Flavopiridol Enhances the Effect of Docetaxel in Vitro and in Vivo in Human Gastric Cancer Cells¹

Gastric cancer is a leading cause of cancer death throughout the world. In the United States, there will be 23,000 cases of gastric cancer diagnosed this year, and this will be associated with 12,800 deaths (1). These patients invariably die of metastatic disease. Recent randomized studies using 5-fluorouracil-based chemotherapies indicate response rates of 10–20%, with a median survival of 7–8 months (2). In recent years the taxanes, docetaxel (Taxotere) and paclitaxel (Taxol), have been introduced into the treatment of solid tumor cancers. This has resulted in some success, especially in the treatment of metastatic breast cancer, in which response rates of up to 68% in untreated and 57% in pretreated patients have been reported (3). However, despite the highly improved response rates in advanced metastatic breast disease, success in patients with metastatic gastric cancer has been less compelling. The results of the Phase II Eastern Organization for Research and Training Early Clinical Trials Group indicate a response rate to single-agent docetaxel of 24%, with only partial responses and a median survival of 7.5 months (range, 3–11+ months; Ref. 4). These results are comparable with the randomized studies reported with combinations of 5-fluorouracil and cisplatin (2). Thus, new therapies are desperately needed in the treatment of gastric cancer. An advance in this area would have a major impact on the outcome of a large number of patients with this disease.

Further opportunities to enhance the effectiveness of the taxanes are being investigated. One approach is to add classic cytotoxic agents to the taxanes in the hope that this will increase the response rate and result in an overall improvement in survival. However, such standard combination therapies will undoubtedly be associated with increased toxicities that may outweigh any potential benefit of the drug combination. Another approach is to combine the taxanes with cell cycle inhibitors that could potentiate taxane-induced apoptosis and result in an increased antitumor effect. One promising candidate along this line is flavopiridol, a synthetic flavone currently undergoing Phase I and II clinical trials (5). Flavopiridol has been shown in vitro to inhibit tumor cell growth at nanomolar concentrations through blockade of cell cycle progression at G₁ or G₂ (6). It is a potent inhibitor of CDKs³ with respect to the ATP binding site. Inhibition of CDKs including CDK-1 (cdc-2), CDK-2, CDK-4, and CDK-7 and hypophosphorylation of pRb have also been reported (7, 8). We have reported previously that flavopiridol at nanomolar concentrations significantly enhances the induction of apoptosis by mitomycin-C and paclitaxel in gastric and breast cancer and by CPT-11 in colon cancer cell lines (9–11). Synergism between flavopiridol and paclitaxel has also been observed against A549 non-small cell lung cancer cells (12). These studies indicate that the combination of paclitaxel and flavopiridol is highly sequence dependent, such that paclitaxel should precede flavopiridol to achieve the maximal effect (10, 12).

Taxanes promote microtubular aggregation and block cells in metaphase (13). The importance of mitotic block in induction of apoptosis in response to the taxanes has been shown by various groups using antisense of cyclin B1 to abrogate cdc-2 kinase activity (14). The prevention of mitotic block also prevents cell death. However, the mechanism defining mitotic block to induce apoptosis by docetaxel is not clearly understood. Docetaxel has shown superior cytotoxic potency to paclitaxel against murine and human cell lines, which may be due to a higher affinity for microtubules (15). Both taxanes also trigger apoptosis in cancer cells via bcl-2 phosphorylation and inactivation (16). However, docetaxel has been shown to be 100-fold more potent than paclitaxel in terms of the ability to phosphorylate the antiapoptotic target bcl-2 (16). Therefore, docetaxel may have theoretical advantages over paclitaxel in terms of its clinical development. Randomized trials comparing docetaxel with paclitaxel are under way.

We elected to examine the effect of flavopiridol on docetaxel-treated MKN-74 gastric cancer cells. In this study we have characterized the apoptotic and cell cycle events associated with this combination therapy. Our results indicate that flavopiridol potentiates the effect of docetaxel both in vitro and in vivo. This effect is associated with the induction of apoptosis when flavopiridol is administered after docetaxel. Pretreatment of cells with flavopiridol inactivates the cyclin B1/cdc-2 kinase. This prevents the mitotic arrest by docetaxel from occurring in the context of a properly activated cdc-2 kinase, thus rendering the sequence of flavopiridol followed by docetaxel combination inactive. The importance of sequence in flavopiridol’s potentiation of docetaxel was confirmed in mice bearing MKN-74 xenografts.

The human gastric cancer cell line MKN-74 was graciously supplied by Dr. E. Tahara (Hiroshima University, Hiroshima, Japan). Cells were maintained in Eagle’s minimal essential media supplemented with 10% heat-inactivated fetal bovine serum (HyClone), penicillin, and streptomycin at 37°C in 5% carbon dioxide. Cells were tested as Mycoplasma free.

The nuclear morphology of the cell (multinucleated and apoptotic) was determined by staining nuclear chromatin with 4′,6-diamidino-2-phenylindole (Sigma, St. Louis, MO). Apoptotic cells have condensed and fragmented chromatin. The percentage of apoptosis was determined by counting the cells and scoring them for the incidence of apoptosis using an Olympus BH2-DM2U2UV Dichetomic Mirror cube filter (Olympus, Lake Success, NY). The protocol has been described previously (10). Briefly, MKN-74 cells were cultured for 24–48 h (approximately 60% confluent) and treated according to one of the following conditions: (a) drug-free media; (b) 300 nm flavopiridol (graciously supplied by Dr. Edward Sausville, National Cancer Institute, Bethesda, MD) alone for 24 h; (c) 100 nm docetaxel (Taxotere; Aventis Pharmaceuticals, Bridgewater, NJ) alone for 18 h; (d) 300 nm flavopiridol and 100 nm docetaxel together for 18 h; (e) 300 nm flavopiridol for 24 h followed by removal of media containing flavopiridol and addition of media with 100 nm docetaxel for 18 h; (f) the same drugs given in reverse order; and (g) 100 nm docetaxel for 18 h followed by removal of drug media and addition of fresh media without drug. In sequential therapy, the floating cells were collected and added back for subsequent treatment. At the end of treatment, adherent cells were trypsinized, pooled with floating cells, washed with PBS, and fixed in 3% paraformaldehyde for 10 min at room temperature. Cells were stained with 4′,6-diamidino-2-phenylindole for 30 min at room temperature in the dark. The aliquots of cells were taken to prepare slides, and duplicate samples of 400 cells each were counted and scored for the incidence of apoptotic chromatin condensation.

The MKN-74 cells were treated as described above and fixed and labeled with MPM-2 antibody as described previously (10). The MPM-2-positive (mitotic) cells show increased green fluorescence; thus shift above the baseline of the dot plot. Mitotic index was defined as percentage of MPM-2-positive cells.

The MKN-74 cells were treated with docetaxel and flavopiridol as discussed above. The cell lysates were prepared as described previously (10). Soluble protein (200 μg) was incubated with 1 μg of anti-cyclin B1 antibody (catalogue number sc-245; Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 2 h. Immune complexes were then precipitated with 40 μl of immobilized rProtein A (RepliGen) overnight at 4°C, and kinase activity was assayed as described previously.

Protein lysates prepared for kinase assays were used for immunoblotting. Western blots were prepared as described previously. The membranes were probed with mouse monoclonal antibodies specific to ppRb (PharMingen) and mouse monoclonal cyclin B1 (catalogue number sc-245; Santa Cruz Biotechnology). The membranes were treated with a secondary sheep antimouse horseradish peroxidase or donkey antirabbit horseradish peroxidase antibody for 1 h at room temperature. Detection was done by enhanced chemiluminescence reagents (Dupont NEN Life Science Products, Boston, MA) according to the manufacturer’s protocol. The levels of expression were quantified using a densitometric scanning system.

The general procedure used in the experiments has been described previously (17). Athymic-NCr-nu male mice (8–10 weeks old) were inoculated s.c. in flank with MKN-74 cells mixed with Matrigel (Becton Dickinson). Treatment was started on the third day after the inoculation of tumor, and animals were treated with the maximum tolerated dose of docetaxel either as single agent or in combination. The maximum tolerated dose of flavopiridol was used as a single agent, and the dose was decreased in combination therapy to minimize cytotoxicity. The average tumor volume at the day of treatment was 27–29 mm³. Mice received docetaxel alone (10 mg/kg), flavopiridol alone (10 mg/kg), or docetaxel (10 mg/kg) followed by flavopiridol (2.5 mg/kg) 0, 3, 7, or 16 h later. One group was treated in the reverse order, such that flavopiridol was administered first, and docetaxel was administered 7 h later. Mice in the control group were given vehicle (PBS) alone. All drugs were administered i.p. twice a week, for a total of four injections. Tumors were measured every 3–4 days with calipers, and tumor volumes were calculated by the formula 4/3 × π × r³ [r = (larger diameter + smaller diameter)/4]. The percentage of tumor regression was calculated as the percentage ratio of difference between baseline and final tumor volume to the baseline volume. These studies were performed in accordance with the Principles of Laboratory Animal Care (NIH Publication No. 85–23, released 1985).

All experiments were done in duplicate and repeated at least three times unless otherwise indicated. The statistical significance of the experimental results was determined by the two-sided t test.

Human gastric cancer MKN-74 cells were treated with 100 nm docetaxel (a clinically achievable concentration) and 300 nm flavopiridol (a concentration established to inhibit CDKs) individually, concurrently, and sequentially as docetaxel for 18 h followed by flavopiridol for 24 h, or the same drugs given in reverse sequence. The treatment times for each drug were chosen based on our previous studies with paclitaxel and flavopiridol (10). As determined by QFM, under these treatment conditions, docetaxel for 18 h or flavopiridol for 24 h (Doc₁₈ or F₂₄) induced apoptosis in 2% of the cells. The sequential treatment of MKN-74 cells with docetaxel for 18 h followed by flavopiridol for 24 h (Doc₁₈→F₂₄) induced apoptosis in 30 ± 2% of the cells (Fig. 1), which was significantly greater than that observed for docetaxel followed by drug-free media (11 ± 2%; Doc₁₈→ND₂₄; P < 0.005). Treatment of Doc₁₈→F₂₄ cells for an additional 24 h in drug-free media (Doc₁₈→F₂₄→ND₂₄) did not further enhance the induction of apoptosis (25 ± 2%) when compared with Doc₁₈→F₂₄ (30 ± 2%). Under these treatment conditions (Doc₁₈→ND₂₄, Doc₁₈→F₂₄, and Doc₁₈→F₂₄→ND₂₄), an examination of cells by fluorescence microscopy indicated that these cells had acquired a multinucleated phenotype, a characteristic that has been attributed to aberrant mitosis following taxane therapy (18). When the cells were treated in the reverse order such that flavopiridol treatment preceded docetaxel (F₂₄→Doc₁₈), only 2 ± 1% of the cells underwent apoptosis. This was no different than the induction of apoptosis by flavopiridol for 24 h followed by drug-free media (F₂₄→ND₂₄; 2 ± 1%). Incubation of cells treated first with the reverse order of flavopiridol followed by docetaxel and then with an additional 24 h in drug-free media (F₂₄→Doc₁₈→ND₂₄) resulted in apoptosis in only 4 ± 1% of the cells. This was significantly less than the 11 ± 2% observed with Doc₁₈→ND₂₄ (P < 0.005), indicating an antagonism to the induction of apoptosis when flavopiridol preceded docetaxel.

Because both flavopiridol and docetaxel affect cell cycle-specific events, we next elected to examine the effects of both agents on cell cycle distribution. To distinguish the G₂ and M cells in the 4n peak, we labeled the cells with MPM-2 antibody, which recognizes epitopes shared by phosphoproteins appearing during mitosis. Two-dimensional flow cytometry was used to quantitate the population of cells in M phase. As anticipated (Table 1), after 18 h of docetaxel treatment (Doc₁₈), the majority of cells (70%) accumulated with 4n DNA content (the G₂-M population). Sixty-two percent of the total cells stained positive for MPM-2 (mitotic index of 62), consistent with docetaxel’s effect on inducing a mitotic arrest. In untreated cells (ND₂₄), only 2% of the population was positive for MPM-2 due to rapid turnover of cells from metaphase to anaphase. Treatment with flavopiridol alone for 24 h (F₂₄) resulted in an increase in both the G₁ (58%) and G₂ (34%) cell populations when compared with untreated controls (43% and 20%, respectively, for ND₂₄). This is consistent with flavopiridol serving as a pan-CDK inhibitor and preventing the entry of cells into S phase. Pretreatment with flavopiridol before docetaxel (F₂₄→Doc₁₈) resulted in 22% of the cells staining positive for MPM-2, with 44% 4n cells. Thus, when compared with Doc₁₈, in which 62% of the docetaxel-treated cells were arrested in M phase, pretreatment with flavopiridol prevented these cells from entering the G₂ and M phase of the cell cycle. When two drugs were given together [(Doc+F)₂₄], the number of cells arrested in mitosis (7%) was also decreased dramatically compared with docetaxel alone. Thus, when compared with Doc₁₈, both F₂₄→Doc₁₈ and (Doc+F)₂₄ exhibited fewer cells entering M phase, with a resulting increase in the G₁ and G₂ cell populations. Twenty-four h after removal of docetaxel (Doc₁₈→ND₂₄), the 4n peak continued to increase (90%). However, the mitotic index decreased from 62% to 22%, indicating that mitotic arrest was transient and that cells exit mitosis after the initial 18-h docetaxel treatment. The addition of flavopiridol for 24 h on docetaxel-treated cells (Doc₁₈→F₂₄) resulted in an even greater decrease in the number of MPM-2-positive cells (12%).

The exit from mitosis and progression into interphase require the inactivation of cdc-2 kinase by degradation of cyclin B1 (19), dephosphorylation of cdc-2 at Thr¹⁶¹ (20), and/or phosphorylation at Thr¹⁴ and Tyr¹⁵ (21). To examine the mitotic arrest and exit, we assayed cyclin B1-associated cdc-2 kinase activity by histone H1 phosphorylation under these treatment conditions in MKN-74 cells. As shown in Fig. 2, 18 h of docetaxel treatment (Doc₁₈) induced cyclin B1/cdc2 kinase activity by more than 10-fold as compared with untreated control cells (ND₂₄). As expected, treatment of cells with flavopiridol alone (F₂₄ and F₂₄→ND₂₄) inhibited the cyclin B1/cdc-2 kinase activity to barely detectable levels. Cotreatment [(Doc+F)₁₈] or pretreatment (F₂₄→Doc₁₈) of cells with flavopiridol prevented docetaxel-induced activation of cyclin B1/cdc-2 kinase (Doc₁₈). Twenty-four h after removal of docetaxel (Doc₁₈→ND₂₄), kinase activity was decreased substantially (more than 5-fold) compared with docetaxel alone for 18 h. Addition of flavopiridol to docetaxel-treated cells (Doc₁₈→F₂₄) resulted in an even higher inactivation of cyclin B1/cdc-2 kinase.

A similar profile was observed for cyclin B1 protein expression. As shown in Fig. 3, with pretreatment (F₂₄→Doc₁₈) and cotreatment [(Doc+F)₂₄] of flavopiridol with docetaxel, there was no increase in cyclin B1 protein expression compared with untreated controls. Cyclin B1 levels increased 2–3-fold after 18 h of docetaxel treatment (Doc₁₈) compared with untreated controls (ND₂₄) and decreased after removal of docetaxel and incubation of cells for 24 h in either drug-free media (Doc₁₈→ND₂₄) or flavopiridol (Doc₁₈→F₂₄). However, the decrease in cyclin B1 protein expression was greater for Doc₁₈→ND₂₄ as compared with Doc₁₈→F₂₄. This did not correlate with the decrease in cyclin B1/cdc-2 kinase activity, which was greater for Doc₁₈→F₂₄ than for Doc₁₈→ND₂₄ (Fig. 2).

To investigate whether the cells exiting mitosis express interphase markers, we examined the phosphorylation status of pRb. As shown for MKN-74 cells in Fig. 4, after 24 h of flavopiridol treatment (F₂₄), the levels of ppRb were lower compared with untreated control (ND₂₄). Twenty-four h after docetaxel removal and addition of flavopiridol (Doc₁₈→F₂₄), the levels of ppRb were decreased compared with Doc₁₈→ND₂₄. Pretreatment of cells with flavopiridol before docetaxel (F₂₄→Doc₁₈) also showed lower ppRb compared with treatment with docetaxel alone (Doc₁₈).

The apoptosis assays indicated that sequential treatment with docetaxel followed by flavopiridol achieved the best results. To investigate whether these in vitro observations are also reflected in vivo and to optimize the time interval between the two drugs, MKN-74 cells were established as xenografts in nude mice (day 0). Mice were treated 3 days later with either each drug alone (Doc or Flavo), concomitantly (DocF0), or sequentially with docetaxel first followed by flavopiridol at 3-, 7-, and 16-h intervals (DocF3, DocF7, and DocF16) on days 3, 6, 10, and 13. Mice were also treated in the reverse order with flavopiridol first followed 7 h later by docetaxel (FDoc7). As shown in Table 2 and Fig. 5, the sequence and the interval between docetaxel and flavopiridol were important determinants of the percentage of tumor regressions in xenografts. By day 17, the greatest tumor regressions (the percentage decrease in tumor volume) were observed if the intervals between docetaxel and flavopiridol were 3 and 7 h. There was a 394 ± 25% increase in mean tumor volume in mice treated with docetaxel alone; whereas in mice treated with docetaxel followed 7 h later by flavopiridol, the mean tumor regression was 18 ± 4% (P = 6 × 10⁻⁴). If the interval between docetaxel and flavopiridol was decreased to 3 h, the tumor volume decreased by 1 ± 9% by day 17. This was statistically superior to docetaxel alone (P = 0.002), although it was not statistically different than docetaxel followed 7 h later by flavopiridol. As shown in Fig. 5, a difference in tumor growth was noted as early as day 6, which was 3 days after the first day of therapy (day 3). These reductions in tumor volumes on day 6 were statistically significant for both the 3- and 7-h intervals when compared with docetaxel alone (P = 0.002 and P = 0.0006, respectively, versus docetaxel alone on day 6).

Even when flavopiridol was administered 16 h after docetaxel, the effect of the sequential therapy was still superior to treatment with docetaxel alone (P = 0.019), although the effect by day 17 was less apparent than the effect at either the 3- or 7-h interval. However, when docetaxel was given concomitantly with flavopiridol, the effect of the combination therapy was statistically no different than that of docetaxel alone. Similarly, when flavopiridol was administered 7 h before docetaxel, the tumor regression was similar to that seen with docetaxel alone. In general, when flavopiridol followed docetaxel treatment, the tumor regressions were higher as compared with treatment with docetaxel alone, docetaxel and flavopiridol together, or flavopiridol before docetaxel. However once treatments were stopped (day 13), all tumors began to regrow by day 20. Mice that received only flavopiridol showed an increase in tumor growth that was similar to that seen in untreated controls.

Mice that were treated with only docetaxel lost an average of 6 ± 9% body weight, whereas mice treated with docetaxel followed 3 or 7 h later by flavopiridol lost an average of 10 ± 6% body weight, indicating that the time intervals between therapies did not render the combinations more toxic. This would indicate that a difference in weight loss between these treatment conditions does not explain the significant difference in response. No animal deaths were observed with docetaxel alone or with docetaxel followed 0, 3, 7, or 16 h later by flavopiridol.

We have previously reported that flavopiridol enhances the induction of paclitaxel-induced apoptosis in vitro in both breast and gastric cancer cells and that this effect was sequence dependent, such that paclitaxel therapy needed to precede the flavopiridol treatment (10). In this study we elected to determine whether these observations could be extended to another taxane, docetaxel, both in vitro and in vivo. Our studies indicate that the flavopiridol enhanced docetaxel-induced apoptosis by 2–3-fold in MKN-74 gastric cancer cells. This effect was highly sequence dependent, such that docetaxel needed to precede flavopiridol treatment to achieve this effect. In addition, pretreatment with flavopiridol before docetaxel antagonized the docetaxel effect. The downstream events that execute the docetaxel-induced cell death involve activation of caspases. In combination therapy of docetaxel followed by flavopiridol, there was enhancement of caspase-3 activation and poly(ADP-ribose) polymerase cleavage (data not shown).

We have extended these observations to an identical in vivo model. Using MKN-74 xenografts, our results again indicate that flavopiridol enhanced the effect of docetaxel. This effect was also highly sequence dependent. In fact, when flavopiridol was administered concomitantly with docetaxel or in the reverse order, the changes in tumor regression were not significantly different than those seen with docetaxel alone. It was only when flavopiridol followed docetaxel that a decrease in tumor volume relative to docetaxel was noted. This effect was both sequence and time dependent, such that the maximum effect was observed when flavopiridol followed docetaxel by 3 or 7 h. However, even with a 16-h interval, the effect of the combination was still greater than that seen with docetaxel alone.

The results presented here indicate that docetaxel (Doc₁₈) as a single agent induces mitotic arrest (i.e., increase in expression of mitotic markers such as MPM-2 labeling, cyclin B1 expression, and cyclin B1/cdc-2 kinase activity) and increased phosphorylation of pRb. We do not observe significant apoptosis at this time. Flavopiridol on untreated cells directly inhibits CDK2 and CDK4, the two kinases critical for pRb phosphorylation. This accounts for the marked decrease in pRb phosphorylation and the decrease in the S-phase population observed on the MKN-74 cells treated with flavopiridol alone (F₂₄). In contrast, when cells arrested in mitosis by docetaxel exit the M phase and enter G₁ (Doc₁₈→ND₂₄), there is a decrease in mitotic markers and a decrease in cyclin B1/cdc-2 kinase activity, and the cells do not undergo cytokinesis (i.e., cells remain with 4n DNA content). We refer to this stage as pseudo-G₁ (10). pRb is found in the dephosphorylated form in the preceding hours of G₁ and needs to undergo phosphorylation to initiate the G₁-to-S-phase transition (22). Thus, pRb dephosphorylation is a marker of G₁ entry. This explains the decrease in pRb phosphorylation observed with Doc₁₈→ND₂₄, as compared with Doc₁₈ alone. We also observed a significant increase in apoptosis during Doc₁₈→ND₂₄ compared with Doc₁₈ alone. This may be due to abnormal exit of cells from mitosis.

However, with the addition of flavopiridol on docetaxel-treated cells (Doc₁₈→F₂₄), there is a greater decrease in both cyclin B1/cdc-2 kinase activity and MPM-2 labeling (22% versus 11%) when compared with docetaxel followed by drug-free medium (Doc₁₈→ND₂₄). This results in an increased accumulation of hypophosphorylated pRb, indicating greater mitotic exit and entry into pseudo-G₁. This greater degree of inhibition could not be explained by the decrease in cyclin B1 protein expression, which, in fact, was greater for Doc₁₈→ND₂₄ than for Doc₁₈→F₂₄. Instead, this must be due to direct inhibition of cyclin B1/cdc-2 kinase by flavopiridol. Collectively, these results indicate that flavopiridol after docetaxel treatment enhances the exit of cells out of mitosis with more rapid entry into a pseudo-G₁ state.

With the reverse combination of flavopiridol followed by docetaxel (F₂₄→Doc₁₈), there was no increase in cyclin B1 expression or in the activation of cyclin B1/cdc-2 kinase when compared with docetaxel alone (Doc₁₈). This resulted in a lower level of phosphorylated pRb. This is consistent with the arrest in G₁ (4% in G₁ with Doc₁₈ alone versus 50% with F₂₄→Doc₁₈) and a decrease in both the 4n (G₂ + M) and M-phase population (70% and 62%, respectively with Doc₁₈ alone to 43% and 22% with F₂₄→Doc₁₈). Docetaxel-induced apoptosis is dependent on activation of cyclin B1/cdc-2 kinase and mitotic block. It has been shown that the prevention of mitotic block during paclitaxel treatment by use of an antisense to cyclin B1 also prevents cell death (14). Thus, pretreatment with flavopiridol antagonized the effect of docetaxel on these cells. Furthermore, the inhibition of docetaxel-induced cytotoxicity by flavopiridol (by cdc-2 kinase inactivation) confirms the importance of mitotic block and exit in docetaxel-induced apoptosis and provides an explanation for the ineffectiveness of the reverse combination (F₂₄→Doc₁₈). Despite this antagonism in vitro, we did not observe this type of antagonism in vivo when flavopiridol was administered 7 h before docetaxel. This may simply reflect differences observed when tumor cells are exposed to drug in monolayers as opposed to tumor xenografts. Alternatively, in vivo we may have missed the time window in which antagonism of flavopiridol could have been observed (for example, flavopiridol followed by docetaxel in 3 h rather than 7 h). However, this was not tested in this tumor model because our goal was to maximize the effect of the drug combination.

A Phase I study of weekly, sequential docetaxel followed by flavopiridol therapy is now under way at our cancer center. From the animal studies, it would appear that both the sequence of drug administration and the time between drug administrations are crucial in order for flavopiridol to augment the effect of docetaxel. It is clear from the studies that giving the two drugs at comparable times will lead to effects that are no greater than docetaxel alone. It seems from the animal data that an interval of 3–7 h between sequential docetaxel and flavopiridol is needed to produce this effect. Interestingly, in our evaluation of sequential irinotecan and flavopiridol, the maximum enhancement was again observed with a 7-h interval between the initial administration of irinotecan and the subsequent treatment with flavopiridol (11). Admittedly, it is difficult to extrapolate animal data to the development of human clinical trials. Although there may be some specific metabolic effect in the mice that necessitates these drug intervals, the data do suggest a unique biological window in which the application of flavopiridol relative to the chemotherapy is critical to achieve a chemopotentiating effect. This may be based on the specific cell cycle effects of each drug under investigation. We do not believe this was due to the i.p. administration of flavopiridol because the drug is rapidly absorbed when given by i.p. administration in vivo. Therefore, in our Phase I study, we plan to use this preclinical data, such that patients with advanced gastrointestinal cancers will be treated with a fixed dose of weekly docetaxel followed in 4 h by escalating doses of flavopiridol administered over 1 h. We hope that this will develop into a new approach in the treatment of gastric cancer, which is in desperate need of new therapeutic approaches.