ASCO 2009, Melanoma Meetings|May 15, 2011 9:43 pm

Melanoma Oral Abstracts ASCO 2009

Each year at ASCO there is one and only one session where selected abstracts are presented via oral, Powerpoint, talks. The following is a summary of the papers presented this year. As always, this represents my opinion and interpretation of what I heard, both during the talk and afterwards in the conference hallways.

First paper, Cancer Testis Antigen (CTAg) as a biomarker–These are a set of proteins expressed almost exclusively by cancer cells including melanoma and/or testicular cells, hence the name. This study looked at the potential of a few CTAgs as biomarkers (see ASCO summary). The bottom line was that patients who were “CTAg-negative” had improved relapse free survival (RFS) but not overall survival (OS). CTAg can also cause an immune response by patients but it wasn’t clear if this was good or bad. Summary: not sure if CTAg expression is always good or sometimes good–therefore, don’t see it as ready for primetime at all.

Second paper, long term follow-up of EORTC (European cooperative cancer group) data on sub-microscopic sentinel lymph node metastases. The basic question here is whether there is an almost incidental level of sentinel node involvement, where the patients do not require additional surgery in the form of complete lymph node dissection or removal. This data continues to mature without change and I believe it to be very impressive. It should immediately change one thing: all sentinel node pathology reports, if the sentinel node is involved with metastatic disease, should report the measured maximal transverse diameter of the largest lesion in the lymph node. The data continue to show that patients with sub-microscopic metastases, less than 0.1 mm in size, behave identically to patients without any metastatic disease detected in the sentinel nodes, with 94% 5 year survival. About 3-5% of those patients have additional positive non-sentinel nodes detected on completion lymph node dissection but this doesn’t seem to affect survival (although statistically it’s a small group). The basic question is, who should get completion lymph node dissection with positive sentinel nodes? The answer seems to be all patients with more than one sentinel node involved, or any patient where the tumor burden in the sentinel node is >0.1 mm. Apparently nobody has seen any patients with more than one sentinel node involved if the tumor burdnen is < 0.1 mm. Summary–nice paper with ‘clean’ data from a well run cooperative group. They are beginning an international registry of these patients (which the Atlantic Melanoma Center has been invited to join by Dr. Eggermont) and I think this should change surgical therapy for these patients.

Third paper, micro RNA or miRNA as a potential biomarker in metastatic melanoma. This study from NYU examined tissue from 59 patients with advanced melanoma and identified six apparently predictive miRNA that could be measured. With such a small sample, and the way they processed the data (no independent validation set of data to test the proposed model), this is also NOT READY FOR PRIMETIME.

Fourth paper, ulceration of primary melanoma and responsiveness to adjuvant interferon therapy. This was presented by Dr. Lex Eggermont of the EORTC. It was chock full of data and I couldn’t get it all down fast enough (as Dr. Eggermont said, “fasten your seat belts”). However, the data were very consistent. Basically, in reviewing the recent EORTC experience with adjuvant interferon, either standard Intron or the newer long-acting PEG-Intron (not yet approved in USA), they found that primary melanoma ulceration is very strongly associated with clinical benefit from either interferon therapy. In fact, patients whose melanoma lacked ulceration apparently derived NO benefit from interferon therapy. The numbers: in 2644 combined patients from two most recent trials, ulcerated melanoma patients had a 25% reduction in recurrence risk which was statistically significant. This benefit was even greated for patients with minimal lymph node disease, ie 0-1 lymph nodes involved, with a 31% reduction in recurrence and a 42% reduction in death rate. This is very impressive, since usually the RFS improvement dwarfs the impact on OS! This benefit almost disappeared for ulcerated melanomas with multiple lymph node involvement (N2) and completely disappeared for non-ulcerated melanomas.

So the bottom line is, who should receive interferon therapy? As Dr. Eggermont pointed out, this data should not be considered definitive but is rather hypothesis generating and as a result he is initiating a new EORTC trial examining the potential benefit of PEG-Intron therapy for Stage II (thicker, no lymph node involvement) ulcerated melanomas. In the spirited discussion that followed, the most interesting comment came from Dr. Kirkwood who noted that the ECOG database on interferon trials (that ultimately led to the approval of Intron by the FDA) does NOT show the same benefit segregated to patients with ulcerated primary melanomas. From a surgical perspective, great stuff and we look forward to being a research site on this EORTC trial also.

Next paper: More Ipilimumab data, this time looking at potential biomarkers indicating response to “Ipi”. Nothing big here in terms of changing practice. The real test of Ipi will come when the Phase III trial data is mature and analyzed. Clearly it has activity in melanoma, meaning there are significant numbers of patients whose cancer responds to the drug. The question will ultimately be if the pre-specified statistical endpoints of the main trial will be hit. The worrisome thing is that a slightly different trial in a slightly different drug, Tremelimumab, was completely negative and Pfizer has halted further development of that drug. In any event, although many abstracts about Ipi were presented in one form or another at ASCO, none came close to definitively confirming or denying its potential for FDA approval.

Next paper dealt with predictive modeling of chemosensitivity of melanoma based on gene expression profiling. This is a molecular technique that assesses the various genes that are turned on and off in a person’s melanoma. Specialized computer software groups the data statistically and by patterns that may help classify a melanoma. In this paper, they were seeking to classify melanoma by gene expression profiling of epigenetic genes (ones that control structure and function of the cell’s DNA machinery), and correlating those profiles with response to chemotherapy. They ended up with a “classifier set” of 82 genes. They analyzed them with neural network software, which builds a weighted decision algorithm based on the available data, using a mechanism felt to replicate the way humans learn. Without getting too technical, I believe their neural network was “over-trained” which is what happens when you don’t have enough samples to separate them into various groups to develop, test and validate your neural network. Without such separate groups, the neural network can be trained to always be correct, which really should not be possible, since nature always has a few unexpected outcomes. But this paper’s network was always right (in the extremely small, 21 patient, sample they had data on). Therefore, much more testing needs to be done, on a wide variety of cases, which will take some time. Ultimately this is the sort of testing that has the potential to identify who will benefit from which type of chemotherapy. These sorts of tests will render the current “chemosensitivity” testing completely irrelevant, if they aren’t already.

Next up was a small Phase II trial of a drug called Dasatinib, which is an antibody that functions as a targeted drug inhibiting multiple kinases. They tested 35 patients with previously untreated metastatic melanoma. There were 2 partial responses and 4 patients with stable disease, with a median PFS of 8 weeks. This is very unimpressive. They believe that biomarkers (there’s that word again) may help identify exactly which patients could respond to this drug, but the initial genetic mutations they looked for didn’t help. The audience got restless–why was this drug study being presented in a forum usually reserved for exciting, groundbreaking, results and drugs???

Next paper was a study comparing high dose Interleukin-2 (IL2) to IL2 plus a peptide vaccine. This is a fairly old study, that was begun in 1998.  Patients had to be HLA-A0201 positive, as the peptide vaccine is specific to that immunologic profile. Parenthetically, that peptide vaccine, a very hot topic when the trial was initiated in 1998, has basically been discarded as a potential therapeutic agent as multiple studies showed it had no benefit. There was more toxicity in the combination arm, particularly increased amounts of cardiac arrhythmias, although nobody knows why. There were also more responses in the combination arm, although it wasn’t a very high amount. This was statistically significant, although the response rates were under 20% in both groups. There was no survival benefit.

Next study was the Synta sponsored SYMMETRY trial. This was a randomized Phase III trial of weekly Taxol with or without Elesclomol. This Phase III trial was based on a positive similar Phase II trial announced in late 2006. In February 2009 the trial was stopped by the independent Data Safety Monitoring Board for reasons not clear at the time; this was the first public update of the data and results.

This was a large trial enrolling over 620 patients total. It was well designed in terms of stratification of patients and statistical considerations. However, when the interim data was reviewed by the DSMB in February, they found that the combination arm had signficantly more deaths than the control arm (80/309 vs 53/312). In addition, the primary endpoint of PFS could no longer be statistically positive, with a hazard ratio (HR) of 0.84. Interestingly, PFS became positive when only patients with normal LDH levels were looked at (the trial allowed up to 2x normal LDH blood test levels). This is similar to the post hoc analysis of the original Genasense trial from a few years ago. When the study was stopped, the HR for overall survial was 1.62 meaning that patients receiving combination therapy were 62% more likely to die than those receiving taxol only. When the data was updated in April, two months later, the HR was now 1.31, showing a 31% increase which is no longer statistically significant. There is nothing in the data to explain why more patients died on the combination arm. SUMMARY–It didn’t work but nobody knows why. Sad and mysterious. Another candidate for the infamous “melanoma graveyard”.

The last oral abstract presented was a randomized trial of adjuvant radiotherapy for node positive melanoma, from an Australian/New Zealand consortium. Patients had to have palpable nodal disease, which is unusual in today’s world where sentinel node biopsy is performed so widely. By explanation, at least 80% of metastatic lymph node disease from melanoma, when identified, is microscopic and non-palpable (ie can’t be felt on examination).

This study’s primary endpoint was relapse within the lymph node basin at risk. The secondary endpoints were OS and PFS.

The results were interesting in that of the 161 patients (of 250 in the study) who relapsed, only 62 relapsed in the lymph node area at risk. As a result, the primary endpoint was positive in that there was a 53% reduction in lymph node basin relapse, BUT neither OS nor PFS were improved. In the discussion that followed, it was pointed out that given the data it was difficult to recommend adjuvant lymph node basin radiation therapy if it only had a very local benefit. Parenthetically, in the US, adjuvant radiation therapy is usually not given, especially in the axilla (armpit) or groin since those two sites have such a high risk of permanent lymphedema after radiation. The presenter was unable to give the exact lymphedema rates when asked.

BOTTOM LINE; after oral abstracts, there were a few potential changes to surgical therapy and also a new trial with PEG Intron but both of those changes had been well known since at least the World Melanoma Congress a few weeks ago. There were no earthshattering results from new agents or new tests. There were certainly no drug regimens or tests that are ready to be implemented on any level outside of clinical trials.

Eric D. Whitman, MD

Senior Editor


  • Share this post:
  • Facebook
  • Twitter
  • Delicious
  • Digg
Shopping Cart (0 Items)
Your cart is empty!

Subtotal: $0.00
Total: $0.00