cisplatin and Taxol in human ovarian car­cinoma cells in vitro. Proc Am Assoc Cancer Res 34:AI 773, 1993 (abstr)

13. Rowinsky EK, C itardi MJ, Noe DA, et al: Sequence­dependent cytotoxic effects due m combinations of cisplatin and the ancimicrotubule agents Taxol and vincristine. J Cancer Res Clin Oncol l 19:727-733, 1993

14. Liebmann]£, Fisher J, Teaugue D. et al: Sequence de­pendence of paclitaxel (Taxol) combined with cisplatin or alky­lators in human cancer cells. Oneal Res 6:25-31, 1994

15. Vanhoefer U, Hamrick A, Wilke H, er al: Schedule­dependent antagonism of paclitaxel and cisplatin in human gastric and ovarian carcinoma cell lines in vitro. Eur J Cancer 31A:92-97, 1995

16. Rowinsky EK, Gilbert M, McGuire WP, et al: Sequence of Taxol and cisplatin: A phase 1 and pharmacokinetic study. j Clin Oneal 9:1692-1703, 1991

17. Rowinsky EK. Chaudhry V, Forastiere AA, et al: Phase I and pham1acokinetic srudy of paclitaxel and cisplatin with granulocyre colony-stimulating factor: Neuromuscular toxicity is dose-limiting. J C lin Oncol 11:2010-2020. 1993

18. Eisenhauer EA, ten Bokkel Huinink WW, Swenerron KD, et al: Europc.an-Omadian randomized trial of paclitaxel in relapsed ovarian cancer: High-dose versus low-dose and long versus short infusion. J Clin Oncol 12:2654-2666, 1994

19. Kennedy MJ, Armstrong D. Donehower R, et al: The hematologic rox icicy of the Taxol/Cytoxan doublet is sequence­dependenc. Proc Am Soc Clin Oncol 13:74, 1994 (abstr 342)

20. Pestalozzi DC, Sotos GA, Choyke PL, et al: Typhlitis resulting from treatment with Taxol and doxorubicin in pa· tients with metastatic breast cancer. Cancer 71: 1797-1800. 1993

21. Lopes NM, Adams EG, Pitts TW. ct al: Cell kill kinetics and cell cycle effects of T axol on human and hamster ovarian cell lines. C',ancer Chcmother Pharmacol .32:235-242, 1993

22. Liebmann JE. Cook JA, Lipschultz C, et al: Cytotoxic studies of paclitaxcl (T axol) in human tumour cell lines. Br J Cancer 68:1104-1109, 1993

7 of 7 Celltrion, Inc., Exhibit 1013