Search tips
Search criteria 


Logo of ccLink to Publisher's site
Cell Cycle. 2016; 15(14): 1801–1802.
Published online 2016 April 22. doi:  10.1080/15384101.2016.1176414
PMCID: PMC4968897

Modeling lymphangiosarcoma in mice

Lymphangiosarcoma is a rare but aggressive tumor of endothelial origin. Currently there is no effective treatment for the disease, and the prognosis for patients is very poor, with a reported 5-year survival rate of approximately 10%.1 Lacking an animal model for the disease with molecularly defined pathogenesis is a major obstacle for our understanding and development of potential therapeutics of the disease. Recently we generated a mouse model with inducible endothelial cell-specific deletion of Tsc1 which recapitulate salient features of human disease, including progression to local invasion and systemic metastasis through the lymphatics.2 We found that hyper-activation of mTORC1 signaling as a consequence of Tsc1 deletion exclusively in endothelial cells leads to malignant cutaneous lymphangiosarcomas. Moreover, sustained mTORC1 activation is necessary for both tumor initiation and maintenance, and mTORC1-dependent autocrine VEGF signaling is identified as a key mediator in vascular tumor development and progression based on our mouse model and human clinical samples.

The etiology of lymphangiosarcoma is largely unknown. Chronic lymphedema is a widely recognized risk factor for lymphangiosarcomas. Typically, lymphangiosarcomas occurs in women who have undergone radical mastectomy for breast cancer with or without radiation therapy and have had chronic lymphedema for many years, i.e., the so-called Stewart-Treves syndrome. Other chronic lymphedema resulting from a congenital, idiopathic, traumatic, or infectious cause also predisposes the patients to lymphangiosarcoma. The rationale for this association is a matter of controversy. Some have suggested that lymphostasis in the lymphedematous regions may produce local immunodeficiency, which is unable to perform immunologic surveillance of normally occurring mutant cell populations.3 Interestingly, in our mouse model, as early as one month after Tsc1 gene knockout, some mice showed signs of edema with swelling in the paws and tails, which recapitulate lymphedema in the human syndrome. Pathological analysis of mouse skin sections showed that the number of lymphatic vessels was dramatically increased in mutant mice, compared to that in control mice. The growth and proliferation of obstructed lymphatics is due to the constitutive activation of mTORC1 in lymphatic endothelial cells. So our results suggest that lymphedema could be the consequence of obstructed irregular lymphatics, instead of a cause of lymphangiosarcoma. Nevertheless, lymphedema could still be considered as an early sign and/or risk factor of lymphangiosarcoma for early diagnosis.

We also observed that at the early stage of tumor progression, the prominent cutaneous vascular lesions are vascular malformations, which are thin-walled, well-differentiated lymphatic channels, containing various amount of blood. Moreover, these malformations had a low Ki67 proliferative index of 2%–3%, compared to nearly 0% for normal endothelial cells. They may have been lesions that would develop into malignant lymphangiosarcomas at a later time in mice, with solid tumor masses and an elevated Ki67 index of about 30%. The early vascular malformed endothelial cells still showed the normal flat shape, whereas those at the late cancerous stage have an atypical plump appearance with vesicular nuclei and prominent nucleoli. These data from our mouse model suggest that there is a developmental scheme for the disease from atypical lymphatic hyperplasia to lymphatic malformations and onto lymphangiosarcoma. Our data also suggest that the establishment of lymphangiosarcoma accompanied by a second hit, i.e., other mutations may play roles for the transition. Similar transformation from vascular malformations to malignant vascular tumors have been seen in some clinical observations.4, 5

Our data indicate that sustained hyper-activation of mTORC1 signaling is necessary for the vascular tumor growth and maintenance in the mouse model. Rapamycin, a potent mTORC1 inhibitor, effectively leads to the regression of established tumors in the mutant mice. These mechanistic insights from our model raise the immediate question of whether rapamycin or other mTORC1 inhibitors will be effective for the treatment of human vascular tumors. Several clinic trials showed that rapamycin is safe and potentially effective for treating patients with vascular malformations and tumors.6,7 Interestingly, the patients who responded well to rapamycin in trials had vascular anomalies with significant lymphatic components. In our mouse model, we found all the cutaneous vascular anomalies were from lymphatics, although the recombinant Cre used in our study targets both blood endothelial cells and lymphatic endothelial cells in cutaneous region. Nevertheless, the mutant mice also developed liver vascular anomalies composed with pure blood endothelial cells, and the liver vascular lesions responded to rapamycin treatment too. Whether rapamycin could only be used for lymphatic derived cutaneous vascular anomalies still remained as an important question.

Our studies provide significant insights into the molecular and cellular mechanisms of lymphangiosarcoma. Additionally, the establishment of a mouse model with a well-defined, inciting molecular alteration of relevance to human disease provides a powerful tool for testing novel therapeutic methods.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.


[1] Pawlik TM, et al. Cancer 2003; 98:1716-26; PMID:14534889; [PubMed] [Cross Ref]
[2] Sun S, et al. Cancer Cell 2015; 28:758-72; [PMC free article] [PubMed] [Cross Ref]
[3] Goldblum JR, et al. Enzinger and Weiss's soft tissue tumors. Philadelphia, PA: Saunders/Elsevier; 2014
[4] Quarmyne MO, et al. Soc Clin Oncol 2012; 30:e294-8; [PubMed] [Cross Ref]
[5] Jeng MR, et al. Pediat Blood Cancer 2014; 61:2115-7; PMID:24740626; [PubMed] [Cross Ref]
[6] Lackner H, et al. Eur J Pediatr 2015; 174:1579–84; PMID:2604070521445948 [PubMed]
[7] Hammill AM, et al. Pediat Blood Cancer 2011; 57:1018-24; PMID:21445948; [PubMed] [Cross Ref]

Articles from Cell Cycle are provided here courtesy of Taylor & Francis