UCSF specialists in cancer genomics and pediatric oncology describe the process of evaluating and determining treatment options for referred patients, with specifics on how care for certain childhood cancers – such as neuroblastoma, leukemia, brain tumors and hepatoblastoma – can be correlated to individual genetic testing results. The speakers detail the scope of the UCSF 500 Cancer Gene Panel and its real-life benefits – leading to targeted treatments and personalized monitoring – and depict how the UCSF Molecular Tumor Board experts collaboratively arrive at recommendations.
Welcome to our presentation this morning on the pediatric oncology Precision Medicine Program at UCSF Benioff Children's Hospitals. My name is Alejandro Suit Cordero. I'm a professor of pediatrics and the chief of pediatric oncology at UCSF. I'm joined today by my colleagues, Doctor Arun Rangiswami, uh, also professor of pediatrics in the division of Pediatric oncology at UCSF, Doctor Jessica Van Ziffel, associate clinical. Professor of pathology, also a molecular geneticist within the Clinical Cancer Genomics Laboratory in the Department of Pathology at UCSF, um, and, uh, Professor, uh, uh, Doctor James Grenert, uh, who's a clinical professor of pathology, also a molecular pathologist in the department of pathology at UCSF. So our precision oncology program at UCSF really spans the wheel of discovery from clinical care to discovery as shown on this slide. Uh, we begin with patients who come to us uh for evaluation for an oncologic problem. Uh, those patients receive a sequencing essay using the UCSF 500 as we will describe. Presentation of this essay is discussed at a molecular tumor board where a group of clinicians and geneticists and bioinformaticians and um other specialists will evaluate these results and provide a recommendation for the patient. That patient can then be followed uh clinically. We can evaluate whether there was a response based on the analysis and the recommendation that was made. That information is then put into a database that can be used to identify in the future other patients who have had similar alterations and who have either responded or not responded to the recommended therapy and therefore we can learn from prior patients in order to treat our future patients in this wheel of discovery and clinical care. So our process begins with a referral from uh either a physician here at UCSF or an outside provider who uh has a patient with a cancer diagnosis who gets a biopsy um and needs to have, have that evaluated with our UCSF 500 assay. Um, we have a group of patient navigators who can help the clinicians as well as the patients enter our system and be um evaluated in our molecular tumor board as shown here. The molecular tumor Board will then define the relevance of the findings found on the UCSF 500 for the care of the patient. That can either be a diagnostic recommendation based on the molecular findings. Sometimes that will lead to a change in diagnosis. It can also be recommendation in terms of targeted therapy, or sometimes we also identify cancer predispositions that would. Entail that a patient should be referred to the cancer genetics clinic. That patient can then either enter into a clinical trial here at UCSF or go back to their home institution where the recommendations made by the molecular tumor Board can be used to guide therapy in the future. So, uh, we have a very active early phase clinical program here at UCSF. We are one of 21, uh, children's oncology group, uh, early phase, uh, member institutions. We participate in this, the network shown here on the right on this map of other institutions that participate in these children's oncology group trials. We are also one of only 4 Alex's Lemonade Stand Foundation Centers of Excellence in Pediatric Developmental therapeutics. So these two affiliations provide us with an opportunity to bring to our patients the most cutting edge clinical therapy, uh, clinical trials available, particularly those clinical trials that are informed by the molecular findings, uh, that we can determine using the UCSR 500. We are also part of a large number of consortia including the COG PEP CTN as I mentioned, also a consortia that are dedicated to treating specific uh types of cancers, for example, the NAN Consortium which focuses on neuroblastoma, the PNA Consortium which was initiated here out of UCSF which um focuses on brain tumors, the TCO Consortium. For leukemia lymphoma, uh, the NCO Consortium for Histocytic Disorders, as well as a large number of industry sponsored trials, uh, and perhaps most importantly, investigator initiated trials. So these are trials that are, um, not available anywhere else because they were based on discoveries from our, um, uh, physicians, scientists, um, and, uh, other investigators here at UCSF. Um, our phase one program is led by Dr. Qavo, uh, on the extracranial solid tumor side. Uh, the immunotherapy, um, side of our phase one program is led by Michelle Hermanston. Our leukemia trials are led by Elliott Stiglitz and the brain tumors by Sabine Mueller. We also have a large number of clinical trials that are really focused on uh what we call biomarker driven trials. So these are trials where you need to have a molecular alteration identified in an assay such as the UCSF 500. This includes the pediatric match trial which is a national. Effort to match uh pediatric cancer patients to targeted therapies. We also have a number of biomarker selective trials including uh NTA fusion trial, an IDH12 trial, um, and several other molecularly driven um uh clinical trials. And these include a wide range of investigational drug classes including NTRAC inhibitors, PARP inhibitors, RAP makinase inhibitors, ATR inhibitors, and kit PDHFR and this list changes all the time since we have a very active. Um, phase one program where we are always trying to bring to UCSF the most cutting edge, um, trials that will drive, uh, new therapies for our patients. So now I'm gonna talk a little bit about the UCSF 500 NGS panel. I'm going to pass this on to my colleague, Doctor Jessica Van Ziffel, who will explain the, how the, the UCSF 500 was designed, um, and how we utilize it to identify molecular alterations um in, uh, in our pediatric cancer patients. Doctor Van Ziffel. Yeah, thank you. So, our UCSF 500 cancer gene panel is a custom pan-cancer panel. It does include 529 cancer genes and 72 intronic regions of these genes. Um, and the design of this panel was, um, initiated and developed through interaction with UCSF providers, including our pediatric oncologists, um, who provided input into the gene list and continue to provide input as we evolve the panel. So this panel is able to detect the common um or identified variant types seen in cancer. This includes single nucleotide variants or SNVs, which are the most common mutation type that's seen, that affects coating regions or splice site changes. Often they're known hotspot alterations. It also detects indels, which are insertion or deletion variants. Often these lead to frame shifts, sometimes they are in frame and can be hotspot regions also in known genes or known pathways. The assay also has um the ability to detect copy number variants, including gene specific amplifications or deletions, um, as well as chromosomal copy number changes such as whole chromosome or arm level gains or losses. In addition, the inclusion of the intronic regions from those 72 genes allows us to identify some structural variants or gene fusion events. These can result from translocations or tandem duplications, inversions or other indel alterations within the genome. We do offer this test as well as a tumor normal analysis as well as tumor only, and when we perform tumor normal analysis, this allows us to distinguish between somatic alterations which are acquired in the tumor. Versus germline alterations which are um present in the constitutional genome. In addition, our essay identifies tumor mutational burden, um, as well as microsatellite instability, and tumor mutational burden that's high, as well as uh microsatellite instability that's high is associated with immunotherapy and so can have therapeutic um implications. The UCSF 500 assay is a capture-based next-generation sequencing assay. And if we are doing a paired tumor normal analysis versus a tumor-only analysis, the first step is to identify appropriate tumor for sequencing, um, which we recommend is at least 25% of neoplastic cell content. We then isolate high-quality DNA from these samples and separately prepare libraries, which are then captured or um enriched for the regions of interest in our sequencing. And that's those 529 cancer genes and the select introns that I just described. The libraries are then sequenced on aluminous sequencers and um our minimum quality threshold is 500X coverage to make sure we have sufficient sequencing um to allow for high quality variant calling. We do run an in-house bioinformatics pipeline. Um, to allow for this variant calling, and we do report those variants that I just described on the previous slide with a median turnaround time, um, from sample receipt to release of the results of 14 days. Over the past several years, we have seen a large increase in our case volume and the number of orders that we receive, um, and we have now over 10,000 cases reported. Um, we do see this significant growth year over year and we do see really a large breadth of cancer types referred for testing. At UCSF we do have a strong neuro-oncology and neuropathology program that have been very um early and strong adopters. And so, we do see an emphasis on neuro cases submitted, but you can also appreciate a broad diversity of, of cancer types that are referred for our testing. Within pediatric cancers, um, because of the strong neuro-oncology component as well, we do see a high percentage of CNS cases referred. But you can also appreciate on the left, um, that we do see a diversity of different cancer types in our pediatric population. And many of these, um, cases are referred for tumor normal testing in the pediatric population. The reason that that's important is that germline findings of significance are identified in almost 20% of pediatric cases sent for sequencing. And so this is a significant consideration and of importance in these pediatric cancer cases. In addition, over 30% of our cases have a clinically actionable finding, and that includes um FDA approved therapies or other um potentially actionable findings identified by the sequencing assay. Our increased volume is seen both locally at UCSF but also has been seen across the state of California and beyond that, we have seen um increased adoption of our assay for referrals from many other states across the United States. Beyond the United States, we also have seen increased referrals internationally, uh, many from our closest neighbors of Canada and Mexico, um, but also you can appreciate, um, that we do have international orders from across the globe. And with that, I'd like to introduce my colleague Doctor Rango Swami who will present pediatric cancer genetics at UCSF. Thank you, Doctor Van Siffel. As Doctor Ziffel highlighted, one of the unique aspects of the UCSF 100 is its ability to identify germline variants in children. An estimated 15% of children with cancer have a familial or genetic predisposition to develop that cancer. Identifying these predisposition syndromes is critical, not only for the treatment of the current diagnosis, but to prevent other cancers later in life. There are treatment implications which include avoiding toxicity, as in the case of Lee Frain syndrome where we try to avoid ionizing radiation exposure. There are also implications for surveillance um of patients uh in order to identify uh new cancers at an early stage. Their implications for lifestyle choices and prevention, um, as well as potential surgical and medical interventions, and finally, uh, implications for family planning. The uh Kaplan-Meyer curve on the right, which was published by the Sick Children's Cancer Genetics Group, highlights the importance of surveillance of patients with Lee Framini syndrome. As you can see on the upper graph. Uh, the outcomes for patients who have surveillance instituted are significantly better with respect to overall survival than those who do not undergo a surveillance protocol. So at um UCSF we have a pediatric hereditary cancer clinic, which identifies and follows children who are born with a genetic mutation that places them at an increased risk for developing cancer. We see more than 40 different predisposition syndromes, um, and each of these has its own specific set of surveillance guidelines, um, that our clinic helps to coordinate. Our genetic counselors work very closely with families to help them understand the implications of being told that their child is at risk for developing cancer and also in addressing the significant psychosocial and emotional impact of a family who has a child with a predisposition syndrome. And I would like to um turn uh the presentation over to my colleague, Doctor Sweet Cordero, who'll discuss the molecular tumor board. The molecular tumor board at UCSF is a multidisciplinary tumor board that consists of a number of experts and discussants shown on the right of this slide. Um, we have, uh, experts in, um, targeted therapy such as Doctor Beth Winger as well as, um, uh, experts in adult oncology and the precision oncology, Doctors Mark Moresser, Doctors Eric Collison, uh, and, um, uh, as well as several other experts here who have domain expertise in various, um, areas that are important for discussion in the molecular tumor ward depending on the patient and the diagnosis. Our tumor board meets every other week, either on the 2nd Monday or the 1st Thursday of the month. Uh, a requesting oncologist will provide a clinical history, um, that will, and also a reason for why they, they would like this case to be discussed at the molecular tumor Board. That case will then be presented by the domain specific expert, um, and then we will have a very lively discussion that is intended to provide the clinician with, uh, information relevant to the care of their patient. Uh, for example, the molecular findings in terms of, uh, whether it's likely to be a sub-call variant or whether it's likely to be a call variant, uh, whether this is an alteration for which there is a known, uh, drug that targets that alteration, as well as uh uh a number of other findings that that may be relevant to the patient, um, as you will see um in the uh the next slides. Uh, within a week, we then deliver a recommendation to the clinician. Um, and we are always uh available for any uh follow-up questions or discussion. So I'm going to now pass it back to, uh, Doctor Rangeswami, uh, Doctor Granner, um, who will, uh, discuss a specific patient who was referred to us, uh, with a diagnosis of hepatoglastoma with treatment effect, uh, of mixed type. Doctor Ranggaswami, thank you, Doctor Sweet Cordero. Uh, the case that I'm going to discuss today is a 36 month old child with recurrent hepatoblastoma who underwent resection at an outside institution and was receiving salvage chemotherapy. He was diagnosed in August of 2021 with a locally advanced uh pretext 3 annotation factor negative hepatoblastoma after presenting with two days of abdominal pain, fever, and irritability. His initial alpha fetoprotein, which is a sensitive biomarker for hepatoblastoma, was elevated at 110,476, and a liver biopsy was performed which demonstrated hepatoblastoma with mixed epithelial and mesenchymal histology. He was enrolled on the children's oncology group Frontline Trial AHEP 1531 and received two cycles of cisplatinum monotherapy with minimal response by both resist and alpha fetoprotein criteria. He then received one cycle of doxorubicin transarterial chemoembolization uh with a tumor and alpha fetoprotein measured progression. He was deemed unresectable at his local institution and was referred to a second outside institution for consideration for orthotopic liver transplantation. There he underwent a resection in November of 2021. However, unfortunately, his alpha fetoprotein following the resection was actually increased. And he returned to his home institution for continued treatment um as per the AHEP 1531 trial with continued cisplatinum monotherapy. However, due to continued elevation of his alpha-fetoprotein, we recommended reimaging, which demonstrated recurrence at his resection margin and his and in his right hepatic lobe. So molecular profiling was recommended at the outside institution. Their platform demonstrated no pathogenic variants. However, we recommended sending the UCSF 500 on the resection sample for analysis of prognostic markers and potential therapeutic targets. So we performed UCSF 500 on this more recent specimen, and the first alteration I want to point out is an exon 3 deletion in the CTN and B1 gene. So this gene codes for beta-catenin, which is a very important driver in both pediatric. And adult liver tumors and exon 3 is an important regulatory domain of that protein and deletion of exon 3 would be expected to increase beta-catenin activation and subsequent WIN pathway signaling. I also want to point out another gene found to have an alteration here is ARID-1A. This was a nonsense mutation which results in a stop codon and truncation of the protein. ARID-1A is involved in chromatin regulation and through that, uh, gene transcription and is an important tumor suppressor across many cancer types. And finally, there's a frame shift mutation in DDX 3X. So frame shift mutations uh end up resulting in premature truncation and inactivation of proteins. DDX 3X is an RNA helicase that's involved in many different um functions of RNA including its uh transcription, translation, stability, and so it's a very important uh regulator of RNA in the cell. Our informatics pipelines did not show any evidence of microsatellite instability, and they found that the tumor mutation burden was low. Thank you, Doctor Brenner. So, uh, as, uh, Doctor Brenner mentioned, uh, betacatenin, uh, or the CTNNB1 gene codes for the protein betacatenin, which is a key signaling component of the wind signaling pathway, um, which is illustrated, uh, here on the right side. So when betacatenin is, um, is activated. Through the signaling axis shown here that leads to increase in transcription of a number of genes which normally during development are involved in patterning and various aspects of development, but it's well known that this pathway can be hijacked in several cancers including lung cancer. So the exon 3 of this gene encodes the serring 3 phosphorylation site for the protein GSK 3 beta which is involved in degradation of beta-catenin. So loss of this axon leads to lack of degradation and therefore increased the nuclear activity. Um There are approximately 50% of the pyoblastoma cases that have been reported in the GE database have beta-catenin mutations. So the GE beta database is a data sharing effort with which UCSF participates that includes the Dana-Farber. Cancer Institute, uh, the Memorial Sloan Kettering, as well as several international partners and depositing our data in this database allows us to have access to a much larger group of patients that can then be, uh, evaluated and compared to the cases that we see here at UCSF. Uh, there have been several efforts to target the one pathway given its importance in, um, in, uh, cancer therapy. So there, the drug PRI 724 has been tested in phase one dose escalation study, um, in combination with this gemcitabine, um, and there's been some evidence of stable disease, uh, in patients. There are several other, uh, clinical trials, uh, such. The one shown here in adults, uh, which, uh, aim to target the, uh, the beta-catenin pathway, uh, at various levels. And one of the important aspects of the molecular tumor board is because we have a very deep expertise in these molecular pathways, it, it helps us to, for example, determine, uh, based on where, where the, um, the alteration is in the signaling pathway, whether it Drug is likely to work. So obviously if the alteration as in this case is in betacatenin, uh, trying to target the pathway higher up, uh, would not be, uh, likely to be effective in this case. So it's very important to have this kind of expertise to be able to, uh, truly, um, provide the clinicians, uh, with the, uh, the best, um, uh, uh, evidence, uh, so, um. Uh, in, uh, in, uh, pediatric patients there's also another, uh, clinical trial with a, uh, wind, uh, inhibitor called Tevivi, uh, uh, which, uh, uh, uh, would also be, uh, potentially, uh, active in this patient, um, so that brings up another aspect. Of the importance of the molecular tumor board is that we can also determine whether there's likely to be a drug that will be available for pediatric patients. So not all, not all adult trials are for drugs that have been made available for pediatric patients. So given the importance of the wind pathway in a variety of cancers, it's perhaps not surprising that this has been a major focus of the efforts to target cancer. There's been a number of companies who have um Evaluated various ways to target the wind pathway is shown here on the right, starting with the receptor on the cell surface down to the mechanisms of regulation of betacatenin and the cytoplasm as well as the activated betacatenin in the nucleus. So for example, PRI 724 has been tested in a phase 1B dose escalation study in combination with gemcitabine. There's been some um uh signal for stable disease. SMO 8502 is currently being evaluated in a phase one trial. Um, there's also a drug called Tegavivi which also targets the wind pathway and has been, um, and is currently being evaluated for use in pediatric patients. One important aspect of the molecular tumor Board is because of the deep expertise of the people who participate, we are able to give. Our clinicians' advice regarding the most appropriate targeted therapy based on the alteration in the individual patient. So for example, in this patient who has a mutation of betacotenin, it would not be appropriate to select a drug that targets the wind pathway higher up, for example, at the level of the cell surface, because that would be unlikely to work when the alteration causes. Chronic activation of betacotenin, which is already in the nucleus. So in this case you would want a mechanism of action of a drug that would lead to a blockade of betacotenin at the site of transcription. Another alteration that was found in this patient was in the gene ERD 1A. ERD 1A codes for a protein, which is an AT-rich interaction domain protein, part of the SWI sniff chromotome modeling complex, and a well-described tumor suppressor protein. Uh, ERD 1A, um, because it's part of the SWI sniff complex, has multiple effects as shown here on the left. Um, but most importantly, it regulates the chromatin to lead to changes in gene expression of a variety of genes. HERD1A like betagotenin is much less frequently mutated in hepatoblastoma. In our review of the Genie database, we found that it was mutated in 6% of 49 cases of hepatoblastoma. So unfortunately the loss of uh ERD 1A, uh, like many other tumor suppressors, is not directly targetable. However, uh, it is increasingly recognized that loss of tumor suppressors can in some cases make tumors um uh susceptible to uh what is known as a synthetic vulnerability. So loss of a gene can make uh genes susceptible to treatment with another pathway. Uh, that can sometimes, uh, uh, cooperate with the, the pathway that has been lost in the patient. Uh, so for example, it's known that ER1A loss in paracellular carcinoma leads to upregulation of a, a protein called ING2 which uh may in some pre-clinical studies uh suggest uh be susceptible to treatment with uh a receptor terrain kinase inhibitor called sorafenib. Uh, similarly, uh, loss in ovarian clear cell carcinomas is known to lead to sensitivity to inhibition with, uh, uh, EZH2 inhibitor, uh, which is an epigenetic therapy that is currently in pre-clinical studies, uh, and as well, uh, also in clinical trials as shown here. Um, in some experiments, uh, and some data also suggests that in ovarian cancer, at least, ER1A loss is associated with the so-called mutator phenotype, which leads to a higher tumor mutational burden, uh, and potentially sensitivity to checkpoint blockade. This points out the importance of not only looking at the genetic mutations but also the um. The specific uh context as in this patient uh with hepatoblastoma, uh we did not see an elevated um tumor mutational burden suggesting that perhaps this finding is specific to ovarian cancer. And lastly, this patient also demonstrated a very interesting alteration in another gene called DDX 3X, which is an ATP dependent RNA healer case. Uh, the, the, uh, process involved is shown here on the right. So RNA helocases are important in regulating the, um, the structure, uh, of an RNA as suggested by the, by the name. Uh, these proteins have, uh, have been found to be overexpressed in a number of cancers including glioma, medulloblastoma, hepatocarcinoma, and others. Um, and lots of DDX 3X has been suggested to lead to a tumor intrinsic type 1 interferon response, um, and an increased infiltration with T and dendritic cells. Um, and, uh, that's, uh, in the, uh, reference shown here. So interestingly, DDX 3X alterations are quite uncommon in hepatoblastoma. So this slide shows the prevalence of DDX 3X alterations in the GEC portal. We actually could not find evidence of any other patient that had ever been described with a DDX 3X alteration. DDX 3X alter mutations are commonly seen in a variety of other cancers. It's shown here on the right, but not previously demonstrated in hepatoblastoma. So on careful review of the literature, our discussants noticed that TDX 3X. Has been recently associated with cisplatin resistance in oral squamous cell carcinoma. So this is a figure from a paper published recently where the author has used a variety of molecular techniques to demonstrate that. Loss of DDX reacts leads to resistance to cisplatin. This graph demonstrates our patient's alpha-fetoprotein response. Um, the initial data point is at his initial diagnosis. Um, following that, we have, uh, his alpha-fetoprotein following 4 cycles of cisplatin and TACE chemotherapy with his MRI showing progression. And then following his resection, the rise in his alpha-fetoprotein until uh upon, based upon the UCSF 500 results, we recommended a change in chemotherapy to the agents when Christine and. Following the switch in his chemotherapy regimen, he had a very rapid decline in his alpha-fetoprotein, as well as his, in his imaging study. Um, and, uh, following two cycles of VI. Um, he underwent an orthotopic liver transplantation. And these are representative um slides showing an MRI scan that was performed uh prior and after institution of VI chemotherapy. One can see in the image on the left that he has a large tumor with multiple nodules, and on the image on the right, these nodules have shrunk significantly. And that was immediately prior to his more definitive surgical procedure. Our patient received an orthotopic liver transplant in February of this year. He had a second UCSF 500 analysis performed on the explanted liver that demonstrated absence of the DDX3X and ARID-1A variants with persistence of the CTNMB1 variant. This suggests the effectiveness of switching from his cisplatinum-based regimen. To enchrystine and irinotecan with subsequent eradication of the two clones above, he then completed a total of 6 cycles of VI chemotherapy in July of 2022 and remains in a continuous complete remission 5 months from the end of treatment with normal alpha fetoprotein levels and negative imaging studies. Of course, beyond the science is um our desire to impact patient care, and this is a quotation from the family with respect to how the UCSF 500 helped them and he, and I think the most important takeaway from here is the line, today he is cancer free. We are incredibly grateful. She would like to refer a patient to UCSF for UCSF 100 testing. The ordering information is shown here on the left. If you would like to refer a patient to UCSF, please call 877-UCHILD. There's also additional information about our pediatric oncology precision medicine program on our website. You can also contact us by email at Pscancermedicine@UCSF.edu. Thank you.
