DISCUSSION
This report describes a series of patients with no evidence of a primary lesion, and therefore considered as having RMS with unknown primary site. This is a rare condition and represents a diagnostic and therapeutic challenge.
The diagnostic difficulties clearly emerge from our study: despite the usual diagnostic workup (including FDG-PET/CT in 8 cases), clinicians were unable to identify a clear primary lesion in these 10 patients registered in the EpSSG MTS2008 protocol. We know that small tumor lesions localized in the extremities may give widespread dissemination (4), unfortunately 3 patients in our series were not fully investigated in this sense and we cannot exclude they had an unrecognized extremity primary tumor. Careful clinical examination and whole-body MRI or FDG-PET-CT/MRI scanning is suggested for all patients with disseminated RMS if the primary site is not immediately apparent. Another possible limitation of our study concerns the inclusion of a child with disseminated leptomeningeal disease and only a cytological diagnosis: we decided to include this patient in our series because he was considered and treated like a case of metastatic RMS.
The frequent involvement of the bone marrow (it was the only organ involved in 2 cases) may support the hypothesis that it could be the primary tumor site, rather than a site of dissemination. We cannot prove a primary bone or bone marrow origin, but recent studies found that RMS can originate from an aberrant development of non-myogenic cells (5). This would justify the initiation of RMS at sites like the bone marrow and bone.
The treatment of patients with disseminated disease but unknown primary tumor presents a series of obstacles. Local control measures are fundamental in the treatment of RMS, but the absence of a primary lesion often precludes the use of local surgery or radiotherapy, and it is difficult to administer radiotherapy as the disease is frequently widely disseminated and when bone marrow is involved. This is an important aspect as our recent experience suggests that radiation is associated with improved survival in metastatic RMS (6). Chemotherapy was the only treatment modality used in most of our patients, and was unfortunately not enough: only 3 patients in our series are still alive. The predictably poor prognosis may explain why clinicians chose to perform an allogeneic bone marrow transplant in 3 patients, although there is no clear evidence to indicate that it is effective in RMS. Previous trials have shown high dose chemotherapy followed by autologous transplant does not add to survival in metastatic patients (7-8). Since only one of these patients is still alive, we can draw no firm conclusions regarding potential efficacy of this procedure, which carries a small but significant risk of mortality.
The poor prognosis of this group of patients is confirmed by a review of the medical literature. We found 16 publications (from 1976 to the present) describing 22 cases of children with RMS and an unknown primary tumor site (9-24). The characteristics of these patients are summarized in Table 2. As in our series, this condition seemed to be more typical of the fusion-positive alveolar subtype of RMS, and occurred mainly in adolescents. The most often involved sites were the bone marrow, bone and distant lymph nodes. A lymphoproliferative disease was often suspected, and the diagnosis was made on analyzing the bone marrow. All patients were treated with chemotherapy, while radiotherapy and surgery were performed in a few cases. Only 3 patients out of 17 with outcome data were reportedly alive with no sign of disease, but the follow-up was short.
In conclusion, metastatic RMS with no clear evidence of a primary site is a rare condition, more likely to affect adolescents and to involve the fusion-positive alveolar histotype. Current treatments are rarely able to cure these patients, who should be included in investigational trials along with other very high-risk, metastatic patient groups.