Clinical and epidemiological studies of haemophagocytic lymphohistiocytosis

Abstract

Haemophagocytic lymphohistiocytosis (HLH) comprises primary (inherited) and secondary forms. The primary forms typically present in young children and carry a very high risk of mortality. The secondary forms, which are the result of different disorders, can present in all ages with greatly varying symptoms. The findings of HLH are related to an overly active but ineffective immune response, with accumulation of activated macrophages and lymphocytes, and a toxic state of hypercytokinemia. The typical symptoms of HLH are prolonged high fever, hepatosplenomegaly and cytopenia. HLH may also cause a meningoencephalitis and significant neurological late effects. A marker of HLH is impaired or absent function of natural killer (NK) cells and cytotoxic T cells. A major subgroup of primary HLH is Familial haemophagocytic lymphohistiocytosis (FHL). FHL is a rare autosomal disease and thus far three diseasecausing genes have been identified: PRF1, UNC13D and STX11. Untreated FHL is invariably fatal, with a median survival of 1-2 months. The only curable method is currently HSCT. Prompted by earlier treatment failures, the Histiocyte Society initiated a prospective international multi-centre study (HLH-94) that combined two previously reported regimens: chemotherapy and immunotherapy, followed by HCST in known familial and/or persistent or relapsing disease.

Aims: The aims of this thesis were to extend the clinical knowledge and diagnostics of HLH; to evaluate the outcome of the HLH-94 study; and ultimately to improve survival.

Results: The overall survival of the HLH-94 treatment far exceeded previous results. The estimated 3-year probability of survival was 55% (95% CI ± 9%). The HLH-94 initial and continuation therapy was successful in a total of 88/113 children (78%, 95% CI 69-85%), in that they were either admitted for HSCT (n=65) or still alive with at least one year follow-up since onset (n=23). The overall estimated 3-year probability of survival post HSCT was 64% (± 10%). The use of a matched unrelated donor (MUD) gave survival results comparable to those achieved when using a matched related donor (MRD), with a hazard ratio (HR) for mortality of 1.02 (CI=0.39-2.68) for MUD compared with MRD. The adjusted HR for mortality when using a haploidentical donor compared with an MRD was 3.31 (1.02-10.76), and the HR for mortality when using a mismatched unrelated donor (MMUD) compared with the use of an MRD was 3.01 (0.91-9.97). Persistent disease activity at two months after start of therapy appears to indicate a worse long-term prognosis.

The increased risk of mortality post-HSCT for these patients remained statistically significant after adjustment for potential confounding factors (HR=2.75, 1.26-5.99, p=0.011). It is often difficult to distinguish at the onset of disease whether a patient has a primary or secondary HLH. This is a major clinical problem as it affects the decision whether an HSCT needs to be performed or not. Four subtypes of NK cell cytotoxicity deficiency have been described. The cytotoxic deficiency can be restored in all subtypes except type 3.

To study association with clinical outcome, we thus pooled types 1, 2 and 4 together and defined them as being non-type 3. The estimated 3-year probability of survival was 46% for type 3 and 75% for non-type 3 (p=0.012). None of the 36 type 3 patients attained a sustained remission (≥one year) after stopping therapy without receiving an HSCT, as compared with 13/29 non-type 3 patients (45%, 95% CI 28-62%). Finally, type 3 patients were associated with a statistically significantly increased risk of having active disease or not being alive two months after start of therapy, as indicated by an adjusted OR of 4.80 (CI 1.38-16.66). This indicates that NK cell sub-typing may provide a valuable tool for clinicians to determine whether or not an HLH patient requires transplantation. At diagnosis, a high proportion of children displayed neurological symptoms and/or pathological CSF (122/193, 63%) (neurological symptoms only: 72/193 (37%); pathological CSF only: 101/193 (52%)).

An increased risk of mortality for patients with both neurological symptoms and abnormal CSF findings was shown when compared with patients with no neurological symptoms and normal CSF (adjusted HR 2.05, 1.13-3.72). A study of genotype-phenotype associations revealed that the frequency of gene mutations varies with ethnicity. The disease-causing mutations in FHL also display different phenotypes with regard to age at onset and pathological CSF at diagnosis.

Conclusions: In order to perform a meaningful clinical study of a rare disease, a collaborative international effort is required. The multi-centre study HLH-94 provides a successful example of this. Treatment according to the HLH-94 protocol has led to a dramatic increase in survival, and the work presented in this thesis will hopefully have a further positive impact on the outcome of children with HLH worldwide.