Persistent splenomegaly and weight loss in CML despite molecular response - making sure you have all the pieces of the puzzle
A 28-year-old man presented to an outside institution with a 3-month history of fatigue and weight loss (12% of body weight).  Spleen size was 22cm below the costal margin.  The remainder of the patient history and physical exam, other than a mild tachycardia at 110, was unremarkable.  Bloodwork showed a WBC of 145 with left shift, plt of 610 and Hgb of 102.   Diagnosis of CML was made and confirmed with peripheral blood BCR::ABL1 positive molecular testing.  Started on nilotinib 300mg bid and had an initial slow response because of high WBC, but hit appropriate milestones and by 18 months, had MR4.5 (0.0032% IS).  Weight loss did not improve and fatigue continued and was felt to be due to nilotinib.  Reassessed clinically and found to have an additional 2% body weight loss and spleen now into the pelvic brim.  Referred to my centre for discussion of the role of splenectomy and possible allograft as he had an HLA-matched sibling.  Physical exam and molecular testing confirmed with nothing new.  Additional diagnostic tests were done.

Did the center perform a bone marrow biopsy to exclude BM fibrosis and molecular tests for JAK2, MPL, and CALR?

No bone marrow was done at diagnosis at the other centre. Whether fibrosis was present or not is unknown. This is important. Once upon a time, the development of fibrosis on therapy was considered disease acceleration, but one would not know this if there was no baseline for comparison.

Similarly, cytogenetics at baseline were not done. Again, clonal progression on therapy can be considered the first stage of acceleration and warrants thinking about more aggressive management such as allograft. Given a young healthy male with a matched sib, allografting is on the table. PCR for BCR::ABL1 however remained undetectable in his case. It is also a strong recommendation that in cases of refractory cytopenias with initial TKI therapy, a bone marrow be done to rule out fibrosis as the cause.

My first manoeuvre was a bone marrow aspirate and biopsy. The first was difficult. No dividing cells for cytogenetics and morphology thin. Biopsy however, showed significant fibrosis. Molecular testing was negative for JAK2 but positive for CALR leading to a diagnosis of co-existent CML and MF. The co-existence of BCR:ABL1 and either JAK2 or CALR has been reported, so this was not a new finding.

I started the patient on full-dose ruxolitinib in addition to the nilotinib and monitored him clinically for more than 6 months. Other than mild fatigue, no side effects and no bloodwork adverse events. Spleen size did not budge as is sometimes noted with huge spleens and significant fibrosis. The patient was becoming more uncomfortable with the abdominal fullness.

After appropriate counselling, the patient received the usual vaccinations and underwent elective splenectomy by laparotomy due to the spleen size. When the surgery had healed, he underwent a myeloablative stem cell allograft with fludarabine/busulfan/TBI conditioning and cyclosporine/post-transplant cyclophosphamide GVH prophylaxis. He engrafted promptly, within 15 days, and was off immunosuppression by 3 months. No acute or chronic GVHD. Nilotinib and ruxolitinib were discontinued one week before conditioning. He remains well and all disease-free 5 years later, with BCR:ABL1 monitoring.

So we have a number of teaching points with this case which is not so unique.

1. Bone marrow aspirate, biopsy and cytogenetics at baseline are still important despite what has become a "routine" in many places.
2. When things do not appear to be improving in unison - pcr negativity and spleen size in this case - rethink the problem and investigate.
3. Co-existence of CML and other MPDs can occur and in a case like this, should be investigated. As an aside, molecular testing can be misleading. BCR::ABL1 should be considered with any new MPD, as CML can present as thrombocytosis with mild leucocytosis or even fibrosis.
4. It is possible to use two TKIs together safely when trying to treat different diseases. However, in the absence of a reasonable response, move on to something else.
5. Yes, allografting still exists and can be done more safely than in the past. With donor pools expanded to include matched related, unrelated, haplo and rare cord, healthy patient age now even into the 8th decade, reduced intensity conditioning which may be as good as myeloablative for CML, better GVH prophylaxis and supportive care, more people could be eligible and delaying SCT, may take the option off the table.

Absolutely agree with the follow-up, the treatment with two TKIs (some patients treated with the same strategy in my center) and the final choice of allografting.
As other unique comments, I would like to remember the possible use of NGS in this rare context, because knowing the non-driver additional mutations may guide rapidly to a transplant (i.e. ASXL1, RAS).