Impact of Pre-existing Invasive Aspergillosis on Allogeneic Stem Cell Transplantation

Background \& Rationale. In patients with pre-existing invasive aspergillosis allo-HSCT is feasible without progression of fungal infection. However, the influence of invasive pulmonary aspergillosis (IA) on transplant related complications and on long term survival has not been investigated in a larger patient cohort under current conditions.

Recently the IDWP and ALWP performed a retrospective analysis on the impact of preexisting aspergillosis on allo-HSCT outcome.2 In summary, there was a trend towards impaired outcome of allo-HSCT in patients with prior IA but there was no significant impact on important allo-HSCT transplant outcomes, such as survival, GVHD and relapse. The data suggest that a history of IA should not generally be considered a contraindication for allo-HSCT. To be able to more precisely investigate the impact of IA on allo-HSCT, a non-interventional prospective study is needed.

Primary objective:

To determine if pre-existing IA influences non-relapse mortality after allo-HSCT

Secondary objectives:

• To determine if pre-existing IA influences:
• relapse free survival
• overall survival
• incidence and severity of GVHD
• incidence of relapse
• incidence of IA post allo-HSCT for the subgroup of transplant with previous IA
• progression of IA

Research design:

Prospective study. Study recruitment is expected to start by May 1st, 2016. It is expected that recruitment will be closed by October 31st 2017. Follow up will be till one year after transplant.

Items:

Data from Med A form, Med C form for all patients, Aspergillus form for patients with probable/proven IA.

Endpoint(s):

Primary endpoint: 1-year non-relapse mortality cumulative incidence

Secondary endpoints:

• 1-year relapse free survival
• 1-year overall survival
• 1-year incidence and severity of GVHD
• 1-year incidence of relapse
• status of IA (before conditioning and at 1 year)

Study population

• First allo-HSCT in patients with AML or
• First allo-HSCT in patients with ALL or
• First allo-HSCT in patients with MDS

Cohort 1: History of probable or proven invasive aspergillosis Cohort 2: No History of probable or proven invasive aspergillosis

Cases: Prior IA = probable/proven IA according EORTC 2008 criteria (Cohort 1) Controls: No prior proven probable IA = all other patients = those really negative for IA and those with possible IA (Cohort 2)

Sample size Assuming a 20% incidence of non relapse mortality in patients without proven or probable history of IA, a total of 2100 HSCTs will be needed in order to test the hypothesis that the incidence of non relapse mortality in cohort 1 (HSCTs with previous history of proven or probable IA) is not higher than cohort 2 (HSCTs without history of IA or with previous history of possible IA) by more than a specified margin of 10%. The estimated proportion of HSCTs in cohort 1 is 5%, thus 105 and 1995 HSCTs will be needed in cohort 1 and cohort 2, respectively, considering an alpha=0.05, a beta=0.2.

Data Collection \& Statistical Analysis Plan:

(List all research variables to be collected and list all outcome variables to be analysed, give a brief description of the method of analysis (in collaboration with the EBMT statistician) All data collection will be performed by the IDWP Data Office (Leiden) according to EBMT guidelines.

Primary endpoint: non relapse mortality The non relapse mortality will be estimated using the cumulative incidence method. Death due to transplant will be considered as an event, whilst relapse of the underlying disease will be considered as competing events. Patients alive at the end of the follow-up will be censored at this date. Cohort 1, vs. Cohort 2 will be compared by the Gray test.

A cause specific Cox model will be performed in order to estimate the risk of dying for Cohort 1 respect Cohort 2.

The following variables will enter the multivariate model as possible confounders: age (as continuous variable), gender (M vs. F), underlying disease (ALL vs. AML/MDS), status at SCT (1st CR vs. ≥ 2 CR vs. Prim Refr/noCR), time from diagnosis to SCT (as continuous variable), donor type (sibling vs. UD vs. Haplo), source of SCT (BM vs. PB vs. CB), donor age, d/r gender match, d/r CMV status, conditioning regimens (MAC vs. RIC vs. TBI), type of immunosuppression (in vivo T depletion y/n, in vitro T-depletion y/n), DLI post SCT (y/n), centre.

Secondary endpoints Relapse free survival and overall survival will be estimated by the Kaplan-Meier methods. Incidence of GvHD and incidence of relapse will be estimated by the cumulative incidence methods. A cause specific Cox model will be estimated to compare Cohort 1 and Cohort 2; it will be adjusted by the confounders analyzed also in the primary endpoint.

The severity of GvHD will be described by descriptive statistics. Frequencies and percentages will be use for categorical variables, whilst median, range, mean and standard deviation will be computed for continuous variables.

Additional descriptive statistic will be separately performed in cohort 2. Main characteristics of patients will be also described.