Allogeneic Hematopoietic Stem Cell Transplantation in Advanced Stage Mycosis Fungoides and Sézary Syndrome

A Concise Review

William T. Johnson; Reetu Mukherji; Saritha Kartan; Neda Nikbakht; Pierluigi Porcu; Onder Alpdogan


Chin Clin Oncol. 2019;8(1) 

In This Article

What We Have Learned From These and Prior Publications Utilizing HSCT for CTCL

Choice of Conditioning Regimen

As previously described RIC has significantly decreased the NRM in NHL.[17,18] Duarte et al. showed higher NRM and poorer OS with MAC with no impact on relapse-free survival. Using only RIC in their cohort of 47 patients, Hosing et al.[49] demonstrated an estimated 4-year OS of 51%. According to de Masson et al., the choice of conditioning regimen had no statistically significant impact on TRM or efficacy. Ten out of the 12 patients described by Paralkar et al.[45] received RIC with 42% alive with sustained clinical responses at a median 22-month follow-up. At a median follow-up of 32 months Shiratori et al. reported an estimated 3-year OS of 85.7% and PFS of 44.4% in which all nine patients received RIC.[50] Lechowicz et al.[48] reported the data from the CIBMT from 2000–2009. In total 83 patients received RIC (64%). OS was 56% and 41% at 1 and 3 years, respectively for RIC and 51% and 31%, respectively, for MAC. NRM at 1 year was 19% and at 5 years was 22%. NRM did not differ significantly between RIC and MAC cohorts. These data suggest that MAC has not improved outcomes and is not necessary for patients with advanced stage CTCL.

T-cell Depletion

de Masson and colleagues from the FSBMT group reported the use of ATG in 43% of their patients (which was driven by local protocol and was not associated with any specific donor type) was the only factor significantly associated with a decreased TRM (HR 1.10−7; 95% CI, 4.10−8 to 2.10−7; P<0.001) as all 6 of the patients whom died during autologous stem cell transplantation (ASCT) did not receive ATG. On the contrary ATG was the only factor associated with increased risk of REL (HR 4.8; 95% CI, 1.8–12.9; P=0.002). The EBMT groups also showed that receiving T-cell depletion carried a higher risk of REL (HR 2.48; 95% CI, 1.15–5.35; P=0.0207).[54] However, Hosing et al.[49] were not able to show a similar increased risk of REL with ATG. Yet, in their cohort 42 (89%) patients received TBSEB radiation therapy just prior to HSCT, of whom 25 (60%) achieved a CR in the skin. We conclude with the EBMT and French group data that in vivo T-cell depletion may affect the outcome of the disease.

Disease Status Prior to Transplant

Very good PR or CR prior to HSCT was the most important prognostic factor for increased PFS in univariate analysis (HR 0.3; 95% CI, 0.1–0.8; P=0.01) in the report by de Masson. Duarte et al.[47] reported only 10/60 patients had <10% residual disease at time of HSCT. However, they did report patients with a lower disease burden prior to HSCT had a lower cumulative incidence of REL compared with those with a higher disease burden (P=0.04). TBSEB was administered prior to transplantation in their prospective protocol in the reports by Duvic et al.[53] and Hosing et al.[49] Although they did not comment on the statistical significance of having a CR prior to transplant, they did report a post-HSCT CR rate of 58%. With this data we may conclude that minimal disease prior to transplant is most ideal, but not a strict requirement to harness the graft-versus-lymphoma effect.

Timing of Transplant

It is important to note again the additional observations made by Duarte et al.[47] in that patients with what they defined as "advanced disease" phase (the definition of which is detailed above) at HSCT had an increased risk of REL [HR 3.07; 95% CI, 1.15–8.20; P=0.0249], lower PFS (HR 3.26; 95% CI, 1.43–7.47; P=0.0051) and worse OS [HR 3.72; 95% CI, 1.49–9.30; P=0.0049] compared to "early disease" phase. Additionally, Cudillo et al.[52] found benefit with earlier transplants. The OS at both 1 and 10 years was 88% (95% CI, 67–100%) for patients transplanted less than 46 months from diagnosis vs. 37% (95% CI, 15–92%) and 25% (95% CI, 8–83%), respectively for patients transplanted greater than 46-months from diagnosis (log-rank P<0.04) [HR 7.26; 95% CI, 0.86–60.95; P<0.068)]. Likewise, they also found DFS at 1 year and 10 years benefited transplanting at an "early disease" phase. In the meta-analysis by Lechowicz et al.[48] of 129 HSCT recipients reported to the CIBMTR, 49% of patients were transplanted >36 months from their diagnosis. They did not comment on the significance. We believe that early transplant consultation may improve outcome of disease in this specific patient population.

Graft vs. Lymphoma

Paralkar et al.[45] detailed their experience on two relapses in their cohort and the argument of graft-versus-lymphoma (Table 2). The first patient whom relapsed 10 months after HSCT received DLI (2.2×108 nucleated cells/kg) at first relapse and achieved CR that persisted for 26 months. The second patient achieved second CR with discontinuation of immunosuppression and remained in CR. None of these patients had received additional treatments for their CTCL. In the EBMT report by Duarte et al.,[47] 10/17 patients whom received DLI responded including 8 CR, confirming allogeneic response would provide anti-lymphoma activity. Cudillo et al.[52] had a 50% success rate with REL treated with DLI. Hosing et al.[49] observed secondary responses in 8 of 22 relapses with immunomodulation post-transplant. Herbert et al. published their successes in three patients whom relapsed after HSCT with decreasing immunosuppression and the use of DLI.[55] The authors made an important observation in that all their patients relapsed with the high grade LCT-MF disease they had prior to transplant. This observation had not been described previously, and the authors suggest the graft-versus-lymphoma benefits are greatest pre-LCT. This theory is supported by the poor 4-year PFS in patients with MF with LCT of 8.6% in the Hosing prospective cohort, and supports the literature arguing for earlier transplant. Certainly, the use of DLI can be fraught with balancing GVHD in many cases, but this literature is quite clear that immunomodulation can be very successful and should be utilized in patients with REL post-HSCT with no to minimal evidence of GVHD.

Graft vs. Host Disease

Paralkar et al.[45] reported GVHD developing in 9 of 12 patients with 4 developing grade III–IV GVHD (Table 2). Six patents developed GVHD of the skin half of which were grade III, and GVHD was the cause of death in one patient. de Masson reported a 70% occurrence rate of acute GVHD of which nearly half developed grade II or higher disease and chronic GVHD developed in 15 patients. Seven out of the 8 patients described by Molina et al. developed GVHD.[43] Shiratori's group document acute GVHD in 8/9 transplants and chronic GVHD in 7/9 patients. Meanwhile, Hosing et al.[49] reported much less rates of GVHD with an incidence of grade II–IV acute GVHD of 40% with an incidence of chronic GVHD of 28%. The cumulative incidence of grade III–IV acute GVHD was 10%. The cumulative incidence of chronic GVHD was 28%. They hypothesized the routine use of TBSEB may lessen the development of cutaneous GVHD.

Relapse Rates

In the CIBMTR meta-analysis by Lechowicz et al.[48] of the REL rate was 50% at 1-year and 61% at 5 years, suggesting that the majority of REL occur within the first year post-transplant. Similarly, Hosing et al.[49] reported 50% REL with the majority occurring within 6 months of transplant. Duarte et al.[47] showed a 45% REL at a median of 3.8 months post-transplant, but only two relapses occurred beyond 2 years. de Masson et al.[46] totaled 19 (51%) relapses with a median time to progression of 10 weeks, and 90% of all REL occurred within the first year. Relapses occurred in 5/9 patients described in the Shiratori cohort at a median of 45 days.[50] We conclude that as much as half of the patients with CTCL may have a relapse after HSCT in the first year after transplant and a decreased tumor burden prior to the transplant may improve the post-transplant relapse risk.


There are no randomized trials comparing HSCT vs. systemic therapies. We must then assess meaningful OS, PFS, and REL in comparison with historical controls, and few of these retrospective analysis report beyond 5 years. Duarte et al.[47] in their extended analysis reported OS of 46% and 44% at 5 and 7 years, respectively. The Kaplan-Meier estimate of OS at 4 years was 51% reported by Hosing et al.[49] In the CIBMTR analysis by Lechowicz et al.[48] the OS was 56% and 41% at 1 and 3 years respectively for RIC and 51% and 31% respectively for MAC. de Masson et al.[46] estimated 1- and 2-year OS rates were 65% and 57%, respectively. For Cudillo et al.[52] the probability of OS was 61% (95% CI, 40–91%) and 54% (95% CI, 33–86%) at 1 and 10 years post-transplant, respectively. Compared to the most recently published data on 1,275 patients with advanced stage disease (IIB–IV) where the median projected 5-year OS rates was 52%, the results of HSCT with a curative intent are encouraging.[7]