So this is the trial that is most suggestive of a benefit for induction chemotherapy. It required confirmation. One trial is usually not enough, so there was a large intergroup trial that repeated that design, radiation alone versus induction, two cycles, followed by radiation. Bill Soss published this in JMCI. It’s been updated since. But because these were primarily radiation oncologists who were conduction this trial, they were interested in also knowing, could more aggressive radiation just by itself – hyperfractionated radiation – cause similar benefit. You look at median survival; well, this trial confirms the benefit of induction chemotherapy and it does not suggest that, at least hyperfractionated radiation, is of benefit. These are updated survival curves from 2006. Here’s the induction chemotherapy and this is hyperfractionated radiation. So there may be some similar long term benefit, but as median survival goes – you saw the numbers – 14, 12 and 11 months, again favoring induction chemotherapy. Now this trial doesn’t have quite the follow-up yet but it does suggest, at least here on, that the long term benefit – what I showed on the Dillman trial for induction chemotherapy – may be less pronounced here than it was in that trial. These curves are all pretty low out there. So induction chemotherapy is an accepted principle and standard therapy for stage IIIb disease.

What about more aggressive radiation? Well, we mentioned that hyperfractionated radiation didn’t do much, but there is a British trial that looked at something called continuous hyperfractionated accelerated radiation therapy, CHART. What that consists of is 1.5 gray, given three times daily for 12 days in a row. Patients sort of lived right next to the hospital in what they called the CHART house, and got 54 gray and after that of course had mucositis. But that didn’t come until the treatment was completed. And compared that to standard fractionation radiation. Now this trial is a little bit different from most others in that it’s a British trial and many patients didn’t have un-resectable disease, be it IIIb or IIIa, but earlier stage disease which in the United States would be treated with surgery. Furthermore, if you look at the histology; squamous cell, well, poorly, moderately or not specified differentiated, but you add all that up and you come to 80% squamous cell. And we mentioned that that’s no longer representative. And squamous cell probably of all histologies tends to spread a little bit later. So if you select for squamous cell carcinoma, as you may have, that histology which is most amenable to better local control with higher and more aggressive radiation. Large cell adeno, 15%. So it’s a somewhat different trial but within those limitations CHART did better than radiation alone, in terms of survival – median and maybe long term. So there may be a role for more aggressive radiation therapy in non-small cell lung cancer.

CHART meant 12 days in a row, so that included weekends and that is not necessarily appealing to radiation oncologists, or for that matter patients. So in the United States the C was dropped. It’s no longer continuous. And what you have then is hyperfractionated accelerated radiation therapy. Given here Monday through Friday on three consecutive weeks, slightly higher dose but still three daily fractions with four hour interval. This was a pilot trial, objective response rate 54%, median survival 13 months. So this may be good, and when something may be good it ought to be examined in a randomized trial. Two cycles of carbo, Taxol, so standard induction chemotherapy but not using an 80’s regimen, using a 90’s regimen, followed by either conventional radiation – so this is the standard arm – versus followed by hyperfractionated accelerated radiation therapy. Now you can see the conceptual beauty of this because we know that induction chemotherapy improves survival by helping with distant disease and that concomitant or very aggressive single modality radiation therapy improves survival by improving local control. So this just might work. And ECOG is currently putting this into randomized clinical trial.

So where is the field going? This is an important trial by ECOG that looks at several of these drugs as cisplatin-based doublets, versus carboplatin and paclitaxel. This will establish the relative contributions or the sameness of these relative contributions, but maybe alter differing toxicity profiles for these four treatment arms. The trial is completed but not yet evaluated or analyzed. SWOG, we mentioned, completed a trial of cis, vinorelbine versus carbo, paclitaxel. They then decided to look at a phase II trial, looking at gem, carbo followed by a single agent, versus cis, vinorelbine followed by docetaxel. This is not yet active, I think, but it’s a concept. EURTC in Europe is looking at cis, paclitaxel as their standard now versus cis, gemcitabine. And then interesting, in a regimen that no longer contains either cis or carboplatin, in this case. Just simply looking at these two together, Taxol, gemcitabine. CALGB finally is looking at what’s the minimum amount of therapy one can get away with. Do patients really need a doublet such as carbo, Taxol or could the single agent be just as good and as end points of survival and response, quality of life and economics. So you see that this group of drugs, these five drugs, that have come in – with the exception currently of CPT-11 – are really quite established, that they are a little bit better than the regimens of the 1980’s in terms of improving the quality of life and survival a little bit more, and that at this time we are just sort of looking at the fringes of what other kinds of combinations might there be and how can be improve survival a little bit more using these tools. But really looking forward. What’s going to be necessary from here on, are new active principles and therapies and that means another generation of new drugs.

So that is for stage IV disease. Chemotherapy is accepted and if you want to have a standard regimen, I think it could be cis or carbo and Taxol. It could be cis with vinorelbine, cis with gemcitabine and Taxotere maybe, although the definitive randomized trials are not out there yet. The same would be true for CPT-11.

So let’s move on then to the other common stage. Stage IV is about 50% of patients. Stage IIIb is defined as either very large primaries or contralateral mediastinal nodes, or supraclavicular nodes. So look at this picture and you will see, because it is M0, that this can be treated with curative intent because you can still slap a radiotherapy field over all known disease. So you have a modality that can address all macroscopic disease. But it is no longer feasible to do surgery, not curative surgery on this kind of picture. If you treat this IIIb disease with radiation therapy alone, what is regionally apparently confined, two things happen. Patients fail at a rate of over 90%, either in the ipsilateral chest that’s within that very irradiated volume, or systemically because even though we don’t detect it, the majority of these patients actually do have micro-metastatic disease. This has been looked at. This is the French group. Arm A is radiotherapy only. Look at the failure at five years. Distant failure is 70% – that’s probably an underestimate – and local failure, 90%. So this should let us think about how good is radiation to the chest, because it can shrink tumors but it usually does not eradicate macroscopic disease. Out of this then is our charge. We know from head and neck cancer that this is going to lead to multi-modality therapy. Our charge in doing so is to improve both distant and local tumor control.

So we have the same models that we had in head and neck cancer, so let’s look at induction chemotherapy first. Remember, induction chemotherapy primarily targets micro-metastatic disease because it is early chemotherapy, systemically active. It may down-stage or shrink the local tumor, but what it may do best really is eradication of micro-metastatic systemic disease. Here’s a summary in non-small cell lung cancer, what induction chemotherapy does. It increases median survival so it is a standard treatment; from 10-14 months. It may increase long term survival. This is less firm, and it does eradicate micro-metastatic systemic diseases. Does not improve local control. I’ll show you an example of this. This is the classic trial. Dillman et al in the New England Journal, updated in national JMCI in 2006. This is the design; radiation alone, 60 gray – these days you’d give a bit more than that – versus two cycles only of platinum Velban followed by that same radiation therapy. These were unresectable stage III or IIIb disease but good performance status and limited pre-treatment weight loss. So it’s a selected group of patients that’s in this trial. This is the long term outcome, and it establishes two principles. One is, an increase in median survival by about four months, 10-14, and more importantly, at least the possibility of long term benefit. You look at five years, 7% with radiation alone, 17% with chemotherapy followed by radiation. Of course, somewhat flattened curves in both arms after 2-3 years. So patients fail early with lung cancer and then after that actually may live for quite some time.

Now have any of these new regimens been compared amongst themselves? This is a SWOG trial that was presented at ASCO this year, looking at cisplatin, vinorelbine versus carbo and Taxol in this case. And what it showed is that response rates were similar at 28% and 25%. This is cisplatin, vinorelbine. Carboplatin, Taxol one year survival 36% and quality of life improved or stable also in about 60%. So this established, as some of us have felt for awhile, that several of these new regimens are probably going to emerge as of equal efficacy. So cisplatin, vinorelbine, carbo, Taxol, cis, Taxol is shown here, but they differ a little bit in their toxicity spectrum. If you look at neutropenia, that was more common with cisplatin, vinorelbine than with carbo, Taxol. If you look at neuropathy, then that was a little more common with carbo and Taxol. So that can guide you sometimes maybe to the choice of regimen.

What about gemcitabine, then? Gemcitabine is also a new active drug. This is the definitive trial for gemcitabine that got that drug approved for non-small cell lung cancer. I failed to mention, Taxol-2 is approved. A large multi-center randomized trial of single agent cisplatin versus cisplatin plus gemcitabine. Not fully published yet but this is data from the abstract, looking at grade II or IV neutropenia, more common with the combination. Same for thrombocytopenia. Response rate was 31% versus 9%, and median survival 8.7 versus 7.6 months with a statistical trend. This has not yet been fully published.

So what about docetaxel then? Docetaxel; there are no completed randomized trials yet as a first line agent but what’s interesting about docetaxel is that it has been evaluated as a second line treatment agent, which is quite unique for non-small cell lung cancer. Remember, we just established that chemotherapy is worth doing first line. Well, second line; it’s been evaluated in a number of phase II trials where response rates were seen as high as about 17%. A multi-center phase II was done and presented at ASCO in 2006 by Dave Kendara, and the response rate there was 15% in patients who had been previously treated, usually with cisplatin-based therapy. So there are now two randomized trials, second line therapy randomized trials that have been presented. They are not fully published yet. This was shown by Dr. Pasella at ASCO this year and they looked at either docetaxel at two doses, high and less high, versus either vinorelbine or ifosfamide and the choice here had to do with prior chemotherapy in that specific patient. When you look at the one year survival, then the low dose Taxol, Taxotere, was superior to the other two arms. That suggests that if you too toxic that can be detrimental but certainly if you look at response rate at 1%, vinorelbine or ifosfamide, docetaxel seemed to be the more promising approach. There is a second trial, docetaxel versus best supportive care in previously treated patients, because we now feel that most patients should get chemotherapy up front. But second line best supportive care trials are probably okay to do. You look at median survival time; 7 months with Taxotere versus 4.7 months with best supportive care. Quality of life – in particular, pain and fatigue – better. So here Francis Shepard who presented this recommended that second line therapy could be given, but again, at the lower Taxotere dose of 75 mg per meter squared. So this is an emerging approach. It is still hard to make generalizations because the prior therapy for many of these patients was not yet what frequently would now include actually a taxane as first line therapy. So if you have it first line, would you use it second line? Probably not.

This is the second trial, a French study. Control arm here was cisplatin with vindesine, an 80’s type of platinum Vinca, versus cisplatin, Navelbine versus Navelbine alone. And this is the outcome: median survival 40 weeks with cisplatin, vinorelbine versus 30 weeks with cisplatin, vindesine. And of note, vinorelbine as a single agent very similar. Of course, vinorelbine as a single agent, I just mentioned in the elderly, coming to similar outcome of about 27 weeks median survival. So this is the second trial that established cisplatin and vinorelbine as a possible standard regimen, slightly better than those of the 1980’s.

There is a third trial that did not show that difference but it really didn’t have a control arm. The only control arm with this one was vinorelbine versus cisplatin, vinorelbine and it showed a median survival for both of about 33 weeks.

So let’s move on to Taxol then. This is probably the best of the U.S. trials. This was an ECOG trial early on. The control arm was cisplatin, etoposide. It was compared with cisplatin, Taxol low dose versus cisplatin, Taxol high dose with GCSF. Of note, and I think this is the only such trial, the Taxol here was given as a 24 hour infusion. Not as we would do now, one or three hours. Median survival: 7.6 months with cisplatin, etoposide and close to 10 months with the two Taxol arms, and there was no dose response curve. Meaning that there is probably a threshold value for this drug but not a dose response curve. So this is a trial that suggested that cisplatin, Taxol could also be a standard regimen. This trial has not yet been published, even though it is a few years old. But supposedly it is going to come up pretty soon.
Cancer
This is a European trial, cisplatin, Taxol versus cisplatin, VM-26. This is a drug that is not available in the United States, teniposide, but these two are also a standard doublet. Published in JCO by _ in 2006. It’s a large study. You see the doses here; cisplatin 80 plus either teniposide or paclitaxel as a three hour infusion. Look at response rate; 28% versus 40% with paclitaxel. And median survival not different but very good on both arms at almost 10 months. One year survival of 41% and 43%. So this trial was actually not positive for survival, although encouraging on both arms, and here the Taxol was chosen as the future arm because it was felt to be less toxic and provide better quality of life, but had some late neuropathy. This is the outcome. You see no difference between the two but somewhat better median survival.

And you see there are at least four, all asking basically the same thing and all coming to the same conclusion. This is the best one, British Medical Journal, 1995. Median survival six months with supportive care, eight months with chemotherapy. One year survival, 16% versus 26%. All four concluded that chemotherapy improved survival by a little. This is a more recent example. MIC, standing for mitomycin, ifosfamide, cisplatin; the British trial, published this month in JCO. And here four cycles of MIC versus best supportive care: median survival, five versus seven months, statistically significant. I show this trial and the data are in the Journal because they looked at quality of life and it was improved with chemotherapy over best supportive care. This had also been addressed in previous trials, but addresses this issue where frequently it was meant that yeah, you only improve survival a little and are patients really benefiting from it? Well, they do.

What about elderly? This is a disease that frequently occurs in the elderly. So patients 70-years-old, or in the past it would be argued, “Well, let’s not treat that patient anymore. It’s too old.” Well, 191 patients were randomized to either vinorelbine single agent or best supportive care. Median survival with chemo: 27 weeks versus 21 weeks. Sort of showing the same thing in this group of patients. Six months survival: 54% versus 39%, statistically significant. One year: 27% versus 5%. Quality life here too was analyzed – this was published in JMCI last year, or early this year rather – improved with chemotherapy. Finally, cost-effectiveness has been looked at, particularly with one of these trials, the so-called Canadian trial. And to not go into great detail, just to make the case, that even though the chemotherapy costs some money they felt that this was well within the range of any other interventions and that cost or cost-effectiveness should not be a barrier to offering chemotherapy. So looking at this information, you can summarize that chemotherapy is an active modality in stage IV disease, that it prolongs survival by a little, that it improves quality of life – and it does so at acceptable cost – and as a consequence should be offered to these patients. The debate is done. That’s the literature of the 80’s.

So what we want to move onto then is in the 90’s, not much any more do we want to use chemotherapy but can we identify better chemotherapy? I mentioned that there were five drugs that came out during this decade. They have now been somewhat evaluated and let’s go over that a little bit, in sequence at least. The first two trials will focus on vinorelbine. That was the first drug that came out. It’s a Vinca alkaloid. This is the definitive trial done by SWOG, Tony Wosniak published in JCO. Cisplatin alone is the control arm, cisplatin, vinorelbine was the experimental arm. Median age 63 years as a standard. Most patients male, as is the case, and most had stage IV disease. So this is a poor prognostic group of patients. Some other trials have more patients with stage IIIb disease in there. You can look at response. So what’s the single agent response rate to cisplatin? It’s 12% and to cisplatin and vinorelbine it’s 26%. Here’s the overall survival: median survival six months with cisplatin alone and about eight months with cisplatin and vinorelbine and statistically significantly improved. So this is the trial, or one of them, that got vinorelbine approved by the FDA for non-small cell lung cancer, and cisplatin, vinorelbine then as a possible standard regimen.

There are some other pretreatment prognostic factors, and most of these are established in patients with stage IV disease, so those are treated with chemotherapy. There are some large trials that have looked at retrospective evaluation of large numbers. There are newer references for this but they all show the same. Stage is the predominant clinical pretreatment prognostic factor. Performance status is a very strong next, pretreatment weight loss and then adverse effects of male gender and age. For patients with stage IV disease in particular, the number of metastatic sites matters and the use of chemotherapy, in particular cisplatin-based chemotherapy. Now in earlier stage disease, those who have surgery we can also look at molecular abnormalities. These are the two that over the years have been established as negative prognostic factors. That’s the K-ras mutation that’s frequently seen in adenocarcinomas, and absence of blood group antigen A which probably correlates with something else missing. It’s not the absence of the antigen that is harmful to the patient, but there may be loss of a tumor suppresser gene associated with that.

So let’s focus then, for non-small cell lung cancer, on therapy and since most of you are medical oncologists that’s what we want to focus on. About 50% of patients present with stage IV disease. They are treated for palliation, and as I will show you, chemotherapy improves survival and quality of life at acceptable cost. It is now accepted therapy for stage IV disease. If you look back to the 1980’s, then the platinum and Vinca regimens have all similar activity, and the same is true for cisplatin and carboplatin which have similar single agent activity. So this is a disease where either one could be used. The current dogma is that doublets are as good as anything. Adding a third or fourth drug has not been shown to be beneficial. There may be exceptions but that’s the rule. There is probably no dose response curve for survival but there may be threshold doses and that’s coming out particularly with the novel agents that you need to get to a certain dose, but from then on there is no further benefit. That is, of course, softer knowledge but nevertheless it looks like that, particularly for the taxines. Activity for single agent therapy in non-small cell lung cancer has been defined historically as 15% or more. Well, it isn’t always even 15%. Some of these are more like 10%.

There are two groups. These are the older drugs that were available a decade ago and typical regimens then consisted of cisplatin in combination with vinblastine or cisplatin in combination with etoposide. There was also a so-called MVP regimen of cisplatin, mitomycin and vinblastine. That was a triplet. Since 1990 we have at least five new drugs with significant single agent activity, and I will focus on some of that: vinorelbine, Taxol, paclitaxel, docetaxel, gemcitabine and CPT-11 or irinotecan.
Cancer
So let’s look at the literature of the 1980’s first because what was established during that decade is that chemotherapy is useful for non-small cell lung cancer. Response rates were around 20-30% with most regimens, median survival times around eight months, so how good really was chemotherapy? What was done then is combine, do a randomized trial, comparing chemo alone versus no chemo. Just supportive care. A large number of trials but usually few patients and if you looked at outcome, only three were statistically positive supporting the use of chemotherapy. You can, however, look into this column and what you’ll see – this is median survival in weeks – is usually 25 to the low 30’s and if you look at best supportive care you usually see 15-17. So that’s a poor-mans metaanalysis. Right out of this you could make some conclusions and then have three trials that have statistically positive results and all the others showing the same thing. So metaanalyses were of course done then.

Lung cancer is smoking-related. This is a totally preventable disease if patients do not smoke. And all of what we talked about wouldn’t be necessary if not for cigarettes. So it is the second most common malignancy in both sexes. Head and neck cancer was 40,000, this is 172,000 cases per year and it’s the most common cause of death in either sex. Meaning that in males it is a more common cause of death than prostate cancer, and in females more common a cause of death than breast cancer. Eighty percent of cases at least attributed to smoking. There may be some other risk factors and a genetic predisposition – largely, we think, because there is no familial clustering outside of smoking – in terms of how individual patients metabolize or otherwise handle the carcinogens contained in the smoking.

Let’s look at some of the incidence figures. These are incidence, cancer rates. This is prostate cancer shooting up when PSA becomes available then down. This is lung cancer in terms of incidence, slightly going down in males for the last 20 years or so. This is incidence rates for females and you see it’s still rising, as far as lung cancer is concerned. This is breast cancer – more common. Now let’s look at the death rates. These are the death rates in males. This is lung cancer. Exceeding everything else by far with somewhat of a reverse trend in the last few years. And this is the same in women, exceeding everything else with what still appears to be a rising incidence of lung cancer deaths in women. So that is where smoking prevention is more important than anything we can talk about in this kind of lecture. This is way too late and conceptually of course, not in an epidemiological sense, the primary thing to do.
Pharmacy
Now what about the histologic types? Seventy-five percent are non-small cell, that’s what we will cover here. The others are small cell and non-small cell, divided up into these cancers: in the United States now adenocarcinoma is the most common, squamous carcinoma is next. This used to be different 20 years ago but now adenocarcinomas are the common ones and we are not sure why that is. Then this is stage-specific survival. Stage I is a confined disease. These patients are treated with curative intent by resection. Stage II is very similar, with at most, hilar lymphadenopathy. Many patients can be cured although you are down now to about 30%. Stage III, we will talk about a bit more in detail, includes regionally advanced disease. So that is disease in the chest involving large primaries or lymph nodes, mediastinal lymph nodes or supraclavicular lymph nodes, but no distant metastases. Then stage IV is distant metastases. We can also look at this is a sense of how far along can we use surgery. For metastatic disease we wouldn’t, and for stage I or II we would. Because these patients can be cured surgically, at least in some cases. Then stage III is split. Where usually IIIa is, at least in theory, resectable whereas IIIb is considered un-resectable. So that is the split from a surgical point of view.