Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Cancer. Author manuscript; available in PMC 2011 April 1.
Published in final edited form as:
PMCID: PMC2847062

A phase II pilot trial of low dose, continuous infusion, or “metronomic”, paclitaxel and oral celecoxib in patients with metastatic melanoma



Tumor angiogenesis has been associated with poor prognosis in patients with metastatic melanoma (MM). Microtubule stabilizers and cyclooxygenase (COX)-2 inhibitors, alone and in combination, have shown inhibitory effects on endothelial cells and tumor angiogenesis. Angiogenesis, the growth of new blood vessels, is necessary for tumor growth and progression. Thus, we tested the safety and efficacy of a low dose of paclitaxel and celecoxib in patients with MM.


Patients received paclitaxel 10mg/m2 for 96 hours weekly by CIV and celecoxib 400 mg po/bid. Systemic tumor response was assessed at 6 week intervals. Tumor measurements at the end of Cycle 1 were used as the baseline for assessment of tumor progression. Patients with unacceptable toxicity or disease progression after Cycle 2 relative to the end of Cycle 1 were taken off study.


Twenty patients were enrolled. Twelve (60%) had received 2 or more prior systemic therapies. Three patients did not receive treatment due to rapid disease progression. Treatment related grade 3–4 toxicities were limited to catheter-related complications. One patient achieved a partial response and 3/20 (15%) patients had stable disease for greater than 6 months. Median time to progression was 57 days (95% CI: 43–151) and median overall survival was 212 days (95% CI: 147–811 days.


Low-dose, CIV paclitaxel and oral celecoxib can produce disease stabilization in a significant proportion of heavily pre-treated patients with MM. These findings support a role for metronomic therapy in this disease.

Keywords: metastatic, melanoma, metronomic, Cox-2, clinical trial, phase II


Malignant melanoma is the ninth most common cause of cancer in the United States. Its incidence, however, is rising faster than any other malignancy, especially among younger individuals (1). Metastatic melanoma is relatively unresponsive to chemotherapy and is frequently treated with biologic response modifiers, such as interleukin-2 (IL-2) and interferon (2). Unfortunately, long-term responses are uncommon, and the median survival for patients with advanced disease is 6–12 months (35). The therapeutic options for patients whose disease has progressed following IL-2-based immunotherapy and/or dacarbazine based cytotoxic chemotherapy are limited and investigational. The association of tumor angiogenesis with poor prognosis (6) has encouraged the study of anti-angiogenic therapy in this disease.

Microtubule stablizers and cyclooxygenase (COX)-2 inhibitors, alone and in combination, have been shown to have inhibitory effects on tumor angiogenesis and growth (7,8). Low doses of paclitaxel were found to accumulate preferentially in endothelial cells and to inhibit angiogenesis in vitro (7). Additionally, plasma from patients treated with “metronomic” therapy exhibited antiangiogenic activity in vitro (7). Moreover, a growing body of evidence indicates that cyclooxygenase-2 (COX-2) plays a significant role in tumor progression in malignant melanoma (8). The contributions of COX-2 in tumor growth and angiogenesis include increased expression of VEGF by tumors cells; the production of the eicosanoid products thromboxane A2, PGE2 and PG12 that can directly stimulate endothelial cell (EC) migration and growth factor-induced angiogenesis (8).

Low dose continuous dosing of chemotherapeutic agents has been referred to as “metronomic scheduling”: In this approach there are no rest periods off therapy and the use of lower doses minimizes toxic side effects. Metronomic therapy has been shown to represent a promising anti-tumor strategy (9,10). A major focus of recent clinical trials has been the inclusion of correlative studies to aid in the scientific understanding of the intervention. Prior studies of angiogenic biomarkers have suggested that metronomic therapy is associated in changes in thrombospondin (11). Moreover, circulating endothelial cells (CECs) and circulating endothelial progenitors (CEPs) have also been shown to exhibit changes over the course of therapy which may correlate with clinical benefit (12,13). In this study, we combined COX-2 inhibition with paclitaxel at metronomic doses to determine its effect on patients with refractory metastatic melanoma. We also assessed correlative biomarkers in an effort to understand the mechanism of action of this combination of agents.



Eligible patients had histologically confirmed melanoma with radiographic evidence of metastatic disease and clinically documented disease progression. Patients could have had multiple prior systemic treatments. Inclusion criteria included an ECOG performance status (PS) of ≤2, life-expectancy ≥3months and adequate organ function. Exclusion criteria included history of myocardial infarction, angina or conduction abnormalities within 3 months, concurrent use of NSAIDs, heparin or warfarin. Patients must not have had prior treatment with low dose paclitaxel chemotherapy, either weekly or daily; patients who had received paclitaxel every three weeks at standard doses were eligible. There were no other limitations on previous therapy. Other exclusion criteria included untreated brain metastasis, prosthetic devices, and active/recurrent infections. A washout period of 4 weeks was required after any radiotherapy or chemotherapy prior to the initiation of study treatment.

All patients provided written informed consent. The study was approved by the institutional review board at the Beth Israel Deaconess Medical Center in accordance with international standards of good clinical practice

Procedures and Treatment

Patients were treated with paclitaxel 10mg/m2 for 96 hours weekly by continuous intravenous infusion and with celecoxib 400mg PO BID. This dose of paclitaxel produces a blood level similar to the concentration of paclitaxel that was shown to have antiangiogenic activity when combined with celecoxib. This was determined previously as plasma from 4 patients treated with these doses exhibited inhibitory effects on cultured endothelial cells (7).

Paclitaxel was initiated one week prior to celecoxib to assess the tolerability and biologic effects of paclitaxel alone. A cycle was 6 weeks and systemic tumor response was assessed at the end of each cycle. At baseline and at the beginning of each cycle, patients underwent a history and physical examination, assessment of performance status, blood tests for a complete blood count, liver function tests, serum chemistries including lactate dehydrogenase, and blood and peripheral blood mononuclear cells (PBMCs) for correlative studies. Chest, abdomen and pelvic computed tomography (CT) scans were obtained every 6 weeks. Treatment was continued until patients developed progressive disease or evidence of unacceptable toxicity. Patients without clinically symptomatic disease progression at 6 weeks could continue on therapy for another 6 weeks with the 6 week CT scan serving as the new baseline imaging study.

After enrollment of 10 patients, there were 2 cases of central line-related Staphylococcus aureus infections. Thus, the protocol was amended to screen for Staph aureus in the nares of patients to determine whether they were carriers of this organism. Cultures from the nares were performed at the beginning of the study and every three months thereafter. Patients who screened positive were treated with intranasal mupirocin. Additionally, standard nursing guidelines were also implemented to ensure consistent catheter care.

Clinical endpoints

The primary objective of the study was to evaluate the safety and activity of paclitaxel and celecoxib administered in the particular dose and schedule in this patient population. Assessment of adverse events and toxicities was performed every two weeks during the first six weeks of therapy and every three weeks thereafter while on treatment, as well as four weeks after paclitaxel and celecoxib were discontinued. A 25% dose reduction of paclitaxel (from 10 mg/m2 to 7.5 mg/m2) was permitted if the higher dose level was associated with significant toxicity in a given patient. Activity was measured by clinical parameters and by CT tumor measurements. (deleted a sentence about definition of disease progression) Response rate by CT was evaluated according to Response Evaluation Criteria in Solid Tumors (RECIST). Response was evaluated at the end of cycles 1 and 2 and then every other cycle thereafter.

Because it was assumed that this treatment regimen would be cytostatic, not cytotoxic and that patients could experience an initial period of tumor growth prior to stabilization, disease progression documented on the initial 6 week scan was only considered true lack of treatment efficacy if new symptoms accompanied it. Thus, patients who showed some radiographic disease progression at week 6 but then stable disease at week 12 (relative to week 6) were considered for continued participation on study. For patients continuing on therapy the week 6 scan represented the new baseline for subsequent response characterization. Time to progression was calculated as time from registration to time of disease progression (clinical or radiographic as assessed by RECIST criteria) or time of progression following new baseline at week 6 in patients showing early disease growth. Overall survival was defined as the time from registration to death. Survival was censored at the date last known alive.


Blood collection

Blood specimens for pharmacokinetic and pharmacodynamic studies and to evaluate antiangiogenic biomarkers were obtained at baseline, during the first week of single agent paclitaxel, during the first week of combination of paclitaxel and celecoxib and every 2 weeks on cycle 1 and 3 weeks thereafter of combination paclitaxel and celecoxib.

Plasma cytokine analyse:s

Plasma cytokine levels (interleukin-8, angiogpoietin-2, thrombospondin-1, vascular endothelial growth factor, soluble VEGFR2, and fibroblast growth factor) were measured by enzyme-linked immunoadsorbent assay kits (R&D) at baseline at during cycle 1 and monthly on therapy until disease progression.

Flow cytometry

Circulating endothelial cells were enumerated by four-color flow cytometry analysis as previously described (12). PBMCs of patients were collected at study entry on the first week into paclitaxel infusion, every 2 weeks for the first cycle and every 3 weeks thereafter in CPT citrate tubes (BD Biosciences) and spun to separate the monocyte layer. The PBMCs were frozen in media consisting of 50% RPMI, 40% human plasma, and 10% DMSO, and stored in liquid nitrogen until the day of analysis. Samples were analyzed in batches to minimize interassay variability. Briefly, the cells were washed with PBS containing 1% albumin and incubated with a panel of four antibodies including CD45, CD31, CD146, and CD133 (12) along with the appropriate isotype controls. Mature CECs were defined as negative for the hematopoietic marker, CD45, positive for endothelial markers CD31 and CD146, and negative for the progenitor marker, CD133. CEPs had the same phenotype but were CD133+. Human umbilical vein endothelial cells (Cambrex) were used as a positive control for CD146 staining and WERI cells (ATCC) as positive controls for CD133.

Statistical Analysis

All patients enrolled in the study, including those who progressed prior to start of treatment, were considered evaluable. Median time to progression and survival was calculated using the Kaplan-Meier method, with approximate confidence limits using the Greenwood formula. Individual data and within subject changes observed at baseline and Day 29 are presented. Within subject changes were compared using a paired t-test and overall changes were compared using a Student’s T-test. In the biomarker analyses no correction for multiple testing was performed.

Sample Size

The sample size, 20 subjects, was selected based on practical considerations. With 20 subjects, we would have 95% power to detect a serious AE occurring in 14% or more of subjects We would also have a confidence interval width of approximately 9–49% for an observed rate of 25% with stable disease for 6 months or more, which was considered sufficient accuracy for preliminary data on safety and efficacy.


This trial was activated in 9/2001, and closed to accrual in 10/2005 after reaching target accrual.

Characteristics of Patients and Treatment

In total, 20 patients were accrued. Of these, 3 patients developed rapid disease progression prior to initiating treatment and were never treated. Table 1 summarizes baseline patient characteristics of the enrolled patients: 30% had greater than 3 prior systemic treatments and 55% of patients had greater than 4 sites of disease. Half the patients had cutaneous primary melanomas while the remaining had choroidal, mucosal or unknown primaries. The median total number of weeks on treatment was 6 with a range of 3–92 weeks. 14/17 treated patients terminated treatment because of progressive disease; the remaining 3 due to toxicity.

Activity and Survival

In an intention-to-treat analysis, 3/20 (15%) patients experienced disease stabilization for over 6 months (95% confidence interval (CI) 3–38%). These 3 patients had time to progression (TTP) of 24, 60, and 92 weeks. Additionally, 4 patients had TTP greater than 12 weeks with specific TTP of 13 and 17 weeks and 2 patients with TTP of 18 weeks. (Figure 1). No RECIST defined complete responses were observed but one patient had a partial response. Median time to progression was 57 days (95% CI: 43–151 days) and overall survival had a median of 212 days (95% CI: 147–811 days) (Figure 2). Five patients survived more than one year, including one patient still alive after 40 months.

Figure 1
Progression-free survival (PFS) for all 20 patients on trial. This includes 3 patients who were enrolled but did not start therapy.
Figure 2
Overall survival for the 20 patients on study. This includes 3 patients who were enrolled but did not start therapy.


Treatment was tolerated well in this patient population with the majority of toxicity due to central line related complications (Table 2). These included 2 patients experiencing bacteremia/sepsis, 3 with deep venous thrombosis- 2 of whom also had pulmonary embolism. Because of the cases of central line-related Staphylococcus aureus infections, the protocol was amended to screen for Staph aureus in the nares of patients and treat identified carriers with intranasal mupirocin. Subsequent to this intervention, there were no further cases of line related infectious complications.

Table 2
Grade 3/4 Toxicity Incidence for all Treated Patients

Toxicities related to the paclitaxel and celecoxib were limited to 1 patient with grade 3 anemia and 1 with grade 3 hypertension.

Plasma and Circulating Endothelial Cell Analysis

Nine of 17 treated patients had CEC/CEP analyses performed. We also measured serial levels of circulating plasma cytokines. While there were trends towards increases in CEP number in patients on therapy longer than 12 weeks the number of samples obtained was not large enough to assign significance to this trend. (table 3 was removed) There was no correlation of levels of sVEGFR2, IL-8, Ang-2, FGF, VEGF, or Tsp-1 with therapy or clinical benefit.


There is a need for new effective therapies for patients with metastatic melanoma. Metronomic chemotherapy is a promising strategy for inhibition of angiogenesis and is associated with lower toxicities than conventional chemotherapy dosing regimens. The current study was designed to study the effect of celecoxib and low dose, continuous intravenous paclitaxel in patients with metastatic melanoma. The treatment regimen was well tolerated with the majority of toxicities being related to the central venous catheter needed to administer the paclitaxel. While the median PFS for this cohort was short, evidence of potential antitumor activity was observed with stable disease greater than 6 months in 15% of patients. In particular 2 patients exhibited prolonged disease stabilization remaining on the study for greater than one year (60 weeks and 92 weeks). In these patients, one patient acheived a RECIST defined partial response with 38% shrinkage of target lesions and the second patient exhibited a 12% shrinkage of target lesions over the treatment course.

The mechanism underlying this activity is uncertain. Although the effects would be consistent with antiangiogenic activity, we cannot formally rule out that a prolonged use of paclitaxel even at the low doses used may have produced direct tumoricidal effects. Moreover, it is conceivable that paclitaxel and celecoxib may have influenced the immune response to the tumor in these patients since metronomic therapy with cyclophosphamide can decrease Tregs as can Cox-2 inhibitors (14).

CECs and CEPs have been found to be elevated in patients with several malignancies. Moreover, in a study by Shaked et al., therapy with vascular disrupting agents led to acute mobilization of CEPs which could be blocked by VEGFR2 inhibition (15). VEGF administration has also been described to increase both mature CECs and CEPs and this was reversed with a VEGFR2 inhibitor. Only a portion of the patients on this study had correlative biomarker studies. With such small sample numbers, we did not find significant changes but in the future, larger studies could examine whether there is an increase in CEPs in patients who receive clinical benefit from therapy during treatment which is consistent with an effect of therapy on blood vessels. Changes in CEPs theoretically, could also be a result of treatment with paclitaxel as described in another recent paper by Shaked et al (16). Overall, in patient plasma there were few robust biomarker changes that could serve as predictive biomarkers of effectiveness. Whether this is a consequence of plasma not reflecting changes in the tumor or the minimal efficacy of treatment or the limited sampling is uncertain. The utility of CEP levels as predictive of stable disease requires prospective validation in future studies.

Since this study was initiated, COX-2 inhibitors have been tested in metronomic dosing in two recent trials. In a study of pioglitazone and refecoxib prior to trofosfamide treatment in patients with advanced melanoma, 4/22 patients received clinical benefit as defined as disease stabilization and tumor responses (17). Additionally, in a study of low dose, metronomic treosulfan and rofecoxib, 4 of 12 patients achieved disease stabilization (18).

Consistent with these and our findings, antiangiogenic therapies have recently been shown to benefit patients with melanoma. Bevacizumab has been evaluated in combination with paclitaxel and carboplatin as well as with weekly paclitaxel and these combinations have shown high rates of disease stabilization (19,20). Additionally, the VEGFR-2 tyrosine kinase inhibitors have shown promise in the treatment of melanoma. (statement about sorafenib removed) Axitinib was recently reported to have single agent activity in stage IV melanoma (22). Our study supports the use of antiangiogenic strategies as a potentially promising therapy for patients with metastatic melanoma.


NIH R21 CA 097730-01A1, Grant from Pfizer Inc., RB was funded by the Clinical Investigator Training Program: Beth Israel Deaconess Medical Center - Harvard/MIT Health Sciences and Technology, in collaboration with Pfizer Inc. and Merck & Co.


No financial disclosures


1. Jemal A, Siegel R, Ward E, Murray T, Xu J, THun MJ. Cancer Statistics, 2007. CA cancer J Clin. 2007;57:43–66. [PubMed]
2. Ives NJ, Stowe RL, Lorigan P, Wheatley K. Biochemotherapy versus chemotherapy for metastatic melanoma: a metaanalysis fo the randomized trials. J Clin Oncol. 2007;25(18S):8544.
3. Balch CM, Reintgen DS, Kirkwood JM, et al. Cutaneous melanoma. In: De Vita VT, Hellman S, Rosenberg SA, editors. Cancer: Principles and Practice of Oncology. ed 5. Philadelphia, PA: Lippincott; 1997. pp. 1935–1993.
4. Balch CM, Buzaid AC, Soong S-J, et al. Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol. 2001;19:3635–3648. [PubMed]
5. Thompson JF, Scolyer RA, Kefford RF. Cutaneous melanoma. Lancet. 2005;365:687–701. [PubMed]
6. Ugurel S, Rappl G, Tilgen W, Reinhold U. Increased serum concentration of angiogenic factors in malignant melanoma patients correlates with tumor progression and survival. J Clin Oncol. 2001 Jan 15;19(2):577–583. [PubMed]
7. Merchan JR, Jayaram DR, Supko JG, He X, Bubley GJ, Sukhatme VP. Increased endothelial uptake of paclitaxel as a potential mechanism for its antiangiogenic effects: potentiation by Cox-2 inhibition. Int J Cancer. 2005 Jan 20;113(3):490–498. [PubMed]
8. Jones MK, Wang H, Peskar BM, Levin E, Itani RM, Sarfeh IJ, Tarnawski AS. Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing. Nat Med. 1999 Dec;5(12):1418–1423. [PubMed]
9. Ng SS, Sparreboom A, Shaked Y, Lee C, Man S, Desai N, Soon-Shiong P, Figg WD, Kerbel RS. Influence of formulation vehicle on metronomic taxane chemotherapy: albumin-bound versus cremophor EL-based paclitaxel. Clin Cancer Res. 2006 Jul 15;12(14 Pt 1):4331–4338. [PubMed]
10. Lennernäs B, Albertsson P, Damber JE, Norrby K. Antiangiogenic effect of metronomic paclitaxel treatment in prostate cancer and non-tumor tissue in the same animals: a quantitative study. APMIS. 2004 Mar;112(3):201–209. [PubMed]
11. Bocci G, Francia G, Man S, Lawler J, Kerbel RS. Thrombospondin 1, a mediator of the antiangiogenic effects of low-dose metronomic chemotherapy. Proc Natl Acad Sci U S A. 2003 Oct 28;100(22):12917–12922. [PubMed]
12. Norden-Zfoni A, Desai J, Manola J, Beaudry P, Force J, Maki R, Folkman J, Bello C, Baum C, DePrimo SE, Shalinsky DR, Demetri GD, Heymach JV. Blood-based biomarkers of SU11248 activity and clinical outcome in patients with metastatic imatinib-resistant gastrointestinal stromal tumor. Clin Cancer Res. 2007 May 1;13(9):2643–2650. [PubMed]
13. Shaked Y, Bertolini F, Man S, et al. Genetic heterogeneity of the vasculogenic phenotype parallels angiogenesis; implications for cellular surrogate marker analysis of antiangiogenesis. Cancer Cell. 2005;7:101–111. [PubMed]
14. Lutsiak ME, Semnani RT, De Pascalis R, Kashmiri SV, Schlom J, Sabzevari H. Inhibition of CD4(+)25+ T regulatory cell function implicated in enhanced immune response by low-dose cyclophosphamide. Blood. 2005 Apr 1;105(7):2862–2868. [PubMed]
15. Shaked Y, Ciarrocchi A, Franco M, Lee CR, Man S, Cheung AM, Hicklin DJ, Chaplin D, Foster FS, Benezra R, Kerbel RS. Therapy-induced acute recruitment of circulating endothelial progenitor cells to tumors. Science. 2006;313:1785–1787. [PubMed]
16. Shaked Y, Henke E, Roodhart JM, Mancuso P, Langenberg MH, Colleoni M, Daenen LG, Man S, Xu P, Emmenegger U, Tang T, Zhu Z, Witte L, Strieter RM, Bertolini F, Voest EE, Benezra R, Kerbel RS. Rapid chemotherapy-induced acute endothelial progenitor cell mobilization: implications for antiangiogenic drugs as chemosensitizing agents. Cancer Cell. 2008 Sep 9;14(3):263–273. [PMC free article] [PubMed]
17. Reichle A, Vogt T, Coras B, Terheyden P, Neuber K, Trefzer U, Schultz E, Berand A, Bröcker EB, Landthaler M, Andreesen R. Targeted combined anti-inflammatory and angiostatic therapy in advanced melanoma: a randomized phase II trial. Melanoma Res. 2007 Dec;17(6):360–364. [PubMed]
18. Spieth K, Kaufmann R, Gille J. Metronomic oral low-dose treosulfan chemotherapy combined with cyclooxygenase-2 inhibitor in pretreated advanced melanoma: a pilot study. Cancer Chemother Pharmacol. 2003 Nov;52(5):377–382. [PubMed]
19. Perez DG. Phase II trial of carboplatin, weekly paclitaxel, and biweekly bevacizumab in patients with unresectable stage IV melanoma; ASCO Annual Meeting; 2007. Abstract - No. 8560. [PMC free article] [PubMed]
20. González-Cao M, Viteri S, Díaz-Lagares A, González A, Redondo P, Nieto Y, Espinós J, Chopitea A, Ponz M, Martín-Algarra S. Preliminary results of the combination of bevacizumab and weekly paclitaxel in advanced melanoma. Oncology. 2008;74(1–2):12–6. [PubMed]
21. McDermott DF, Sosman JA, Gonzalez R, Hodi FS, Linette GP, Richards J, Jakub JW, Beeram M, Tarantolo S, Agarwala S, Frenette G, Puzanov I, Cranmer L, Lewis K, Kirkwood J, White JM, Xia C, Patel K, Hersh E. Double-blind randomized phase II study of the combination of sorafenib and dacarbazine in patients with advanced melanoma: a report from the 11715 Study Group. J Clin Oncol. 2008 May 1;26(13):2178–2185. [PubMed]
22. Fruehauf JP. Axitinib (AG-013736) in patients with metastatic melanoma: A phase II study; ASCO Annual Meeting; 2008. Abstract - No. 9006.