Thyroid lymphomas in human thyroid tissue with autoimmune thyroid disease xenografted in severe combined immunodeficient mice

Malignant lymphoma of the thyroid (MLT) frequently arises in patients with a background of Hashimoto's thyroiditis (HT); however, the mechanisms underlying this chain of events are unknown, and there has been no experimental model. Recently, the development of malignant lymphoma has been reported to occur in peripheral blood lymphocytes engrafted into severe combined immunodeficient (SCID) mice. We xenografted human thyroid tissue from patients with HT or Graves disease (GD) into SCID mice to determine the frequency and nature of MLT in these grafts. Human thyroid tissues ( 12 HT, 1 GD, and 15 from normal [paranodular] tissue) were xenografted into 72 mice (43 mice with HT or GD tissue) within 2 hours after human surgery. Human peripheral blood mononuclear cells (PBMC; 4 autologous HT, I allogeneic HT, and 1 allogeneic GD) were injected intraperitoneally into 6 of the latter 43 mice. In addition, 16 additional SCID mice received normal PBMC injections (alone). The mice were killed 6 to 20 weeks after xenografting. In 4 of 33 SCID mice bearing HT thyroid grafts (without addition of PBMC), MLT developed in the HT graft between 8 and 16 weeks after xenografting. In addition, one spleen of a mouse xenografted with GD tissue alone developed a human malignant lymphoma, although the xenografted thyroid in that mouse did not manifest lymphoma. One additional mouse xenografted with HT thyroid tissue and allogeneic HT PBMC developed malignant lymphoma of both the xenografted thyroid and the mouse spleen. In this mouse, the clonality of these lesions in the two organs was different: the thyroid showed restricted expression of immunoglobulin A (IgA) kappa, whereas the spleen exhibited lambda light chain restriction. One human MLT was removed from a SCID mouse, and equal halves were rexenografted into a nude mouse and another SCID mouse. Thyroid antibodies and IgG levels increased in the second SCID mouse, and the MLT survived; in the nude mouse, however, thyroid antibodies and IgG gradually disappeared, and the MLT regressed, virtually to normal. No MLTs were found in the normal human thyroid xenografts. In SCID mice receiving normal PBMC alone, lymphomas tended to develop when more than 35 x 10⁶ cells were engrafted (a number similar to that of the lymphocytes in the HT xenografts); thus, the MLTs may reflect merely the numbers (and perhaps density) of human lymphocytes present in the xenografts. It is possible that committment of many of the HT-infiltrating lymphocytes to the thyroid might add an additional factor. However, whether this model will prove useful to study the possible transition of HT to MLT remains problematic.