SLIDES & TRANSCRIPTS
Friday, December 5, 2003

Chemotherapy for Clinical Stage I Germ Cell Tumors

George J. Bosl, M.D., F.A.C.P.

First of all I have to integrate all three approaches. If I am going to talk about chemotherapy for stage 1, I have to think about the other two options of observation and surgery and how they integrate with this.

So, our goals here are low relapse rate, low treatment toxicity, few long-term treatment sequelae, avoid retrograde ejaculation in the non-seminomas and the options are these three with the considerations being slightly different for non-seminomas and seminomas. And I only have one slide on seminomas here, very much toward the end. I pressed the wrong button.

TOP

Critical questions are what is the risk of retroperitoneal disease; what proportion really need chemotherapy? I am going to keep coming back and harping on this issue. What proportion really need chemotherapy? Randy Millikan's comments last time having to do with chemotherapy and bladder cancer, I think, very much pertain in this disease. Is the histology predicted by primary tumor? Where and what is late relapse? How frequent is malignant transformation and what, if any, are the late sequelae of chemotherapy?

To some extent we must draw some conclusions from advanced disease, since the clinical stage 1 series are not old enough. But, there are good data in advanced disease, and I think a few particular cases make the point about certain things that will happen.

TOP

First, if we take a look at what is the risk of relapse, if we look at four studies with no vascular invasion, clinical stage 1B with observation, relapse rate about 17 percent and about 98 percent are cured.

There are several studies with 1B disease and observation. About half of them relapse. These will also be cured with appropriate subsequent management, surgery, chemotherapy or some combination thereof. I chose one study but there are a few others with elevated markers in the clinical stage 1S. If you don't treat them they all relapse; and so, if we take this, we can draw a couple of obvious conclusions.

Clinical stage 1S, as Joel said before, requires chemotherapy; and in 1A or 1B, the chemotherapy question is restricted to 1B because of the high relapse rate. And I may touch on this just a little bit more, if we have a few minutes to talk about seminoma at the end.

TOP

Let us remember these relapse rates. We come back here and look at the risk of relapse in observation versus node dissection, 1B observation. I mentioned 45 percent; [in] big surgical series with node dissection, about 13 percent. Again, everybody is being cured here. So, this is roughly what one can expect from observation and from a thorough node dissection. Modified node dissection reduces relapse rate by eliminating retroperitoneal relapse. I think everybody would agree with that general rule.

TOP

If we then add the third option of chemotherapy which has been usually BEP (Bleomycin, Etoposide, and Cisplatin) times two, relapse rates do drop in these series to around 2 percent with relatively short-term follow-up compared to these other studies. And again, the cure rates all look to be about the same. The question is, how many really need chemotherapy? And we need to take a look at all three of these to be able to draw some conclusions about that.

TOP

So, if we have the three options of observation, node dissection and chemotherapy, and look at those relapse rates from the last slide, the number that would relapse out of 1000 hypothetical patients is about 500 in observation, about 150 after node dissection, maybe 5 or 10 after chemotherapy.

Therefore the number of patients who get chemotherapy are these numbers here: 500 who underwent observation and relapse, 150 after node dissection, and fully 1000 patients who got chemotherapy (number of cycles three to four) for observation and node dissection and BEP times two. We are going to come back to this issue of BEP times two versus three or four. The conclusion: retroperitoneal node dissection leads to the lowest chemotherapy requirement.

TOP

If we then take a look at this same slide, looking at the number that relapse, number with chemotherapy, number of chemotherapy cycles, and extrapolate out again from these 1000 hypothetical patients the assumption of clinical stage 1 chemotherapy approach is that retroperitoneal disease is dead after chemotherapy in the clinical stage 1 setting and teratoma is not a serious consideration.

TOP

This is straightforward. The biology tells us that carcinoma in situ, intertubular germ cell neoplasia, seminoma, non-seminoma can be either extra-embryonic differentiation or somatic differentiation down the three germ layers to mature and immature teratoma and teratoma with malignant transformation.

TOP

So, if we say that the hypothesis is absence of teratoma in the primary means no teratoma in the retroperitoneal nodes, and I just picked a few series here because I needed to keep the slide short, after chemotherapy, two very large series by Toner and Beck - number of patients without teratoma in the primary - one-third to one-half had teratoma in the retroperitoneal nodes. This is after chemotherapy. You say, "Okay, clinical stage 1 to pathologic [stage 1] about 11 percent," as Joel has pointed out before. And if it is EC (embryonal carcinoma)only, you still have it. So, even if you have pure embryonal carcinoma, you may find in the retroperitoneum mature teratoma.

The conclusions are that teratoma occurs in metastasis, even if there is none in the primary, no matter what setting that you examine. It cannot be predicted. Chemotherapy does not treat [teratoma], and among 100 clinical stage 1 patients, 9 to 10 will have teratoma in the retroperitoneum.

TOP

The normal abdominal CT may harbor disease. Here, [referring to slide] is a patient with a retroperitoneal node from a right-sided tumor post chemotherapy. [He] had two nodes with mature teratoma. Here is a clinical stage 1S after chemotherapy. This node appeared - pure mature teratoma. There is no setting that you are going to be able to find in which this won't occur.

TOP

Joel touched on late relapse, by definition, beyond 2 years after chemotherapy. In point of fact, three-quarters of them occur beyond 5 years. Most of them occur in the retroperitoneum; a smaller percentage in the pelvis or in the retrocrural area; very seldom in other sites.

Here [referring to slide] is an example of [relapse at] 19 years. This one happened to be a mixed histology mature teratoma, but it is just to make the point that it occurs many years later.

TOP

If we take a look at what is late relapse, most of the time it will, in fact, be a germ cell tumor; but fully 30 percent of the time it will be either teratoma and/or malignant transformation. That is a lot compared to what you see earlier on in the course of the disease.

Here [referring to slide] is an example [of relapse] 14 years afterwards that turned out to be undifferentiated sarcoma. Malignant transformation occurs in 10 to 15 percent of late relapse and in 3 to 10 percent of first presentations. This will occur.

TOP

Here is a worst case scenario. You would probably be hard pressed to prove that that [referring to slide] is a small positive node in a patient with an HCG of 1200. However, it did shrink. So, therefore in retrospect one can say it was a node. There is undifferentiated, there is a tumor, [and] 2-1/2 years later there was pure undifferentiated sarcoma.

TOP

This [referring to slide]is 38 years later in a patient who decided he wasn't going to undergo therapy. This is a malignant transformation and the patient died of metastatic adenocarcinoma.

TOP

So, we come down to a key question. What if any are the late sequelae of chemotherapy? These are the acute sequelae [referring to slide]. You are all aware of them.

TOP

However, in 1985, Nick Vogelzang reported that hypomagnesemia was associated with Raynaud's phenomenon and Stu Leitner, a medical oncologist in the audience working with me at that time, showed that most patients after platinum-based chemotherapy were hyperreninemic and hypoaldosteronemic; and he stated that patients should be carefully observed for delayed cardiovascular toxicities.

TOP

Unequivocally. Raynaud's phenomenon, first reported by my good friend Nick Vogelzang in 1981,

TOP

has now been followed by vascular effects, including myocardial infarction at a 7 to 1 excess risk, in patients receiving BEP-based therapy. And, in an additional group of patients with exercise-induced ST depression, reported by Meinardi in JCO (Journal of Clinical Oncology), a slightly [higher] incidence in England reported by Huddard.

This has been also associated with a significant increase in cardiac risk factors, with elevated hypercholesterolemia, hypertension that is very significantly in excess; and all the Raynaud's phenomenon occurs in patients who received Bleomycin. This is also associated with erectile dysfunction in a population of patients, along with peripheral neuropathy that remains persistent 10 to 15 years after platinum-based chemotherapy.

TOP

So, if we come back to this hypothetical 1000 patients, what is the excess risk? You can see [referring to slide] that there is an excess predicted number of patients with myocardial infarction, hypertension, cholesterolemia and probably acute leukemia, based upon the extrapolations from the paper by Meinardi et al.

TOP

So, in conclusion, risk of retroperitoneal disease is about 50 percent; with a node dissection only 15 percent ever need chemotherapy.

If you give up-front chemotherapy, 85 percent are treated unnecessarily. Nerve dissection, node dissection, as Joel said, there is a very high likelihood of anterograde ejaculation.

Histology of the primary does not predict metastasis. The majority of late relapse is retroperitoneal and late. [Using] up front chemotherapy and no node dissection, late relapse will happen. Malignant transformation will happen and late toxicity, especially vascular, will happen from BEP-based chemotherapy.

Late relapse, plus malignant transformation, plus myocardial infarction, is excess death that won't be measured until 10 to 15 years after the patient is treated. Proper node dissection causes none of these and has low retrograde ejaculation. Nerve dissection and node dissection, nerve dissection RPLND, is the standard of care in my opinion. I believe, just like what Heidi Nelson has done in colon cancer, the randomized trial is necessary. Up-front chemotherapy remains investigational and the randomized trial is necessary with long-term follow-up.

TOP

[No text is associated with this slide]

TOP

I will remind you that the lower bound has been reached in metastatic disease and fewer patients are cured with three cycles of EP, carboplatin versus cisplatin, carboplatin versus EP in seminoma and lower dose EP16 compared to high dose. Collected Phase 2 trials can't answer these questions. They are hypothesis generating, not standard setting, and modifying the standard must be based upon improved patient outcomes not available technology or chemotherapy Phase 2 trials.

Late effects and late relapse are real problems, and the biology, I don't believe we are taking into account. I don't think we should alter the standard of care when cure rates are in the 98 to 99 percent range, when patients are handled methodically. And I think this requires Phase 3 trials.

Thank you very much.

TOP