The classic situation in which I would choose to use an immunomodulatory drug (IMiD) would be for a patient who has moderate to severe anemia or transfusion-dependent anemia without symptomatic splenomegaly or bad MF-related symptoms. The anemia is an important consideration, and for these patients I would use lenalidomide combined with steroids. I dose the lenalidomide depending on the patient’s platelet count. If the platelet count is higher than 100,000, I use 10 mg a day 3 weeks on, 1 week off. If the platelet count is less than 100,000, then I tend to start with 5 mg a day 3 weeks on, 1 week off.

Lenalidomide helps improve the patient’s anemia in about 20% to 30% of cases. The responses are durable for a fair amount of time, lasting for months to a year or longer. Clinical trials are ongoing with another IMiD, pomalidomide. One has to be careful with pomalidomide because it is myelosuppressive, so the dose is around 0.5 mg per day. Results of the Phase II trials suggest a 20% to 30% benefit. I believe we should reserve judgment about the ultimate role of pomalidomide until the results of the Phase III trial of pomalidomide are published. Another interesting approach currently being evaluated in clinical trials is the combination of an IMiD with a JAK inhibitor.

I would use an IMiD for a patient who is anemic with, say, a hemoglobin level of 7.5 g/dL, who is not very symptomatic, has thrombocytopenia and whose spleen is not enlarged. I would use lenalidomide combined with steroids instead of thalidomide for these patients. In our experience, the combination of lenalidomide and steroid is better, with a higher response rate, more sustainable responses and a better safety profile. I usually see a 35% to 40% response rate in terms of improvement in a patient’s anemia.

I would also use an IMiD for patients with chromosome 5q deletion. Overall, however, if the patient has an overlapping syndrome, I will look to see which is the most pronounced feature. If I observe more dysplastic changes, then I start treatment with 5-azacitidine or decitabine. If I see the opposite — more proliferative features, enlarged spleen — then I’ll start with ruxolitinib and add decitabine down the road. I think that the combination of an IMiD with a JAK inhibitor can be feasible but not concomitantly and not as up-front therapy. It is not a combination that should be used outside of a trial setting at the moment.

In my mind, IMiDs are appropriate at this point only for patients with anemia. I used IMiDs much less frequently this year than I would have 3 or 4 years ago. If I have a patient who has anemia and a platelet count of 40,000, the patient doesn’t have diabetes with neuropathy and ruxolitinib doesn’t seem appropriate, then I would consider administering thalidomide. Occasionally I will have a patient with a deletion 5q abnormality and I’ll use lenalidomide. I’m not fond of using pomalidomide and putting patients with JAK2-positive disease and anemia on the study.

I base my decision of whether to use thalidomide or lenalidomide on the patient’s platelet count. If the patient has bad cytopenias and I’m trying to address the anemia, then I won’t use lenalidomide. For a patient with bad neuropathy, I will probably not use thalidomide. It depends on the patient’s profile.

If anemia is the only difficulty, then I will probably be more likely to start with an IMiD or a clinical trial addressing the issue of anemia. My choice of an IMiD would depend on the patient’s platelet count. If he or she also has thrombocytopenia, we probably will start with thalidomide and prednisone. That probably has a higher response rate than lenalidomide and is less expensive. If the patient’s platelet count looks good, then I will probably select lenalidomide with prednisone.

I would use thalidomide for a patient with significant splenomegaly, thrombocytopenia or anemia.

About a third of the patients with MF present with a hemoglobin level of less than 10 g/dL, and if they don’t have constitutional symptoms or bothersome splenomegaly, ruxolitinib is not going to help.

This is a setting in which IMiDs would be a preferred therapy with or without corticosteroids. I usually coadminister IMiDs with corticosteroids. Danazol, or one of the other androgens, can also be helpful for some patients in this setting. I tend not to use erythropoiesis stimulating agents (ESAs) in this group of patients because there have been reports of splenomegaly worsening during epoetin therapy and I have seen it in my own practice. Also, a couple of papers have suggested, although they were not convincing, worse outcomes in patients treated with ESAs.

I have been using thalidomide primarily unless the patient has a deletion 5q abnormality, in which case I would use lenalidomide. I’m not sure that should change now with the arrival of pomalidomide. It has not been compared head to head with thalidomide so it’s not clear that pomalidomide is any better than thalidomide. I am comfortable with thalidomide at this point, and unless it’s convincingly shown to be better, I’ll likely stick with thalidomide in combination with corticosteroids.

This scenario is one of the few indications where thalidomide still has a place. I tend to use thalidomide at a fairly low dose, 50 mg, and the patients seem to tolerate that well. I’m not convinced that lenalidomide is better, and in my own experience it seems that the cytopenias may be a little worse with lenalidomide.

For a patient with anemia or who develops anemia while on ruxolitinib therapy, treatment with IMiDs can be considered. Anywhere from a 20% to 30% response rate in correcting anemia, and occasionally thrombocytopenia, has been reported with IMiDs. I seldom will opt to use IMiDs preceding treatment with ruxolitinib. I usually use IMiDs if a patient’s disease has failed to respond to JAK2 inhibitors. However, there is one setting in which I will use IMiDs first, and that is for patients with anemia who start with low counts, such as those who a priori have low bone marrow cellularity and start with a very low white cell count, such as 3,000, and a platelet count of less than 100,000. I will not use JAK2 inhibitors in these patients because of the myelosuppressive nature of these drugs. In my experience, the use of JAK2 inhibitors in these patients worsens their condition. The white cell count tends to drop further to a potentially dangerous myelosuppression level, the platelets drop, and the anemia becomes transfusion-dependent. This group of patients does not benefit from treatment with JAK2 inhibitors.

I primarily use lenalidomide, although others prefer thalidomide. Lenalidomide may initially aggravate myelosuppression and the anemia. It may aggravate the thrombocytopenia and even neutropenia. You should allow the patient to remain on lenalidomide for 2 to 3 months. If the patient is tolerating treatment and you modify the dose accordingly, then occasionally you have a response. The duration of the response to lenalidomide varies. I have seen responses lasting for 2 years.

Patients who have significant anemia and are transfusion dependent or are about to be transfused still deserve to be treated with ruxolitinib. Either you start with the lower dose and go high in monthly increments as tolerated, or you add another medication for anemia, such as danazol, epoetin alpha or low-dose thalidomide. I use thalidomide because it’s not myelosuppressive. No overlapping toxicity occurs with danazol, erythropoietin and thalidomide, and they can improve the anemia in 20% to 25% of the patients. Otherwise, there are no other medications that I can combine safely with ruxolitinib.