Table 4 Genetic association studies of common complications after chemotherapy and/or HSCT

Infection

Mullighan et al⁶⁹

MBL2

Multicenter review of allo-HSCT patients and donors analyzed MBL2 coding or promoter-region mutations in relation to neutrophil recovery, febrile days, culture-confirmed infection; in multivariable analysis, MBL2 (d) coding mutations and absence of high-producing HYA promoter haplotype (r) associated with major bacterial infection after neutrophil recovery; this effect was independent of d/r genetic matching at MBL2 loci; MBL2 genotypes were not associated with aGVHD. No adjustment for factors other than genotype

Rocha et al⁶⁸

F _c γRIIa, IIIb MPO

Retrospective, single-center study of HLA-identical HSCT patients studied several host-defense and inflammatory-gene polymorphisms and risk of all infections within 180 days of follow-up; multivariable analysis (adjusting for factors other than genotype that affect hematopoietic recovery) found associations between F_cγRII a (r) genotype and risk of overall infections, MPO (d) genotype and severe bacterial infection, F_cγRIIIb (d) genotype and time to neutrophil recovery

Mucositis

Ulrich et al⁵⁰

MTHFR

Single-center study of CML patients who underwent allo-HSCT (Cy/TBI or Bu/Cy) and MTX prophylaxis for aGVHD analyzed oral mucositis index (OMI) score, engraftment time for platelets and neutrophils, bilirubin, and MTHFR activity in relation to MTHFR genotype; lower MTHFR activity (TT genotype) associated with slower platelet recovery and higher mean mucositis score

Robien et al⁵¹

Single-center study having overlap with above study population⁵⁰; excluded patients who required leucovorin rescue; multivariable analysis adjusting for conditioning regimen, pre-transplant vitamin use, age, BMI⩾25 showed higher mean mucositis scores associated with MTHFR genotype (TT)

Kalayoglu-Besisik et al⁵³

(Letter) Single-center Turkish study in HLA-identical HSCT recipients who got Bu/Cy conditioning, CSA and short-course MTX GVHD prophylaxis. Follow-up to engraftment or day 28 showed no difference in incidence of mucositis grade IV or less or in seventy of mucositis by MTHFR genotype.

Toffoli et al⁴⁹

Single-center study in refractory ovarian cancer patients treated with chronic low-dose, oral MTX±carboplatin analyzed plasma homocysteine and MTX levels and mucositis severity in relation to MTHFR polymorphisms; patients with grade III or IV mucositis had ↑ prevalence of the MTHFR low-producing genotype and hyperhomocysteinemia compared to baseline and patients with other MTHFR genotypes. MTX levels were similar across genotypes, but not evaluable in >50% of patients with toxicity.

Toffoli et al¹²⁸

(Letter) Case series in breast cancer patients undergoing adjuvant CMF chemotherapy showed increased toxicity in patients with MTHFR TT genotype

Chiusolo et al⁵⁴

Single-center review of MTHFR genotype in patients receiving maintenance chemotherapy including MTX 15–30 mg/m² weekly for ∼2 years for acute leukemia; 61 patients evaluable for MTX intolerance, defined as neutrophil count <0.5 × 10⁶/l, >2-fold ↑ in bilirubin and liver enzymes, and/or mucositis (assessed by OMI), nausea, vomiting, diarrhea, or fever requiring dose reduction or treatment delays; significantly more MTX intolerance seen in patients with TT vs other genotypes

HVOD

Kallianpur et al^32,33

HFE, CPSI

Single-center case-cohort studies including auto- and allo-HSCT recipients, in which HVOD (Baltimore criteria) was assessed prospectively. Genetic analyses were blinded to outcome. Multivariable analysis showed progressive ↑ in adjusted relative risk of HVOD with ↑ number of HFE C282Y alleles present (0, 1, or 2 C282Y alleles); stratification by CPSI genotype suggested effect modification by CPSI AA genotype. Iron stores were not measured.

Summar et al²⁴

Srivastava et al³⁴

GST

Single-center, pharmacokinetic study of GST polymorphisms in children with β-thalassemia major who underwent HSCT and Bu/Cy±ATG conditioning; found ↑ incidence of HVOD (Baltimore criteria) and lower Bu concentration in patients with GSTM1 null genotype

Ertem and Akar⁴⁶

Factor V Leiden

(Letter) Pediatric case series: allo-HSCT and Bu/Cy or Cy/ATG conditioning; 3 of 4 HVOD cases had factor V Leiden; no prothrombin mutations

HVOD

Duggan et al⁴⁷

Prothrombin

(Letter) Single-center case–control study in allo-HSCT patients with many different regimens and methods of GVHD prophylaxis. Prevalence of prothrombin mutation was 13% in patients with strictly defined HVOD compared to patients without (P=0.05) and much higher than in the general population (no adjustment for other risk factors). No HVOD cases had the factor V Leiden mutation.

ALI

Summar et al²⁴

CPSI

Single-center studies in the same HSCT cohort found CPSI AA (r) and a triplet repeat polymorphism of GCLC (r) to be protective against fatal ALI in HSCT recipients, adjusting for conditioning regimen and other risk factors

Kallianpur et al^32,40

GCLC

Hildebrandt et al⁴³

CCR-1, MCP-1

Single-center study that correlated BAL findings in mouse models of ALI or chronic noninfections lung dysfunction after allo-HSCT with BAL findings in allo-HSCT patients with ALI and in healthy volunteers; found ↑ MCP-1 and ↑ CCR2 expression in mouse lung after allo-HSCT, reduced ALI severity associated with CCR2−/− donor or pre-treatment with anti-MCP-1. Newly diagnosed ALI in humans also associated with ↑ MCP-1 in BAL fluid

GVHD (Single-institution studies in allo-HSCT, except where otherwise noted)

Lin et al⁶⁰

IL-10

All HLA-identical sibling donors; CSA and MTX used for GVHD prophylaxis; Recipient IL-10⁻⁵⁹²A allele and linked promoter haplotype associated with ↓ risk of grade III or IV aGVHD. This result was confirmed in a separate cohort of 423 patients.

Socié et al⁶¹

IL-10, IL-6

HLA-identical sibling HSCTs; GVHD prophylaxis with CSA+MTX in 90% of patients; independent of other clinical risk factors, aGVHD was associated with polymorphisms in IL-10 (d/r); IL-6 (r) polymorphisms correlated with cGVHD

Middleton et al⁶²

IL-10, TNFα

TNFd3/d3 (r) genotype and ↑ IL-10⁻¹⁰⁶⁴ (r) microsatellite repeat length associated with ↑ grade III/IV aGVHD in matched-sibling HSCT (CSA monotherapy for GVHD prophylaxis)

Cavet et al¹²⁹

IL-10, TNFα

HLA-identical sibling HSCTs; CSA and MTX used for GVHD prophylaxis; TNF d3/d3 (r) genotype and IL-10⁻¹⁰⁶⁴ (r) microsatellite repeat length associated with grade III/IV aGVHD

Cavet et al⁷³

IFNγ, TNFα, IL-6, IL-10

HLA-identical sibling HSCTs; GVHD prophylaxis was CSA±MTX or corticosteroids; IFNγ (r) genotype (r) associated with more severe grades of aGVHD; confirmed association between TNFd (r) and IL-10⁻¹⁰⁶⁴ (r) mutations and aGVHD; IL-6⁻¹⁷⁴ (r) genotype correlated with ↑ risk of cGVHD and with trend to ↑ aGVHD

Remberger et al⁶⁵

TNFα

All matched-unrelated donor HSCTs (molecular typing); TNFd microsatellite genotype (r) associated with ↑ risk of aGVHD grades II–IV and higher TNFα levels during conditioning. GVHD prophylaxis was CSA+MTX

Rocha et al⁶⁸

IL-10, IL-1Ra

IL-10 (r) and IL-1Ra (r) genotypes associated with ↑ risk of cGVHD; IL-1Ra (d) genotype associated with ↑ aGVHD grades II–IV

Nordlander et al⁶⁶

TNFα, IL-10

HLA-identical sibling HSCT in 85 patients, 111 matched-unrelated donors; GVHD prophylaxis mostly CSA+MTX; TNFd (r) and IL-10⁻¹⁰⁶⁴ (r) genotypes correlated with ↑ aGVHD grades II–IV

Takahashi et al⁶⁴

TNFα, IL-10

Related (matched or mismatched) and unrelated-donor HSCT with CSA+MTX for GVHD prophylaxis in most patients; donor-derived TNF⁻³⁰⁸ and IL-10 alleles (TNF2 A, IL-10⁻¹⁰⁸²G) associated with ↑ severe aGVHD and cGVHD

GVHD

Gagne et al⁵⁹

KIR

This study in unrelated (HLA-identical or mismatched) and related (HLA-identical) donor HSCT found 0% aGVHD in sibling HSCT and 100% in unrelated-donor transplants if recipient KIR genotype was included in the donor genotype (ie donor had the same number or more activating KIRs).

Renal failure

Juckett et al⁷⁴

ACE

Single-center retrospective study in 1-year survivors of allo-HSCT whose renal function was followed for up to 3 years; rate of decline in creatinine clearance associated with ACE genotype on multivariable analysis despite no difference in survival across genotypes. Conditioning regimen was the same in all patients but not standard for most institutions; GVHD prophylaxis was T-depletion of graft and CSA; renal shielding changed twice during the course of the study; ACE inhibitor drug use was not analyzed

Overall mortality

Lin et al,⁶⁰ Cavet et al,¹²⁹ Rocha et al,⁶⁸ Srivastava et al,³⁴ Summar et al,²⁴ Kallianpur et al^32,33,40

IL-10 (r), TNFα (r), F_cγR IIIb (d), MPO (d), GST (r), CPSI (r), GCLC (r)

Different polymorphisms in these genes associated with ↑ or ↓ survival in single-institution studies (see above)

Cook et al⁵⁸

KIR, HLA-C

First study to demonstrate the importance of HLA-C group of recipient in survival in the absence of HLA mismatch, restricted to patients with myeloid leukemias. Donor KIR2DS2 further decreases survival in homozygous HLA-C2 recipient despite no detectable effect on aGVHD. No effect of GVHD prophylactic regimen was noted in the analysis