Thyroid irradiation

Let’s talk just very briefly about thyroid irradiation. Potential sources for thyroid irradiation include external beam irradiation or radioiodine fallout. There was a lot of childhood irradiation being done back in the 40’s and 50’s and there were many many cases of thyroid cancer noted in the adults of children who were exposed, for reasons that would certainly not be considered medically appropriate uses of radiation in this day and age. It is clear to me that children are more radio-sensitive than adults. Now we have radioactive fallout. Catastrophe; I’m sure many of you remember this. This was shortly after the Chernobyl reactor accident, and it truly has been a catastrophe in any number of different ways, not the least being sort of an epidemic of especially childhood thyroid cancer cases with fairly aggressive courses that have been noted. Beginning about four years after – it was a fairly short latency, actually – after the reactor accident. I guess hopefully we will never see another Chernobyl but on the other hand, as the events in Japan last week have proven, by no means are we going to be free from the potential risk of catastrophic nuclear accidents in the future.

Familial syndromes associated with papillary thyroid cancer: there are familial papillary thyroid cancer syndromes but a specific familial papillary gene has not been identified. Papillary cancer is seen in increased frequency with some GI syndromes, Gardener’s syndrome, familial polyposis, Cowden syndrome. Carney complex is the combination of myxomas, nodular adrenal cortical disease, Sertoli cell, testicular carcinomas and thyroid malignancies are also seen with increased frequency in that. A couple of pathologic variants of papillary thyroid cancer may be noted at the time of surgery, and sort of going from less aggressive to more aggressive are an encapsulated papillary thyroid cancer variant, the so-called follicular variant. These are tumors in which there is no gross papillary element present in the tumor, but the cytology of the individual cells is consistent with the papillary form. This has about an equal biologic behavior to ordinary papillary thyroid cancer. Of somewhat greater concern are the diffuse sclerosing variants, tall cell variant. There is another variant known as the columnar variant, which may be sort of a variant of tall cell variant, and an insular form which is fairly aggressive is also seen. So a couple of less common variants can be seen.

Follicular thyroid cancer accounts for about 15% of the thyroid cancer in the United States, and as I mentioned, more common in areas of chronic iodine deficiency. We touched on the issue that follicular carcinoma cannot be distinguished from follicular adenoma by aspiration, and so if that quandary remains, in general I will recommend surgery to the patient. The general spread of follicular carcinoma is by hematogenous spread to lung, bone, liver and brain. A specific variant form that is though to be a variant of follicular carcinoma is the so-called Hurthle cell carcinoma. Hurthle cells, which may be seen in a variety of different thyroid lesions, including many non-neoplastic conditions. So the presence of Hurthle cells is not in and of itself specifically a marker of a thyroid neoplasm. But occasionally one will see neoplasms which consist predominantly of oxyphilic cells with sort of a granular cytoplasm and this defines the Hurthle cell variant. One of the important issues in terms of Hurthle cell carcinomas, is that in comparison to follicular or papillary carcinoma, it is considerably less likely to take up radioiodine and therefore in one sense is effectively a surgical disorder. Many, although not all, investigators report a worsened prognosis compared to follicular carcinoma. It would hardly be surprising if this was in fact the case based on the fact that it appears to be a less differentiated tumor. But it’s not been a universal finding in all studies.

The most common form of thyroid cancer

So with that, let’s talk a little bit about thyroid cancer itself. A variety of different types. Differentiated cancers arising from thyroid follicular cells are of course papillary or follicular cancers.  The less differentiated is the anaplastic, which fortunately we see very rarely. Arising from an embryologically distinct population of C cells is medullary thyroid cancer. I’ll talk very briefly about it towards the end. I don’t know if the speakers, when they were talking about some of the other islet cell neoplasms, discussed this at all or not. Finally, there are sort of non-thyroid thyroid things. Non-Hodgkin’s lymphoma can present as a goiter or a thyroid nodule occasionally. Metastatic disease and some other less common lesions. But we are going to focus on the differentiated follicular-derived thyroid cancers. About 17,000 cases are diagnosed yearly in the U.S. and the incidence of diagnosed cases has been increasing. It’s not clear to me if this represents some true alteration in the incidence of the disease, or whether this is in part a reporting issue of previously undiagnosed cases, since thyroid cancer often will be fairly indolent. The disease is more common in women than in men.

The most common form of thyroid cancer that we see in this country, and in general in areas of iodine sufficiency, is papillary thyroid cancer. Probably accounts for 80-85% of the thyroid cancers that we see in the United States. Papillary thyroid cancer can be noted as an incidental finding. Occasionally one will undergo a thyroidectomy and a microscopic focus or maybe several microscopic foci of papillary thyroid cancer will be found almost as an incidental finding. It would appear that that form of microscopic papillary thyroid cancer is probably of little clinical concern, since the finding is that these things are found on autopsy series in unsuspected patients much more frequently than they present clinically. So there are probably several people in this room who are harboring microscopic foci of papillary cancer and very likely you will never know about it. Papillary cancer can be multifocal throughout the thyroid gland, or bilateral. The general spread of thyroid cancer is local invasion spread to cervical lymphatics and then distant metastases, particularly to the lung.

Risk factors for papillary cancer; thyroid irradiation. Papillary cancer is seen with increased frequency as a second malignancy in individuals with a history of childhood cancer. This in part may be related to the irradiation issue. It may be in part related to general genetic factors with regard to tumor predisposition. There is some argument as to whether or not Grave’s disease, because of its sort of the endogenous thyroid stimulation is a risk factor. Dietary iodine increases the relative risk of papillary thyroid cancer but decreases the risk of follicular cancer. If you look at areas of the world that are iodine deficient you tend to see more follicular thyroid cancer relative to the United States.

Thyroid Cancer

Thyroid cancer is most frequently going to present as a solitary thyroid nodule. These are by no means uncommon findings. In the Framingham study, just the health surveillance study, the prevalence of these nodules was close to 10% in women. The important figure that I tell my patients when I first see them is that the a priori odds of a solitary thyroid nodule being a cancer is actually not that high; only about 5-10% of thyroid nodules are ultimately going to prove to be malignant.

Clinical risk factors: the important risk factors, issues that would make one highly suspicious, a nodule that is growing rapidly. Say, over a period of weeks to months. A family history of a familial syndrome, multiple endocrine neoplasia or perhaps medullary thyroid cancer. Not by any means a common situation and in fact I would hope not to be seeing anybody with that history at this point, for reasons that we’ll discuss. Among the physical characteristics, a nodule that is very hard, that appears to be fixed to adjacent structures. Very rarely there will be vocal cord paralysis present. Obviously the presence of clinically suspicious lymph nodes or the occasional patient who would present with frank metastatic disease. But in fact, most of the time when patients come to see me with a thyroid nodule they are asymptomatic and there is very little a priori reason to feel that they either do or do not have a thyroid cancer.

This is kind of a busy algorithm. It’s reproduced in your notes. The important point here is that from my perspective the single most important test, in terms of evaluating solitary thyroid nodule, is a thyroid aspiration. Frequently I will do that at the time of the initial visit, even if I know nothing else about laboratory studies or anything like that. Oftentimes a biopsy will provide enough information to make a treatment decision right off the bat. If the biopsy is highly suspicious for being a papillary cancer, you don’t need at that point to do a lot of scanning or imaging studies. On the other hand, if the biopsy is very reassuring about this being a colloid nodule, for example, or perhaps chronic thyroiditis or something like that, you also don’t need to do a whole lot. The one area to keep in mind in which the biopsy can be difficult is where the cytologist reports back the biopsy findings as being suspicious for a follicular neoplasm. The problem here is that on an aspiration cytology level, a follicular neoplasm, that aspirate will not distinguish a follicular carcinoma from a follicular adenoma. So those are the areas that can be a difficulty. It’s in those patients that I’ll get a scan because occasionally the nodule will be hot and if we know the nodule to be hot on scan, we know it’s not a follicular carcinoma. So that’s the one situation in which a scan is particularly useful in distinguishing a cancer from an adenoma. Otherwise, if the scan shows it to be cold then most of those patients are going to be indeterminate and I would recommend surgery for most of those patients because there won’t be any other less invasive procedures that will be definitive.

A second trial that’s been reported is GOG protocol

 A second trial that’s been reported is GOG protocol 114, also an Intergroup study between SWOG and GOG. This randomized patients to IV Taxol/IV cisplatin or a dose-intense approach of two cycles of IV carboplatin dosed to AUC9, followed by IV Taxol and IP cisplatin. This study showed no difference in overall survival. This is not statistically significant, in fact the P value is now 0.11 for this. But a statistically significant progression-free survival, which was not seen in 104. So we have here a slightly different result from a second study that was not really a pure IP versus IV comparison. What we are doing currently is a third trial trying to definitively answer the question. This study is going to randomize patients to either IV Taxol/IV cisplatin or Taxol IV on day one followed by cisplatin IP on day two and Taxol IP on day eight. The reason for giving Taxol both IV and IP is that Taxol is not taken up to any great degree from the peritoneal cavity. So if you are going to get systemic exposure to Taxol you have to give it both IV and IP. This is a tolerable regimen and the study is ongoing. At the current time we have no definitive answer but at least for right now there is not enough evidence in the literature to warrant the use of intraperitoneal therapy outside of clinical trials.

Now what about high dose chemotherapy with stem cell support? The largest data base reported was reported in 2006 at ASCO, 390 women treated over an eight year period and reported to the North American Bone Marrow Transplant Registry. Median age 48, 63% platinum sensitive and overall response rate of 72%. Very impressive, until Dr. McGuire discusses it. What Dr. McGuire did was to take the GOG data using single agent Taxol as a salvage regimen in platinum resistant patients and plot the disease free survival and overall survival curves on the same graph with the survival curves for the high dose chemotherapy. Remember, the high dose population, 63% of them were platinum sensitive, so it’s a better risk population. The red curve is the disease free survival for use of Taxol alone as salvage. The white curve with the big early drop-off is the high dose chemotherapy and there is no place along here that you will see the red curve below the white curve. These are the overall survivals. Taxol is the blue curve on top followed by platinum sensitive high dose chemotherapy, and then platinum resistant high dose chemotherapy. Now this sort of analysis doesn’t prove that high dose chemotherapy doesn’t work. It also though doesn’t support its use outside of clinical trials. We tried to do a randomized phase III trial in the United States. We can’t accrue patients to this because of bias about the appropriate approach. Not everybody, though, believes that high dose therapy ought to be used. About half believe that it’s unethical to use high dose therapy and half believe it’s unethical not to use high dose therapy. So you are unlikely to see a randomized phase III trial come out of the U.S. in the near future. There are ongoing trials in Europe that hopefully will provide us with some answer to this question. But as of right now, there is no clinical rationale to use high dose therapy plus stem cell support outside of clinical studies to treat ovarian cancer.

The drugs that would be chosen as alternative therapy

Let’s look at the drugs that would be chosen as alternative therapy. First there are a group of four drugs that have been shown in published studies to be active in patients who are platinum resistant by the definitions I showed you; Navelbine, about an 18% response rate in platinum resistant patients, tamoxifen, 13%, ifosfamide, 12%, and 5FU leucovorin 17%. All of these responses are in the mid-teens but there is some activity. The better group of drugs though are those drugs that have been shown to be active in Taxol/platinum resistant patients, resistant to both of those agents. These include oral etoposide, which has the highest raw response rate in patients with Taxol/platinum resistant disease, 32% in a GOG trial. Eight responses among 25 patients. Topotecan has a 13% response rate in Taxol/platinum disease. Doxil a 19% response rate. If you went to ASCO you saw a study presented showing 23% that had 63 patients on it. The full data base has 110 patients in it. It was collected by a pharmaceutical company which is now Abbott, and the overall response rate was 19%. Gemcitabine 14% response rate in Taxol/platinum resistant patients. My own personal order of choice here is oral etoposide as the first salvage. If and when they relapse after that, Doxil and then topotecan in that order. Though both topotecan and Doxil have FDA approved indications in ovarian cancer, etoposide is in the compendium as a salvage drug but the problem with oral etoposide is you have to go through the durable medical equipment provider to be reimbursed. So it’s a little more complicated, though I have never viewed pills as durable medical equipment. So that’s the situation with regard to salvage therapy.

Right as the last thing I want to comment very briefly on dose-intensive approaches like intraperitoneal therapy and bone marrow transplant. What we are looking for is a second threshold on that dose-intensity curve. We’ve shown you evidence that the dose-intensive curves flattens after a threshold of roughly 60 per meter square for cisplatin, AUC5 for carbo and 175 per meter square for Taxol. With these dose-intensive approaches we are looking for a second threshold. Because of time I will go through these quickly. There have been two randomized trials looking at IP therapy as front line therapy. The first was published in December of 2006 in the New England Journal of Medicine. This is Southwestern College Group, GOG Intergroup study, randomizing patients to IV Cytoxan/IV cisplatin, versus IV Cytoxan/IP cisplatin. That article will tell you that this makes intraperitoneal therapy the treatment of choice for small volume residual disease. It does not.  The study was flawed. The accrual was extended to 650 total patients. The original goal of the study, a hazard ratio of 0.67 was never reached. There was a difference of 49 versus 41 months median survival favoring IP therapy, hazard ratio of 0.77. But the study really gave us a counter-intuitive result because the 0 to 0.5 cm subset, the subset we thought would benefit the most because it had the smallest volume of disease, actually didn’t benefit. Benefit was seen in the larger nodule group which doesn’t make any sense. So for that reason this study served to keep the issue alive.

The patients who have grade II or grade III disease

Now what about the patients who have grade II or grade III disease? These patients were randomized to either observation or cisplatin as a single agent. The disease-free survival at five years for the cisplatin group was 83%, for the observation group was 64%. That’s highly statistically significant. The overall survival at five years was 87% versus 81%. Better with cisplatin than with observation, but not statistically significant. Now some people accept disease free survival improvement as a reason for treating, others don’t. My own opinion is that these patients ought to be treated with some form of platinum based therapy because of the difference that we see in the disease-free survival. But there is still an ongoing debate about this group.

The other group, those at high risk, because of other factors, were randomized to either intraperitoneal P32, which was a nice placebo, or cisplatin based therapy. Actually when this study was started this was considered the standard but we have reasonable evidence to suggest that it really has very little impact. At five years the patients who got cisplatin had an 81% disease-free survival compared to 66% with intraperitoneal P32 and there have been too few events in the overall survival at five years to permit any kind of reasonable analysis, but there is a small difference favoring the cisplatin based therapy. Again, the same reasoning applies here. Either you accept disease free survival improvement as a reason for treating or not. The consensus among the gynecological oncology community at least is shown on this side. That is, that everybody ought to have a hysterectomy, a bilateral salpingo-oophorectomy and a careful surgical exploration to document the exact stage of disease. Those deemed to be at low risk for recurrence, the low risk group, should receive no further therapy. And those deemed to be at high risk either because of grade or because of one of the other factors, should receive some form of platinum based therapy with expected improvement in disease free survival to 82% at five years as compared to 60% with not treatment. If I were picking the therapy I would use a combination of Taxol and platinum in this group but we have no studies as yet showing that that has any advantage over single agent cisplatin.

Now what do we do in the salvage setting? There’s actually on the ASCO site a summary of a lot of data on salvage therapy if you are interested. It’s in the section on the educational material from the 2006 meeting. But very briefly, what we are talking about here is a population of patients who either recur after or progress on front line therapy. And the general principle or approach to these patients is to look at what their response was to front line therapy and then base your decision on that. In that regard, what you do is try to divide the patients up into two groups. A group of patients that almost certainly are still sensitive to the drugs they received as a part of front line therapy, and a group of patients who almost certainly are going to be resistant to those drugs. Those who should be considered sensitive are those who responded to front line therapy and had a significant treatment-free interval before they required further treatment. Now the current definition the cooperative groups are using is six months. There is nothing magic about six months. What that real number should be, we don’t know for sure. It’s probably a continuum. Six months has been selected as the break-point by the cooperative groups. For chemo-resistant disease; this includes those patients who either progress while receiving front line therapy or whose best response to front line therapy was stable disease, or who had a short treatment-free interval before they recurred. That is, less than six months. In those who are chemo-sensitive, they ought to be retreated with a platinum-based regimen. And if you use Taxol/platinum front line, Taxol/platinum is what ought to be used second line. The best data we have suggests that the expected response rate here is between 50-60% and the median survival will approach two years. Those patients who were resistant to front line therapy should be treated with alternative drug therapy. Now the list of drugs actually includes a total of seven.

The two regimens that are acceptable

So by 2006 our standard of care has changed in regard to the two regimens that are acceptable. I personally think that there are two regimens that are reasonable. You could add a third. The two that most people would mention would be Taxol as a 24-hour infusion at a dose of 135 per meter square followed by cisplatin at 75 per meter square, as tested in GOG111. Or Taxol 175 per meter square over three hours followed by carboplatin, dosed to an AUC of 4-5 based on the best evidence that we have. Some would add also a third arm, a 24-hour infusion of Taxol at 135 per meter square followed by carboplatin dosed to an AUC of 4-5. Those would be reasonable approaches to the treatment of ovarian carcinoma at the present time, for advanced disease.

Now let’s look briefly at limited disease, stage I or stage II disease. The way we approach this disease really is to divide the patient population into two groups; a low risk group and a high risk group. These are the current GOG definitions of these two groups; patients at low risk are those who have all of the following features: grade I disease that is intracystic in nature. That is, it is contained within the substance of the ovary, there is no tumor on the surface of the ovary, so that one way you might picture it is that the tumor does not have access to the peritoneal cavity to seed. There is no evidence of disease outside of the ovary. In other words, this is not a stage II patient. Peritoneal cytology is negative and there is no ascites. If all of these features is present that patient will be at low risk for recurrence. Patients are at high risk for recurrence if any one of these features is present: grade II or grade III disease, extracystic disease – that is, disease on the surface of the ovary so it has access to the peritoneal cavity – disease outside the ovary, stage II disease, positive peritoneal cytology or ascites. If any one of those factors is present that patient will be at high risk for recurrence and their approximate relapse rate will be 40%. Quite high. For the low risk group the approximate relapse rate will be 10%.

Now how do we manage these? Well, the best evidence we have to date comes from an Italian trial run by the GCOG group where they looked at three groups of patients. One group they labeled low risk, and uses the exact definitions that I just gave you – that is grade I disease, intracystic, no ascites, no peritoneal cytology and no extra-ovarian disease. Group number two were deemed to be at what they called intermediate risk, and this increased risk was associated with either grade II or grade III disease. That was the only reason for putting these patients into a higher risk group. And in the third group, patients were at higher risk because of one of the other factors we talked about, either ascites, positive peritoneal cytology, extra-ovarian disease or disease on the surface of the ovary. In the first of these groups, those at low risk, they simply observed these patients. They had 92 patients on study, at a median follow-up of 45 months there have been five relapses. The five year disease-free survival is 90%, the five years overall survival is 92% and with continued follow-up there have been virtually no relapses beyond this point. So it would appear that with just simple observation you are going to have a cure rate of 90% or better. For that reason our standard of care in this group of patients is no further therapy.