Targeted Oncology Virtual Tumor Board&reg;

, which is focused on ovarian cancer. In today&rsquo;s

presentation, my colleagues and I will review 4 clinical cases. We will discuss an individualized approach to treatment for each patient and will review key trial data that impact our decisions.

I am Dr Robert Coleman, professor in the Department of Gynecologic Oncology and Reproductive Medicine, and executive director of MD Anderson Cancer Network Research, and the Ann Rife Cox chair in gynecology at The University of Texas MD Anderson Cancer Center in Houston, Texas.

Today I&rsquo;m joined by Dr Susana M. Campos, assistant professor of medicine in the Division of Medical Oncology at Dana-Farber Cancer Institute at Harvard Medical School in Boston, Massachusetts;

Dr David O&rsquo;Malley, medical director of gynecologic oncology, director of clinical research of gynecologic oncology, co-director in the Gynecologic Oncology Phase I Program, and professor at The Ohio State University and the James Cancer Center in Columbus, Ohio;

Dr Gregory Riedlinger, assistant professor of pathology in the Division of Translational Pathology at Rutgers Cancer Institute of New Jersey and Rutgers Robert Wood Johnson Medical School in New Brunswick, New Jersey;

and Dr Shannon N. Westin, associate professor and director of early drug development in the Department of Gynecologic Oncology and Reproductive Medicine at The University of Texas MD Anderson Cancer Center in Houston, Texas.

Thank you for joining us. Let&rsquo;s get started with our first case.

Our first case will be discussed by Dr Susana Campos. Why don&rsquo;t you take it away, Susana?

Thanks. I&rsquo;d like to share with you guys a case that&rsquo;s quite interesting. It&rsquo;s a 44-year-old premenopausal woman who presented to her gynecologist complaining of weakness and vague abdominal discomfort for a duration of about 4 months. Her past medical history is mainly unremarkable. She has 2 children and has always had a normal menstrual cycle. In terms of her family history, both of her parents are living. There is a history of lung cancer on her father&rsquo;s side. On a physical examination, what was dominant was really right-sided abdominal tenderness and swelling. Imaging studies were done, which showed a right-sided solid cystic adnexal mass, measuring about 8.8 centimeters in size. There was periaortic lymphadenopathy, omental caking, a solitary liver lesion, and a splenic mass. A percutaneous biopsy did show high-grade ovarian adenocarcinoma.

OK, so it was a good case. It is one we see frequently. First, talk a little about the pathology. Give us your thoughts on what you are thinking about this pathology.

Particularly as a pathologist, we&rsquo;re just looking at tissue initially on an H&E [hematoxylin and eosin stain] slide. There are certain morphologic features for which we&rsquo;re thinking about, depending on the specific diseases. In this case, they&rsquo;re saying high-grade ovarian cancer. So we&rsquo;re usually looking for more of a serous morphology. And then it&rsquo;s an adenocarcinoma, so it&rsquo;s a gland-forming tumor. The pathologist is not going any further than that without immunohistochemistry. So you&rsquo;re going to be flying a bit blind.

So typically, a pathologist will use a panel markera PAX8 transcription factor&mdash;to kind of confirm that it&rsquo;s of Mullerian origin. And then, with the high-grade serous, they tend to uniformly actually have a p53 mutation. So on immunohistochemistry, there are actually 2 phenotypes. There are the high expressors, where all of the tumor cells would have p53; or the null phenotype, where none of the cells have p53. And actually, when we do further molecular testing, this correlates with the actual underlying p53 mutation we see. With the

missense mutations, these are always the high expressors that actually stabilize the p53 protein. Whereas the p53 frame shifter truncating correlate with loss of function. And so, that&rsquo;s kind of consistent with the high-grade serous.

In ovarian, a lot of times we&rsquo;ll use a marker like WT1, and that&rsquo;s kind of what tends to be seen a little bit more somewhere in the peritoneum, indicating a peritoneal origin. But there is kind of extensive literature now on what&rsquo;s called high-grade serous ovarian carcinoma really originating in the fallopian tubes. So women had been getting prophylactic BSOs [bilateral salpingo-oophorectomies], who were germline

carriers. And a number of groups set up protocols to really section through the end of the fallopian tube and looked for lesions there. They found these serous tubal intraepithelial carcinomas [STICs], which were believed to potentially be the origin for these ovarian carcinomas.

. Even before the serous tubal intraepithelial carcinoma, which would be a morphologic diagnosis, you can perform a p53 immunostain and see a precursor lesion. And so, it was really beautiful work. They went with a number of diagnosed ovarian cancers that had actually metastasized and went from these p53 precursor lesions, performed laser capture microdissection on p53 immunostain, the STIC lesions, or the ovarian cancers, and then the metastasis, and showed that these basically all have the same truncal origin. As they spread, they acquire additional mutations showing that the high-grade serous ovarian cancers are originating in the fallopian tubes.

That&rsquo;s a really important distinction. This is an area that&rsquo;s evolved. Everybody puts up a textbook. They talk about ovarian cancer and they still talk about the origin on the surface of the ovary. I think it&rsquo;s an important discussion and I appreciate that feedback.

Case 1: Stage IV Ovarian Cancer With Multiple Metastatic Sites

Now this patient, she&rsquo;s young, right? She&rsquo;s 44 years old.

And I&rsquo;m just thinking, what goes through your mind when you see a 44-year-old who has a supposed ovarian cancer?

She&rsquo;s very young and we haven&rsquo;t any history in terms of extensive comorbid conditions, so that&rsquo;s actually quite important. We really don&rsquo;t have a family history to draw from on here, but you&rsquo;re still going to be wondering, does she carry a

mutation, whether it be a germline or a somatic mutation? So that&rsquo;s quite important, certainly as these studies have evolved at this point in time. And given the fact that she has a liver lesion, she has a splenic mass, the question is, how is she best served? So this is really put in the hands of a multidisciplinary team, including our surgeons, our medical oncologists, and the like.

Presuming this patient came and saw you first, probably by a referral from somebody who kind of did the reverse  they saw a patient with symptoms, got a CT [computed tomography] scan, and then referred for a pelvic exam, and then you saw that this is probably of GYN [gynecologic] origin &mdash; what is your natural process?

The natural process is always to work in conjunction with our gynecological surgical oncologists. At the end of the day, the opinion that&rsquo;s rendered in terms of whether it be adjuvant, primary debulking surgery, or neoadjuvant therapy is really made by the medical oncologist, maybe sometimes radiation, or radiologists, as well as the medical oncologist.

This is a tough case. You know, she&rsquo;s only 44 and we don&rsquo;t know if she has resectable disease or not.

You can do a spleen. You can have a solitary liver lesion. You have peri-aortic lymph node. These may all be resectable. It&rsquo;s dependent on what her peritoneal disease looks like. At The University of Texas MD Anderson Cancer Center, I know you are doing a lot of assessments by laparoscopy. We have a much more difficult time doing that at The Ohio State and the James Cancer Center, just because of the way our practices are set up. But in a 44-year-old, I really want to make sure that we find a reason not to operate on her. I think that is important. Now, saying that, this is going to be a big debulking.

Right. I was going to say, Shannon, wasn&rsquo;t that really kind of the whole point of doing the scope  to try to find the patients for whom you could get a complete resection in?

Absolutely. We use a validated score and the whole goal of the score is to give you an idea of what your likelihood is of achieving no gross residual disease, which we know, based on multiple retrospective analyses of prospective data, is where you get the best bang for your buck, right? Those are the patients who live the longest. What we don&rsquo;t know is, is it really the surgery that does it, or are we just picking those right patients and they just are going to do well no matter what we do? We haven&rsquo;t answered that question, I don&rsquo;t think.

That&rsquo;s a great question. That&rsquo;s a great point. I think Dr Neil Horowitz&rsquo;s paper looking at some of the GOG [Gynecologic Oncology Group] data is probably one of the most underquoted papers, in regard to looking at the complexity of the surgery and the score with regard to their disease. And if you have both a high tumor volume and a high &hellip;.

Complexity of surgery, they don&rsquo;t do as well. I know she&rsquo;s 44, right?

I often use that paper to justify the question you just asked, which was, is this biology versus just surgical ability? In that paper, the more aggressive the surgery, if you get that patient to the same disease volume they still do not as well as if they had not had the surgery at all.

A good point, also, is we&rsquo;re told that she has a solitary liver lesion. We don&rsquo;t know that it&rsquo;s a solitary liver lesion.

Oftentimes, this is just a CT scan. We&rsquo;ve all known that CT scans can often underestimate what volume of disease is there. So that&rsquo;s something to be mindful of.

It&rsquo;s a good point. So in this case, you send her to Dave, and Dave says, &ldquo;I&rsquo;m going to operate.&rdquo; Are there other procedures that you would recommend that she have before you just release?

You could always do a PET [positron emission tomography] scan, or an MRI [magnetic resonance imaging] of the liver if you&rsquo;re really looking for more disease. You&rsquo;re right. The splenic mass is a doable thing. It&rsquo;s the liver lesion that I think turns me on a different scale. But I would do an MRI of the liver just to be sure that we&rsquo;re not really dealing with more than multifocal disease.

And one thing we started doing in recurrent disease, not up front, is the MRI of the peritoneal cavity.

With proper protocols, you could really see the peritoneal disease. Because what&rsquo;s probably going to define this is really the peritoneal...

There are a group of things that are really hard to assess. The mesentery is the other thing that I ....

I was going to ask you about that. We recently undertook a study among our group and compared it against the radiologists to see who could do a better job of defining who these people are. What&rsquo;s your sense of that?

I think it&rsquo;s a crapshoot based on who you see.

There are going to be some people who are exceptional at reading the radiology, and some people who aren&rsquo;t. We have an opportunity, as surgeons, because we can look at the radiology and then we can go in and operate on the patient and can really correlate. The question is, how often do you really do that in an organized manner?

I think that really is a testament to the minimally invasive evaluation in these patients.

In having either a 2-step process or time set aside to do a combined procedure  combine the minimally invasive to conversion to maximally invasive.

Right. We still struggle with that though because patients would like to have everything done at one time. But there is some benefit to doing a 2-step procedure, when you do the laparoscopy and then subsequently operate on the patient. And not the least of those is the ability to counsel the patient. If you&rsquo;ve identified colon disease that&rsquo;s going to require a potential colostomy, it&rsquo;s really great to be able to tell a patient up front.

We&rsquo;ve had people who actually declare and want to go to neoadjuvant therapies specifically because they thought they were going to be ....

Right. We have had patients for whom we thought, &ldquo;This patient could be debulked.&rdquo; But they absolutely said, &ldquo;No way am I having a colostomy,&rdquo; so they went to neoadjuvant.

Case 1: Considerations for Debulking Surgery

So this is a great segue into the next question, which is, what is the treatment approach, right? What we are talking about here is the surgical approach, and obviously the alternative approach would be to go with chemotherapy first. Susana, maybe you can give us a little bit of an idea of what that literature is telling us about now?

Sure. By definition, we have a stage IV ovarian cancer at this point. We&rsquo;ve had many different studies, in terms of neoadjuvant therapy versus adjuvant therapy: The EORTC, the CHORUS study, the JCOG, and, of course the SCORPION trial; and soon, hopefully, to be the true test will be the TRUST data coming up.

The ones that we concentrate on the most are the EORTC and the CHORUS data. And just recently published in the

in 2018 was a pooled analysis of both the EORTC and the CHORUS trials. It was 1200 patients, so it was a sizable group of individuals. They looked at the data and the data showed that the end result didn&rsquo;t change very much. Neoadjuvant therapy was really similar to that of primary debulking surgery in the group as a whole.

But you know, they made some very important comments: Patients had 3 centimeters and less than 5 centimeters of extrapelvic disease, versus stage III with greater than 5 centimeters extrapelvic disease. And they made some distinctions as to which patients really would do better with primary debulking surgery. It&rsquo;s 1,200 patients, but only 19% really had stage IV disease. But even in that stage IV, granted it&rsquo;s 19% of individuals, if you looked at median survival, the median survival was better with neoadjuvant therapy in stage IV in that particular review  about 3 months &mdash; and so was progression-free survival. So one could be mindful that this is a stage IV ovarian cancer.

So Dave, what&rsquo;s your experience with the impact of neoadjuvant therapy on surgery?

It&rsquo;s pretty clear that you&rsquo;re going to have less bowel surgery, less colostomies, less ICU [intensive care unit] admissions, less blood transfusions. So when we&rsquo;re really looking at those patients, it clearly is going to be a less morbid surgery. I always talk to my fellows and say, &ldquo;But it&rsquo;s always tougher than you think it&rsquo;s going to be. You have a bunch of fibrosis, you end up digging off the omentum for forever and trying not to do a transverse colectomy.&rdquo;

I think it&rsquo;s a really important aspect. Now, one thing you just brought up is, again, one of the most underquoted statistics is that in less than 5 centimeter disease, particularly if it&rsquo;s isolated, those patients did significantly better with primary debulking.

So identifying those patients who we shouldn&rsquo;t be operating on, and on the same side, identifying those patients who we should absolutely be operating on.

That&rsquo;s right. One of the prerequisites of the 2 neoadjuvant trials was they had to have some convincing evidence if the path wasn&rsquo;t non-GYN [gynecologic]  that it was GYN &mdash; that double-negative there. What kind of things can we use to help convince us that this biopsy we took of the omentum is actually ovarian cancer?

But actually, in the follow-up with that, not that long ago we would do cytology via the ascites and we thought that was good enough.

Do we need core biopsies on these, or FNAs [fine needle aspirations], or what?

I think we&rsquo;ll definitely get to that. During one of the other cases, I want to bring that question up. Thanks for bringing it up. But if we have a specimen, let&rsquo;s say we biopsy the omentum, how can you convince us that it&rsquo;s of Mullerian origin?

There are certain markers, and we kind of touched on the PAX8, which seems to be &hellip; it&rsquo;s positive in thyroid, and in some kidney, but it tends to be mainly positive in Mullerian. Especially with the so-called high-grade serous, it&rsquo;s one of the diseases that&rsquo;s defined by a specific molecular alteration where all of the high-grade serous have p53. With the immunohistochemistry, if we feel confident that it&rsquo;s going one way or the other, that all of the cells have high expression of p53, or it&rsquo;s the complete null, that certainly goes with ....

What about for mucinous? We see this in mucinous cancer as well  the same kind of disease distribution, but the markers are all different, aren&rsquo;t they?

Yes. I guess I&rsquo;m not as familiar with how that would be different.

Usually, from what I understand, the CK .&hellip;

And you know, the other thing in the trials, just to be fair, is that the way we did it is we also kind of broke down the CA 125 [cancer antigen 125], CEA [carcinoembryonic antigen] and they had a ratio of 25 to 1. And so, that was another piece. We do our best. Sometimes we&rsquo;re wrong. Even in the study that we did, we found that about 10%, 11% of the time, when we did the scope and got tissue biopsy from the ovary, we were actually wrong. It was actually a non-GYN primary.

But I think if you use that criteria, you can be fairly confident that you&rsquo;re treating a Mullerian or GYN primary.

Yes. And I would guess that this disease type is probably the better one to at least, potentially, see a response to therapy.

Case 1: Considerations for Neoadjuvant Therapy

Okay, so now we have a patient and we&rsquo;ve come up with a surgical plan. Let&rsquo;s say we decide to go with a neoadjuvant approach, even though Dave really wanted to operate on her. What&rsquo;s our chemotherapy choice?

We have many different data. We&rsquo;ve looked at ICON8. We&rsquo;ve looked at GOG 262. We looked at GOG 0218 and ICON7. At the end of the day, it really boils down to a paclitaxel and carboplatin. The question of bevacizumab is a good question here. We have a 44-year-old, stage IV, high-risk, generally healthy individual. So we would ask ourselves the question: Is this the person for whom you might want to incorporate bevacizumab? And if you did, what data could you draw on to actually substantiate that opinion?

This was a good question, to be honest. If you look at GOG 0218 and ICON7, we had a progression-free survival benefit in GOG 0218. But if you look at more of the modified ICON7, and looked at specific cohorts  for example, suboptimally debulked stage patients, all stage IV patients &mdash; they had a concomitant patient population of about 502 patients. And they did actually see a benefit of bevacizumab in that cohort.

So here you have a 44-year-old woman who was young with no other comorbidities, by presumptions, for whom we&rsquo;re going to undergo neoadjuvant therapy despite the fact that maybe there&rsquo;s a difference of opinion.

Would you give her the benefit of the doubt of putting her on bevacizumab, relying and leaning on the data of the modified ICON7 trial? I don&rsquo;t think it&rsquo;s unreasonable if it&rsquo;s done right.

The problem in the United States is it&rsquo;s not approved unless they&rsquo;ve had surgery. So I think for those patients, I&rsquo;m more apt to start the doublet. And then, if they&rsquo;re not responding or their ascites isn&rsquo;t drying up, then it may potentially be appealing to have that approved. So normally when we do that with neoadjuvant, then we&rsquo;re going to bring the bevacizumab in afterward.

I agree. It gets very complicated, in terms of when to start, when to hold. And how many cycles are you really giving in the neoadjuvant setting?

At the end of the day, you&rsquo;re really looking at 1 or 2 cycles. You might find yourself in more complications than it&rsquo;s worth.

I like what Dave said. I do that often, where I start the chemotherapy and then, if I&rsquo;m not seeing the benefit that I want, maybe add the bevacizumab. And I would just add the subanalysis of GOG 0218 was pretty compelling where you saw that the stage IVs looked similar to stage IIIs, in those patients who received the bevacizumab, as far as survival was concerned. But still, you&rsquo;re looking for something to help guide your decision.

That is really a very important point. You have to give time after the bevacizumab before you take them to surgery, and make sure you&rsquo;re holding that prior to the cycle. I really like to give them 8 weeks without bevacizumab, but I think 6 weeks is probably more realistic.

Yes, especially with the every-3-week dosing.

Right. So how many cycles before, and how many cycles after?

I generally give 3. But you know, certainly 3 to 4 is reasonable. At that point, you want to take a look. You want to see if you&rsquo;re getting any benefit. If you&rsquo;re getting past 3 or 4 and you&rsquo;re still not going to be able to take that patient to surgery, the chance that she&rsquo;s going to have a good outcome has really been diminished.

In our database, when we look at the 3 versus the 6, there wasn&rsquo;t as much of a difference as I had thought there was going to be.

No, there wasn&rsquo;t as much, but there was a slight difference. In some other retrospective studies, it&rsquo;s been a significant difference.

That&rsquo;s right. Dr Nicoletta Colombo did this nice study and actually looked at the number of cycles, pre, and basically showed that there was this kind of law of diminishing returns. The more pre-cycles that you gave, the less well the patients did. But some of that&rsquo;s also selection, right? Because the patients who immediately respond, you take the surgery after 3.

Well, somewhat selection, but a take home point is unless there&rsquo;s a reason, 3 to 4 cycles prior surgery. You know? We&rsquo;re seeing a lot of our patients in consult but they&rsquo;re already &hellip;.

And so, your options are really limited. They have to be evaluated by a gynecologic oncologist prior to the 3 to 4 cycles to get the interval debulking.

After the 3 to 4 cycles, what do you use, as a surgeon? What criteria do you use in terms of taking them to surgery? We look at chemotherapy responsiveness. That&rsquo;s first and foremost. But are there any other criteria? Do you look at CA 125 [cancer antigen 125] trends and the like?

We tend to use a lot of RECIST [response evaluation criteria in solid tumors], just because that&rsquo;s what we do with all of the other trials. And so we are using that type of idea, but that can be difficult because sometimes it&rsquo;s just that peritoneal carcinomatosis that&rsquo;s not really measurable. So you just have to get a sense of, why didn&rsquo;t you take her to surgery up front? And are those problems gone now? So if it was a performance status issue, how does she look in front of you right now? If it was disease location, have things shrunk? Have they shrunk enough? Is there still quite a bit of disease that would be residual after surgery? Let&rsquo;s say they had bulky adenopathy in the mediastinum. When do you take that patient to surgery?

But those are the kinds of things you have to weigh when trying to decide.

Case 1: Bevacizumab in Stage IV Ovarian Cancer

Let&rsquo;s move on to another case, similar in nature, but I think it&rsquo;ll take us a little bit further into that treatment paradigm. I&rsquo;m going to have Shannon go ahead and just give us the context of this case, and then we&rsquo;ll move on.

Absolutely, no problem. So this is a 55-year-old African American woman who presented to her gynecologist complaining of abdominal pain, bloating, and weight gainthe usual suspects. Her past medical history is significant for mild hypertension, which is managed with hydrochlorothiazide, and type 2 diabetes, managed with metformin. She does have a family history of a mother who died of breast cancer at age 54, a maternal aunt who died of breast cancer at age 65, and a father who is alive with prostate cancer. Her physical exam is significant. Her abdomen has a bilateral lower-quadrant tenderness. She does have a positive fluid wave for ascites. Her pelvic exam demonstrates a large nodular mass on the left side.

Imaging is obtained. A CT [computed tomography scan] of the abdomen and pelvis demonstrates bilateral adnexal masses. The left side is 8 cm, and the right side is 3.5 cm. She has ascites and diffuse peritoneal carcinomatosis. She does undergo a paracentesis of 4.5 L, and her cytology is significant for high-grade adenocarcinoma. A core biopsy of the omentum is then obtained with a high-grade serous carcinoma. Her CA 125 [cancer antigen 125 level] was elevated, at 305 [U/mL].

She gets treatment. She undergoes an exploratory laparotomy followed with an omentectomy, BSO [bilateral salpingo-oophorectomy], appendectomy, and resection of all gross pelvic nodes. She does have microscopic residual disease of R1, and germline molecular testing demonstrates a

alteration. Her final pathology is high-grade epithelial ovarian cancer involving the omentum, ovaries, and 5 lymph nodes.

OK, great case. So let me ask you a couple of things. Obviously, her past medical history is screaming for a

She is young, with a huge family history. It&rsquo;s almost overboard. But we looked at this, and I understand it has been looked at in many places as well. How reliable is the family history? In the first case, we had a 44-year-old with a negative family history. How reliable is family history?

It&rsquo;s fairly reliable, but you&rsquo;re going to miss a big proportion. About 45% of patients who have a

mutation don&rsquo;t really have any significant family history and are not what we would consider of young age. So I think over age 50. It&rsquo;s really interesting, so that&rsquo;s why we really push for universal testing of all patients with any high-grade ovarian, fallopian, or peritoneal cancer.

I was impressed by that. You know, back in the day we used to use these computer algorithms. We would bring in bits of family history, and then it would spit out a likelihood for a

mutation. The ASCO [American Society of Clinical Oncology] policy at the time was that it had to be 10% or higher. And so we did a study and saw that in that group of patients that were predicted to be 10% or lower, we end up missing, like, 25% of our mutations.

And I think it just drills home the point regarding, as you mentioned, should we just be testing everybody? What&rsquo;s your policy?

We are trying to test everybody with paired testing up front. So germline and somatic. You know, currently the commercially available entities will leave some holes and opportunities. So we&rsquo;re trying to figure out the best way to do it. We&rsquo;re actually in the process of changing our algorithm right now. If we have someone on neoadjuvant chemotherapy, we&rsquo;ll often wait to test them until their surgery so that we&rsquo;ll have more tissue rather than just a core biopsy. Because often the core biopsy is not adequate to perform somatic testing on. So we&rsquo;re trying to test everybody with germline and somatic at the time of a surgical diagnosis.

Surgical diagnosis, OK. Shannon, I was going to ask you this before, but let me get to it now. This patient had a paracentesis done, and then she had to have a core biopsy. First of all, why did she have to have both? And how confident are you on the cytology from the fluid?

This is probably another case where you&rsquo;re going to stump me, but it sounds as if it&rsquo;s something that you guys all know about. Maybe the paracentesis is unreliable. But obviously, as a pathologist, you always hear that the tissue is the issue. You always want more tissue. And so, ideally for the patient, you don&rsquo;t want to have to go back in if you can do it off, you know, whatever is a paracentesis. And do the immunohistochemical work-up that way. For this patient, it seems that they have to go back in there for some reason and obtain another biopsy, and maybe they just didn&rsquo;t have enough tissue up front. I&rsquo;m not sure.

It seems that a fair amount of times with the paracentesis alone, you just get an adenocarcinoma diagnosis, which is fine, because generally for the up-front treatment we&rsquo;re going to treat these patients the same. But you know, personally, you want to know if they have clear cell or endometrioid, because there&rsquo;s going to be such a different molecular milieu in those patients.

It&rsquo;s become standard in our institution that we rely on the core biopsy.

The problem is the timeliness of the core biopsies from our interventional radiology colleagues. Now, I would want it on this patient because her CA 125 is only 350 [U/mL]. We don&rsquo;t know what her CEA [carcinoembryonic antigen level] is, but it does bring up the question of a nonserous histology, which may or may not make any difference in your management, but it would be nice to know.

Yeah, and I think that given they had such a large volume of ascites, I wonder if that was done just for that purpose. It was symptomatic. At the time of surgery, when we have a patient with a large amount of ascites, we&rsquo;ll end up draining at the time that we do the incision, and then we go from there. But you know, however the sequence is done, they end up getting both the tissues and then ultimately make that diagnosis. But yeah, I was also thinking, with the 44-year-old, that clear cell or endometrioid are tumors I&rsquo;d be thinking about as an alternative histology.

Case 2: Newly Diagnosed Ovarian Cancer, gBRCA-Mutated

So this patient, Shannon, as you said, underwent a primary cytoreduction. It looks as if they got her resection down to very small, microscopic residual disease. Then the germline data became available, and she showed a

mutation. So tell us what your thoughts are with primary therapy now. You&rsquo;ve got a young, relatively healthy patient. She had a cytoreduction, obviously large-volume disease now down to no visible growths. What&rsquo;s the best treatment for her?

You just basically throw your dart at the board. You can do whatever you want, right? I think we&rsquo;ve come full circle. We originally had paclitaxel-carboplatin intravenously every 3 weeks. And we&rsquo;ve gone through a number of different modality assessments looking at dose-dense paclitaxel and dose-dense carboplatin, or intraperitoneal [IP], and then the addition of bevacizumab. This is why it should be like a 20-year fellowship instead of just 3 years.

But I think for somebody like this, who has minimal residual disease and is known to have a

mutation, I&rsquo;m going to already start thinking maintenance. That&rsquo;s going to help inform what I&rsquo;m going to do for their treatment. For a patient like this, I know that I&rsquo;m going to go ahead and give olaparib down the line. And so I&rsquo;ll look at how I&rsquo;m going to give that in terms of modality. For the patient who I&rsquo;ve essentially debulked, I would likely just give it every 3 weeks. That&rsquo;s just based on the most recent data that we have. Despite all these different modalities that were utilized, essentially the progression-free survival [PFS] and overall survival [OS] were the same, or in the ones who are reported overall.

What do you think about this dose-dense idea? It&rsquo;s like the world shifted on us.

It did, and it shifted back, you know? It was just Q3-week scheduling [every 3 weeks], and that&rsquo;s actually what we&rsquo;re back to. But this case is interesting, because you know she&rsquo;s a

mutation carrier, and you know very well there are clinical trials that have now emanated and that she could be an eligible candidate for that. Not only would you present to her a PARP [poly (ADP ribose) polymerase] inhibitor, in this case, olaparib, because it&rsquo;s actually approved, but you could ask whether or not immunotherapy has any role. You would probably shy away from giving her anything that wouldn&rsquo;t guarantee her a PARP inhibitor.

I think the majority of up-front trials now are guaranteed, right? So there&rsquo;s the ATHENA trial. And as you mentioned, there are the FIRST and DUO-O trials. I think the

-mutant population is going to be assured to get a PARP on those trials.

But I think we have so many up-front trials right now. Every single patient has to be discussed to go on a clinical trial. We have to do better. I was so happy to hear you say

Oh, I was just going to ask you that. I didn&rsquo;t even hear her say that.

We have retrospective subanalyses of our GOG [Gynecologic Oncology Group] data. Our friend Wendel Naumann, MD, just had a publication with regard to

-positive patients and how much better they have done with intraperitoneal chemotherapy. Once again, my first choice is always going to be a clinical trial. None of the clinical trials are enrolling intraperitoneal chemotherapy, but if she was not interested and I knew she had a

mutation, I would have a serious conversation about intraperitoneal chemotherapy in this patient.

The rationale is that&rsquo;s platinum based and it would be targeting the...

Yeah, there are retrospective data that demonstrated a pretty impressive progression-free and overall survival benefit in those

s that had staining. It was kind of, it wasn&rsquo;t really...

But again, I think we&rsquo;re starting to see some really interesting data. We need to do this more prospectively and we need to identify those patients.

So let me take it down 1 more even further controversial topic. If you think that the intraperitoneal chemotherapy is actually causing all that damage in the

patients, to make them have such long outcomes, what about HIPEC [hyperthermic intraperitoneal chemotherapy]?

Should it do more? I don&rsquo;t know. We have 1

article on HIPEC in the neoadjuvant interval debulking setting. So at the time of interval debulking, they had HIPEC. Again, they showed us a progression-free survival improvement. We really need to look at it as an option. But I think the most important aspect to that paper is that you&rsquo;re not going 6 or 8 weeks, maybe even 10 or 12 weeks, without chemotherapy. So you&rsquo;re doing that chemotherapy between.

Absolutely. I think there is tremendous regrowth in that short period of them. We did a small stud, and then all these patients had to get CT [computed tomography] scans before they enrolled in a certain portion of the study. These were optimally debulked patients, and yet they had measurable disease. You could argue the point that this is just inflammation, what have you, postsurgical. But to your point, are we seeing reemergence of disease so soon, and that&rsquo;s why HIPEC actually has gained the popularity or the interest that it has?

You also wonder, though, is it HIPEC, or is it just IP after surgery? I think that&rsquo;s another big&hellip;

Yeah, we&rsquo;ve mentioned this many times. The quality control here is we don&rsquo;t really know that we&rsquo;re actually heating the cancer. We know we&rsquo;re putting hot stuff into the abdominal cavity, but how many tumor probes do we have that actually say that we&rsquo;re actually heating the tissue?

And for the people who have done HIPEC, to have a trial in the

that has no difference in their toxicities, for those who have done HIPEC, there is a significant difference of toxicity.

That gives me a lot of pause to see that.

Yeah, and then it&rsquo;s the old argument with the neoadjuvant versus primary surgery, right? They look at the absolute benefitthe 12-month PFS and the 30-month OS&mdash;and they go, &ldquo;All our trials are better than that.&rdquo; And this was the same thing. The numbers were not that impressive. It was that good. It&rsquo;s just better than placebo or the nontreatment arm. So, OK, we got HIPEC on the table. Everybody can be happy.

The good news is, there are lots of trials ongoing now. Thank gosh they finally got this sorted out. So hopefully we&rsquo;ll get some more information as that rolls out.

Case 2: Primary Treatment Options

OK, so what did we decide on treatment? Bevacizumab? No bevacizumab?

I think we decided that every 3 weeks, and no bevacizumab.

OK, so every 3 weeks, paclitaxel with carboplatin, and no bevacizumab.

All right. If you go to the NCCN [National Comprehensive Cancer Network] Guidelines, no matter what answer it was, I think every answer, even the 1 you came up with, would be right on target. OK, so they&rsquo;re all candidates. Now what?

We&rsquo;re likely going to treat about 6 cycles, and then we typically reassess and make sure we&rsquo;ve gotten what we wantedmeaning there&rsquo;s no evidence of new disease.

So you&rsquo;re going to go 6 cycles and then reimage?

Or just test for CA 125 [cancer antigen 125]? Do you reimage after 6 cycles of chemotherapy?

I do. Many of my partners do not. I think it&rsquo;s a nice baseline.

You have a group of patients who are going to have lymphocysts, which in the future may be misinterpreted disease.

So prior to all this treatment, I&rsquo;ve already been counseling her. This is, again, a known

patient. We&rsquo;ve been talking about the fact that this is not going to be the end of therapy. So at the completion of chemotherapy, give her a few weeks off. And then, typically within about 3 to 4 weeks, start her on olaparib.

OK. What&rsquo;s your standard approach to that? Regarding that first cycle, what do you do? Do you do it differently?

Well, I think a lot of this is counseling, right? So first, of course, the goal is for the patient to live longer without disease. But you don&rsquo;t want to be adding additional toxicity. This is a population for which typically, before this, it was surveillance. They were on nothing. They were feeling good, and they were off treatment. This is their window. This is kind of their honeymoon time. So you want to make sure that you make it as much of a honeymoon as possible. And so it&rsquo;s really important to counsel these patients and make sure they know that even though this isn&rsquo;t chemotherapy, there are adverse effects. But they&rsquo;re manageable.

We generally see our patients once a month when they&rsquo;re on olaparib maintenance, but we do make sure they have ample access to the clinic team so that if they start to have nausea or any symptoms that make us concerned, we can intervene before that 1-month time point. You don&rsquo;t want someone to be struggling through that first month with fatigue and nausea and things like that and then be done with treatment.

And we know that treatment modifications are frequent with these drugs.

So that&rsquo;s part of our counseling. Around half of patients will require a dose reduction. They all go home with antiemetics, and we follow up with them within the first week with a telephone call, but then see them monthly. Part of that education is right from the beginning. But managing often includes dose reduction and antiemetics or other medications for fatigue.

We often send people home with a prophylactic antiemetic. We don&rsquo;t necessarily tell them they have to take it, but we counsel them just like you did with chemotherapy. &ldquo;If you find that you&rsquo;re starting to have nausea, then go ahead and take that 30 minutes to an hour before you take your drug, and see if that doesn&rsquo;t get you where you need to be.&rdquo;

So I cheat a little bit. I actually do the reverse, OK? Let&rsquo;s pick a PARP [poly (ADP ribose) polymerase] inhibitor like olaparib. I start at an intermediate dose, like 200 [mg], you know, twice a day. And then I give it for a week. I give it for a week and a half, and I acclimate the dose to 300 [mg], which actually saved me a lot of dose reductions in the course of time. I know that&rsquo;s not how you&rsquo;re supposed to do it, but that&rsquo;s how we do it.

No. Actually, in practicality, we were doing that as well, absolutely.

And the dosing and the availability to escalate or deescalate is sometimes easier at those doses.

Case 2: Treating BRCA-Mutated Ovarian Cancer

So you mentioned olaparib. Maybe you can just, for the audience, kind of run through the data that support that indication?

Yeah, absolutely. So the indication was based on the SOLO-1 data, which was a randomized controlled trial, a large study, in which they got people to a responsethe majority of the patients had a complete response, and a proportion had a partial response after 6 cycles&mdash;and then, within an 8-week time frame, randomized them to either olaparib or placebo. And of course this population was all

mutations were germline, although there was a proportion that had somatic mutations. And so they were treated for just 2 years, and that&rsquo;s an important point. They stopped treatment at 2 years, and when they presented the data last year at ESMO [European Society for Medical Oncology Congress] and then with the subsequent paper, they demonstrated a clear progression-free survival benefit and a hazard ratio of 0.3. And importantly, this 3-year progression-free survival was 60%, compared with 27% in the population that got placebo. So it does seem that you&rsquo;re maintaining kind of the slope of the curve or the shape of the curve beyond when they stopped the maintenance therapy at 2 years, which is really exciting.

And it&rsquo;s interesting that we don&rsquo;t have that drop-off like we see in the bevacizumab trials, in which you have that significant change in progression-free survival after stopping therapy. It&rsquo;s really maintained in a nice linear tail.

Here&rsquo;s a question for you, because this comes up often. What if this patient came to you a year and a half ago, and this data had not emerged, right? A patient comes to you. You&rsquo;re 6 months out from finishing chemotherapy. She has a

mutation. You have SOLO-1 data. You have the approval. Would you start her at that point in time? How much of a gap goes by before you say yes or no?

I&rsquo;ll give you my opinion, but let me ask you this first.

I haven&rsquo;t. What we don&rsquo;t know is, do they get a similar amount of benefit later on down the line? It will be interesting to see when the overall survival reads out, because eventually these patients are going to get a PARP [poly (ADP ribose) polymerase inhibitor], at some point. And so the question is, do you get the most bang for your buck up front? I think my gut feeling is maybe you do, but if you&rsquo;ve got somebody who&rsquo;s already out 6, 9 months&hellip;

That&rsquo;s exactly what we&rsquo;ve done.

That kind of an image has already been done.

Well, I think there are a couple of points, 1 of which is: We thought these patients did so well. I would also not start this patient if she came in 18 months later. But we thought these patients did so well. Twenty percent of these patients will be platinum-resistant from the SOLO-1 data. About 10% to 20% are disease-free at 5 years.

So here are my thoughts on this. I&rsquo;ve been toying around with this idea in my mind, so I&rsquo;ve got to put it onto paper. If you look at the curves carefully, 1 of the things you notice is that at about a year out, they start to parallel each other. What that tells me is that in this patient population of

patients, you know the patients who are left with disease at the end of chemotherapy, whether it&rsquo;s the second look or not, are essentially looking for something to be treated with in maintenance.

If you think about using a PARP inhibitor in patients with known disease at the completion of therapy, you&rsquo;re getting a treatment effect. And then when we talk about the next case, which is about a patient with recurrent disease, we know that there is a treatment effect. I think what you&rsquo;re seeing is that early on there is this treatment effect that is maintained. But you start to lose the ability, or the proportion of patients who are going to ultimately respond. You&rsquo;ve weeded them out, and essentially everybody then is basically on a natural kind of disease-progress curve. And so, I would say that if you actually look at the slope between these 2 curves, they start to flatten out and be more like each other, probably somewhere between 6 or 9 to 12 months.

It&rsquo;s a really interesting thought. We do know that if these patients&rsquo; CA 125 [cancer antigen 125] levels start to go up, we&rsquo;re going to be quick to look to see if there is a PARP option after or before a platinum doublet.

Case 2: Olaparib in BRCA-Mutated Ovarian Cancer

OK, I&rsquo;m going to present another case. We&rsquo;re going to move now into the recurrent space. I&rsquo;m interested in hearing your thoughts on this particular patient. So this is a 68-year-old Caucasian woman who was seen by her primary care physician for a recent onset of chest pain and shortness of breath, and she reported some recent vaginal bleeding, early satiety, and a weight loss of 12 pounds. She has a past medical history that&rsquo;s significant for hypertension and arrhythmias, both of which have been medically managed and are stable. Her family history is significant for cardiac disease history on her mother&rsquo;s side, and her father died of pancreatic cancer.

She&rsquo;s morbidly obese. There&rsquo;s dullness to percussion on the right, and she has a kind of asymmetrical chest expansion. And on pelvic exam, there&rsquo;s a mass or a suspicious mass in the pelvis. Laboratory findings are significant for a CA 125 [cancer antigen 125] of 445 [U/mL], and on imaging she has a CT [computed tomography] scan of the chest that shows this effusion, which confirms what we found on the chest exam. Her abdomen shows this 4-cm mass in the area of the left ovary with some residual omental caking and peritoneal implants, and there appears to be some involvement of the mesentery.

She has a thoracentesis, 1500, and that cytology was positive for high-grade adenocarcinoma; and then ultimately has a pelvic mass biopsy with a core specimen that shows a high-grade serous carcinoma that has immunohistochemistry, which was nicely presented in some of the other cases that seem to be consistent wit

positive, CK7 positive, CK20 negative. And in this patient, it is noted that she has a germline mutation that&rsquo;s wild type for

The patient has undergone some treatment before. Her treatment history consists of debulking in primary surgery, which was incomplete. She had residual disease of 1.5 cm and was given the program of GOG-0218 with paclitaxel, carboplatin, and bevacizumab followed by bevacizumab maintenance for 15 months, and she achieved a complete clinical response to that.

Then she undergoes nearly or a little over 2 years in which she&rsquo;s disease-free but comes back with symptoms that are very similar to what we see in patients with recurrent disease. She has severe abdominal bloating and fatigue, an elevated CA 125 of 255 [U/mL], and at least 3 sites of recurrence on imaging, and her performance status is 1. So she&rsquo;s given carboplatin-docetaxel. After 6 cycles of therapy, she has a partial response and her residual disease is at 1.5 cm.

So this case has a lot of discussion items. I hope we can get through many of them, and we want to talk about them. But before we go too far, I want to talk about the potential molecular information we might want to gain. So we know that she is a germline

wild-type patient. Would you guys get any somatic testing at this point? She is platinum sensitive and recurrent. Dave?

So we try to test up front, when we do the somatic and the germline, as we talked about before. However, many of our patients, particularly within this time frame, would not have had somatic testing. So if they had not had somatic testing, we would absolutely look to 1 of the commercially available tests and/or internal next-generation sequencing of the tumor.

Same thing. We have, in our institution, a clinical trial in which we do next-generation sequencing up front. And so if we did not have that Foundation Medicine, Inc test, MSK- IMPACT, or anything of the sort, we would absolutely get somatic testing on her, yeah.

Yeah, the same. Previously, we were really focusing on doing the somatic testing in the platinum-resistant setting, because we felt like that was really where you utilize that information. But now, as we&rsquo;re moving these targeted therapies up earlier in the line, we&rsquo;re trying to do it as soon as possible.

Case 3: Recurrent Ovarian Cancer, Platinum-Sensitive, BRCA WT

Greg, maybe you can tell me a little bit about what the difference between somatic and germline is? And then, also, what is next-generation sequencing?

A lot of people don&rsquo;t like the term

. You&rsquo;ll hear it called second-generation sequencing, and there&rsquo;s potentially third- and fourth-generation sequencing. Some people like the term

. You can do DNA or RNA, but mainly DNA, and make a fragment of what you&rsquo;re going to sequence, but many millions of fragments. And so that&rsquo;s why it&rsquo;s the massive amplification. It has the power, compared with traditional Sanger sequencing, to sequence millions of reactions at the same time. So you can sequence with some of the commercial assays. Most people aren&rsquo;t necessarily doing the entire exome or the genome. There obviously are some companies that do that, but a targeted panel in which you can look at like 320 genes or things like that on MSK-IMPACT or Foundation Medicine tests. As the cost of this has decreased, and then companies like Foundation Medicine, Inc, have gotten FDA approval for a lot of the advanced cancers, we&rsquo;re seeing more and more of it being used routinely.

And so the important thing to know with somatic testing is that it&rsquo;s really done on the tumor. Most of these companies, when they&rsquo;re sequencing the tumor, it&rsquo;s from an FFPE [formalin fixed-paraffin embedded] block. If it&rsquo;s ever insufficient, it&rsquo;s typically because there&rsquo;s not enough tumor in the block compared with background lymphocytes or a stromal tissue. But you&rsquo;ll see a cutoff typically of about 20% for that type of sequencing, to make sure you&rsquo;re actually sequencing the tumor, which is the somatic as opposed to what would be in the germline. And with most germline testing for solid tumors, that&rsquo;s typically going to be performed on blood. Where it&rsquo;s really becoming useful is when you&rsquo;re seeing a lot of treatments that are not necessarily as much derived from tissue of origin but underlying molecular alterations.

, obviously those aren&rsquo;t the only genes involved in the homologous recombination repair. You have other genes, like

and things like that, which could be targeted with similar therapies, potentially. But there are many others. With larger-panel sequencing, where it&rsquo;s not necessarily even exome but where you&rsquo;re looking at 320 genes, you can pick up a lot of other things. If the patient didn&rsquo;t have mismatch repair testing, you can actually see that signature that we do see more, or the microsatellite instability that we see more often in the endometrial or the colon cancers. But there obviously are certain percentages of cancers that have that.

Then there are other driver alterations in specific genes that are seen much, much less frequently in Mullerian origin, but there are trials with the NCI-MATCH in which patients can match on therapy with that. And there are other rearrangements that we typically think about in lung adenocarcinoma, with

Advanced Ovarian Cancer

Case 1: Stage IV Ovarian Cancer With Multiple Metastatic Sites

Case 1: Considerations for Debulking Surgery

Case 1: Considerations for Neoadjuvant Therapy

Case 1: Bevacizumab in Stage IV Ovarian Cancer

Case 2: Newly Diagnosed Ovarian Cancer, gBRCA-Mutated

Case 2: Primary Treatment Options

Case 2: Treating BRCA-Mutated Ovarian Cancer

Case 2: Olaparib in BRCA-Mutated Ovarian Cancer

Case 3: Recurrent Ovarian Cancer, Platinum-Sensitive, BRCA WT

Case 3: Molecular Testing in Recurrent Ovarian Cancer

Case 3: PARP Inhibition in Recurrent Ovarian Cancer

Case 3: Selecting Therapy For Recurrent Disease

Case 4: Recurrent Ovarian Cancer, Platinum-Resistant, BRCA WT

Case 4: I-O Therapy for Platinum-Resistant Recurrent Disease

Case 4: Treating Platinum-Resistant Recurrent Disease