Donor Bone Marrow-Derived Stem Cells Contribute to Oral Squamous Cell Carcinoma Transformation in a Recipient After Hematopoietic Stem Cell Transplantation

R ecently, the increasing number of cases suffering from secondary solid malignancies developing in recipients after allogeneic hematopoietic stem cell transplantation (allo-HSCT) has attracted more attention, as less donor-derived solid malignancies after allo-HSCT have been reported [1,2]. To date, such kind of malignancy is regarded as a rare event due partly to the lack of attention, but this model implies that bone marrow-derived stem cells can contribute to carcinogenesis, which greatly alters our recognition of cancer origination and formation as well as therapeutic and preventive strategies [3]. More accumulated clinical reports and tissue samples are especially desirable for further systemic analysis on such rare malignancies. However, the paucity of appropriate detecting system currently renders screening for donor-derived solid malignancies difficult in clinic, which ultimately impedes our understanding of the disease. Here we describe a case of donor-derived oral squamous cell carcinoma (OSCC) in a young adult patient after 2 years of allo-HSCT for acute leukemia treatment.

A 20-year-old Chinese single woman was diagnosed with acute lymphoblastic leukemia (ALL) in July 2004. The patient achieved complete remission after 2 courses of induction chemotherapy. Because of high-risk ALL, she ultimately underwent allogeneic peripheral blood stem cell transplantation (allo-PBSCT) from an human leukocyte antigen (HLA)-matched unrelated male donor in August 2006. The donor was from Buddhist Tzu Chi Stem Cells Center, Taiwan, China. He has no history of malignancy and is in good health. The myeloablative conditioning regimen consisted of busulfan and cyclophosphamide. Graft versus host disease (GVHD) prophylaxis consisted of cyclosporine A (CSA) 2.5 mg/kg/24 h, methotrexate 10 mg on days 1, 3, and 6, and mycophenolate mofetil 0.5 per day. A successful engraftment was documented on day +30 with both chromosomal and polymerase chain reaction–short tandem repeat (PCR-STR) analyses, and from then on the complete chimerism of peripheral blood mononuclear cells (PBMCs) and remission of the disease have been maintained. However, on day +12 the patient developed skin grade II acute GVHD. No specific treatment was given for her. Subsequently, she complicated with intestinal grade II acute GVHD characterized by recurrent diarrhea on day +49. Methylprednisolone 2 mg/kg/day and CSA 4 mg/kg/day were administered. Seven days later, acute GVHD was controlled gradually. After that, chronic GVHD with skin rash and oral mucositis accompanied by leukoplakia occurred after 5 months postengraftment. Prolonged immunosuppressive agents containing prednisolone and CSA were administered. In April 2008 (20 months postengraftment), the patient presented with a symptom of pain in the tongue, which worsened in the following 2 months. A firm mass was seen in the tongue and a biopsy was performed. The tissue specimen with hematoxylin and eosin staining showed a well-differentiated OSCC. The neoplastic cells were pleomorphic, with prominent nucleoli. Cell nests and keratinous pearls could be seen clearly (Fig. 1A, B), confirming the diagnosis of OSCC. Then, she was treated with mass excision in addition to supraomohyoid neck dissection immediately. The patient was in complete remission for her primary and secondary malignancy when she last followed up in the outpatient department 4 years post-transplantation.

OPEN IN VIEWER

FIG. 1.

Pathological and FISH examination. (A) A well-differentiated OSCC with cell nests and keratinous pearls can be seen clearly (HE,×100). (B) Higher magnification microscopic findings (HE,×400). (C) In the OSCC region, 87.6% of cells are P-CK (red) positive and CD45 negative (double-SP immunohistochemical staining). (D) In the adjacent normal region, no cells with positive P-CK and negative CD45 can be found (double-SP immunohistochemical staining). (E) In the OSCC region, 85.6% of cells contained Y (orange) chromosomes (FISH). (F) In the adjacent normal region, 89.6% of cells contain XX (green) chromosomes. Yellow arrowheads indicated CD45-positive (brown) white blood cells, which were from donors. White arrowheads in F indicated cells containing XY chromosomes, which were donor-derived leukocytes. Red arrowheads in E indicated tumor stromal cells of donor origin. FISH, fluorescence in situ hybridization; OSCC, oral squamous cell carcinoma; HE, hematoxylin and eosin; P-CK, pan cytokerin; SP, streptavidin perosidase. Color images available online at www.liebertonline.com/scd

We then further analyzed the paraffin-embedded neoplastic tissue sample to identify the carcinoma origin after informed content. Double-streptavidin perosidase (SP) immunohistochemical staining [CD45 and pan-cytokerin (P-CK)] and fluorescence in situ hybridization (FISH; centromeric X- and Y-chromosome probes) were performed to preliminarily identify the carcinoma origin. CD45 is a specific surface marker for leukocytes, which are all from donors after a successful myeloablative allo-HSCT, while well-differentiated OSCCs are positive for P-CK. Thus, CD45-negative and P-CK-positive cells are regarded as tumor cells. Additionally, in our case the donor is male and the recipient is female. Hence, cells containing an XX chromosome pattern are identified as recipient-derived cells and cells containing an XY chromosome pattern are identified as donor-derived cells. Thus, double-SP immunohistochemical staining and FISH analysis of the tissue specimen showed that in the tumor region the overall frequency of P-CK-positive, CD45-negative tumor cells and cells containing an XY chromosome were 87.6% and 85.6%, respectively (Fig. 1C, E), but in the adjacent normal region P-CK-positive, CD45-negative tumor cells and cells containing an XX chromosome pattern were 0% and 89.6%, respectively (Fig. 1D, F). These results preliminarily revealed that the OSCC tumor region including tumor cells, inflammation cells, and stromal cells was composed of transformed cellular components of both recipient and donor origins. Subsequently, we performed laser-captured microdissection (LCM) and PCR-STR analysis to confirm the carcinoma origin. Two serial paraffin sections were cut at 8 μm thickness. The first section was stained with methyl green and used for LCM based on the second section, which was double-stained with P-CK and CD45 antigens. A minimum of 5000 CD45-negative and P-CK-positive malignant cells were successively and separately laser-microdissected. PCR-STR was performed after DNA extraction. Fourteen highly polymorphic STR sequences were amplified: D3S1358, vWA, FGA, AMEL, D8S1179, D21S11, D18S51, D5S818, D13S317, D7S820, D16S539, TH01, TPOX, and CSF1PO. At the same time, archival PCR-STR analysis of PBMCs pretransplantation and 30 days post-transplantation as well as PBMCs from donors were obtained from Blood Center of Zhejiang Province, China. PCR-STR analysis of microdissected malignant cells compared with archival analysis of PBMCs pre- and post-transplantation as well as PBMCs from the donor confirmed the donor origin of OSCC cells (Fig. 2). When the patient was diagnosed with OSCC, PCR-STR analysis of PBMCs of the recipient showed full donor's chimerism. Additionally, morphology examination of bone marrow aspiration as well as minimal residual disease with flow cytometry indicated complete remission of the primary disease.

OPEN IN VIEWER

FIG. 2.

PCR-STR archival analysis of PBMCs pre- and post-transplantation as well as PBMCs from the donor (A). PCR-STR analysis of microdissected tumor cells (B). PCR-STR, polymerase chain reaction–short tandem repeat; PBMCs, peripheral blood mononuclear cells.

Donor-derived solid malignancies post-transplantation in clinic furnishes a unique disease model for carcinogenesis research. OSCC transformation of transplanted healthy allogeneic donor bone marrow-derived stem cells in recipient strongly implies that recipients might provide specific microenvironments supporting the malignancy transformation. Several factors maybe affect the host microenvironments. First, the patient suffered from severe oral injury caused by GVHD and drug toxicity resulted in persistent local chronic inflammation. Recent researches have highlighted the role of chronic inflammation in promoting marrow incorporation into cancer. Inflammation, tissue injury, and repair are the most possible mechanisms [4]. As a result, donor-derived stem cells can repopulate the injured oral tissues taking part in tissue repair. Second, prolonged immunosuppressive agents administered to the patient caused weakened immune surveillance. Altered immune status would increase not only the risk of primary disease relapse [5] but also the risk of solid malignancies. Third, it is possible that, after allo-HSCT, when an undifferentiated donor stem cell fuses with a mature differentiated cell under certain microenvironment, the resultant cell might retain the mature cell phenotype or lead to genetic instability, forming cancer-initiating cells or developing cancer stem cells [6]. We have not found any fusion karyotype by FISH analysis or mixed chimerism by PCR-STR analysis. Thus far no evidence of cell fusion has been confirmed in donor-derived solid malignancies. Donor-derived OSCC in our observation also indicates a novel malignant cell origination. The allograft in our patient is PBSCs, where the major cells are HSCs and the number of MSCs is very low [7]. HSCs are considered to have committed differentiation characteristics only to blood cells, whereas MSCs are multipotent cells, which can differentiate into a variety of cell types [8]. Therefore, in our case HSCs are possible to differentiate into OSCC cells. Up to date, altogether 14 cases of donor-derived carcinomas have been reported [9 –14]. Among them, the allograft of only one case was PBSCs. Additionally, in most reported cases only a small proportion of cancer cells were donor-derived, whereas in our case, we found that cancer cells constituting the OSCC were completely donor-derived. Factors involved in different proportions of donor-derived cancer cells in the carcinoma are elusive. Further, such donor-derived solid malignancies will also provide new therapeutic and preventive strategies for malignancies. Host microenvironment is an important factor associated with carcinogenesis, such as chronic inflammation, injury, altered immune status, and so on. Thus, recovery from the altered microenvironment or targeting the abnormal microenvironment would be of high significance for tumor prevention and therapy. On the other hand, the novel malignant cell origination that donor bone marrow-derived stem cells contribute to the transformation of malignancy would bring early monitoring methods and potential therapeutic targets for malignancy and lay the foundation for new targeted therapy of malignancy. However, it is still not certain that the detailed kind of stem cells and altered microenvironment contributed to malignancy transformation. Moreover, most bone marrow-derived solid malignancies have been reported only in allo-HSCT models. Other disease or animal models, more accumulated clinical reports, and tissue samples are needed in the future.

In conclusion, our study provides a novel mechanism involved in carcinoma transformation of donor bone marrow-derived stem cells after allo-HSCT. Detailed evaluation of such patients would provide valuable clues regarding malignancy origination and relapse, let alone a further knowledge of such solid secondary malignancies would be of high significance to develop strategies for cancer treatment and prevention.