GIST Support Wiki

ASCO Convention - 2010, Chicago, Illinois

It was a whirlwind (Chicago, the windy city... ha!) at ASCO this year for GSI representatives Carolyn Grobe and Becky Bensenhaver. We covered an immense amount of information regarding drugs being developed for GIST and other cancers, but we had a lot of fun too. We were able to cover most of our GSI agenda, and often needed to split up to cover different topics at the same time... like novel therapies for GIST and the PI3K/mtor inhibitors. Links will be made to the meeting abstracts now available at and when the presentations become available online, we will update with those links.

Carolyn and Becky

Short description of abbreviations you will see in some summaries:

TKI- tyrosine kinase inhibitors

IM - Imatinib (Gleevec)

SU - Sunitinib (Sutent)

N - Nilotinib (Tasigna)

SDH - Succinate dehydrogenase

AE - adverse events

PD - progressive disease

PRpartial response

RFSrecurrence free survival means how long after surgery before your visible disease returned on a CT, and it is not particularly linked to how long you lived.

PFSprogression free survival means how long your tumors were not growing and this may not necessarily be linked to how long you lived, OS. You can have a progression event and yet live a long time if you regain control of the disease.

OSoverall survival means how long you live. This is the most important finding. Most patients care about how long they live more than they care about anything else.

SURVIVAL is an over used term, and is not used in the abstracts in the same way it is commonly used in everyday life. Progression Free INTERVAL and Recurrence Free INTERVAL would be better terms. Survival when used in an abstract does not necessarily mean how long you lived.

SIGNIFICANCE In statistics SIGNIFICANCE only means that the observed results were not likely to have happened by random chance. Statistically signficant means that the result probably happened as a result of the treatment and not from quirks of random chance in the patient sample. Statistical SIGNFICANCE does NOT imply one way or another whether the benefits were PRACTICALLY IMPORTANT.

In everyday usage, signficance implies that something was practically important. This is not the case in statistics. A drug, for example could with statistical signficance extend the overall survival of a patient by 5 weeks. The patient may not think the extra five weeks was practically important in the grand scheme. Remember this about the term "signficance" when you read the abstracts.

b.i.d.medication taken twice a day

Glossary of other clinical trial terms:


Lessons Learned from Clinical Trials In Gastrointestinal Stromal Tumor

Presented by Drs. Jean-Yves Blay (University Claude Bernard), Ronald DeMatteo (MSK) and Michael Heinrich (Oregon Health and Science Univ. OHSU)

Combination Therapy with Established and Novel Targeted Agents to Improve Outcomes in Patients with Advanced GIST Speaker: Jean-Yves Blay, Centre Leon Berard

At present, Gleevec is considered the first-line defense in the treatment of GIST. However, the median time to progression for high-risk GIST patients treated with Gleevec is in the range of 20 – 24 months. Dr. Jean-Yves Blay from France presented a discussion at the 2010 ASCO meeting of the rationale for use of drugs in combination for the treatment of Gleevec-resistant GIST. A 2006 study at OHSU indicated that 70% of metastatic Gleevec-resistant tumors contain secondary mutations on affecting additional exons. This is especially true pertaining to exon 11 mutant primary tumor types. The secondary clones vary in their drug sensitivity, with mutations located on exon 13 or exon 14 being likely to demonstrate sensitivity to Sutent and Tasigna. Secondary mutations on exon 17 are more likely to be resistant to Sutent, but have sensitivity to Nexavar and Dasatinib (Sprycel). Exon 18 mutations tend to be resistant to both Gleevec and Sutent.

The outcome of treatment is dependent not only upon the nature of the primary mutation but also on the nature of the secondary mutation indicating the benefit of study of use of a combination of drugs in the future.

It has been determined that there is often reactivation of the AKT mTOR pathway with the emergence of resistance. The concept of pathway reactivation led to the study of the combination of Gleevec and an mTOR inhibitor. The results of this clinical trial showed promise in reverting resistance in laboratory testing.

One of the theoretical benefits of combined drug therapies would be the ability to address not only the activity of the tumor cells with a tyrosine kinase inhibitor (TKI) but also the extracellular supportive structural framework of GIST tumors with blocking of vascular endothelial growth factor (VEGF) receptors. Sutent is based upon this concept. Perhaps combinations of other TKIs and VEGF drugs could hold promise for future investigations.

Designs for future clinical trials could also attempt to answer treatment questions such as: Is A + B > A alone? Is A + B > A then B? Is there an advantage to use of 2 TKIs simultaneously? Is there an advantage to use a TKI + VEGF simultaneously? Is there an advantage to using a TKI with a cytotoxic chemotherapy agent?

The following are some additional examples of drug combination trials:

Gleevec plus Midostaurin: A review of the Phase I/II trial of Midostaurin and Gleevec in patients with refractory GIST was presented. Midostaurin is an oral broad-scale tyrosine kinase inhibitor including KIT PDGFRa + b, VEGF2, FGFR, and FLT3. Midostarurin had been shown to have significant in vitro activity against KIT and PDGFRa mutations. There were significant drug-drug interactions noted with this combination of Gleevec and Midostaurin. Midostaurin induced decreased Gleevec blood levels requiring use of alternating of the two drugs. This combo did not fare well secondary to drug-drug interactions and is closed.

Gleevec + Sutent: This study is being conducted by the Spanish Sarcoma group but the results haven’ t been fully reported.

Gleevec + RAD001 (everolimus)(mTOR inhibitor): This combination was investigated several years ago in a US and recently in Europe in Phase II studies. The combination was based on the concept that there could be a positive synergistic effect with combined use of a tyrosine kinase and an mTOR inhibitor. Patients were randomized into 2 groups. Group I had failed Gleevec only. Group II had failed Gleevec plus other drug. Results: Gleevec increased the bioavailability of RAD001 resulting in a need for lower RAD001 dosing. However, RAD001 did not affect Gleevec plasma levels in a converse fashion. The combination of the two drugs were reasonably well tolerated. 37% of the patients demonstrated a 4-month PFS by radiographic evaluation, with some patients demonstrating a substantial improvement. The results of this study provide justification for further evaluation of the Gleevec + RAD001 combination in patients after failure of first-line Gleevec therapy or second-line therapy with a second drug.

Gleevec + Doxirubicin (cytotoxic drug): This study published in 2008 by the Spanish Sarcoma Group (GEIS) evaluated the effect of the combination treatment of Gleevec plus Doxirubicin at 2 different dose levels on 25 progressive GISTS with previous Gleevec failure. Radiographic evaluation revealed two of the trial patients (9%) had a Partial Response and 10 (43%) had Stable Disease for greater than 16 weeks. The median Progression-Free Survival was 84 days. It was of note that one Wildtype GIST patient and one other mutant GIST patient had Stable Disease for a substantial period of time. The conclusion is that selected groups need to be further explored. Gleevec + IGF1-R inhibitor: This combination could prove beneficial for treatment of Wildtype GIST and trials are being sought on this question.

Sequential therapies/rotation: This is an area that warrants further investigation as this approach limits drug interactions, improves tolerance, and may be an improved method of controlling emergent resistant clones. Examples of success with treatment rotations in CML may be interesting in this perspective.

Conclusions of the presentation:

Frequent emergence of Gleevec resistance justifies the need for further research in combined or sequential therapy approaches.

Future treatment concepts to watch: the potential role of switch pocket kinases to modulate the behavior of kinase targets in a unique way. Presented by Dr. Michael Heinrich, OHSU

Most GIST tumors harbor mutant KIT or PDGFRA kinases, which are targets of imatinib (IM) or sunitinib (SU). Resistance to IM or SU is commonly due to the acquisition of secondary mutations that markedly reduce the potency of kinase inhibitors. Both IM and SU bind to the KIT/PDGFRA ATP binding pocket. KIT and PDGFRA kinases possess negative and positive regulatory domains that modulate kinase activity by competitive binding into the switch pocket domain.

Switch pocket kinase inhibitors (SPKIs) occupy this pocket and block enzymatic activity. We tested the activity of three SPKIs (DP-2976, DP-3636, DP-4444) in enzymatic assays using various wild-type and recombinant KIT kinases. In addition, a panel of kinase mutants was biochemically profiled for sensitivity to these SPKIs using several cell line models, including GIST cell lines derived from imatinib-resistant tumor clones. All three SPKIs were very potent against wild-type KIT and PDGFRA.

In addition, these compounds maintained their potency against representative GIST-associated drug resistance mutations (V654A, T670I, D816H, D816V). DP-2976 and DP-3636 were also potent against the pan drug-resistant PDGFRA D842V mutation. To confirm these observations in a GIST cellular context, we tested the relative potency of these compounds against two previously described IM-resistant GIST cell lines (exon 11 + V654A, exon 11 + D820A). All three compounds were significantly more potent than IM or SU for both cell lines. These results also correlated with enhanced inhibition of downstream signaling pathways, cellular proliferation and survival.

Conclusions: Switch pocket kinases demonstrated markedly superior in vitro potency compared with IM or SU against a panel of GIST-relevant mutant kinases. Based on these results, we hypothesize that these compounds might have useful clinical activity against drug- resistant GIST. Further preclinical studies are indicated to develop these compounds for human clinical studies.

The Role of Adjuvant Imatinib after Primary Resection of Localized Primary GIST: An Update on the Z9001 Trial - Optimal Duration of Therapy Speaker: Ronald DeMatteo, Memorial Sloan-Kettering Cancer Center

GIST recurrence factors (size, location, mitotic rate and whether it has ruptured) from this study have been calculated in a nomogram. Dr. DeMatteo revealed studies indicate that mitotic rate seems to be the most important factor for GIST recurrence: the hazard ratio for recurrence for tumors that have mitotic rates over 5/50hpf was 11 (11 times greater then for tumors with lower mitotic rates) whereas the hazard ratio for large tumor size was only 2 (2 times greater) and the hazard ratio for non-gastric tumor location was also 2. In one example shown a tumor with a mitotic rates over 5/50hpf had an 80% rate of recurrence compared to a recurrence rate of 20% with mitoses under 5/50hpf. This information will help to decide which patients are at high risk and should receive adjuvant therapy with first-line drugs. The nomogram for estimating the probability of recurrence-free survival after surgical resection of primary GIST is located on the GSI website at:

Using GIST Genotype to Optimize Kinase Inhibitor Therapy and Dosing the Unusual Patient with Non-exon 9/11 GIST Speaker: Michael Heinrich, Oregon Health and Science University, gave a presentation on the impact of GIST genotype upon the selection of the optimal treatment choice.

Small molecule kinase inhibitors such as IM and SU have transformed the medical treatment of Advanced GIST. However, at the molecular level, GIST is a heterogeneous group of diseases which can be categorized in different ways, for example: Adult vs. Pediatric GIST, KIT mutant vs. PDGFRa vs. Wild-type (WT) GIST, and even within WT GIST now we’re beginning to see additional heterogeneity. Biochemical and clinical data suggest that treatment can be optimized using genotyping results to guide TKI dosing.

The results are based on examination of 2700 GIST tumors in his laboratory with Chris Corless. It investigates metastatic tumors and the in vitro potency of IM against mutant kinases to find the dose needed to control tumor growth. This information will also help to determine the dosage in adjuvant therapy.

The following summarizes their recommendations for target starting doses of treatment for metastatic GIST based on primary tumor genotype:

Exon 9 - (9.1% of GISTs) Start at 800 mg IM to inhibit mutations(similar to wild-type). SWOG North American Phase III study indicated 17% inhibition at 400 mg and 67% with 800 mg. However EORTC study indicated no effect of IM dose on response rate. The overall survival was increased, although not statistically significant.

Exon 11 - (59.6% of GISTs) has extreme sensitivity to imatinib (IM) Start at 400 mg - no difference in PFS and OS between 400 and 800 mg of IM

Exon 13 - (1.9% of GISTs) Start at 400 mg dose IM

Exon 17 (.08% of GISTs) Start at 400 dose IM

PDGFRA Exon 12 (2.5% of GISTs)and Exon 14 (.5% of GISTs) may be more sensitive to IM than exon 11 and often have reduced risk of recurrence. Start at 400 mg dose IM

PDGFRA Exon 18 (7% of GISTs) D842V resistant to IM -800 mg/day IM or clinical trial; For not D842 IM sensitive - 400 mg/day dose IM; For not D842V, IM resistant - 800 mg/day dose IM or clinical trial with switch pocket drug.

WILD-TYPE - usually includes BRAF or RAS mutation, SDHB deficient and NF-1 associated.

Adult wild-type GIST is rare but has kit mutations in 80-85%, usually in the stomach (70%) with spindle cell histology and aggressive behavior of metastatic disease. Possibly igher dose sensitivity to IM - start at 400 mg dose IM to 800 mg dose IM.

No known KIT or PDGFRa mutations - 400 mg dose IM.

BRAF mutant WT GIST - Clinical trial or await BRAF inhibitor.

SDHB deficient WT GIST (+) SDHB Immunohistochemical staining - Future role for front-line SU, Sorafenib, or renal cell carcinoma therapies.

NF-1 related GIST - MEK inhibitor.

Pediatric wild-type GIST is very rare and has KIT/PDGFRA mutations in less than 10%, often multi-focal primary of epitheliod histology, 90% in stomach and metastatic disease can behave indolently. Pediatric GIST has some connection to renal cell carcinoma due to derangement in succinate and fumarate.

Often SU is a better drug choice than IM for pediatric GIST and RCC.

Optimize surgery, consider SDH complex genotype testing, close surveillance indicated, and consider enrollment in clinical study.

Pediatric and Adult Wildtype GIST

Presented by Dr. Michael Heinrich, OHSU

Adult GIST is rare, making PED GIST very, very rare. In adult GIST more than 80% of patient’s tumors have either a KIT or PDGFra mutation. Whereas in PED GIST the incidence of a KIT or PDGFra mutation is very unusual, less than 10%. There is a slight male predominance of GIST in Adult WT patients. However, in PED GIST it’s a very extreme difference with girls or young women being 8-9 times more common than boys or young males. In adult GIST there is usually a unifocal primary compared to PED GIST, which often has a multi-focal primary presentation. In the adult form spindle cell histology is most common whereas in PED GIST epithelioid is most common. The stomach is the most commonly involved site in the adult, about 70%. In the PED GIST population the stomach is the site of the primary tumor 90% of the time. And in most cases adult WT GIST metastatic acts very aggressively, whereas in most cases of PED GIST metastatic disease can behave indolently in that they may survive many decades with the help of a good surgeon in managing their metastatic disease.

Now, even within PED GIST there are discreet syndromes. There is Carney’s Triad, which is the complex of GIST, Paraganglioma (PGL), and Pulmonary Chondroma. There is also Carney-Stratakis Syndrome, which presents as the combination of GIST and PGLs. We have now recently learned that this particular type of GIST is associated with germline loss-of-function mutations of the Succinate-dehydrogenase (SDH) B, C, or D gene. There is also a sporadic Pediatric GIST that doesn’t fit into either of the above 2 categories, and there is NF-1 associated GIST. NF-1 GIST is rare with around 6-25% of neurofibromatosis patients developing one or more GIST during their lifetime.

In the case of Carney-Stratakis syndrome (GIST tumors and PGLs) the underlying biology involves the SDH composed of 4 members: SDHa, SDHb, SDHc, and SCHd which forms something called mitochondrial complex II. This complex links SDH activity in the KREBS cycle in the cytoplasm with the mitochondrial electron transport chain. When you disrupt this complex at SDHb, c, or d the SDH activity can remain normal but the electron transport chain is disrupted. Deficiency of this complex causes both reactive oxygen species generation and up-regularities of hypoxia-induction factor (HIF-1a). We are now beginning to see a curious overlap of the biology of PED GIST with renal cell carcinoma (RCC). In RCC a deficiency of the Von-Hippel Lindau (VHL) protein leads to up-regulation of HIF-1R and it’s downstream targets such as Glutamine (GLUT1) and VEGF. There are other familial RCCs that involve deficiencies in SDH or Fumerate. It is now clear that with PED WT GIST, some are caused by derangements of this pathway suggesting a common biology between PED GIST and RCC suggesting the possibility that maybe everything we’ve learned about RCC in terms of therapeutics might be applicable at least to SDH-deficient PED GIST.

If you take a group of PED GIST patients you can actually find that even though they apparently didn’t have Carney-Stratakis syndrome (because they didn’t present with PGLs), if you sequence their SDH b or d genes you actually find mutations in 11%. You can also do immunohistochemistry(IHC) for SDHb deficiency (PGL with SDHb deficiency stains differently in the PGL with VHL deficiency). Using their IHC analysis they found that the SDHb protein was absent in 100% of PED GIST patient despite the absence of mutation, again suggesting that defects in cellular respiration my be a central oncogenic mechanism in PED GIST.

What about adult WT GIST? Adult WT GIST, present in 15-20% of adult GISTs, likely represents a heterogeneous group of disease. We can find KIT activation at least biochemically in most cases, however, the underlying biology is not understood. There are identifiable subsets in adult WT GIST. There are BRAF mutant GISTS which are rare, RAS mutant GISTS which are very, very rare, and SDHB deficient GIST which are not being identified immunohistochemically. We’re not sure why they lack the SDHb protein because they may not have SDH mutations, but this is an important lead for us. And there is also NF-1 associated GIST. Again, 6-25% of patients with NF-1 will develop one or more GISTS during their lifetime. And then there are some Adult WT GISTS with none of the above, which means we really know what is going on.

In terms of WT KIT going back to kinase sensitivity, it’s sensitive to IM but requires a higher dose (1000 nMolars) than a KIT exon 11 mutation. In the Phase II clinical studies with the adult WT patients treated with 400 vs. 800 mg/day IM there was no difference in PFS or OS between the 2 dose arms. So, for Adult WT GIST as a group, not considering the subgroups that we can’t routinely clinically identify, the data suggest that IM 400 mg/day is the recommended starting dose.

What about BRAF V600E mutant GIST? In three reported series, overall 7% of WT GISTs were reported to have BRAF. BRAF mutations were not found in KIT or PDGFra mutant GIST. In our series of WT GIST we found 1 BRAF mutation out of 99 patients. Also there were 2 cases with RAS mutations. And so there may be a potential role for BRAF or MEK inhibitors for treatment. In GIST BRAF and RAS mutations are rare, whereas in melanoma exactly the opposite is true with BRAF and RAS mutations being common and KIT mutations being rare. How does this play out in terms of treatment? At least in the case of Pediatric GIST it’s interesting that the front line response to IM is not that great. It was noted in a study that the difference between time-to-progression (TTP) on SU vs. IM showed a trend toward longer TTP for 5 out of 7 patients using second-line SU than using first-line IM. It was speculated by Dr. Heinrich that possibly this goes back to SU being a better RCC treatment. Perhaps this can be attributed to the concept that PED GIST patients may be be more biologically similar to RCC than adult GIST.

Abstract # 10008 - Succinate dehydrogenase in KIT/PDGFRA wild-type gastrointestinal stromal tumors. Presented by Dr. Katie Janeway on behalf of the Consortium of Pediatric and Wildtype GIST Research (CPGR).

85% of GISTs in children are wildtype, lacking either KIT or PDGFRa mutations. A sub-group of these pediatric GIST patients are classified as having Carney-Stratakis Syndrome. Carney–Stratakis Syndrome infers an inherited predisposition to develop GIST tumors and paragangliomas (PGLs) caused by a germline mutation of the succinate dehydrogenase (SDH) gene subunits B, C or D. These germline mutations result in dysfunction of Complex II of the electron transport chain and defective cellular respiration.

In this study it was found that 16% of the pediatric wildtype patients without a family history of PGL had germline mutations of the SDH gene. Thirteen pediatric KIT- / PDGFRA- wildtype GISTs lacking somatic mutations or deletions in SDH were evaluated for SDHB expression by immunohistochemistry (IHC) and western blotting. Interestingly, despite the fact that these 13 did not harbor an SDH mutation, they nevertheless all demonstrated a complete loss of SDHB protein expression. In contrast to this, only 25% of KIT-mutant and 0% of NF1-mutant GIST demonstrated a complete loss of SDHB protein expression.

In addition, Complex II activity was measured in a subset of the pediatric KIT- / PDGFRA-wildtype GISTs and in KIT-mutant GISTs. Complex II activity was absent (comparable to levels seen in SDH mutant paragangliomas) in pediatric KIT- / PDGFRA- wildtype GISTs lacking somatic SDH mutations or deletions.

Conclusions: A germline mutation in SDHB, C or D can cause KIT- / PDGFRA- wildtype GIST without an associated family history of paraganglioma. In the absence of SDH mutations or deletions, pediatric KIT- / PDGFRA-wildtype GISTs have complete loss of SDHB protein expression and loss of complex II activity. These findings indicate that SDHB loss and defective cellular respiration may be central mediators of oncogenesis in GISTs lacking receptor tyrosine kinase mutations.

Imatinib (Gleevec)

Gleevec combined with other drugs is discussed above in the summary of Dr. Blay's presentation: Combination Therapy with Established and Novel Targeted Agents to Improve Outcomes in Patients with Advanced GIST. One of the more promising combinations with IM in mouse models is GDC-0941, a PI3K/PTEN/AKT inhibitor.

Abstract #10020 Activity of GDC-0941, an inhibitor of phosphoinositol 3 kinase (PI3K), in gastrointestinal stromal tumor (GIST) xenograft and duration of response after discontinuation of treatment in combination with imatinib. The efficacy of GDC-0941 (GDC) was tested alone and in combination with imatinib (IM) in IM-sensitive GIST (exon 11) xenograft (mouse model). Oncogenic signaling in GISTs is sustained mainly via PI3K/AKT rather than RAS/MAPK pathway. IM and GDC+IM reduced tumor volume by 86% and 95%, respectively, while GDC had no significant effect. IM and GDC treatments showed only grade 1 or grade 1&2 histologic response (HR), while grade 3 HR was observed in 75% of tumors under GDC+IM. Compared to control, mitotic activity decreased 2-fold under GDC, while it was not detectable in the other arms. Apoptotic activity was respectively 3- and 15-fold higher than control under IM and GDC+IM, being almost unaffected under GDC alone. After 24 hrs, SUV decreased in all treated groups, and remained low particularly under IM. The highest reduction in tumor/lesion glycolysis (TLG)was observed under GDC+IM.

In GDC and IM groups, inhibition of AKT was visible but incomplete. Only GDC+IM induced complete PI3K/AKT pathway knockdown. After treatment discontinuation, immediate tumor re-growth was observed in IM group, whereas GDC+IM treatment yielded long-lasting volumetric effect (98% and 15% of baseline, respectively, after 4 weeks of treatment end). SUV and TLG increased constantly under IM, whereas same parameters remained low in GDC+IM. Mitotic index in GDC+IM was still 2-fold lower than in controls. The combination of GDC-0941 and imatinib has extensive antitumor efficacy in GIST xenograft, inducing more substantial apoptosis and durable effects than imatinib alone. This effect was sustained even after treatment withdrawal.

Abstract #10035 - Combination of targeted therapy (IM) with immunotherapy (peginterferon a-2b) to overcome drug resistance and tumor stem cells in GIST Lymphocytes and interferon (IFN) collaborate to protect against development of carcinogen-induced sarcoma and spontaneous carcinoma. Impaired interferon signaling is a common immune defect in human cancer. Results state the combination showed promising efficacy in 8 GIST patients (three exon 11, one exon 9 and two wild type). A second partial response after progression (due to imatinib resistance) by re-initiation of PegIFNa2b is unprecedented.

Though this study was small, Dr. Chen was excited about the ability of stem cells to mount even a temporary immunity defense against resistant mutations to increase progression free survival. Immunotherapy may improve the benefit of molecular therapies. The trial was closed early in anticipation of a larger future study.

Abstract #10054 The development of a nomogram to predict overall survival (OS) in patients with locally advanced or metastatic GIST receiving first-line treatment with imatinib. Dr. Goldstein from Australia explained how his proposed nomogram was for metastatic GIST (as opposed to locally advanced GIST). It indicates high mitotic rate, high neutrophil count, and absence of an exon 11 mutation significantly and independently predicted reduced OS. He stated it would be ideal if overall survival could be calculated by incorporating specific mutations but more information will need to be shared and collected.

Nilotinib (Tasigna)

Abstract #TPS332 - Preliminary Data of Nilotinib in the First-line Treatment of Patients with Metastatic or Unresectable GIST Twenty one patients were enrolled with median follow up of 176 days. The study included 19 patients with metastatic GIST, 11 male and 8 female. Best overall response included 6 patients with PR, 6 with stable disease and 2 with progressive disease. Rate of PFS at 6 mos. was 85.7%. Early analysis of this pilot study indicated substantial clinical benefit and a favorable safety profile in the first-line treatment of patients with metastatic or unresectable GIST.

Abstract #10090 - - Evaluation of nilotinib in advanced GIST previously treated with imatinib mesylate and sunitinib. Patients with advanced GIST progressing following IM and SU often continue with good performance statuses and need additional therapy. This study tested the benefit of N, a second generation tyrosine kinase inhibitor that functions via ATP-competitive inhibition, developed to optimize binding to BCR-ABL for the therapy of chronic myelogenous leukemia. In vitro the agent can inhibit KIT and PDGFR phosphorylation, is active against IM-resistant cell lines and achieves higher intracellular concentrations compared to IM.

PR was experienced in patients with exon 11 (dup 567-576) and exon 17 (D820G); stable disease over 24 weeks in exon 11 (del 552-558) and exon 17 (N822K) had SD. SD over 24 weeks in exon 11 (V555) and exon 17 (Y823D). Progressive disease in exon 11 (K558) and exon 17 (N822K). N was well tolerated in these patients with advanced GIST. Accrual was halted due to insufficient clinical benefit. However, N may provide benefit to specific subsets of advanced GIST with exon 17 mutations.

Abstract #10016 - Nilotinib for patients with advanced GIST who failed imatinib and sunitinib: Negative effect of prior major gastrectomy on exposure to nilotinib. Nilotinib, a novel tyrosine kinase inhibitor (TKI), has been suggested to be active for advanced gastrointestinal stromal tumors (GISTs). N was active and safe in patients with advanced GIST resistant to both imatinib and sunitinib. Prior gastrectomy seemed to decrease the bioavailability of N significantly at least in some patients. Monitoring of plasma drug concentration may be particularly important for proper N therapy in GIST patients with major gastrectomy.

Abstract #10017 - Phase III Trial of Nilotinib in Patients with Advanced GIST: First Results from ENEST g3 (a phase III, randomized, open-label, multicenter trial was performed to test N vs best supportive care) with physician choice to continue or stop IM or SU in patients with advanced/metastatic GIST who had failed IM and S. N is a new selective oral TKI that inhibits the KIT, PDGFRa and BCR/ABL kinases and has shown anti-proliferative activity against both IM sensitive and resistant GIST cell lines. N uses different transport mechanisms which enable it to reach higher intracellular concentrations. Passive cellular uptake of N may limit the impact of efflux pump-driven drug resistance. In two early clinical trials, N was generally well tolerated and provided clinical benefit to patients with IM and SU resistant GIST.

The trial consisted of 248 GIST patients randomized 2:1 (N:control). The most common sites were small intestines (39.5%) and stomach (33.5%). AEs were anemia, asthenia, lipase increased,

The study did not meet its primary endpoint and there were no statistically significant differences in PFS or OS found between the N and control arm. However, based on local investigator radiologic assessments, significantly longer median PFS was observed in patients in the N arm than in the control arm (median PFS: N, 119 days; control 70 days). Although the study was not powered to identify statistically significant differences in median OS, a trend toward a 52 days improvement in favor of N was noted in spite of 64 (77%) patients crossing over from the control to N arm.

The ad-hoc exploratory subgroup analyzed true third-line patients who had progressed after only one prior progression of IM and SU, and were intolerant or had confirmed progression on prior SU therapy within 14 weeks of the end of SU treatment. These 51 patients experienced a significantly longer median OS in the N arm than in the control arm - median 405 vs 280 days.

Even though the study did not meet its primary endpoint, N appeared to provide clinical benefit in true third-line patients with an improvement of four months in median OS. Additional studies in more homogenous patient populations are needed.

Abstract #10015 - Phase II Trial of Nilotinib as Third-line Therapy for GIST Patients in Japan This study investigated efficacy and safety of N as a third line therapy for Japanese GIST pts who have shown resistance and/or intolerance to IM and SU.

Thirty five IM and SU resistant patients with the most common sites of sites of small intestines (60%) and stomach (31.4%). Median PFS was 113 days, and median OS was 310. Twenty three patients had stable disease as the best response, and one patient with KIT mutations in exon ll (dup 567-576) and exon 17 (D820G) experienced PR. N at 400 mg b.i.d. was well tolerated with most common treatment-related events as nausea, vomiting, hyperbilirubinemia and appetite loss (28.6% each) No patients developed grade 3-4 hematological toxicities except for anemia (5.7%).

PI3K/mTOR inhibitors

Paths for Clinical Development of PI3K Inhibition

Targeting the PI3K Pathway in Human Cancer presented by Neil Rosen, MSK

This pathway is mutated in the majority of human cancers. PI3K and mTOR are important components of an intracellular signaling network, the PI3K-Akt-mTOR pathway is an essential step towards the initiation and maintenance of tumors. This signaling cascade is also a key regulator of angiogenesis and upregulated metabolic activities in tumor cells. Because this pathway is constitutively activated in many tumor types, blocking the pathway could simultaneously inhibit the proliferation and growth of tumor cells and sensitize them toward programmed cell death (apoptosis), decrease angiogenic factors (VEGF).

Because of the major physiologic dependence of this pathway, you might expect major toxicity, but the grade 2 AEs from the various trials are nausea, dysgeusia, diarrhea, rash, fatigue, decreased appetite, increased alanine aminotransferase (ALT) and decreased carbon monoxide diffusing capacity (DLCO). Incidences of drug-related hyperglycemia were seen at the 32 mg dose level in GDC-0980 trial, but no significant effect on glucose levels was evident with increasing dose.

Scientific and commercial viability of PI3K/mTOR drugs is heightened; many are currently being developed:

1) Upstream inhibition of PI3K activation (HER-kinase, IGF kinase inhibitors);

2) Downstream inhibition (mTOR, rapamycin, mTOR kinase inhibitors);

3) PI3 kinase inhibitors (class I or isoform specific, promiscuos PI3K inhibitors, dual-specificity mTOR/PI3K inhibitors);

4) AKT inhibitors (perifosine, AKT kinase, PH domain, kinase inhibitors, selective and non-selective)

Tumors that indicate mutational activation of the PI3K pathway were sensitive to PI3K inhibition. The hormonal dependent cancers of breast and prostate seem to have very good sensitivity to PI3K inbhibitors because they have activation of the AKT and mTOR pathways, and not mutations to ERK, MEK, KRAS and BRAF, which are insensitive to PI3K inhibition.

For the tumors that do exhibit ERK, MEK, KRAS and BRAF activation, a PI3K inhibitor alone will not work. Inhibiting PI3K and mTOR together have a synergistic action that have proven anti-apoptopic and anti-angiogenic action in these activated pathways. This is the rationale for these combination trials.

Abstract #3079 - A first-in-human, phase I study to evaluate the dual PI3K/mTOR inhibitor GDC-0980 administered QD in patients with advanced solid tumors or non-Hodgkin's lymphoma.

GDC-0980 is a potent, highly selective, oral, dual inhibitor of class I PI3K and mTOR kinase which demonstrated broad preclinical activity in breast, ovarian, lung, and prostate cancer models. AEs were nausea (25%), fatigue (50%), diarrhea (42%), and flatulence (25%). No drug-related grade 3 or higher events have been reported to date. Potential signs of antitumor activity have been observed in a mesothelioma patient who had a decrease in target lesions of ~32% by RECIST after one cycle of 8 mg GDC-0980 QD and a soft tissue sarcoma patient who had a decrease in tumor mean SUVmax of ~31% assessed by FDG-PET after 2 weeks of 16 mg GDC- 0980 QD. Dose escalation continues.

Abstract #3003 A first in-human phase I study of BKM120, an oral pan-class I PI3K inhibitor, in patients with advanced solid tumors Presented by J. Baselga, Barcelona BKM120 is a potent and highly specific oral pan-class I PI3K inhibitor. Unlike other inhibitors, it does not inhibit the related kinases—mTOR and Vps34. BKM120 was evaluated from 12.5-150 mg. Daily oral dosing at 100 mg (MTD) showed clear signs of clinical activity.

Abstract #3004 A phase I dose-escalation study of XL147 (SAR245408), a PI3K inhibitor administered orally to patients with advanced malignancies Presented by G. Edelman XL147 is a selective inhibitor of class I PI3K isoforms. In preclinical cancer models, XL147 is cytostatic or cytoreductive as monotherapy and enhances the efficacy of targeted agents and chemotherapeutics. This open label, phase I dose escalation study assesses the safety, pharmacokinetics, pharmacodynamics and activity of XL147 in 68 patients with advanced solid tumors and lymphoma. XL147 reduced PI3K pathway signaling and phosphorylation of MEK and ERK in tumor tissues exhibiting diverse molecular alterations, with associated reductions in Ki67 and increases in apoptosis.

One lung cancer patient(225 mg dose, 21/7) had a confirmed PR with a 33% reduction in the target lesion. Thirteen patients continued on treatment ≥ 16 weeks, with nine patients on treatment ≥ 24 weeks. XL147 was generally well tolerated up to 600 mg 21/7 and CDD. The most common drug-related toxicities were skin rashes, nausea and diarrhea. Inhibition of PI3K and ERK pathway signaling has been demonstrated in solid tumors, and prolonged stable disease has been observed.

Abstract #3005 - First-in-human phase I study of the oral PI3K inhibitor BEZ235 in patients (pts) with advanced solid tumors. Presented by H. Burris. BEZ235, a dual PI3K/mTOR inhibitor was found to inhibit VEGF-induced cell proliferation and survival in vitro and VEGF-induced angiogenesis in vivo. Because RAD001, a specific mTOR inhibitor, was ineffective in preceding experiments, it was concluded that the effects observed for BEZ235 are in part driven by PI3K target modulation. 18 of 35 evaluable patients had detectable decreases of 18FDG uptake on PET scans. Available PD and efficacy data show that BEZ235 is active in patients (especially in those with PI3K pathway dysregulated tumors).

BEZ235 was effective in KRAS mutant breast cancer patient but not effective in a KRAS mutant colorectal patient. It was effective in an exon 9 patient. Metformin (insulin inhibitor) with PI3K was more effective in pre-clinical models. Future studies will use a new formulation of BEZ235 with improved bioavailability and PK properties.

BEZ235 was well tolerated with a favorable safety profile. AEs were fatigue, diarrhea, and increased plasma c-peptide tied to AE of hyperglycemia. Next step in PI3K clinical trials --- may have activity in tumors positive for PTEN and without PI3K mutations. PTEN and PI3K testing feasible in formalin fixed tissue sections.

Abstract #10048 - Multicenter, single-arm, two-stage phase II trial of everolimus (RAD001) with imatinib in imatinib-resistant patients (pts) with advanced GIST. RAD001 or Everolimus- an orally available mTOR inhibitor. Due to the absence of any partial responses and an increasing number of available and potential treatment options for imatinib-resistant GIST patients, the second stage of the trial was not initiated. Nevertheless, given its low toxicity profile and the fact that one-third of the patients had stable disease as best response, with one patient experiencing a response for over three years, the everolimus plus imatinib drug combination may still be considered in select patients who progress after exhausing all other approved treatment regimens.

Abstract #10038 - Multicenter, triple-arm, single-stage, phase II trial to determine the efficacy and safety of everolimus (RAD001) in patients with refractory bone or soft tissue sarcomas including GIST (after IM and SU) This trial showed promise with 6 patients showing response to the drug at week 16. RAD001 shows acceptable toxicity and displays limited clinical activity in heavily pretreated patients with bone and soft tissue sarcomas. The efficacy in imatinib- and sunitinib-refractory GIST is promising.

Abstract #2541 - A first-in-human phase I study to evaluate the pan-PI3K inhibitor GDC-0941 administered QD or BID in patients with advanced solid tumors. GDC-0941 was well tolerated up to doses of 330mg daily and its effects on insulin levels were evident starting at 245mg. Antitumor activity has been noted in several cancers other than GIST, including a partial response in an ER+PR-HER2 breast cancer patient, decreases in CA-125 in ovarian cancer patients and decreases in tumor "FDG avidity observed on PET scans of several tumor types. The GDC4225g Phase Ia trial is ongoing.


Abstract #3035 - -Phase I dose-escalation study of continuously administered regorafenib (BAY 73-4506), an inhibitor of oncogenic and angiogenic kinases, in patients with advanced solid tumors. Regorafenib, similar to sorafenib, is a novel diphenylurea oral multikinase inhibitor of angiogenic (VEGFR1-3, TIE2), stromal (PDGFR- β, FGFR), and oncogenic kinases (KIT, RET, RAF). It was well tolerated with most frequent AEs stated as rash, hand/foot syndrome and fatigue. Two out of 36 patients achieved a PR (6%; neuroendocrine pancreas carcinoma and squamous-cell carcinoma periorbital) and 22 patients (61%) had stable disease ≥6 weeks, including one hepatocellular carcinoma patient for 90 weeks.

Abstract #7585 - Phase I study of regorafenib (BAY 73-4506)administered continuously in patients with advanced refractory non-small cell lung cancer Drug was well tolerated at 100 mg with skin toxicity being the most common reason for drug reduction or interruption. Of 17 patients, 13 had stable disease at least 6 weeks (76%), and 4 had SD for 12 weeks.


Abstract #2529 - A phase I dose-escalation study of the Hsp90 inhibitor STA-9090 administered once weekly in patients with solid tumors.

STA-9090 is a potent, next-generation Hsp90 inhibitor that is structurally unrelated to the first-generation ansamycin class of Hsp90 inhibitors, such as 17-AAG and IPI-504, and has shown superior activity and an improved safety profile relative to these agents in preclinical studies. 35 patients received STA-9090 in one hour indfusions once a week for a 28 day period. AEs reported in ≥ 20% of patients were diarrhea, anemia, fatigue, abdominal pain, nausea, elevated alk phos, constipation and dyspnea; most of these AEs were mild to moderate in severity. Preliminary signs of clinical activity have been observed including a durable PR in a patient with rectal carcinoma, as well as several cases of tumor shrinkage and prolonged SD in patients with tumors including NSCLC, renal cell carcinoma and GIST, who were treated for 16 to 48 weeks.

Abstract #3083 - A phase I dose-escalation study of the Hsp90 inhibitor STA-9090 administered twice weekly in patients with solid tumors.

29 patients received one-hour infusions twice weekly for three weeks of a 28-day cycle. AEs reported in >20% of patients were fatigue, nausea, anemia, diarrhea, vomiting, anorexia, and headache; the majority of these AEs were mild to moderate in severity. Dose escalation continues with a total of 32 patients enrolled at doses up to 50 mg/m2. Preliminary signs of clinical activity have been observed and include a durable PR in a patient with melanoma and tumor shrinkage with prolonged SD in several patients with different tumor types, including melanoma and NSCLC.


Having coffee with Dr. Annette Duensing was enlightening! She is humbly brilliant, and spoke openly about her research testing a variety of GIST cell lines against specific miRNA to determine why imatinib works the way it does to stop tumor growth. One roadblock is getting fresh (non-frozen, non-preserved) GIST tumor tissue to grow a variety of cell lines - there is quite a limited supply. Maybe GSI can post a webpage for those willing to donate fresh tumor tissue to her lab. Note these arrangements need to be made BEFORE surgery! GSI offers information on donating tissue on wild-type tumor samples to Dr. Kim to further his work with pediatric, Carney triad and wildtype GIST:

Abstract #10007 - In vitro activity of novel KIT/PDGFRA switch pocket kinase inhibitors against mutations associated with drug-resistant GI stromal tumors. The company Deciphera is developing new KIT inhibitors that use a different approach. All current KIT inhibitors depend on binding at the ATP binding site, where they inactivate the KIT by competing with AKT to bind at the location encoded by exon 13-14. In contrast, the drugs in peclinical development as "switch pocket" inhibitors bind at the location encoded by exons 17-18, the activation loop of the KIT receptor.

None of the current drugs are effective against mutations of the activation loop (whether rare primary mutations or common secondary mutations). The activation loop is like a gate that swings, and mutations there keep the gate open, allowing cell growth signaling. The switch pocket inhibitors lock the gate closed. in the inactive position. So far preclinical testing on various GIST cell lines with different mutations has shown 3 different Deciphera drugs are promising.

The one with the most promise at this stage is currently called DP2976, which succeeded in inhibiting all mutations it was tried against, and which did not lose effectiveness over time in cell line experiments. In addition, these Deciphera drugs are highly selective, without off-target actions that would produce side effects. Hopefully a Phase I clinical trial can be stated before the end of 2010.

Abstract #10019 - Correlation of the topography of KIT exon 11 mutation with primary GIST location and predictive value for PFS in patients with advanced GIST: Results from the BFR14 randomized phase III trial of the French Sarcoma Group. All exon 11 mutations are not equal in GIST tumors: There is a substantial level of heterogeneity within them. The impact of this heterogeneity on clinical presentation and outcome of advanced GIST treated with imatinib was investigated in the BFR14 series. The study concluded that the site of mutation on KIT exon 11 varies according to the primary site of GIST and is an independent prognostic factor for PFS.

Of the total 434 GIST patients, 107 patients with KIT exon 11 were put into four groups (G): G1 whose first mutated codon was <555 (n=16, 15%) - more frequent in gastric GIST; G2 whose first mutated codon was from 555 to 559, excluding those in Group 3 (n=40, 37%) with 559 observed almost only on non-gastric sites; G3 composed of the deletion of C557 and C558 only, i.e.WK deletion, (n=22, 20%); G4 whose first mutated codon was 560 and beyond (n=23, 21%) - more frequent in bowel sites.

The most frequent point mutations were on 557, 559, and 560. PFS differed significantly in the 4 groups: (G1)median PFS not reached; (G2)22 mos; (G3)55 mos; (G4)58 mos. OS was not significantly different in the 4 groups. PFS in advanced GIST with mutations starting between 555 and 559 is poorest when treated with IM. Questions are: Should the approach to these mutations be modified? Are IM levels relevant for this group (like exon 9 which requires higher drug level to bind? Is there an alternate therapeutic approach that is superior?

Neither the length of the mutation (from 1 to 20 codons), type of mutation, nor individual point mutation correlated to PFS or OS. These results deserve confirmation in other prospective series in advanced GIST and may be worth exploring in adjuvant series.

Abstract 10084 - Gastrointestinal stromal tumors (GISTs) and other cancers: Cambridge GIST Study Group experience. Dr. V. R. Bulusu of Addenbrooke's Hospital in Cambridge U.K. sees all local GIST patients and participated in this study of 150 GIST patients between 2005-2009. Fifteen percent of all GIST patients had another type of cancer during or before GIST diagnosis. 62% of GISTs were found incidentally in patients (and were considered low or very low risk), and 28% had another cancer diagnosis before GIST. All were spindle cell.

Sites of other cancers: one new metastatic pancreatic cancer with liver metastases diagnosed during imatinib treatment for inoperable GIST; esophagus 4.5%; stomach 14%; colorectal 19%; prostate 14%. Rare associations include one patient each with NF1 syndrome-somatostatinoma with duodenal GIST, chronic lymphocytic leukemia and gastric GIST, male ER+ve breast cancer and gastric GIST, abdominal soft tissue sarcoma with gastric GIST, paraganglioma with MIBG avid Gastric GIST. Four patients developed other cancers (breast, lung, ovarian and pancreatic) during the follow-up of a pre-existing GIST.

Ongoing Phase II Masitinib (AB1010) Trial - Dr. Jean-Yves Blay from the French Sarcoma Group did not have new information to divulge about his this trial other than there is a delay due to stalled resources of the lab and it will be two to three years before the accruing results are made public. When asked specifically about whether masitinib has been tested against certain difficult GIST mutations, like D842V, he replied that he personally had one patient who was stable for 2 years with that mutation on masitinib. ;

Abstract #TPS277 - The OPAL study: A phase II study to evaluate the efficacy, safety, and tolerability of tosedostat (CHR-2797) in elderly subjects with treatment refractory or relapsed acute myeloid leukemia Tosedostat is a novel oral aminopeptidase inhibitor. This means that it is expected to block the activity of aminopeptidase, an enzyme that is involved in the breakdown of proteins in the body. By blocking this enzyme, tosedostat is expected to decrease the level of ‘amino acids’, the building blocks of proteins, in the body’s cells. With fewer amino acids available for the production of new proteins, cells are expected to find it harder to grow and multiply. The Oncology Medical Director for Chroma Therapeutics, UK, Dr. Martin Toal, will be forwarding GSI the Phase I results which included solid tumors (he mentioned renal cell carcinoma).

Abstract #6511 - Phase I trial of AP24534 in patients with refractory chronic myeloid leukemia and hematologic malignancies AP24534 is an oral multiple tyrosine kinase inhibitor and potent pan-BCR-ABL inhibitor, including T315I mutation. It was well tolerated until 60 mg. and there was evidence of antileukemia activity in patients resistant to second generation TKIs. Speculation exists that it may be effective against the imatinib-resistant mutation T601I mutation.

Abstract #e13585 - Anti-VEGF drugs, endothelial dysfunction, and hypertension: Is there a relationship? Hypertension is one of the most frequently reported adverse event with the use of anti-vascular endothelial growth factor drugs (VEGF inhibitors like sunitinib or sorafenib) during cancer treatment. The underlying mechanisms are unclear but the results of this German study indicate that anti-VEGF drugs could cause endothelial dysfunction by decreasing L-Arginine bioavailability for nitric oxide synthesis and increasing sympathetic tone thereby resulting in hypertension. They do not alter the levels of endogenous nitric oxide inhibitors in short-term. There were six patients in this study and this information needs to be confirmed in an ongoing study with larger patient population.