The incidence and clinical characteristics of liver cGvHD onset during calcineurin inhibitor tapering

Chronic graft-versus-host disease (cGvHD) remains a prominent issue in patients with hematologic malignancies receiving allogeneic hematopoietic stem cell transplantation (allo-HSCT), and the liver is particularly affected. The use of calcineurin inhibitors (CIs) has proven to be effective in combating the onset of liver cGvHD, however little information is available on the incidence or characteristics of cGvHD during the period of CI tapering or withdrawal. As such, Yoshimura et al. performed a retrospective study, recently published in the International Journal of Hematology, in which they investigated cGvHD of the liver in patients withdrawing from cyclosporin-A or tacrolimus. We summarize key results below.¹

Methods

This study investigated patients who underwent a first or second allo-HSCT between 2007 and 2016. Tapering/withdrawal of CIs was defined as a reduction of cyclosporin-A to ≤40 mg/day or a reduction in tacrolimus to ≤0.4 mg/day. The tapering schedule for CIs included a 5% dose reduction every week starting from Day 50 and concluding with complete withdrawal by Day 180. In high-risk or relapsed patients, a dose reduction of 10% was applied.

Exclusion criteria

The following patients were excluded from the study:

• Patients who died before Day 100 following allo-HSCT.
• Patients who did not have their CI dose lowered.
• Patients who were switched to alternative immunosuppressant therapy.

Defining cGvHD

Liver injury was defined by elevation of any liver enzymes to more than two times the upper limit of the normal range. Alternatively, if the initial liver enzymes at the start of the low-dose CI period were more than two times the upper limit, liver injury was defined by elevation of enzymes by five times the upper limit. Liver injuries were defined as proven, probable, and possible, with the first two labels used to define cGvHD and suspected cGvHD, and the latter classified as no cGvHD.

Results

Baseline characteristics of the patients eligible for analysis (N = 242) are summarized in Table 1.

Table 1. Baseline characteristics of eligible patients*

AA, aplastic anemia; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; ATL, adult T-cell leukemia/lymphoma; CML, chronic myeloid leukemia; MDS, myelodysplastic syndromes; ML, malignant lymphoma; MM, multiple myeloma; MPN, myeloproliferative neoplasms. *Adapted from Yoshimura et al.1
Characteristic	N = 242
Median age, years (range)	45 (15–68)
Female, %	42.1
Diagnosis, %
AML	41.3
ALL	19.4
CML	2.5
ATL	3.3
AA	6.6
MDS	14.0
MPN	1.2
ML	5.0
MM	2.9
Other	3.7
High-risk disease, %	22.7
Donor type, %
Matched sibling	28.9
Matched unrelated	38.4
Mismatched sibling	3.3
Mismatched unrelated	9.9
Cord blood	6.2
Haploidentical sibling	13.2
Myeloablative conditioning, %	59.9
Female donor to male recipient, %	23.6

A total of 182 patients had no liver cGvHD, including 55 patients with possible cGvHD and 127 with no liver injury. While 60 patients experienced either proven (n = 2) or probable cGvHD (n = 58). Liver cGvHD occurred at a median of 187.5 days after HSCT.

Risk factors for cGvHD during CI tapering

Following a multivariate analysis, the authors identified that the risk factors for developing liver cGvHD during the low-dose period of CI included a donor age >40 years, myeloablative conditioning, female donor to male recipient, and recipient seropositivity for herpes simplex virus (Table 2).

Table 2. Multivariate analysis of risk factors associated with liver cGvHD during the low-dose CI period*

cGvHD, chronic graft-versus-host disease; CI, calcineurin inhibitor; HR, hazard ratio; HSV, herpes simplex virus. *Adapted from Yoshimura et al.1
Risk factor	HR	95% confidence interval	p value
Donor age >40 years	2.2	1.16–4.19	0.02
Myeloablative conditioning	2.19	1.12–4.30	0.02
Female donor to male recipient	2.53	1.25–5.09	<0.01
Recipient seropositivity for HSV	2.98	1.33–6.70	<0.01

cGvHD onset during CI tapering versus outside of tapering

A total of eight patients had onset of cGvHD in periods outside of the low-dose CI tapering. Yoshimura et al. compared these eight patients with the 60 patients who had cGvHD onset during CI tapering and found that patients with cGvHD during low-dose CI had a higher peak of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST), and generally higher levels of alkaline phosphatase (ALP). Such elevation is in line with hepatitis-like changes observed in liver GvHD. There were no significant differences for other liver enzymes, such as total bilirubin, or in the ratio of ALP/ALT, which was in contrast to the investigators’ hypothesis for an increased ratio associated with low CI dose.¹

Treatment response

Out of the 60 patients with low-dose CI-induced cGvHD, 18 received no treatment, two received chemotherapy for leukemic transformation, 27 were given further CI treatment, and 13 were given corticosteroids.

• In patients given CI treatment, 21 responded to therapy but five required additional systemic corticosteroid therapy, while one patient was unavailable for analysis.
  • A univariate analysis for patients receiving further CI therapy showed no risk factors were significant associated with treatment failure.
• For patients given corticosteroids, eight patients responded. For the five patients who did not respond, two patients died from infection related to cGvHD treatment and three had disease relapse during corticosteroid treatment.
  • A univariate analysis for patients receiving corticosteroid treatment showed total bilirubin level at the initiation of corticosteroid administration was a risk factor for treatment failure (hazard ratio, 1.5; 95% confidence interval, 0.55–18.04; p = 0.029).

Conclusion

This retrospective analysis identified donor age >40 years, myeloablative conditioning, female donor to male recipient, and recipient seropositivity for herpes simplex virus as significant risk factors for developing cGvHD during low-dose CI periods. The elevation of AST/ALT was associated with cGvHD onset during low-dose CI. Finally, elevated total bilirubin at the start of corticosteroid treatment for cGvHD was significantly associated with increased risk of treatment failure. Yoshimura et al. highlighted limitations of the study, including its retrospective nature, liver cGvHD not being diagnosed by the National Institute of Health criteria, lack of CI level monitoring, and the low sample numbers in the treatment analysis.