By Drs. Sebastian Bauer, West German Cancer Center, University of Essen, Germany
and Jonathan Fletcher, Brigham & Women’s Hospital, Harvard University, LRG Research Team


Researchers talk plainly about microGISTs, how this cancer develops and what all this might tell us



Dr. Jonathan Fletcher


Dr. Sebastian Bauer

Most of us have moles somewhere on our bodies. Yet, although many moles are akin to benign neoplasms, and are exceptionally common in the general population, we all know they very rarely transform into malignant skin cancers. What few people know is that one could describe GISTs as the “moles” of the stomach. Every third person has a tiny GIST in the stomach, so the incidence of GIST in the readers of the LRG newsletter is probably not substantially different from the incidence of GIST in the readers of “Sports Illustrated”. We know this (in part) thanks to the diligence of Drs. Kaori Kawanowa, Shinji Sakurai and their colleagues at several medical centers in Japan, who carefully studied 100 whole stomachs that were surgically removed from patients with non-GIST gastric cancers. The average length of an adult stomach is 10 inches (25 centimeters) and Dr. Kawanowa and colleagues sectioned these stomachs at 5mm intervals. This must have amounted to many hundreds, or even thousands of stomach slices to examine under the microscope, from each patient. To many people’s surprise, Dr. Kawanowa identified 50 very small GISTs in 35 of the stomachs.

However, essentially none of these tiny GISTs would be deemed clinically relevant or dangerous (1). What is striking and relevant for patients affected by metastatic GIST is the fact that a considerable number of the harmless tiny GISTs do have mutations of the KIT or PDGFRA genes – ie, the mutations that are the major tumorigenic mechanisms in malignant GIST (2).

Fortunately, a single gene mutation in the KIT or PDGFRA genes alone does not turn a normal gastrointestinal cell into a malignant GIST; otherwise we would have two billion GIST patients in the world. Indeed there is strong epidemiologic and laboratory evidence that multiple mutations involving crucial genes are required to create most cancers. These concepts were first popularized by Dr. Carl Nordling from Finland, who was an architect and historian, but who spent a lot of his time thinking about problems outside his own professional field. In 1953, he published a multi-mutation theory, aiming to explain why cancers became progressively more common as people aged (3). He proposed that the increase in cancer with aging could be explained by assuming that most cancers required six sequential mutations, and that therefore considerable time was required for these random mutations to develop in any given pre-cancer. His theory was later refined by Dr. Alfred Knudson, who believed that most cancers require at least two crucial genetic mutations (4).

Dr. Brian Rubin from Cleveland Clinic, a member of the Life Raft Group research team, made a related observation in his mouse models of GIST. In order to recapitulate human GIST he introduced a mutation of KIT into the mouse genome – a so-called germline knock-in model (5). Interestingly, Dr. Rubin’s mice did not develop typical malignant GIST but mostly just a hyperplasia (overgrowth of a specific cell type) of the gastrointestinal ICC-cells, which are the normal cell counterpart to GIST. Similar observations were obtained by Dr. Peter Besmer at Memorial Sloan-Kettering Cancer Center, who discovered the KIT gene, and who has created another mouse model of GIST, in which a KIT mutation was introduced into the mouse DNA (6). Notably, Dr. Rubin found that cross-breeding his KIT-mutant mice with mice that carry another cancer gene mutation resulted in much more malignant GIST – suggesting that Dr. Knudson’s hypothesis applies to GIST.

We know that patients with metastatic GIST can respond for years to imatinib, but this treatment generally does not fully cure the disease. The importance of the above-mentioned tiny GISTs, a.k.a. “microGISTs”, is that they give us a starting point, extremely early in GIST development, from which to identify the various additional crucial gene mutations that are required to create a malignant GIST. Furthermore, because we know that virtually all microGISTs persist as small benign tumors, or even undergo autodestruction (rather than progressing to a malignant cancer), these studies should enable us to identify biologic barriers to malignant progression in GIST. By understanding such barriers, one might reveal novel therapeutic targets in GIST, including abnormal proteins that prevent GIST cells from dying, even in the face of effective KIT/PDGFRA inhibition.

What do we already know of the additional genetic mutations – ie, those above and beyond KIT or PDGFRA mutations – that are required to create a malignant GIST? One traditional way to identify these is by “karyotyping”, in which GIST cells from surgical biopsies are grown in the laboratory setting, and the chromosomes from the cells are then identified using special stains and visualized using a microscope. Chromosomes contain the cells’ DNA, and karyotyping is the same approach that many of us are familiar with from amniocentesis studies, in which the chromosomes from fetal cells are assessed for disease-associated abnormalities, and to determine the gender of the fetus. A normal cell has 46 chromosomes, but the karyotypes in many malignant GISTs show losses of several chromosomes, and often losses of particular parts of a chromosome. These highly recurrent chromosome abnormalities identify the locations of genes that control growth and other important properties of the GIST cells: genes that normally keep the GIST precursor cells in line. When the GIST control genes are lost by chromosome deletion or mutation, this releases a brake on the benign GIST cells, enabling them to grow and behave more aggressively. GIST karyotyping and other chromosome or DNA studies, to date, have localized the general regions containing more than five crucial GIST-causing genes in the human genome, and several of these genes have been identified (see below). Identification of the other GIST genes is a major research priority, because these insights will likely lead to advances in GIST therapies.

Several researchers have observed that deletions of part of chromosome 9 are rarely seen in low-risk GISTs but are common in high-risk, malignant GISTs. The gene targeted by these chromosome deletions is CDKN2A (7, 8), whose function is to inhibit the cell cycle in Interstial Cells of Cajal or ICCs (the cells from which GIST arise) and other cells. CDKN2A ensures that the normal cell counterparts of GIST do not grow in an unregulated manner. Patients whose GISTs have lost CDKN2A have a substantially higher risk of developing metastatic disease than those who don’t. As part of the Life Raft Group’s research initiative a number of GIST have been analyzed not only by looking at the chromosomes but also by sequencing every gene (whole genome and whole exome sequencing). One major observation from these studies was that most malignant GISTs have genetic mutations, involving CDKN2A and other related genes, leading to dysregulation of the cell cycle. These studies suggest that nearly all GISTs require one or more mutations that increase cell cycle activity, in order to progress from a low-risk GIST to a high-risk GIST. Efforts are ongoing to corroborate the findings in a larger group of GIST samples using a comprehensive panel of cutting-edge research assays. Notably, mutations affecting cell cycle related genes do not seem to make GIST cells less sensitive to KIT/PDGFRA inhibitors such as imatinib. Nonetheless, these findings might be relevant clinically, given that various therapeutic inhibitors of the cell cycle, including CDK4/6-inhibitors, could restore cell cycle control in GISTs with CDKN2A deletions, or with other cell cycle defects.

Some of the GIST cell cycle defects result from suppression of a protein considered to be the master regulator and watchdog of the cell cycle – p53. Biologists tend to be nerdy people and devote some of their long late-night lab hours to naming newly discovered genes in a manner that stimulates the fantasy of those who work on them. Interesting examples are MerlinTeashirtSpätzle, Van GoghBrainiac, Hamlet and INDY (the last one stands for: I’Not Dead Yet because it is a gene which prolongs the life of fruit flies when mutant). By comparison, “p53” (also known as TP53) has a less provocative and bureaucratic-sounding name, referring to the size of the encoded protein on a kilo-dalton scale. One could argue that the all-important p53 gene warrants a name such as Cerberus or Archangel (gene names already claimed by some less-studied genes), because p53 very often decides whether a genetic insult to the cell can be repaired or – alternately – whether the cell should undergo apoptosis (i.e. go belly-up).

Treating GIST cells with imatinib and then adding drugs that “wake up” the mighty guard, p53 greatly increases the number of apoptotic cells (9). However, meddling with p53 may not be the least dangerous thing to do in patients – because p53 tweaking may also send normal cells to heaven. Current clinical trials are very cautiously investigating whether such treatments can be safely given to patients.
We hope the next few years will shine more light on the genetic and biologic differences between the “mole-y”-microGISTs vs. malignant GIST, and thereby bring us closer to the ultimate goal of curing GIST.


     1.    Kawanowa K, Sakuma Y, Sakurai S, et al.: High incidence of microscopic gastrointestinal stromal tumors in the stomach. Hum Pathol 37:1527-1535, 2006
    2.    Corless CL, McGreevey L, Haley A, et al.: KIT mutations are common in incidental gastrointestinal stromal tumors one centimeter or less in size. Am J Pathol 160:1567-1572, 2002
    3.    NORDLING CO: A new theory on cancer-inducing mechanism. Br J Cancer 7:68-72, 1953
    4.    Knudson AG, Jr.: Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 68:820-823, 1971
    5.    Rubin BP, Antonescu CR, Scott-Browne JP, et al.: A knock-in mouse model of gastrointestinal stromal tumor harboring kit K641E. Cancer Res 65:6631-6639, 2005
    6.    Sommer G, Agosti V, Ehlers I, et al.: Gastrointestinal stromal tumors in a mouse model by targeted mutation of the Kit receptor tyrosine kinase. Proc Natl Acad Sci U S A 100:6706-6711, 2003
    7.    Schneider-Stock R, Boltze C, Lasota J, et al.: High prognostic value of p16INK4 alterations in gastrointestinal stromal tumors. J Clin Oncol 21:1688-1697, 2003
    8.    Lagarde P, Perot G, Kauffmann A, et al.: Mitotic checkpoints and chromosome instability are strong predictors of clinical outcome in gastrointestinal stromal tumors. Clin Cancer Res 18:826-838, 2012
    9.    Henze J, Muhlenberg T, Simon S, et al.: p53 modulation as a therapeutic strategy in gastrointestinal stromal tumors. PLoS One 7:e37776, 2012

Learn more about the treatment of GISTs on our page about Initial Treatment.