: Dr Sitlinger, you can tell us about CD22-targeted immunotoxin, the mechanism, and how it works with moxetumomab?

: This is now approved in the second line or second relapse. It’s an interesting kind of first-in-class type of drug where it’s fused to a pseudomonas exotoxin, and so once it is digging into the cell, the toxin base inhibits protein synthesis and ultimately triggers apoptosis. There have been several studies, and recently in ASH , they presented data on long-term follow-up with this drug. Patients are having long-term response. We’re still waiting for years out like we have for some of the other drugs, but in patients who are having those earlier relapses, it’s something to consider.

: Dr Ravandi, I know you participated in the Kreitman published study. You obviously had good experience with these agents. What are your experiences with these agents efficacy wise?

: My experience in hairy cell has been limited. We have treated a couple patients. In terms of hairy cell, as you have seen the reports in publications, it is effective in the approved setting, and that’s the second relapse of patients. Dr Kreitman has shown that in many of these patients, you can get MRD –negative responses. The question of MRD negativity comes back, but in the relapse setting, it’s more useful to have MRD negativity because the patient has already shown that the disease has a tendency to keep coming back. If you’re treating a relapse patient, it might be more prudent to try to achieve an MRD-negative response.

Obviously, the drug has a couple adverse effects that are well publicized. Hemolytic uremic syndrome tends to happen in about to 5% to 10% of patients. Also, capillary leak syndrome requires significant monitoring and asking the patient to drink a lot of fluid during the course of therapy. Other things also need...carefully monitored for hemolytic uremic syndrome, including renal function...during the course of therapy. But again, with monitoring and with appropriate management, these are generally limited in terms of the degree of incidents and are also generally reversible. I do believe there is a need to use aspirin prophylaxis when you get this drug, so then some patients with low platelets, we would need to be a bit careful. Overall, it is an active drug, and it does produce MRD-negative responses. You heard about the long-term follow-up data.

: I participated in the original trial. It has an interesting history. The first iteration of the drug did cause quite a bit of hemolytic uremic syndrome,...but the patients on that trial...were truly refractory patients. These are patients that you had to demonstrate that after they got cladribine for the fourth time or they didn’t respond at all that they were without treatment options at that time. The drug is quite effective, as Farhad mentioned, quite prolonged responses in that difficult-to-treat population. It clearly has a role.

: Can someone talk more specifically about the response rate that has been observed in this particular study or to a particular agent? I know the one benefit of this agent is the MRD negativity, as all mentioned, the deeper response they were getting for patients, which hopefully will lead to more durable remissions and less likelihood of needing another line of therapy. For those who are more familiar with it, any one of you, maybe talk about what CR [complete response] rates we see in this type of patient and then the MRD negativity.

: With the original trial, 41% of patients achieved a CR, and 27 patients, so 34%, achieved MRD negativity. Of those patients, 30% or 24 patients maintained a remission for more than 180 days. These were the truly refractory patients. These patients have done quite well compared to what they had done with other treatments.

: There is some anecdotal experience with a few patients being retreated with the drug. These are patients who first got it as an investigational agent, and now that it’s commercially available have been retreated many years later. I have one such patient, for example, who retreated successfully and achieved a nice complete response once again.

Videos

The Role of CD22-targeted Immunotoxins in HCL

: In the relapse setting I would use the combination. We’ve been using cladribine and rituximab, and I’ve used pentoxide and rituximab once or twice, and that’s quite effective. In the few patients that we have had that we used a second-line therapy, the duration of remission has been longer than the first remission, which is against all the published literature...mentioned earlier—that with the single-agent nucleoside analog therapy, second and third uses of nucleoside analog are usually associated with lower response rate and shorter response duration.

But when we introduced the rituximab to cladribine in the relapse setting, the second remission has been longer than the first remission in the few patients that we have had enough time to follow up. This has also been shown by Dr Claire E. Dearden at The Royal Marsden in the UK. I don’t think she has published that, but they also show that the combination produces second remissions that are longer than the first remission.

One other area that I wanted to touch that we didn’t cover is in variant hairy cell. I do believe that in variant hairy cell, the combination cladribine plus rituximab—or as Dr Robert J. Kreitman has published, bendamustine plus rituximab—is the way to go because we all know that single-agent nucleoside analog therapy in variant hairy cell is associated with lower response rate and also definitely shorter response duration. The addition of rituximab does not cure these patients, but it does produce higher response rate and longer response duration.

Landscape Change in the Treatment of Relapsed/Refractory HCL

 Let’s discuss the relapsed and the refractory patients with the hairy cell leukemia after they get treated with their first-line therapy of choice. Andrea, what defines a relapse? At which point should we consider treating the patients again?

 Various papers have looked at the rates of relapse. It’s usually 30% to 50%, depending on what you’re looking at. Technically, true laps are not someone who has disease that still had a PR [partial response] to initial therapy, but at least a clinical CR [complete response] by counts. A true relapse does not need to be re-treated again until they meet the criteria for treatment. They’re at treatment-naïve hairy cell leukemia because of pan cytopenias or other symptoms. That’s because, while absolutely treatable, it’s not curable in the long run. So you don’t need it if the patient is doing well and has some mild cytopenias, and you detect that it is present, treating them doesn’t necessarily change their course, and you’re exposing them to treatment that they may not need and all those adverse effects. Additionally, this is where the MRD [minimal residual disease] debate comes in, we do know that with each subsequent treatment at least with the pairing analogs, your remissions are likely to get shorter in between. Exposing to treatments when they don’t need them may end up using up treatments faster.

For all those reasons, we do not need to treat them until they’re showing symptoms that would require treatment as outlined before. One thing that’s really important is the timing. If you relapse with less than 2 years versus relapsing after 2 years, and part of that is basically because if you’re relapsing in a longer response to appearing analog, you may respond and often do respond with a long period of duration. If it was 10 years ago and it relapses again, then you certainly can get another long duration with rechallenging versus a patient who is relapsing less than 2 years. The likelihood of having a durable response when they didn’t initially is much shorter. Those are important keys to think about when you’re picking what you do next and when to treat.

 Dr Coutre, what are the options, with the second- or third line therapy, for relapsed or refractory hairy cell leukemia? What are some of the guidelines? I know NCCN has listed some of the drugs, and Dr Sitlinger also mentioned retrying cladribine or pentostatin, especially to late relapsed patients. How has the landscape changed over the last several years with research that’s been done in this area?

 Similar to Andrea, for somebody who’s had a long, durable response, I would re-treat with cladribine or cladribine with rituximab. I would not advocate changing nucleoside analogs. I don’t see the point for a patient like that. Then of course there are our clinical trial options that we’ve touched on. Ibrutinib is in trials. The BRAF inhibitor mentioned and other BRAFinhibitors are in trials. Those can be perfectly appropriate options for patients too. We’re always trying to develop new and more effective therapies. Then we have approved drugs, but I believe that’s for second-relapse patients in their first few months. As with many very uncommon hematologic diseases, we have a wealth of options, which is always nice.

 Dr Saven, if the patient has received a cladribine or pentostatin as their first-line therapy and they relapsed, when you re-treat them with cladribine, for example, would you add rituximab the second time if they didn’t receive it the first time?

 I add the combination to patients when they’ve relapsed unless there’s some reason why I wouldn’t add the rituximab. I do combination therapy for patients who’ve relapsed. If a person’s first remission is 12 years, it could be debated. But if the patient’s remission is 3 or 4 years, and they didn’t do as well as they were supposed to do, then I’d be inclined to add the rituximab. Andrea does quote that 2-year interval, which I always found a little humorous because I was once asked to write a review, and I came up with a No. 2. Then I noticed it was added to all consensus and it came into NCCN Guidelines. I’m not sure that 2½ is less than 1.8 years. If the patient doesn’t do like everybody else and has a brief remission, whether it’s 2 years or 2½ years, you’re told that he’s not that sensitive to purine nucleoside analogs. I don’t think the 2 years is on the Ten Commandments. You can use your clinical judgment about the advisability of reintroducing the purine nucleoside analog. It was also thought that pentostatin and cladribine were almost exactly the same drug, but they aren’t. When I was a filler, I wrote a paper in 

. There were 5 patients who were refractory to pentostatin a number of times, so they were clearly refractory. When we introduced cladribine, they actually responded. Then the papers from the United Kingdom in which patients got cladribine first and later got pentostatin and then responded. There are mechanistic dissimilarities between the purine nucleoside analogs. They’re not exactly the same drug, although they’re very similar. What is concerning is the repeated use of a purine nucleoside analog and damage to the bone marrow in and of itself. Subsequently, you make clear that hairy cell may induce hyperplastic changes in the bone marrow. Also, there is the cumulative effect of profound immunosuppression. They’re not the same drug, but they have similar long-term consequences.

Historical Treatment for Relapsed/Refractory HCL

 It’s an interesting perspective here, about rituximab, the cladribine-plus-rituximab issues. Farhad, do you give all your patients cladribine plus rituximab? Is there any other downside that you see? Are there patients for whom you probably wouldn’t consider adding rituximab?

 We have continued this study. This is probably the longest-lasting study in our department and in our institution. It was opened in 2003 and still active. We still enroll to the study, and for that reason we still use it sequentially. Because we have had such good results, we didn’t feel the absolute need to do it concomitantly.

Now, I will tell you that more recently, in the last couple of years, because the NCCN Guidelines don’t even mention clinical trial as an option for frontline therapy of hairy cell leukemia specifically since cladribine or pentostatin, we’ve had some problems with insurance companies approving it, so we have treated some patients with cladribine alone. More recently, I have treated a couple of patients with cladribine alone because of COVID-19. They didn’t want to get the additional immunosuppression, and they weren’t keen on getting vemurafenib for whatever reason. So yes, we are still using that strategy, and I can tell you the responses continue to be durable many years later.

 How about the goals of a therapy for these patients? Obviously the first goal is the reversal of their cytopenia. Some of the questions that occasionally come up after they get therapy, and they get a bone marrow biopsy after 4 months. Or does everybody even need a bone marrow biopsy after getting this type of a therapy if they have a normal blood count? Then the second is that when they do get it, despite the fact that their blood count is normal because of the persistent presence of a hairy cell leukemia, some of these patients may end up getting a recommendation for another line of therapy, which may be important for another topic to just clarify here. Are there any comments on the need for bone marrow biopsy and what to do with the bone marrow biopsy information?

 In my practice I don’t do post-treatment bone marrow as long as the blood counts are normalized. If there are ongoing cytopenias, I elect to defer that bone marrow, but I don’t do it routinely. I don’t treat the minimal residual disease [MRD]. If I’m going to use the Rituxan, I’m going to use it up front. I’ll wait and see, but I don’t treat people based on minimal residual disease. Just to go back to Dr Ravandi-Kashani, I tend to use the combination more in younger patients than older patients. I’m not sure they need to be disease-free for longer periods of time, because chances are they’re going to die of something else. I don’t routinely do a post-treatment bone marrow. Dr Coutre mentioned that you could do a bone marrow at 4 months. I have data that patients can be positive at 3 months, negative at 6 months; positive at 6 months, negative by 9 months; positive at 9 months and gone at 12 months. It’s important when you do your minimal residual disease. Six months may not be long enough. I was the filler doing all the bone marrows in all these patients, and what struck me was that we didn’t know exactly when to do that, and some people will convert only at 1 year. I’m not quite sure what it means. Dr Ravandi-Kashani’s study I didn’t think required a bone marrow at 4 weeks. You just did it.

If you’re going to use Rituxan, just do it. I’m not sure a follow-up bone marrow is going to provide all that much information outside a clinical research study. I noticed there are consensus reviews in everything else. Always call for a post-treatment marrow, but I’m not sure what you do with that information except upset the patient.

 I fully agree. I don’t do routine bone marrow in practice. Occasionally, you do see a patient referred to you who’s had a post-treatment bone marrow and still has some detectable disease, and they want to know what to do, or that it’s already been acted on, for example. That is a slippery slope, I must say. I always teach the fellows that if you’re going to do a task, ask yourself what you’re going to do with that information. Right now I would not act on detectable minimal residual disease if a bone marrow was done.

 I agree. Outside the setting of a clinical trial, I don’t think there is any need to do a repeat bone marrow except, for example, if you have a patient who is 6 months out and still has significant cytopenias. That’s the situation that I would do a bone marrow. I agree that probably 3 months is not the best time to do the bulk bone marrow. You can be positive at 3 months and become negative at 6 months or, as Alan said, even further out. And also in terms of MRD, Dr Saven has a paper that looked at some patients 17 years out. They still had some hairy cells in their bone marrow, and they were not having any problems. This is a huge debate, and it’s not something that can be settled in a short period of time.

 I agree. I quote that paper a lot, Dr Saven. It is quite interesting because you actually do a bone marrow biopsy in patients who have completely normal blood counts. Then in a serial bone marrow biopsy you actually detect them, with no clinical consequence, in need of a therapy. But when they do, it does upset the patients and raises all sorts of questions from their perspective, understandably. What does that mean? What do you do with that information?

Having good expert opinions and groups giving the thoughts [is important] because some may get another therapy unnecessarily that they don’t need for many years or ever. It’s an important point to bring out and encourage, obviously to discuss with their physicians and/or second opinions if that ever comes up.

When to Not Use Combination and the Need for Bone Marrow Biopsy in HCL

 Another big question pertains to adding rituximab. We talked about this MRD [minimal residual disease] with deeper remission. Should it be a goal for everyone, or a particular patient population, that you would like to get a deeper remission?

 We all primarily use cladribine as frontline. I can’t remember the last time using pentostatin, for example, with hairy cell. Patients do exceedingly well for long periods of time, but for young patients of course, you anticipate having to treat them again and again, probably throughout their life. Now, what about rituximab? Seeing it work initially was intriguing, and what I’ve been really struck by is Bob Kreitman’s recent publication; it’s an interesting study design. If we’re doing a response assessment for hairy cell, and choose to do that with a bone marrow exam, you should wait at least 4 or 6 months because you could see ongoing responses later. In his study he randomized patients to cladribine plus up-front rituximab, starting on day 1, and then weekly for 8 weeks And then he looked at the 6-month response, including minimal residual disease versus patients getting standard cladribine and then doing the response at 6 months and then the delayed rituximab infusions. It wasn’t a direct randomized trial with or without rituximab, but it was effective. The striking finding to me is, not surprisingly, a much higher percentage of patients who are MRD undetectable, but with long-term follow-up, was maintained. The responses were maintained for a longer period of time. In that group the median response wasn’t even reached yet. It goes along with the theme we’ve seen in other diseases: Deeper responses give you longer response durations. We have to be critical and ask, “Well, does that matter?” Can you simply get that response, or do you do just as well even if you need to be treated earlier for a disease like this, which we can’t cure? Because of that, I would say that certainly for the younger patients and select older patients, I’ll use the combination now.

 I am a big proponent of adding rituximab. Hairy cell leukemia has probably the highest expression of CD20 on its cell surface. Rituximab was developed in Southern California. We all know that rituximab is beneficial to virtually every other lymphoid malignancy when it was added to the classical or standard therapy for the other malignancies. For example, there have been studies in lymphoma and CLL and even in ALL. It was a randomized study from the French group that showed the rituximab addition to the classical treatment benefits patients in terms of most studies or even overall survival. Conceptually, for me it’s a no-brainer that we have a disease where there is a very high expression of CD20. There had been a number of studies that have used rituximab as a single agent in the relapse setting. Although it’s not as dramatic, highly effective agent as a single agent in the relapse setting, it does produce responses, so we know that it is active. For me, it has been a no-brainer to use the combination.

Originally when we designed the study that was mentioned, that was 15 years ago. Hairy cell leukemia was a disease that was thought to have been captured by cladribine. We didn’t want to make any mistakes or individual problems. We decided to do sequential, and we’ve kept up the same way, but I totally agree that if I want to combine the two drugs, I also would give it concomitantly as was described by my colleague, and he wrote a long-term follow-up of the combination of cladribine followed by Rituxan, which was published in the 

 on essentially a very long follow-up of over 10 or 12 years.

We have had 1 relapse, which wasn’t even a full-blown relapse. The physician decided that they wanted to treat the patient again. This is a very durable response. Now, Dr Saven will come back to me and say that I have had the hairy cell patient, received single agent cladribine in the nineties, and they’re still in remission. That’s obviously the core of this discussion. Are we going to treat the entire population of hairy cell leukemia patients with the addition of rituximab, where at least—Dr Saven will correct me—about 50% or 60% will have a very long response and will never relapse. I don’t even think it’s a medical discussion. That’s a socioeconomic discussion. Also just to discuss other frontline combinations, Dr Saven mentioned there is an interest in combining vemurafenib with CD20 antibodies. Jae has a vemurafenib plus obinutuzumab study. The Italian group presented data on vemurafenib plus rituximab. It seems that adding rituximab to vemurafenib also significantly improves the response rate, and likely will significantly improve the response duration. In a drug situation where you have an oral drug that you take daily, it is important to highlight a strategy that will improve your response duration that would obviate the need to continue to take an expensive oral daily for a long period of time.

Combination First-line Therapy in Hairy Cell Leukemia

 Let’s begin with Andrea, perhaps with just a brief overview of the indications for treatment. Do all patients with hairy cell leukemia need therapy? Can some of them be observed with monitoring alone?

 Hairy cell is treatable, and we can get very long remissions with cladribine and pentostatin. It’s not curable, and most patients will relapse at some point. You don’t necessarily need to treat right away. There are 3 big reasons to treat. One would be significant cytopenias. If you have a neutrophil count that’s less than a 1000 mm3, if your hemoglobin is less than 11 g/dL, if your platelet count is less than 100,000 mm3, those are all indications that the hairy cell is causing significant issues, and you’re at higher risk for infection. If you’re having symptomatic splenomegaly or other symptomatic adenopathy, and lymph nodes are not the most common thing but certainly do occur, and constitutional symptoms if they are significantly impairing activities of daily living. Those could all be reasons to treat.

 Dr Saven, can you start us off with some of the first-line therapy options for patients with hairy cell leukemia. I know we talked about some pure analog, and there are a couple of choices there. When would you choose 1 versus another, and perhaps a little about your schedules and how is it administered?

 In ideal circumstances that purine nucleoside analog is the treatment of choice. You’ve got 2 real choices: 1 is cladribine, or 2-chlorodeoxyadenosine, and the other is pentostatin. Cladribine historically was given as a single week of infusion. There was in vitro data showing that prolonged exposure was better than brief exposure, and it’s the continuous infusion for a week.

There was 0.1 mg/kg per day by continuous infusion for 7 days. Ultimately, that became a hardship because people needed a catheter and a PICC [peripherally inserted central catheter], and the catheter got infected, so we converted to 0.14 mg/kg per day every 7 days over 2 hours for 5 days, and that represents 0.14 mg/kg over 2 hours each day. There hasn’t been a direct comparison, but if you look at the results, they’re almost identical.

We give it now as a 2-hour infusion over 5 days, and it’s 1 cycle of therapy. We give it intravenously, but there are preparations in which you can give it subcutaneously. Years ago we did a PK study showing that you can actually give it orally. If you double the dose and give it orally, you can achieve the same levels. But at Scripps [Health in San Diego, California] in 2020, we give it as a 2-hour infusion times 5 days. Pentostatin is another treatment available as first-line therapy. It’s given every other week, 4 mg/m2 IV, but it’s more protracted therapy. It’s given every other week for 3 to 6 months, so not surprisingly, cladribine has emerged as the treatment of choice because of its convenience. At the beginning of our studies we’re going to exclude actively infected patients. We appreciated the immunosuppression of this drug, and people had to have the infection stricken before they started.

People who champion pentostatin initially didn’t have that exclusion, and you could receive the drug on this study. I still wouldn’t use either purine nucleoside analog in the presence of active infection. There have been a number of publications about whether you can combine cladribine with anti-CD20 rituximab. The way Dr Ravandi-Kashani did it was you give the cladribine, you wait a month, and then you give 8 weekly doses of rituximab. The NCI [National Cancer Institute] has given it on day 1 and then weekly times 8, together with the 5 initial doses of the cladribine.

You can achieve a deeper level of response, ie, lower minimal residual rates. Whether that’s clinically important or not, and if it’s worth the extra immunosuppression that’s going to follow is the subject of debate in the age of the coronavirus and the fears about immunosuppression. If you elect to use a purine nucleoside analog, you may or may not want to add the rituximab given the prolonged immunosuppression that is likely to follow.

Interferon, the magic product that was on the cover of 

magazine many years ago, was approved to treat hairy cell leukemia. It subsequently had other uses. I haven’t used interferon for many decades, but there is a role for it, probably in actively infected patients. If you can get hold of vemurafenib, a BRAF inhibitor, it may be a role for interferon. If you have a pregnant patient and she’s not a candidate for splenectomy in the second trimester, maybe there’s a role for interferon. But the remissions from interferon are not very deep. Most of them have partial remissions, taking retracted therapy over the course of the year and self-injections of a medication. It’s not commonly used anymore. Also, if you use interferon after a purine nucleoside analog, it’s very unlikely to work. Other safety concerns, not really. Blood counts may get worse before they get better, and with cladribine it has about a 40% incidence of neutropenic fever. But despite the incidence of neutropenia, at least at Scripps, we’ve seen rare infections.

That’s how I approach new patients with hairy cell leukemia. There are emerging data about the role of vemurafenib as frontline therapy. There are studies at Memorial Sloan Kettering [Cancer Center] in combination with obinutuzumab. But that’s in the realm of research rather than standard of care. But in the day of the coronavirus, when you don’t want to immunocompromise patients, that is a consideration.

Determining When to Treat and First-Line Therapy for HCL

: With differential diagnosis of hairy cell leukemia, the bone marrow biopsies, and cell marker and flow findings, and perhaps some of the lymphadenopathy issues may point in one direction or another. When we see patients like this, similar to leukemia or lymphoma, what other blood cancers should we be considering or rule out, for suspected hairy cell leukemia, because it's a rare disease? Dr Ravandi, please comment on some of the differentials for hairy cell leukemia.

: The morphology of hairy cell is characteristic. They have these standard projections from the cells, but there are some other conditions that can be fairly similar in terms of morphology. For example, a splenic marginal zone lymphoma, which also has these projections from the cells. Then there is the very uncommon hairy cell variant, which is perhaps 5% to 10% of all the cases of potential hairy cell leukemia diagnosis. In terms of morphology, it’s typically characteristic for the pathologist, and the flow cytometry markers that were described by a previous presenter are also typical in hairy cell leukemia. As mentioned, the B-cell markers as well as CD11, CD103, and CD25 are positive, and there are a number of markers that we explained are negative. In variant hairy cell leukemia, CD25 and sometimes also CD103 are negative. Variant patients tend to present with a higher white blood cell count, so at presentation hairy cell leukemia patients tend to be pancytopenic, although, you can have classical hairy cell leukemia with a higher white count as well. But the ones who have a high white count tend to be variant hairy cell, and patients with variant hairy cell also tend to be older. The median age for classical hairy cell is about mid-50s, but in variant hairy cell they tend to be older, middle-to-late 60s typically, and even older. This is important in terms of management because variant hairy cell doesn't respond as well to nucleoside analogue, so if you do come up with a diagnosis of variant, you need to think about your treatment strategy to be a bit different.

Another thing with variant hairy cell is they are not 

 mutated. That's a good marker to distinguish variant from classical hairy cell leukemia. The use of flow markers is not limited to diagnosis. We can monitor hairy cell leukemia clearance in a number of ways. For example, we have been using a soluble IL-2 [interleukin-2] receptor, which is soluble CD25. Also, you can use soluble CD22. But you can also use flow markers for looking for minimal residual disease. Now, the debate about the relevance of minimal residual disease clearance in hairy cell is extensive. The flow markers are important to use both for diagnosis and for follow-up, and 

 mutation is important in terms of distinguishing classical hairy cell from variant hairy cell.

 mutations? Because I think sometimes we have a targeted gene panel dedicated for 

, and then some are more NGS [next-generation sequencing] panels, or targeted gene sequencing. I’m curious whether we routinely do it for all patients with hairy cell leukemia, and then come to what assays you are using.

: Increasingly, we tend to use our institutional NGS panels on most of our hematologic malignancies, for better or worse. Sometimes a patient with hairy cell leukemia has that done for that reason. In the past, we've simply used immunohistochemistry, for example.

: We use a PCR for specific patients and then complement that with immunohistochemistry.

: We use Sanger sequencing, but I'm not going to go through all the differences between that and next-generation sequencing. My understanding is it's a little simpler and it’s more targeted.

Differential Diagnosis and Molecular Tests in Identifying HCL

: What other testing should we be performing when you see a patient who has a suspected diagnosis or clinical presentation of hairy cell leukemia? I’ll start this section with Andrea.

: The first part of the diagnosis after you've had your CBC [complete blood count] with differential and taken a look, is a bone marrow biopsy. Often it is extremely difficult or impossible to get an aspirate, and that's due to the fibrosis that's induced with the hairy cell as well as probably some interlocking of those hairy cells with the projections. It's important that you still attempt it. But still you'll get a good biopsy to then look under the microscope. You'll see the classic fried egg appearance that still comes up on boards for many people, and it's usually a hypercellular marrow. There is a percentage, less than 10%, of hypocellular marrows that coexist, but it's usually hypercellular. We were just talking a lot about 

 testing. I will note it has not been incorporated as an NCCN [National Comprehensive Cancer Network] automatic diagnostic criteria, but I think most of us find it's important. One, because it will help if there’s any question with the immunohistochemistry to classic variant vs other variants of hairy cell. It also potentially does guide treatment down the road.

 testing, the other thing that's important is your immunohistochemistry testing. Classically, I'm sure we all have seen variants, hairy cell is pan-B cell antigenic expression, CD19, CD20, CD22. And importantly, in distinguishing from other indolent lymphomas, it lacks CD5, CD10, CD23, and CD27, also CD21. CD103 is a very sensitive marker for hairy cell, and classically you have CD11c and CD25. Immunohistochemistry testing is extremely important. Another thing that comes up is when to do imaging, and I open that up to the rest of the panel, but it's certainly not necessary. If you have questions about the spleen or if other things are going on, you’re palpating lymph nodes and you're wondering, it can be useful, but imaging is not a mandatory aspect of the hairy cell diagnosis.

: I agree. I don't routinely do CT scans or ultrasounds of people's abdomens. If the patient’s on a clinical study and you want to know the depth of response, I think it's a necessary step to do to see if you can resolve splenomegaly. About 20% of patients will have lymphadenopathy on CT scan, but in terms of routine care of patients, I don't see any advantage, especially if you can palpate the spleen. Even if you can’t palpate the spleen, I'm not sure it moves the ball very far forward to be able to know whether there’s imaging of a spleen. It's not clinically significant.

: Dr Saven, you spoke about the prognosis of the patients with hairy cell leukemia. We spent time on clinical presentations and diagnostic work-ups. When we see them, is there anything that we can share with our patients that you know about how they are likely going to do in terms of the genetics, clinical age, or anything you consider?

: The first thing I tell the patient is, “You've got the best of the bad bunch, and your life expectancy is very close to that of the normal population.” Despite the rarity of this disease, there’s a plethora of treatments. Hematologists are squabbling about what's the best of the combination of drugs and which is the least toxic. The overwhelming majority of patients seem to do very well. If you're on the younger side, below 40, when you get the disorder there seems to be a slightly more aggressive nature to the disorder, but it’s still very indolent. If your flow cytometric pattern doesn’t express all the markers that Andrea alluded to, is CD25 negative, but you got everything else, you may not do as well as everybody else. If you’re 

 positive, it may be that you don't have typical hairy cell leukemia and you’re not expected to do as well. We spoke about the MAP kinase expression. They tend to do worse. If you’ve got the 

, they tend to do worse. The size of your spleen, severity of your cytopenias, tend not to affect your prognosis. Patients who present with typical or atypical hairy cell leukemia who have a high white count tend not to be as sensitive to the purine nucleoside analogues. But the overwhelming majority of patients do exceptionally well and can expect a near-normal life expectancy.

I don’t know if second malignancies is a topic for discussion today, but the incidence is increased. There was a long debate as to whether it's treatment related because the drugs are very immunosuppressive, or whether it's the disease itself, or combinations of the above. Often patients get second malignancies 10, 20 years later, and that's actually the cause of their demise.

: How do you frame the conversation with the 35-year-old with hairy cell leukemia? They're looking at a horizon of about 50 years.

: Let’s start with, “You should do very well. You’re going to become an expert on this disease. You’re going to read all the articles. Your disease may be a little bit more aggressive.” We haven't gotten to treatment yet, but I tend to give them combination therapy with a purine nucleoside and an anti-CD20, just so I can try to get closer to minimal residual disease. They tend to do very well despite the young age, but you're right, they want to know what's going to happen when they are 60, 70 years old. They should do well, but not necessarily so because we don't have a curative approach. We don't think that we can cure this disease, even with all the therapies available, so at some point they're going to relapse.

Signs and Tests for Diagnosis and Prognosis of HCL

: Dr Coutre, perhaps you can enlighten us about the molecular genetic changes that have been identified in hairy cell leukemia.

: This is a fascinating disease because of its molecular basis. In 2011 we had the publication from Italy, from doing whole exome sequencing of patients with hairy cell leukemia. They published a report that said we sequenced these patients, and essentially every patient had a 

 mutation, V600E. It was striking, and it is in the case of hairy cell, a disease defining mutation. In fact, those who don't have it, who have a hairy cell-like presentation, are probably the patients with hairy cell variant, which largely we consider a distinct disease these days, including how we might approach treatment. The importance of that mutation is that it leads to constitutive activation of the MAP kinase pathway, which controls cellular processes like proliferation and cell death. You get phosphorylation of ERK and MEK downstream, and that gives this cell an advantage if you will, and essentially drives the transformation from a normal cell to a leukemic cell.

We'll talk more about how we take advantage of that singular mutation from a therapeutic perspective. There have been other molecular abnormalities described, some of the cyclin-dependent kinases, for example, that typically act as tumor suppressor genes. Those are actually inactivated in some patients with hairy cell, admittedly the minority, less than 20%, but they coexist with the 

 mutation, and they play some role in those patients. It’s acting as a suppressor gene, it’s inactivated, and the cell has an advantage. They also tend to have the same allele frequency as the 

 mutation, suggesting that it's really an early event, a clonal event if you will, that is important to the transformative process.

: With hairy cell sometimes you get a dry tap when you do a bone marrow. I've seen this frequently, that they send a sample for 

 mutation, and it comes back as negative. This has to be well-publicized because, again, in most of our experience classical hairy cell almost always, perhaps even 100% of time, does have the 

 V600E mutation. But the sampling is very important, and there are other ways of eliciting that 

 mutation, for example, immunohistochemistry is a way of showing its existence. The reason I bring this up, I have had consultations about a patient who morphologically was classical hairy cell. Flow markers were classical hairy cell. This was a multiply relapsed patient, and they were reluctant to give the patient the BRAF inhibitors because they felt that there was no 

 mutation. I think that point has to be emphasized, that you can actually miss this if you just use the simple dry tap from the bone marrow because you won't get adequate specimen.

: Dr Saven, have you ever seen a patient with classic hairy cell where you've done testing for 

, and you're absolutely convinced that it wasn't present?

: Yes, it’s very uncommon. But like everything in medicine, nothing is 100%, but I think it's 99.9%. I have seen 2 or 3 patients where I repeated the bone marrows and they were aspirated. I did all the next-generation sequencing, it was negative, and they didn't have hairy cell variant. They tended not to do as well as the patients who were 

 positive, but it's only anecdotal. I've only got a few cases, the overwhelming majority are positive, but it's not 100%.

: No, they were negative for that. They exclude each other. If you’re 

: That’s an important point that you raised there. I had another extreme case, where everything else, immunophenotypically, morphologically, [suggested] hairy cell leukemia, but because of a negative 

 which was missed or there was sensitivity of the assay perhaps, the patient ended up getting treated not for hairy cell leukemia, but for another indolent, non-Hodgkin B-cell lymphoma-type of a category. The patient ended up getting complications from the therapies, and so forth. It’s important to look at all factors, not just 1 particular molecular feature, but immunophenotypically and the morphology, and then obviously clinical presentation. I do think it’s a very important point that you raised there.

Molecular Genetic Changes for Hairy Cell Leukemia

entitled “Updates in the Management of Hairy Cell Leukemia.” I'm Dr Jae Park from Memorial Sloan Kettering Cancer Center in New York. Joining me today in this virtual discussion are my colleagues, Dr. Steven Coutre from Stanford University Medical Center; Dr. Farhad Ravandi-Kashani from MD Anderson Cancer Center; Dr Alan Saven from Scripps Health; and Dr Andrea Sitlinger from Duke University Cancer Institute. Hairy cell leukemia is a rare, slow-growing cancer of the blood involving excess abnormal B cells. Today we're going to discuss a number of topics pertaining to the management and practical application of the most recent data on the disease. Our first topic is disease state and biology of a hairy cell leukemia. Dr Saven, can you share with us an overview of the hairy cell leukemia epidemiology, risk factors, and clinical presentations?

: Hairy cell leukemia is a relatively uncommon hematologic malignancy. There are about 600 to 800 cases annually in the United States. I think that's underestimated. That was before the introduction of flow cytometry, where it's easier to detect hairy cell leukemia circulating in the peripheral blood. It's a disease of males rather than females. The ratio’s about 4:1, usually middle-aged patients, but there’s great variability. The youngest patient I’ve seen is 26, and I've seen the disease in 90-year-olds. It's mostly a disease of Caucasians. Why it doesn't affect other ethnic groups to the same degree I’m uncertain, but it certainly occurs in Hispanics. Of the 800 patients I've seen, probably only 1 has been African American, and I've seen a handful of Asian Americans with this disorder. What predisposes patients to this disorder? No one really knows. There are various theories about pesticides, organic solvents, but I don't think they've been definitively proven. Among the Caucasians, there’s a higher incidence among Ashkenazi Jews. The reason for that is unclear. It is a disorder of the memory pre-plasma cell, and circulating lymphocytes can accumulate in the spleen, less commonly in lymphocytes in the bone marrow, and hence the clinical presentation of splenomegaly, cytopenias. Monocytopenia is a hallmark of untreated hairy cell leukemia. This is a disease with a very indolent and protracted clinical course, and we'll get into the various factors associated with it.

: Sometimes we see very young patients with hairy cell leukemia. It is a disease of relatively younger patients compared to other blood cancers that we may see, or the leukemias. Do any of you have thoughts about patients in their 20s and 30s, which are very rare, but any differences in terms of the clinical, not so much presentations, but in terms of the prognosis? Are there any other factors to consider for younger patients?

: When it occurs in younger patients, it tends to be somewhat more aggressive. The prognosis in younger people isn't as good as in older people, but older patients can also die of unrelated causes. It's probably a little bit more biologically active in younger patients. I rarely see patients in their 20s. The youngest patient I've ever seen is 26, and I'm not sure if it occurs in childhood.

: Dr Sitlinger, what are the laboratory findings for these patients? How do they determine it’s hairy cell?

: Similar to other indolent lymphoproliferative disorders, there are 3 big buckets in which patients are diagnosed. One is that they present to their primary care doctor and have laboratory tests drawn for whatever reason, screening, and are found to have significant cytopenias. About 60% to 80% of people are found to have pancytopenia, and neutropenia particularly, as well as monocytopenia. They're both in the range of 80% of patients who will have this disorder. They will often present to us because they were found to have lab abnormalities. Neutropenia and pancytopenia are by far the predominant lab abnormality. There are about 10% of patients that present with a leukocytosis, and these tend to be the more aggressive and maybe even patients with variant hairy cell leukemia. For the general practitioner, or general oncologist, just because you're seeing leukocytosis doesn't rule out hairy cell.

One of the very profound features would be splenomegaly. Up to 90% or more of patients have splenomegaly. Some patients can present with profound, even splenic rupture just with how big their spleen is. That's less common now with imaging, but that has been a problem in the past. What's interesting with hairy cell compared to other leukemias and lymphomas is it's much less common to have lymph nodes involved. Clear splenomegaly is a clue that hairy cell is in the differential, among other things as well.

Then finally, people also present with systemic complaints. It's usually fatigue, weakness, sometimes weight loss, particularly a large spleen was causing early satiety and people are not eating as well. They also can present with profound thrombocytopenia, leading to bleeding. The other thing is there are either recurrent infections or recurrent autoimmune complications that will also alert patients. What’s interesting with hairy cell compared to other leukemias and lymphomas is it's not as common to have fevers and night sweats with hairy cell. Usually if I see a patient with fever or night sweats, I'm either starting to think something else is going on, or it’s just the infections that come along with these patients since they are immunocompromised. When I talk to patients, those are the 3 main presentations that we see initially that alert us to that hairy cell diagnosis.

: Anecdotally, the uncommon patient who presents with the leukocytosis, I've been impressed that they often have extremely high fevers.

: I also wanted to mention that the situation of massive splenomegaly and significant constitutional symptoms is becoming less common, mainly because people are getting routine blood screening. Frequently patients are found to be slightly cytopenic on a routine blood evaluation, and then further work-up shows hairy cell leukemia. There are unusual presentations. Lymphadenopathy is not that common, but I've had patients with very unusual presentations. I had a young lady who had a paravertebral mass, which turned out to be hairy cell leukemia. I had a gentleman who had a renal mass. The thinking was renal cell carcinoma. It turned out to be hairy cell leukemia. I had another gentleman who was managed by orthopedics for almost a year because of back pain and severe bone pains. Eventually, he had a blood count and he had massive involvement of all of his skeletal system with hairy cell leukemia. Unusual presentations do exist.

: Especially when you're seeing a lot of hairy cell, you're going to see a lot of the unusual and uncommon variants. That's an excellent point. It is interesting, and certainly the vasculitis and some of the autoimmune manifestations that go with it as well can present in other organ manifestations. There are a lot of different ways, but it's interesting to highlight that in a majority of patients, you will see less of some of the other common things. It’s fortunate that we are catching these patients a lot earlier, and we’re not having the massive splenomegaly and splenic ruptures that used to be seen.

Steven Coutre, MD, Stanford University Medical Center

Articles