Up Island Eggs
Ruminations 23. Steady As She Goes
posted Monday, 6 July 2009
Rumination 23. Steady As She Goes
By
Thomas P. Vogl
July 6, 2009
By
Thomas P. Vogl
July 6, 2009
Last week's CT scan, seven weeks after I started on Sutent, showed that I had stable disease; no new lesions and the extant lesions had not changed in size (within image mensuration error). While I have been described as having stable disease before, the increase over two months has never been this close to zero. So, I expect to be around at least another six months. What is particularly pleasant about this is that I am experiencing no side effects from the drug. I did have some minor side effects, mainly tiredness, for the first week or so, but that has completely cleared up. So, stable disease, no side effects, and having to go in to Boston only once a month and overnight only every other month -- what more could one possibly ask?
What is different about Sutent that it works on me? There is no definitive answer, but I can speculate. In contrast to the other drugs I have been on that regulate a very specific cellular pathway or process, Sutent targets multiple cell surface receptor tyrosine kinases (RTK). Of particular interest in my case is that it inhibits the RTK KIT. The drug that Dr. Hodi first proposed when my metastases were detected was Gleevec, which is also a KIT inhibitor. However, tests at that time concluded that while I did have KIT changes, they were not the mutations that Gleevec addresses. Sutent has broader KIT inhibitory effect. Sutent also inhibits most receptors for platelet derived growth factor as well as VEGF (Vascular Endothelial Growth Factor) receptors. By inhibiting VEGF, which is necessary for the formation of new blood vessels that tumors need to feed their growth, tumor growth is fully or partially stopped. It appears that when VEGF is inhibited the immune system is stimulated, which is a good thing. So, although Sutent is composed if a single chemical entity, Sunitinib (1,1-dimethylbiguanide - if anyone cares), it attacks a tumor several different ways simultaneously. For that reason I can hope that it will continue to work on me for a longer time than some of the single pathway inhibitors we have previously tried. Given the current discussions of health care in Washington, I feel obliged to mention that were it not for the excellent drug coverage provided by my former employer, I could not afford Sutent, whose actual price significantly exceeds my total income from all sources -- it costs an astounding (one might say obscene) $8259.11 a month!
This calm time is my opportunity to talk about some peripheral issues and perceptions. Possibly the most interesting one is what causes some people's tumors to grow rapidly and other people's slowly. The bare outline of a study to address this question is presented below the break.
There are, in addition, two personal observations that are appropriate in this Rumination. One is a consequence of its date of publication falling very near to my 80th birthday. When I was a teenager, 65 years ago, I found it unfair that I was born just a few years to soon; I anticipated a life expectancy of less than 70 years and would therefore miss the millennium. I certainly never expected to see 80. Two years ago, when the liver metastases were first discovered, I was reliably (and reasonably) informed that I would be symptomatic within a few months and, by implication, gone within a year. Again, little, if any, chance of reaching 80. Yet, here I am. An amusing sidelight is that 30 or so years ago I decided that I would declare myself to be middle aged half way between whatever my present age was and 65. Xeno's paradox to the contrary not withstanding, I became 65 (and middle aged) after all. I then decided that I would consider myself 'old' halfway between my present age and 80. Unfortunately, Xeno didn't help that one either and here I am, about to become 'old'. Ah, well, I don't feel any different, and that is a good thing. As my birthday present to me, I promise myself more postprandial naps.
The other observation is to reassure any of my readers who may be the reluctant owners of a chronic disease, their caregivers, and their family, by making explicit what we all know implicitly but often forget or ignore. Anytime we feel a twinge or catch a bug or a cold, we inevitably worry whether it is related to the chronic condition and whether it suggests a relapse or exacerbation. This suspicion is perfectly natural but can easily get out of hand and interfere with getting on with life. Since everyone is subject to such twinges and bugs, it can easily drive individuals with a natural inclination to worry to distraction. I think the best remedy is to constantly remind oneself that such symptoms are not related to the chronic disease unless the relationship is proven. Innocent until proven guilty is the way to go - Kathy and Katherine, please take note.
***********************************************************************************************
It should come as no surprise that I am fascinated by the question of what mechanisms modulate the aggressiveness of cancer. So the question that immediately occurred to me is what is the statistical distribution of survival times. (Is it random – flip of a coin, or is there a consistent pattern). While my access to literature is somewhat limited, I am discouraged by the fact that while I found that considerable effort has gone into surrogate predictive models and measures such as five year survival statistics, Kaplan-Meier survival curves, progression-free survival, and time to progression, I could find no information regarding the statistical distribution within any of these measures of survival - are they normally distributed, skewed, bimodal, or ...? Because it is not possible to know exactly when a metastases first developed (it can be weeks, months, or years between that point in time and when it is detected) directly determining the underlying distribution is a daunting problem; however, the distribution of the surrogates can reasonably be expected to reflect the distribution of the underlying reality, sufficient at least to explore the general shape of the distribution. I could not find any data on the distribution.
What follows, although cast as a broad outline of a study, is really a plea to start now to collect the specimens and relevant patient data so that as large a specimen library as possible be readily available when cost effective technology is ready for the task. The study that I am outlining is best carried out in an institution with a large population of cancer patients, such as Dana-Farber, Sloan-Kettering, etc. Although it will not be able to be carried out for a few years until genomic, epigenomic, and proteonomic testing of biopsy specimens becomes cheaper and routine, that time is close enough [e.g., S.P. Shay et al., Mutation of FOXL2 in Granuloma-cell Tumors of the Ovary. NEJM 360: 2719-2729 (June 25, 2009) and J. Shendure and C.J. stewart, Cancer genome on a Shoestring Budget, NEJM 360: 2781-2783 (June 25, 2009)] that now is the time to begin the collection of data and specimens.
Consider the totality of the cancer patient population, excluding those who had successful resections of their tumors and sufficient time has elapsed to consider them cancer-free. For each of the commonly accepted classifications of cancer (prostate, pancreatic, colon, GIST, HER-2 positive breast, etc.) data on individual survival is available. That some patients progress quickly and others slowly is well known. From the patients in each classification, select the top and bottom quartiles or quintiles of survival time as the study population. Of course, the time spans involved will differ markedly with the cancer involved; none the less, for each tumor there are patients who progress very quickly and those who progress very slowly (relative to the mean). Consider this subset of the total population as the subjects of the study whose underlying hypothesis is that there are detectable differences between the fast and slow progressors (FPs an SPs). The task is to define and quantify the similarities and differences among clusters of FPs and SPs. Equally important is to determine the cluster characteristics that best characterize / classify / group cancers.
I am using the term 'cluster' in the statistical sense of a group that segregates a property. Each study subject is a member of many clusters. (As a specific example, I cluster with males, melanoma, mucosal melanoma, 75-85 year age group, married, hepatic metastases, c-KIT replication positive, etc., etc.) Of course, many of the possible clusters are irrelevant to the study at hand and others will turn out to be so.
The first task of the study is to define as many potentially relevant clusters of the whole study population, FPs and SPs combined. I would expect these clusters to include ( and also to include many others) primary tumor site, common clinical and lab characteristics, specific aberrant cellular pathways, genomic anomalies, specific protein excess and insufficiency, which drugs have been tried with what results, etc. It is important to note that these clusters contain both SPs and FPs and that each cluster includes tumors from all organs and tissues that fall within the cluster definition. E.g., the c-KIT positive cluster includes all organ and tissue types that are c-KIT positive. It may well be desirable to define several different c-KIT positive clusters to separate out different alleles or mutations and replication. (Cluster compactness and correlation analysis will reduce the number of clusters that are retained.)
Once the clusters and their members have been selected, the next task is to identify dissimilarities between the SPs and FPs in each of the clusters identified in the first step. Of course the same database can be used for other analyzes such as the relationship of cluster characteristics to drug response. At this time, what is important is the establishment of a sample and data repository. The details of any study proposed now will certainly be significantly modified by the ongoing flood of experimental results and their interpretation.
If nothing else, I suggest that any serious contemplation of a study of this kind will go a long way to removing the blinders resulting from the organ based tumor classification and treatment (particularly chemotherapy) that is a consequence of historical imperatives that focus on organs but may be of marginal significance in the light of current oncological science. The painfully slow pace of testing and approving Sutent for KIT positive tumors other than renal cell carcinoma and gastrointestinal stromal tumors (GIST) is a worthwhile case study this issue.
The collected Ruminations may be found at http://upislandeggs.com/Ruminations.htm
and my e-mail address at the bottom of the page at http://upislandeggs.com/
comments