My job was to present the pro position for detection of minimal residual disease in high risk patients.

TOP

This self evident truth was provided by the very quotable, Dr. Sondak regarding minimal residual disease in melanoma: Melanoma markers are good.

He gave me very specific instructions for this presentation: Be brief.

Actually, I didn't receive any of your instructions, Vern, and one of our e-mails must have been down.

TOP

As you just heard from Dr. Buzaid, the staging system has now given us some increased ability to discriminate levels of risk among patients with melanoma.

The entire discussion has been about how to assess the level of risk for our patients so that we can then present treatment options and assess the potential benefit.

You can see that patients with stage one melanoma have a better prognosis than those with stage two, three, or four, but still some patients with stage one disease -- those between here and here -- die of their cancer, the same with the stage two disease, and some patients with stage three disease, of course, are cured, and were cured without any adjuvant therapy at all, by surgical treatment, even some with stage four disease.

How do we find which of these patients have a good prognosis and which have a bad prognosis, and how do we determine treatment for those patients?

TOP

You might ask, why do we need markers, beyond all the things we just talked about, in a very complex staging system, to assess this risk? Why not just use this new staging system to assess the patient's risk?

I think the staging system does break patients into very broad categories of risk, and molecular or other markers may assign a better assessment of risk, identify patients who may benefit the most from therapy, the patients with a better prognosis, for whom therapy may not be warranted at all, and especially identify more homogeneous patient populations for adjuvant therapy trials, and maybe even monitor the response to therapy.

TOP

As you have just seen, there is a big diversity, even among patients with stage three melanoma, in terms of their prognosis.

There is 10-year survival with ulcerated patients with more than one positive node versus those with microscopic disease in the lymph nodes, big difference in survival, even among patients with stage III disease.

TOP

So, there are some other diseases where we have markers of minimal residual disease where it affects treatment options. I have listed a few here. Whether or not those treatment options are carefully considered and are beneficial is sometimes not always clear.

The fact is that the markers may affect our therapy decisions, and that is where we want to be with melanoma.

TOP

So, when we are looking at all these different prognostic factors that we can determine, it is important to think of these as diagnostic tests.

Like any other diagnostic test, we must consider the specificity and sensitivity, positive and negative predictive values, and false negative and false positive rates.

A very sensitive test may not be very useful, if there is a high proportion of patients with false positive results. That wouldn't be very helpful at all.

All these things need to be established before these markers will be useful, and they all need to be reproducible.

TOP

So, in terms of prognosis, I can think of several situations where measurement of minimal residual disease might be helpful, such as in the prognosis of patients with stage I and II melanomas, to determine who are at greatest risk for recurrence and death.

The pick up is very high risk stage I and II patients because most of them, as you know, will do well long term.

How about the stage III melanoma patients who have residual or persistent disease after they have lymph node dissection? That might be useful information.

The stage IV patients rendered NED by surgery, some of those patients do well long term, some of them are potentially cured. It would be nice to give them some idea about their risk.

TOP

Eventually, some of these markers that we have talked about today and the new ones that will be developed may help us to actually get to the point of predicting the patients who will benefit from adjuvant therapy, from interferon, from vaccines, from immunotherapy and new forms of therapy yet to be developed, and perhaps predict the small fraction of patients who are capable of undergoing a complete pathologic response to the immunotherapies that we have right now.

If we could understand those mechanisms, it would go a long way toward trying to induce complete responses in more patients than we can do now.

TOP

Dr. Reintgen already talked about the sun belt melanoma trial that is ongoing now, looking at PCR staging of sentinel nodes to determine those patients who might benefit from more aggressive therapy.

All of these patients have just a positive sentinel node by PCR testing, and are randomized to either observation lymph node dissection, or lymph node dissection plus interferon, to see if more aggressive therapy is warranted in these patients whose only evidence of disease is a positive molecular staging test.

Only time will tell. Some people might say that we have put the cart in front of the horse here by randomizing patients to treatment options before we fully know the prognostic significance of PCR testing but, again, time will tell.

TOP

You can see here data similar to what you have already seen before, with PCR testing of sentinel nodes. We can pick out a population of patients that seem to do worse in terms of disease-free survival, even with early analysis.

TOP

We can find circulating blood cells, melanoma cells in the peripheral blood. As several institutions have already shown, these markers may impart a worse prognosis to these patients.

Whether or not this is useful information that we can use to predict which patients will recur, to monitor the course of their therapy and the effectiveness of that therapy, remains to be seen.

TOP

It is also important, I think, to not just rely on any one marker, but to see which independent prognostic factors combine best to give us the patient's assessment of risk, and the best combination of these factors still need to be identify.

TOP

For example, when we look at data from the Sun Belt Trial, we can see that ulceration, which is now featured very prominently in the new staging system, combined with the sentinel node status, gives us an increased level of discrimination of the patient's risk in terms of disease-free survival.

Here are patients with histologically negative sentinel nodes, ulcerated, not ulcerated.

TOP

Histologically positive sentinel nodes, ulcerated, not ulcerated.

TOP

Here are patients with sentinel nodes positive only by PCR testing. Even in this group, ulcerated patients do worse than those who are not ulcerated.

I think in the future we will be able to combine several of these very important prognostic factors to determine the level of risk for these patients.

TOP

I think it is very important that in all future trials of adjuvant therapy we need to consider the most important prognostic factors that we have at our disposal.

Breslow thickness, of course, ulceration and pathologic staging of the regional lymph nodes, I think, are important in order to get the most homogeneous patient populations for study.

If we have learned anything, I think, in adjuvant therapy trials so far, it is that it is very important to have very homogeneous groups of patients that are comparable, in order to make sense of our results.

I think that it is very important in the future that all clinical trials have meaningful basic science correlates and study of these promising markers of minimal residual disease, not just little side studies where we throw a bone to our basic science colleagues or people in our own lab to do a few studies, but real, meaningful studies built into the clinical trials.

TOP

This has led me to postulate that the need for prognostic markers of minimal residual disease increases as the toxicity of the adjuvant therapy increases.

It is important to select patients most likely to benefit from the therapy, again, from the patients unlikely to benefit from that therapy or any therapy, and perhaps the patients who don't need therapy at all. That is our charge with these markers for minimal residual disease.

TOP

In conclusion, I would say that the level of risk determines the magnitude of potential benefit from therapy.

All of these markers can assess the potential risk. Conventional TNM staging is a good start, but I think additional markers may improve the ability to determine the level of risk and the potential level of benefit, and eventually we will be able to predict the response to specific treatments. Thank you.

TOP