Discussion
In this retrospective study, we present our single center long-term experience of allogeneic HSCT in six patients with MNGIE disease, including the first patient (patient 1) who underwent HSCT. Despite obvious statistical limitations of this relatively small cohort and heterogeneity of pre- and post-HCST characteristics, important lessons can be learned. Detailed assessment of the long-term follow-up of our first patient as compared to the other five as well as to previously reported patients allows us to reinforce previously suggested guidelines and recent position paper on diagnosis, prognosis and treatment of MNGIE, when carefully considering selected patients for this procedure.
The first multicenter experience of allogeneic HSCT for MNGIE published by Halter summarized HSCT in 24 patients with a follow up ranging from 27 months to 8.5 years. The mortality of 15 patients out of a total of 24 was substantial. Overall, they concluded that allogenic HSCT can alter the natural course of MNGIE and therefore should be considered as a therapeutic option for carefully selected patients and that the clinical status of the surviving patients improved significantly over time, even in patients with very severe disease manifestations. This is not inline with our experience showing that patients who exhibited symptoms of progressive disease at the time of transplantation did not improve to any significant extent following HSCT and their GI-related symptoms and complications continued to progress. An important parameter to be followed is weight gain, which occurred very slowly even in patients who resumed oral feeding and were able to ingest high caloric nutrients. In fact, only patient 1, who was transplanted early presumably before the development of nonreversible changes, has started gaining weight years post-transplant.
Age at diagnosis and timing of transplantation are of paramount importance. HSCT should be carried out as early as possible, to maximize the potential for recovery as well as minimizing the risks associated with HCST. In our cohort, the best outcome was achieved in patient 1 who was diagnosed early at age 8 years owing to a family history of two affected cousins (fig. 1A) and had only mild symptoms at the time of HSCT. Patient 3, who was 27 years of age and had severe GI manifestations at the time of transplant, died 12 days post-transplant. This is keeping with the perception that late transplantation is a significant risk factor for HSCT. It is possible, as was previously suggested, and as featured by our patient 1, that longer follow-up is required to evaluate the outcome of these patients. Since patients with MNGIE are born with this progressive mitochondrial disorder, and age of HSCT is of paramount importance, including MNGIE in panels of newborn screening should be considered.
According to the previous consensus proposal for a standardized approach for allogeneic HSCT, careful donor selection is important, with an HLA-identical sibling as the first preferred option, with both non-carriers and heterozygous TYMP mutation carriers as potential donors. All our patients were transplanted from a fully 10/10 HLA-matched related donor. The family trees of the two kindred, depicted in Figures 1A and 1B, highlight the role of consanguineous marriages in the causation of fatal, ultra-rare autosomal recessive disorders such as MNGIE. On the other hand, the detailed plotting of the consanguineous pedigrees enabled tracking and identification of 10/10 HLA-matched donors for all transplanted patients. In general, carrier status of the HSCT donor in inborn errors of metabolism is considered to be a significant factor by influencing post-transplant enzyme levels which appear to be important in determining long-term results, including neurocognitive outcome in some disorders. In our group, four patients were transplanted from heterozygous carriers (patients 1,2,5,6). It is possible that the slow recovery of patient no 1 is related to the fact that she was transplanted from her heterozygous mother, but on the other hand her favorable outcome over time could indicate that carrier status is only one of several factors to be taken into consideration. Therefore, whilst a fully matched related donor who is not a carrier is the preferred donor, our experience suggests that in selected patients, for whom a carrier is the only available fully-matched related donor, it is acceptable to proceed with HSCT from a carrier.
Furthermore, based on published data regarding matched unrelated donors (Tables 3 and 4), 10 out of 12 published patients with MNGIE who underwent matched unrelated donor HSCT engrafted and developed GVHD, 50% of them having grade 3-4 acute GVHD and dying from GVHD or infections. Another known factor for developing GVHD is the source of cells. Only patient 1 received peripheral stem cell collection, which probably contributed to her GVHD, whereas 4/5 patients who received bone marrow as a source of stem cells (excluding patient 3 who died on day +12 post HSCT) did not develop GVHD.
Toxicity of the conditioning intensity regimen needs to be balanced against the risk of secondary engraftment failure while using reduced intensity and taking into consideration the toxic effect of specific medications on the mitochondria. Favorable results can be achieved by careful selection of HLA-matched donors, an adequate number of cells in the graft and strict monitoring of chimerism to detect graft failure early and guide therapeutic approaches to prevent graft loss. In our cohort, four of our patients (patients 3-6) received a reduced to treosulfan-based conditioning regimen instead of a busulfan-based regimen to try and minimize GI toxicity for symptomatic patients. Patient 3 started engraftment with full donor chimerism but died on day 12 post-transplantation from multiple organ failure. Patient 4 had full donor chimerism 3 years post-HSCT and did not have any significant organ toxicity. In contrast, our young patients 5 and 6 engrafted with mixed chimerism and during the first year post-transplant had progressively decreasing donor cells leading to secondary engraftment failure despite repeated infusions of donor lymphocytes. Five out of six published transplanted patients who underwent a non-busulfan-based conditioning regimen or cord blood source had primary or secondary engraftment failure (Tables 3 and 4) in contrast to patients who were conditioned with a busulfan-based regimen. Thus, according to published literature and our own experience, reduced intensity conditioning is not sufficient for these patients with normal bone marrow and a healthy immune system. The conditioning regimen for patients with MNGIE should be busulfan-based with addition of anti-thymoglobulin to achieve a higher percentage of donor chimerism. We conclude that the risk of secondary engraftment failure while using a reduced toxicity conditioning regimen outweighs the risk of organ toxicity associated with a busulfan-based regimen, bearing in mind that a second HSCT might not be possible due to disease progression, as happened with two of our patients (5 and 6).
According to published data, overall survival of transplanted patient with MNGIE is about 37%, but almost all these patients were older than 16 years of ageand had severe manifestations of their disease (Table 3 and 4). Survival in our cohort was 66%.
Overall, our experience reinforces the significance of HSCT timing with regard to HSCT-related risks, as well as a prognostic factor for transplant outcome. We conclude that severe GI symptoms are mostly irreversible and constitute a poor prognostic indicator both short and long-term. HSCT should therefore be carefully considered in those patients.
In conclusion, our present study emphasizes the short and long-term outcome of allogenic bone marrow transplantation as a curative option for patients with MNGIE if the following factors are carefully followed: early diagnosis before irreversible GI symptoms occur, careful selection of HLA-matched related donor and the use of a busulfan-based conditioning regimen. Further studies might lead to other promising therapeutic options and future developments such as newborn screening may contribute to early diagnosis.
Conflict-of-interest disclosure :
All authors declare no conflict of interest.