Hydropic leiomyoma: a distinct variant of leiomyoma closely related to HMGA2 overexpression

Uterine leiomyoma (ULM) is the most prevalent benign tumor of the female reproductive system, occurring in up to 70% of pre-menopausal women. Several different subtypes have been identified, including hydropic leiomyoma (HLM). Certain aspects of HLM’s morphology make it difficult to identify – particularly, cases with extensive hydropic change can hide features typical of leiomyoma, making diagnosis tricky. HLM genetics are also poorly understood, and, not surprisingly, large-scale studies are all but absent from the literature, with most reports focused on individual cases.

Dr. Brannan B. Griffin, a pathologist at Northwestern’s Feinberg School of Medicine, saw these many missing pieces in HLM’s clinicopathological profile. Along with his team, Griffin went on to conduct the largest HLM series study to date. They raked through the university’s database for leiomyoma cases from the past 15 years, pulling samples that fit HLM’s description. Histological, immunohistochemical, and genetic analyses were then performed on the tumors, and compared alongside each patient’s medical history. I was lucky enough to be able to discuss this groundbreaking study with Dr. Griffin. Below are excerpts from our interview (edited for length):

Me: Your approach to retroactive analysis of these tumors – combing through Northwestern’s database to find cases that fit your criteria – sounds painstaking, but effective. Since genomic profiling seems to be becoming more standardized in cancer research, do you think studies like yours will become more common?

Dr. G: Yes…a strong incentive exists to find the genomic profiling of certain diseases because the basic science can ultimately lead to a clinical impact. This idea is the whole principle behind targeted therapies, which occur once someone identifies the crucial components in mechanisms of disease.

Me: You ended up finding that HLM was more distinct from typical ULM than prior studies had suggested; HLM’s morphology, genetics, and clinical correlations were all pretty unique. Were you surprised by this, or did you suspect it already?

Dr. G: Yes, when we looked at our overall experience with cases of HLM and its mimics, we appreciated these tumors to be different histologically in that they just did not look like ULM tumors. We were not that surprised by our findings ultimately, because they confirmed what we had suspected. We were grateful that our morphologic suspicions correlated appropriately with both the clinical data and the immunohistochemical and molecular data.

Me: You found that HMGA2 was rearranged in 32% of your samples. Did the study shed light on anything new about the gene’s role in HLM?

Dr. G: Yes, our results allowed us to conclude HMGA2 acts as a disease driver in HLM. We found tumors that possess HMGA2 overexpression had high pAKT levels and increased Ki-67 proliferation index. This finding is something that was novel and indicated HMGA2 had a strong contribution to AKT pathway signaling. While we did know about HMGA2’s role as a regulator of transcription factors/DNA synthesis, the downstream effect in pathway signaling was enlightening.

Me: Reading this made me realize that I’d always thought of benign tumors as relatively symptom-free and non-concerning, but you showed that HLM can present with some pretty nasty (even cancer-mimicking) side effects. Do you think benign uterine tumors like HLM deserve more attention?

Dr. G: Arguably, I do believe benign tumors in general deserve more attention in the right clinical context. When you compare benign tumors to poorly behaving malignancies, no, the focus should definitely be on helping patients suffering from those malignancies. But when considering benign neoplasms that are extremely common, cause significant disease burden, and propose diagnostic or clinical management difficulty, then one should very much focus on the disease.

Our investigation is a great example of a benign disease worth studying because we had significant findings with clinical implications. We filled in a knowledge gap missing from the literature, as well as highlighted how HMGA2-driven tumors have morphologies that bring about diagnostic difficulty and likely require stronger clinical management.

This publication is currently available on PubMed:

CLICK TO VIEW PUBLICATION