Edited excerpt from the discussion:

DR McCARTY: This 78-year-old white woman was originally diagnosed in 1991. She was treated with left mastectomy followed by five years of tamoxifen. I first saw her in January of 2004. At that time, she had developed two lesions on her chest wall and left axilla. A biopsy was performed and confirmed adenocarcinoma, consistent with the breast primary. The tumor was ER-positive and HER2-negative. The patient had surgical resection, and I then recommended chest wall irradiation, which she refused.

We did a bone scan, which revealed widespread bony metastases which were asymptomatic. I then put the patient on letrozole and also recommended bisphosphonates intravenously. Again she refused, saying, “I don’t want anything IV.” So I followed her on letrozole.

DR LOVE: Can you talk a little bit more about this woman and her life history?

DR McCARTY: At first, she would come to the office by herself and later with her daughter. She was very set in what she would and would not accept. She had been working all her life at a desk job. She had a good performance status and was totally asymptomatic from the bone metastases. It was quite a surprise when we found disease on the bone scan.

DR LOVE: What was her concern about receiving intravenous therapy?

DR McCARTY: She just didn’t want to be stuck with a needle, and she didn’t want to receive anything that in her mind might be very “strong” and cause more side effects. Even though I tried to educate her otherwise, I wasn’t able to sway her. We did do a bone density scan that showed some degree of osteoporosis. She finally agreed to take an oral bisphosphonate, and she was treated with risedronate. Interestingly, when we repeated a bone scan a year later, her bone density had improved significantly, despite the aromatase inhibitor.

However, by October of 2004, she started presenting with multiple skin lesions. These were red and raised, initially on her chest wall and upper abdomen. Then they began to appear on her neck, and she then developed one on her right eyelid that grew to three centimeters. During this time, her CA27-29 had gone up steadily to 729, and then for the first time, her liver function tests came back as abnormal.

Due to the abnormal liver function tests, I ordered a CT scan that revealed two lesions in the right lobe of the liver. She wouldn’t even consider doing anything locally there, and I talked to her again about chemotherapy.

DR LOVE: Before we move on to that, did you try any other hormones?

DR McCARTY: We did not.

DR LOVE: How long had she been on adjuvant tamoxifen? And then how long after that was her relapse?

DR McCARTY: She was on tamoxifen for five years, and then nine years later relapsed.

DR LOVE: Do you think she responded to the letrozole?

DR McCARTY: I don’t think she did. She seemed to have a slow, steady progression from January to October with no significant evidence of a response. That is one of the reasons that I was not enthusiastic about going to another hormonal approach.

DR LOVE: Joyce, considering this woman’s attitude toward therapy, would you have tried additional hormones? And once you had decided hormones were not an option, how you would have thought through chemotherapeutic options?

DR O’SHAUGHNESSY: The fact that she didn’t respond to letrozole and progressed in a fairly short period of time is very interesting. In a recent issue of the JCO, there was a nice paper about the IMPACT trial comparing tamoxifen versus anastrozole preoperatively. Mitch Dowsett provides a detailed analysis of the strength of ER, PR and HER2 in terms of reduction in Ki67. Basically, anastrozole was definitely superior to tamoxifen, particularly for patients with low ER positivity (Dowsett 2005). So it’s interesting that she didn’t respond.

DR LOVE: Specifically, what chemotherapy would you be thinking about in this situation?

DR O’SHAUGHNESSY: I believe that we will ultimately discover that different chemotherapeutic agents work better or worse in different metastatic sites. I have been impressed with the combination of a taxane and capecitabine in patients with the bone and liver metastases (Blum 2004; Gradishar 2004; Mackey 2004; O’Shaughnessy 2002).

Of course, this raises the issue of combination versus sequential single-agent chemotherapy, and in a patient who is 78 years old, I wouldn’t necessarily recommend a combination.

However, for younger patients in which the tragedy of metastatic breast cancer has just taken over their lives, and they want the best chances of response for as long as possible, I usually use either weekly paclitaxel or every three-week docetaxel with capecitabine.

I think the capecitabine dose of 825 mg/m² bid, 14 days on and seven days off, is now pretty well established for combination therapy. I will usually treat for six or seven cycles with the combination, stop the taxane and keep going with capecitabine. The duration of response that I have seen with some patients has been remarkable. That is not to say that you wouldn’t have seen the same if you had used sequential therapy, but the duration of responses in a number of patients is impressive, and it’s gratifying.

We don’t have data yet on the combination of nab paclitaxel and capecitabine, although we are interested in conducting a Phase II trial with that combination. I have no reason to suspect it would be any different, if not better based on the increased tumor penetration with nab paclitaxel.

In this month’s JCO, an Italian group reported a Phase II first-line metastatic trial of capecitabine in patients with a mean age of 73 (Bajetta 2005). Patients were started on capecitabine 1,250 mg/m² bid. Two deaths occurred, so they reduced the dose to 1,000 mg/m². The report included about 40 patients treated with 1,250 mg/m² and another 43 patients treated with 1,000 mg/m². In the trial, they observed an acceptably low rate of Grade III toxicities with the lower dose in under 10 percent of patients. In both cohorts, the response rate was 35 percent, which is pretty impressive, and another third of patients treated with 1,000 mg/m² had stable disease for more than six months. That is getting remarkably high.

Therefore, I would be comfortable with single-agent capecitabine for this patient. I don’t know if I would start at 1,000 mg/m2. Usually, as people get older, they experience a little diminution in renal function, so I would probably approach the dose gingerly. You can always go up a little bit later.

DR LOVE: Cliff, could you comment on this case and some of Joyce’s points?

DR HUDIS: I would treat her with hormone therapy. She’s 78 years old and walked in and said, “I don’t want IV therapy.” I read that as, “I don’t want to be sick.” She went more than a decade from her presentation until she recurred and now has incurable, bone-dominant disease, albeit with liver metastases. I’m not convinced that there is enough of an upside to chemotherapy to justify recommending it to her, let alone early in the course. But you have to see the patient, sit across the desk from her and talk to her.

DR LOVE: What sequence of hormones do you think you’d be likely to start with?

DR HUDIS: The fact that she did so well after tamoxifen argues that she may be somebody who is sensitive to a SERM or a SERD as opposed to pure estrogen deprivation. I would consider tamoxifen if she refused the parenteral administration of fulvestrant. Either way, I have no firm dogma about that.

DR LOVE: Assuming she reached the point where you considered her hormone resistant, still in the same situation, what would you have done?

DR HUDIS: Exactly as Joyce described. I would use single-agent capecitabine as my first chemotherapy for her. It might be the only chemotherapy she ever receives.

DR LOVE: Where do you see nab paclitaxel fitting in? And what are your thoughts about the recent press release describing an advantage for bevacizumab added to paclitaxel as first-line therapy for metastatic disease (Miller 2005a)?*

DR HUDIS: Let’s take the second question first. As it applies to this case, we have a 78-year-old woman, and many patients in this age group are going to be screened out of the routine use of bevacizumab because they are going to be found to have unsuspected ischemic changes in their brains and other risks for bleeding. They also often have hypertension, renal insufficiency and other risk factors. Those are going to reduce the population of potential candidates.

If this patient were willing to take firstline chemotherapy, and if your conclusion was that she should receive paclitaxel, then I think one would have to consider adding bevacizumab once we see the full data and can interpret it. However, given what I learned from the Wall Street Journal this week, which is that the addition of bevacizumab provides a four-month prolongation in time to progression, I believe it certainly will be motivating, especially if it comes at an acceptable toxicity price.

With regard to nab paclitaxel, I have great enthusiasm for that as a therapeutic advance in the area of the taxanes. The one randomized trial we have tells us that it’s probably a better taxane in terms of activity, but frankly, even if you threw out that advantage, the fact that it’s a halfhour infusion, doesn’t require specialized IV tubing and causes no hypersensitivity reactions makes it more practical.

That’s why nab paclitaxel doesn’t have to be better than paclitaxel to be very useful. It only has to be safer and equally effective. I have real confidence that this drug will be popular in many communities.

DR LOVE: Joyce, what are your thoughts about Cliff’s points?

DR O’SHAUGHNESSY: Because the tolerability is extremely good, I tend to use nab paclitaxel at 100 mg/m² weekly. The data right now are in taxane-pretreated patients, so most of the patients that I’ve used it with have been heavily pretreated. In my own practice, it’s been another nice treatment option for patients.

Ongoing trials, including one at Memorial, are evaluating weekly front-line nab paclitaxel. I want to see more data, but it has been more force of habit that I am using it at the same place we used the drug in our trial.

DR LOVE: Joyce, what are your thoughts about the bevacizumab data?

DR O’SHAUGHNESSY: It’s not too surprising that it was a positive trial (Miller 2005a; [6.1]) because as a single agent, it has activity. It also had activity in the capecitabine randomized Phase III trial (Miller 2005b). That was a late-line population, and it’s difficult to change the median of anything when only a small percentage of patients benefited.

I guess the number one thing I want to see are some interesting exploratory subset analyses. For example, is the benefit of bevacizumab going to be largely seen in the higher-grade tumors like ER/PR-negative or perhaps HER2-positive tumors?

What about indolent disease? You might have patients with some indolent biology, so I need to see the data. However, this sounds like it is a real advance for select patients in whom you believe it will be safe to administer bevacizumab.

DR LOVE: It’s interesting that both of you came up with capecitabine as this woman’s next chemotherapy. It sounds like that would have been your choice regardless of her attitude toward IV chemotherapy. Is that correct? Cliff, if this woman had no barriers to getting IV chemotherapy and was 55 years old, would that have changed your approach?

DR HUDIS: Globally, the answer is it might have changed my approach. I don’t harbor a firm belief that the order of single agents matters as much as people believe it does. For patients reluctant to receive chemotherapy for reasons that are largely emotional, capecitabine allows them to make that transition and say to themselves, “I’m not getting intravenous therapy. It’s not so bad.”

The fact that capecitabine is as active as many of the intravenous therapies that we routinely use makes this sort of a silly point, but it is one that people buy into. So if I have a patient who is reluctant to start intravenous therapy, I will look toward capecitabine.

If I have an older patient, to tell you the truth, it cuts both ways. This patient was reflective of one attitude, but the truth is that capecitabine does bring up compliance and safety issues related to self-administration. It’s not crystal clear to me that it’s always safe for or better for a person to take medication at home. Sometimes you have a little more control over them if you can give them intravenous therapy and withhold it when you should.

DR LOVE: Can we follow up with this patient, Dr McCarty?

DR McCARTY: This patient was placed on capecitabine 1,000 mg/m² initially. We made some minor dosage reductions, and she did quite well. Her skin lesions gradually disappeared over the next few months, and her CA27-29 came down from 729 to 26, in the normal range. When she came in for her eighth cycle, she was starting to have some moderate hand-foot syndrome at that point. We performed a CT scan, and it showed disappearance of the liver lesions (6.2, 6.3). By this point, all of her skin lesions were completely gone. We repeated a bone scan and there was a significant decrease in both the number and intensity of bone metastases.

We opted to stop the capecitabine because of the hand-foot syndrome (Scheithauer 2004), and she has only been off the drug for a month at this point. I have recommended to her that we go back to hormonal therapy at this time.

DR LOVE: When she recuperates from the hand-foot syndrome, Joyce, what would you think about?

DR O’SHAUGHNESSY: I would treat her with either tamoxifen or fulvestrant. We don’t have any randomized data, but I probably would give her fulvestrant.

I’ve had some success — particularly elderly patients like this — who have had an intervening course of chemotherapy and have then gone back to fulvestrant afterwards.

So just based on my experience, I probably would treat her with fulvestrant.

There will be interesting preclinical data at ASCO this year from Larry Norton and his group. They have done some interesting preclinical modeling about the duration of capecitabine treatment, the duration of time off and the dose (Lokich 2004; Saeki 2005). The bottom line that I took away from a brief presentation that I saw is that if you give animals more than eight or nine days of capecitabine, you receive additional toxicity without any additional benefit.

We have all struggled with how to change the schedule and dose of capecitabine in patients who are clearly benefiting and motivated to keep going with therapy. Recently, I have been emboldened by these preclinical data and have cut down the duration of treatment to approximately a week and increased the dose a little bit, but have given a good amount of time off. Not just a week, but 10 to 12 days off.

That extra time off makes a difference in terms of healing of the skin, and at least in my anecdotal experience, I haven’t lost anybody’s response. Before, I was all over the place with the schedule trying to figure out for each patient what might work, but now I’m going up on the dose, shortening the duration and providing longer time off.

DR HUDIS: Those data are going to be presented this week at AACR by Tiffany Traina — who is one of our fellows — Maria Theodoulou, Larry Norton and me (Norton 2005). The preclinical data are exactly as Joyce described, but what we did was a little different. With preclinical scientists at Roche Japan — that’s where the animal experiments were done — in addition to conventional dose escalation with a fixed schedule, we also looked at a serie s of schedules — a couple days of treatment s, long breaks, intermit tent t her apy and prolonged therapy.

Exactly as Joyce said, at a certain point, using a mat hematical model, we coul d see that we were no longer killing that many more cancer cells with that cycle of therapy but were indeed adding more toxicity with each day of treatment.

Even worse, because you add toxicity, you delay the initiation of the next cycle. So by extension, the argument would be that 14 days on and seven days off of capecitabine is wrong. You may only need the f irst seven days of treatment. The second seven days may add no efficacy advantage and delay the initiation of the next cycle.

We are conducting a Phase I trial eval uating seven days on and seven days off, which is another empiric regimen that people have pl ayed with but now we have preclinical evidence to support.

DR LOVE: Joyce, you have a US Oncology adjuv ant trial analyzing capecitabine and docetaxel. Can you talk a little about that? Also, in view of some of these issues about schedule and dose, is there a future for capecitabine in the adjuvant setting (6.4)?

DR O’SHAUGHNESSY: We are conducting both an adjuvant and a neoadjuvant trial. The adjuvant randomizat ion is AC followed by docetaxel at 100 versus AC followed by docetaxel at 75 mg/m2 with capecitabine. The capecitabine is administered at 825 mg/m2 bid, 14 days on, seven off.

We started wit h a dose of 950 mg/m² bid, 14 on, seven off, based on results from the pivot al Phase III trial that led to the approval of capec itabine with docetaxel. In that trial, the original dose was 1,250 mg/m2 bid, 14 on, but by cycle two, the patients were down and the median delivered dose was actually 25 percent less. So the survival advantage seen with capecitabine in that clinical trial was with a delivered dose of 950 mg/m2 bid over 14 days, starting in cycle two. That was the reason we went forward with capecitabine at 950 mg/m².

After the first 150 patients were accrued to the adjuvant trial, 75 on each arm, the DSMB had a look at the data. We also noted that a number of patients came off of therapy due to toxicity.

Also, more dose reductions and dose interruptions existed in the capecitabine arm.

We reduced the dose to 825 mg/m² bid. The increased toxicity may have been due to the timing of therapy right on the heels of the AC. Particularly, what you see right after AC is a little more mucositis than when you administer the regimen by itself.

The DSMB has now observed 500 patients, 250 on each arm, and they have determined that the 825 mg/m² bid is a safe, deliverable and feasible dose for these patients.

We have about 225 more patients to accrue to that trial and hope to have some preliminary data, perhaps by San Antonio 2007.

We’re also conducting a preoperative trial of FEC 100 times four followed by XT, with the docetaxel 35 mg/m2, day one and day eight, and with capecitabine 825 mg/m² bid, 14 days on, seven days off.

We’re excited today because we are opening another aspect of that trial to add trastuzumab to the FEC, dropping the epirubicin dose down to 75 mg/m².

So it’s FEC 75 with trastuzumab, followed by the XT with the trastuzumab. This follows through on the exciting pathologic complete response rate of 67 percent that Aman Buzdar reported (Buzdar 2005).

DR LOVE: Just to pick up on that question, Joyce, how low can you go with the capecitabine dose and still see efficacy (6.5)?

DR O’SHAUGHNESSY: A Japanese trial presented at ASCO several years ago used a total daily dose of 1,650 mg/m², 21 days on and seven days off (Kusama 2001) in anthracycline- and taxane-pretreated patients.

The response rate was right around 25 to 28 percent, but those are the only data I’m aware of that have gone lower than the 1,000 mg/m² bid.

The benefits of capecitabine can be extremely long lived. I can only remember one patient a couple of years ago who did need to be hospitalized for some diarrhea. She was a frail, elderly patient and didn’t stop treatment appropriately. I believe with vigorous patient education and spelling out specifically — “Do not even consider taking another pill, if you have X, Y and Z” — that the toxicity is manageable.

Select publications

* Note: This recording took place prior to the 2005 ASCO meeting at which the bevacizumab/paclitaxel data from E2100 were presented.