Lung Cancer Update 2007 (3)

DR JOHNSON: In terms of our algorithm for the management of metastatic disease in the clinical setting for patients who are not in the EGFR-enriched populations, we follow the Eastern Cooperative Oncology Group (ECOG) algorithm. For patients with adenocarcinoma without brain metastasis, serious cardiovascular or cerebrovascular problems or clotting, we recommend paclitaxel, carboplatin and bevacizumab. For patients with squamous cell carcinoma, brain metastasis or hemoptysis, we administer paclitaxel and carboplatin without bevacizumab. We try to utilize the same drugs off study as we do on study. For patients with a number of comorbidities, we administer a single agent such as vinorelbine.

For patients treated with paclitaxel, carboplatin and bevacizumab, side effects we see include hypertension and an increased risk of clotting, bleeding and proteinuria, which are all manageable. We also see an increased risk of deep venous thrombosis and pulmonary emboli.

For patients in second-line therapy off study who have been treated with two agents — most commonly carboplatin/paclitaxel in our setting — and have a good response and go off therapy for an extended period, we’ll commonly go to docetaxel as second-line therapy. For a patient who shows a mediocre response to initial chemotherapy, we will generally use erlotinib as the second agent. We often use pemetrexed as the third-line agent. For almost everybody off study, we use one of the three approved agents for second-line treatment — pemetrexed, docetaxel or erlotinib.

Click on the image to enlarge

Lung Cancer Update 2007 (3)

DR WAKELEE: In the setting of first-line metastatic disease, I believe we’re still left with a platinum doublet, potentially even a nonplatinum doublet, as the chemotherapy base. Additionally, some trials are evaluating carboplatin and pemetrexed as another platinum doublet. It’s a reasonable option, and it may be less toxic than some of the other regimens, but it’s not better, and you lose an agent that’s commonly used in the second line. Is that good or bad? I believe it’s a matter of order, and I do not believe it’s a huge step forward; it’s simply a nice alternative.

For my patients in a nonprotocol setting who are not eligible or for whom I don’t want to administer bevacizumab, in the first line I tend to use carboplatin and gemcitabine. I’ll also use carboplatin/paclitaxel. It’s between those two options, and we discuss the toxicity differences and scheduling differences with each patient. Occasionally I’ll use a cisplatin backbone if the patient wants to be extremely aggressive.

Interview, May 2007

DR ALAN B SANDLER: ECOG-E4599 showed that the addition of bevacizumab to paclitaxel/carboplatin in metastatic disease provided a progression-free survival advantage over that same chemotherapy alone. I believe that bevacizumab and other VEGF-mediated agents or VEGF-directed agents have two distinct mechanisms of action.

First, I believe they have an effect on the tumor itself. Tumors have leaky vasculature, and an anti-angiogenic agent such as bevacizumab appears to help prune some of the newer vasculature, diminishing the leakiness and, therefore, allowing for better drug penetration by decreasing the interstitial fluid pressure in the tumor.

Dr Willett at Harvard showed this in patients with rectal cancer, demonstrating decreased interstitial pressure in rectal tumors pre- and postbevacizumab. That’s one effect: providing better chemotherapy penetration to the tumor. That would not appear to be as important in the adjuvant setting, in which there is no tumor.

The other effect relates to the concept of eliminating or reducing the development of new vasculature for the initial microscopic and then small tumors, stopping the new blood vessels from forming and turning the cancer into a chronic disease. It’s hoped that mechanism will play a major role in the setting of early disease.

Click on the image to enlarge

Lung Cancer Update 2007 (4)

DR EDELMAN: In the advanced disease setting, I’ve held fairly closely to the ECOG-E4599 eligibility criteria. We had discussions in which people were concerned because of the neutropenic fever or the hemoptysis seen with the addition of the VEGF inhibitor, but again, viewing this in the aggregate, patients did better with bevacizumab. They lived longer, and so if we have patients who would have been eligible for that study, then we do approach them about the use of bevacizumab.

I have used it pretty much exactly as it was used on E4599. The only difference is that I tend to use less cytotoxic chemotherapy — I use four cycles, not six, and I base that on my belief that the evidence is pretty compelling that pushing the cytotoxics does not aid you after four courses of therapy. I’m an advocate of evidence-based medicine, but here and there one can do an induction. I could certainly be criticized, but I believe it’s a reasonable approach and it’s well tolerated.

Lung Cancer Update 2007 (2)

DR JOAN H SCHILLER: In ECOG-E4599, our statisticians conducted an unplanned analysis evaluating which subpopulations benefited from bevacizumab and which did not. They reviewed all the predefined stratification factors, and none of these resulted in a difference between whether or not patients were likely to benefit from bevacizumab. A difference did appear between men and women, however, which is puzzling. Both men and women derived a benefit from bevacizumab in terms of response rate and progression-free survival, but for some reason the men seemed to have a longer overall survival rate if they were receiving bevacizumab compared to chemotherapy alone. For the women, the survival appeared to be roughly the same.

Granted, this was a retrospective analysis, and the women did extremely well on the control arm. The median survival for women on the control arm without bevacizumab was more than 13 months compared to a median survival of approximately eight and a half months for the men on the control arm. I don’t know why women on our control arm did so well, but that’s one reason why we didn’t see a benefit.

Click on the image to enlarge

One explanation why women on the control arm may have done so well is that they may have received more epidermal growth factor inhibitors. This study was conducted when gefitinib and erlotinib were just coming out. The big news was that women seemed to benefit more than men from those drugs, so it’s possible that the women were more likely to receive those drugs than the men, and that’s why the women on the control arm performed better.

As a practicing oncologist, I have been treating women with lung cancer with bevacizumab, based on the fact that this was not a prospective analysis, and no difference was seen in the colorectal carcinoma data evaluating men versus women. In the breast cancer data, it appears that women are benefiting from bevacizumab. Based on those factors, I’ve continued to treat women. Moving forward, gender will be a major stratification factor.

Lung Cancer Update 2007 (3)

DR JOHNSON: My experience with the carboplatin/paclitaxel/bevacizumab regimen is that it does have added toxicity. We have used anti-angiogenic agents for approximately five to seven years, so we have experience with them. The oral pills that are the VEGF II inhibitors share some of the same side effects, which include high blood pressure and increased risk of clotting, bleeding and proteinuria. The side effects are manageable, as with many of the other agents we use. The hypertension is treatable, and we handle most of it ourselves.

The risk of clot is real, however. We see an increased risk of deep venous thrombosis and pulmonary emboli. When you apply the algorithm of limiting this treatment to the adenocarcinomas with no history of hemoptysis, you don’t see much of a problem with hemoptysis and with other risks of bleeding, although in the randomized trial it clearly runs around two or three percent. This is true for both the US and European trials, even within that selected population.

In terms of the patients who have bleeds, we’ve identified squamous cell cancer and a history of hemoptysis as risk factors.

In my experience, one of the differences with the anti-angiogenic agents, and this is true of both bevacizumab and the small-molecule inhibitors of the VEGF receptors, is that these lesions can cavitate. It’s different than what we’ve typically seen with cytotoxic therapy alone. These spherical lesions hollow out in the middle and develop a cavity, which appears to be associated with the development of hemoptysis.

The assumption is that the antiangiogenic agents block the blood flow to the middle of the tumor, it necroses and you lose some of the structure. The blood vessels can’t regrow, and they bleed into it.

In the trials we designed using agents directed against VEGF, we have not held therapy if it’s an uncomplicated cavitation. With any hemoptysis an oncologist will obviously stop treatment, but so far we don’t have enough evidence to stop treatment for a cavitation.

Click on the image to enlarge

Lung Cancer Update 2006 (4)

DR SANDLER: We attempted to define prognostic variables for pulmonary hemorrhage in patients who received bevacizumab. It was a case control study in which we combined the data sets from a Phase II study with those from the ECOG-E4599 study and attempted to assess a wide range of prognostic variables to see if one could better define which group of patients was more at risk.

We looked at 22 patients with Grade III or higher pulmonary hemorrhage. Not surprising with the limited number of patients, nothing was statistically significant, but there appeared to be trends for patients with baseline cavitation in their tumors and a history of hemoptysis that predated treatment.

Patients with hemoptysis were not allowed in the study. In ECOG-E4599, it was not specifically written into the study at first, but then one or more patients entered the study who had hemoptysis. After the first 60 or so patients, it was put in specifically as an exclusion criterion.

In our study, we had an independent radiology group examine all the individual CT scans, and tumor size and location did not seem to correlate with pulmonary hemorrhage. We saw a hint that endobronchial disease might be an issue, although that was not statistically significant and it is a very difficult interpretation on a CT scan, and the results were inconsistent across all the CT scans and techniques.

Lung Cancer Update Think Tank 2006

DR LYNCH: We’re participating in a trial to answer the question regarding the use of bevacizumab in patients with treated brain metastases.

Patients have their brain lesions radiated first, and then they receive chemotherapy with bevacizumab. Because of the restrictions in eligibility for ECOG-E4599, which did not allow patients with CNS metastases, I believe we have to follow an evidence-based approach, and I have not been using bevacizumab in this setting outside of a protocol.

DR MILLER: We may be amending the current bevacizumab clinical trials to allow patients with previously radiated brain metastases. These contraindications to anti-VEGF therapy have relative degrees. Certainly squamous histology and hemoptysis are much more powerful contraindications. This drug is very active in patients with glioblastoma multiforme — huge tumors with lots of edema — and we’re undertaking approval strategy trials for those patients. We usually obtain the blessing of a neurologist to use bevacizumab in treated brain metastases, but we certainly have done it.

DR ROY S HERBST: I would wait until more data are available to use bevacizumab in patients with CNS lesions, which I expect will be soon. One trial, called PASSPORT, will determine if you can use chemotherapy with bevacizumab for patients with previously treated brain metastases.

Interview, July 2007

DR HANNA: At ASCO 2007, Christian Manegold presented a randomized Phase III study called the AVAiL trial. Patients with metastatic disease received cisplatin/gemcitabine with a placebo, bevacizumab at 7.5 mg/kg or bevacizumab at 15 mg/kg. The primary endpoint was originally overall survival, but it was amended to progression-free survival.

Both the 7.5 mg/kg and the 15 mg/kg arms had a statistically significant improvement in progression-free survival. Although they were not meant to be compared to one another, the two bevacizumab arms appeared to improve the progression-free survival by just about the same amount. The toxicity profiles of the two dose levels were very similar. The 15 mg/kg dose had a little more bleeding than both the control arm and the 7.5 mg/kg arm. It also had a higher rate of Grade III and IV hypertension. There may have been some slight toxicity disadvantages, with no apparent efficacy advantages with the higher dose.

The incidence of bleeding was low overall. The fear was that the cisplatin/gemcitabine chemotherapy would cause more thrombocytopenia, and when combined with bevacizumab, it might result in some higher-risk bleeding.

The rate of fatal pulmonary hemorrhage was one percent or less on all three arms — that wasn’t the concern. It was other Grade III and IV hemorrhages, which trended a little worse for the 15 mg/kg arm. They didn’t assign a p-value, so I am not sure whether or not it was statistically different.

Lung Cancer Update 2007 (3)

DR SOCINSKI: As a purist, I’d point out that the AVAiL trial wasn’t designed to address the dose question.

The way I interpret AVAiL is that it’s a second positive trial evaluating the use of bevacizumab in combination with chemotherapy — in this case, cisplatin/gemcitabine. The regimen appears to be safe, and both the 7.5-mg/kg and the 15-mg/kg doses improved the primary endpoint of progression-free survival. No survival data were presented.

The 7.5-mg/kg dose did not appear to be less toxic, and I have continued to use 15 mg/kg, based on the survival results from ECOG-E4599. I would bet that at least by ASCO 2008, we will see some survival data from the AVAiL trial, and perhaps that will change our minds about the dosing. For right now, in the absence of survival data in that trial, I’ve continued administering the 15-mg/kg dose.

Lung Cancer Update 2007 (3)

DR WAKELEE: In the United States, carboplatin/paclitaxel with bevacizumab is approved. Given the AVAiL data, gemcitabine/cisplatin with bevacizumab would certainly be a reasonable approach now.

We’re conducting an ongoing trial with carboplatin/gemcitabine/bevacizumab. I wouldn’t say that regimen is “ready for prime time” — not until we have the toxicity data, given the increased thrombocytopenia and neutropenia with carboplatin/gemcitabine. Substituting docetaxel for paclitaxel is also reasonable because we don’t have any toxicity differences that would be of concern.

AVAiL was a European study of gemcitabine and cisplatin with or without bevacizumab. It evaluated two doses of bevacizumab: 7.5 mg/kg or 15 mg/kg. The 15-mg/kg dose was the dose used in the ECOG-E4599 carboplatin/paclitaxel study.

AVAiL demonstrated a statistically significant improvement in progression-free survival — not a big difference, but a real difference statistically — with both the 7.5-mg/kg and the 15-mg/kg doses. The trial wasn’t powered to compare 15 mg/kg to 7.5 mg/kg — only both of those doses to placebo. Overall survival data weren’t mature yet.

The big question is whether we can get away with using 7.5 mg/kg of bevacizumab. I’m cautious still. We don’t have the survival data yet. We have no real way of evaluating any difference between 15 mg/kg and 7.5 mg/kg, even if we could do it statistically. I don’t believe it’s wrong to consider using 7.5 mg/kg, but I’m not ready to make the change in my practice. Certainly we won’t be making a change in the ECOG-E1505 adjuvant trial, in which we’re still using the 15-mg/kg dose every three weeks.

Of note, the bleeding risk in the AVAiL trial was lower than expected. They didn’t observe any significant CNS hemorrhages, which is an issue that had been raised in ECOG-E4599.

Approximately nine percent of patients on the trial were on therapeutic anticoagulation. This was an exclusion criterion for people going on, but once they were on the trial, if they ended up needing anticoagulation therapy, they were able to stay on the study. There was no increased bleeding for that group either.

Click on the image to enlarge

Lung Cancer Update 2007 (3)

DR SOCINSKI: The question that I am asked most frequently by practicing oncologists about metastatic disease is regarding how to approach never smokers. The never smokers represent approximately 10 percent of the population. In my experience, if you use the cutoff of 10 to 15 pack years, the oligosmokers comprise approximately another 10 percent. So one in five patients with lung cancer fall into this category. That’s not insignificant when you consider the number of patients with lung cancer.

The one observation I am convinced of in that population is that anti-EGFR therapy seems to be important. The question I struggle with regarding the never smokers is that many of them are eligible for bevacizumab. What do you do in that setting? Are they candidates for erlotinib or bevacizumab? What’s the role of chemotherapy?

One option is to treat these patients with chemotherapy and bevacizumab and then, as we continue the bevacizumab, perhaps add erlotinib. We have a lot of safety information, and I don’t believe we’re going to harm patients with that approach.

If patients are not bevacizumab candidates — let’s say they have brain metastases — then the question is, should we use chemotherapy followed immediately by a maintenance strategy with erlotinib or chemotherapy with erlotinib or erlotinib alone?

We currently have the CALGB-30406 trial that randomly assigns these patients to erlotinib alone versus carboplatin/paclitaxel with erlotinib. It is exploring two of the three possibilities. You might argue that we should have used four cycles of chemotherapy followed immediately by erlotinib or chemotherapy alone as a control arm, but there are only so many questions you can ask in a randomized Phase II trial to sort out these issues.

Lung Cancer Update Think Tank 2007

DR MILLER: I tend to use erlotinib more either in the first- or third-line setting. I don’t have a huge second-line cohort. I’m driven by knowing either the EGFR mutation status or the clinical factors to incorporate erlotinib into therapy early on. If someone has a favorable profile — a 75 percent positive predictive value for a response to erlotinib, for example — those patients live for a long time, and it’s only a matter of time until we establish a survival benefit for patients with EGFR mutations, treated in that fashion, rather than with chemotherapy. We need the trials to be conducted, and they’re ongoing.

Lung Cancer Update 2007 (3)

DR JOHNSON: We believe it’s important to design a clinical trial to ask the question whether patients who have EGFR mutation-positive disease will perform better with front-line EGFR-directed therapy like gefitinib or erlotinib. We believe that for two reasons. One is that time to progression, at least with the axon 19 deletion mutants, is approximately one to one and a half years, which is two to three times longer than with conventional chemotherapy. In addition, the survival with that group in the retrospective studies is approximately three years, and that’s in comparison to 10 to 12 months with other protocol groups.

It’s important to set up the clinical trials to show that that’s the case. However, the process for obtaining the gene sequence is not easy. You need to have 300 to 500 tumor cells, and they have to undergo DNA sequencing, which is currently the approved test. The results of that test take approximately two weeks. People come to the conclusion that we haven’t seen the definitive evidence that mutation testing should be incorporated in practice.

I believe we need to take the steps to show that is the case. Some of us have been able to integrate it into our practices, and we use it for making decisions regarding whether patients should receive initial treatment with gefitinib or erlotinib.

Click on the image to enlarge

Lung Cancer Update 2007 (4)

DR WALTER J CURRAN JR: Erlotinib is generally better tolerated, especially compared to doublet-based chemotherapy. If it provides the same palliation and arrest of symptoms you might see with doublet chemotherapy at the start and you have a never smoker or an oligosmoker in a low PS state at diagnosis, perhaps a bit on the elderly side, I would like to have data to support erlotinib as initial treatment for that patient. We don’t have the data, but I’m hoping to see it because erlotinib is an option that many patients and families would prefer.

Lung Cancer Update Think Tank 2007

DR KIM: We don’t conduct EGFR mutational testing on everyone who walks through the door. But when we see clinical factors predictive of a response to erlotinib, mostly the never smokers or those with adeno-bronchoalveolar features, I offer them the standard option, which would be chemotherapy/bevacizumab. The second aspect would be to consider the nonstandard therapy, erlotinib.

I presented those options to two different patients on the same day. One was a 60-year-old female who was a light smoker, five pack years, in her twenties, and the other one was 30 to 35 years old and a never smoker. Both elected the first-line erlotinib option.

Interview, July 2007

DR HANNA: In approaching a patient with metastatic disease in the clinical setting in the first- and second-line situation, the most important questions are: What is the performance status of the patient? What are the comorbidities? Is the patient losing weight? What’s his or her appetite like? If a patient is PS 3 or 4, clearly, the right thing is best supportive care.

If the patient is PS 2, but in addition is having significant loss of appetite, loss of weight and comorbidities, then I believe the appropriate thing for that patient is best supportive care, unless the patient is a never smoker. Then I would consider single-agent erlotinib.

For patients who are PS 0 or 1 and don’t have contraindications to chemotherapy, I believe a platinum-based two-drug regimen is standard. For patients who are bevacizumab eligible, the addition of bevacizumab is reasonable. That would include patients who don’t have brain metastases, squamous histology, a history of hemoptysis or uncontrolled hypertension. I treat those patients initially with two courses of chemotherapy and repeat the CT scan. If they appear to be experiencing a clinical benefit, I administer four courses of chemotherapy. Because the ECOG-E4599 study continued patients on bevacizumab, I administer that in maintenance until the time of progression.

Click on the image to enlarge

Click on the image to enlarge

Lung Cancer Update 2007 (4)

DR EDELMAN: I approach PS 2 patients by administering a dose-attenuated, platinum-based regimen. I have found the carboplatin/gemcitabine regimen to be extremely well tolerated. We’ve used that, either the two drugs by themselves or sequentially, followed by weekly paclitaxel.

This is also well tolerated. I believe there’s a fair amount of evidence that says that those who doubt the role of a platinum agent in PS 2 patients should consider repenting: 1) the study that was presented by Obasaju of carboplatin/gemcitabine versus gemcitabine, which showed similar results for the overall population and the PS 2 preplanned subanalysis, and 2) the CALGB study that evaluated carboplatin/paclitaxel versus paclitaxel, which showed that if anything, the PS 2 patients probably see the most dramatic degree of benefit from the addition of a platinum agent.

Why is that, and why do we have this significant split? I believe it’s because PS 2 is a heterogeneous group. There are three groups of patients who end up what we call, in our simplistic way, PS 2. There are those whose performance status has decreased as a consequence of their disease. We all see patients who come in, and they have their families who say, “This guy was working, doing manual labor four weeks ago,” and they say, “Yeah, now it’s a pain to get up and walk around.” Reasonably, they’re still walking around, doing normal activities, but they’re too fatigued or they’re weak and they’ve lost weight. That’s one group. Those are the disease-result PS 2s.

You have a second group that’s been PS 2 for 20 years. They have comorbidities. They never get out of bed to begin with. That’s another bunch, and then you also have frail individuals. We all know this type of patient, the little old lady who looks as if she’s going to get blown away in the next wind storm, and they have poor muscle mass — they’re doing fine until suddenly they’re not. I think those latter two groups, the ones who have significant comorbidities and the frail patients, tend not to do well when they receive a two-drug regimen because, for one reason or another, it aggravates their preexisting comorbidity.

Click on the image to enlarge

Lung Cancer Update 2007 (3)

DR SOCINSKI: There was a very interesting trial presented at ASCO this year designed to evaluate the administration of immediate maintenance docetaxel following four cycles of first-line doublet chemotherapy versus second-line docetaxel per the standard approach, which is to wait until time of progression. Their ability to deliver second-line therapy was much higher in the immediate group than in the delayed group. We know if you let the natural history of this disease play out, things happen and patients who are good candidates for treatment become marginal- or notreatment candidates, based on declining performance status and disease-related symptoms.

Click on the image to enlarge

We know that second-line therapy works, but it can only be effective if you can administer it to the patient. So, post-ASCO 2007, this raised the question in my mind: How do you follow patients after four to six cycles of first-line chemotherapy? What triggers you to institute second-line therapy? I do not consider myself an overtester so I do not perform a lot of x-rays and CT scans while following patients, but I do see them every four to six weeks. I can tell a lot just by their appetite, pain level and chest x-ray. But I tend to think that obviously, you’re not going to benefit patients with second-line therapy if they end up not being good treatment candidates.

This trial also suggested that the time to disease progression was improved with immediate docetaxel rather than waiting until disease progression to institute second-line chemotherapy. There was a trend toward improved survival in this setting. We know that second-line therapy improves survival, but if you don’t receive it, you’re not going to live longer.

Lung Cancer Update 2007 (1)

DR RONALD B NATALE: A major study was presented by Roy Herbst at ASCO 2006 evaluating the combination of bevacizumab and erlotinib in the second-line setting. This was an unselected group of patients, and the objective response rate was close to 25 percent, which is considerably higher than the 10 percent or so objective response rate one would expect with erlotinib alone.

That was encouraging and has led to a definitive randomized Phase III trial in which I am participating, and in fact, I am one of the major accruers to the BETA (bevacizumab and Tarceva^®) study. This is a randomized trial in the second-line setting in which all patients receive erlotinib and then either placebo or bevacizumab every three weeks. That study will answer the question as to whether the combination confers a benefit.

Lung Cancer Update 2007 (3)

DR SOCINSKI: The attractiveness of combining erlotinib and bevacizumab is that they target two new, validated pathways. Each agent has been shown to improve survival. It’s a novel targeted approach that breaks away from some of the traditional toxicities we have with regular chemotherapy. It makes biologic sense to combine them.

The initial data we had from MD Anderson and Vanderbilt were encouraging, and a randomized Phase II trial suggested that bevacizumab added to chemotherapy or erlotinib was better than chemotherapy alone. It also suggested that the combination of erlotinib and bevacizumab appeared as good, with less toxicity, than the chemotherapy/bevacizumab arm. The bevacizumab/erlotinib combination opens up the possibility that some patients may be better served with a noncytotoxic approach.

I believe the jury is still out on that issue. Phase III trials are ongoing that will answer the question about combination bevacizumab/erlotinib. We also have to remember that we may be able to identify with various biomarkers patients who, at least from the erlotinib point of view, may be the best candidates for that approach.

Select Publications