4. Discussion
Childhood hemophagocytic lymphohistiocytosis is a rare and aggressive
condition characterized by a high mortality rate. Primary HLH, which
results from genetic mutations, is less common than secondary HLH. In
line with a recent multicenter study in Thailand, this study also found
that IAHS is the most common subgroup within secondary HLH, with EBV
being the dominant pathogen in IAHS. 7 In this cohort,
the distribution of HLH subgroups was 50% IAHS, 32.6% MAS-HLH, 15.1%
MAHS, and 2.3% primary HLH. It is important to note that there were
limitations in genetic testing capabilities at our institute. Tests were
conducted on only three patients: one diagnosed with Chediak-Higashi
syndrome, another with Griscelli syndrome type 2, and the third with
SPTCL.
The common clinical manifestations of childhood HLH in this cohort
included fever (100%), hepatomegaly (88%), anemia (77%), and
splenomegaly (74%). Laboratory parameters, based on the HLH-2004
diagnostic criteria, revealed hyperferritinemia (96%), hemophagocytosis
(83%), bicytopenia or pancytopenia (65%), hypofibrinogenemia (52%),
and hypertriglyceridemia (47%). Non-diagnostic criteria laboratory
findings included transaminitis (83%), hypoalbuminemia (79%), direct
hyperbilirubinemia (63%), and hyponatremia (56%). These findings are
typical in patients with HLH and are similar to those in previous
publications, with some variations in the frequency of findings that
might represent differences in presentation across HLH subgroups.9, 17-19 In developing countries, including Thailand,
several biomarkers such as soluble CD25, NK cell activity, and CXCL9 are
not available. This limitation affects both recent multicenter research
and this study. 7
The major cause of death in childhood HLH as identified in previous
study was R/R HLH, which accounted for 86.4% of the non-survivor cases.17 In this cohort, the incidence of R/R HLH was
34.9%, closely aligning with the range of 36.5 – 40.5% reported in
prior studies. 17, 19 There has been several research
efforts focused on the risk factors for R/R HLH, determined at different
points in post-treatment time. Bin Q et al. reported that platelet
counts greater than 100,000 /μL at 2 weeks post-treatment were more
likely to achieve resolution, with an odds ratio of 18.4 (95% CI, 2.7
– 122.9; p = 0.003). 8 Xu XJ et al. published a
retrospective multicenter study in China, revealing that at the
8th week of induction treatment, being under the age
of 12 months at diagnosis, having platelet counts below 80,000 /μL, CNS
involvement, and not receiving etoposide treatment were independent
prognostic factors for refractory disease, with odds ratios (95% CI) of
3.11 (1.01 – 9.64), 2.63 (1.06 – 6.53), 3.80 (1.27 – 11.38), and 3.12
(1.60 – 6.06), respectively. 9 Yanagaisawa R et al.
compared a group of children with recurrent EBV-IAHS to those in
remission. They found that being under the age of 2 years at diagnosis
and having an EBV load greater than 103 copies/mL at 2
weeks post-treatment were risks for recurrence in children with
EBV-IAHS. 10 In our study, we used clinical and
laboratory parameters at the time of diagnosis to early predict R/R HLH
in children. Multivariable logistic regression analysis revealed that a
pSOFA score of ≥ 8 and age less than 3 years were independent risk
factors for R/R HLH. Although there was no statistically significant
difference in the R/R HLH incidence across each HLH subgroup in this
cohort, the only two cases with primary HLH both experienced R/R HLH.
Several previous publications have described children with both primary
and secondary HLH, noting that an age of less than 2 years is a poor
prognostic indicator. 19, 20 This suggests that even
in patients with primary HLH, which typically presents during infancy,
age remains an independent predictor. Furthermore, a younger age has
been identified as a risk factor for R/R HLH, as mentioned earlier.9, 10 However, in this cohort, an optimal age at
diagnosis cutoff of less than 3 years revealed a significantly higher
AUC of 0.69 compared to an AUC of 0.60 for the age cutoff of less than 2
years (p = 0.033). In terms of organ dysfunction, published data have
shown several specific organ involvements, such as myocardial damage,
respiratory failure, hepatobiliary dysfunction, coagulopathy,
gastrointestinal disorders, and multiple organ dysfunction syndrome, as
independent risk factors for mortality in childhood HLH.21-23 According to guidelines published by the
Histiocyte Society in 2022, the pSOFA score has been implemented as a
severity grading tool in secondary HLH for ICU-admitted patients.15 However, we adopted this score and applied it to
all patients with primary and secondary HLH in this cohort, regardless
of ICU admission status. The pSOFA score appeared to be a valuable
predictor for R/R HLH. A pSOFA score cutoff of 8 yielded an AUC of 0.76.
This closely aligned with the use of a cutoff value above 8 for
assessing mortality risk in critically ill children.12 Moreover, the pSOFA score was also identified as an
independent predictor of mortality in this study.
In this cohort, the 2-year OS rate was 61.9%, which was lower than the
71.3% reported in a previous study conducted in Thailand.7 However, the outcomes for patients without R/R HLH
were significantly better, with a 2-year OS rate of 90.5%, compared to
only 13.5% for patients with R/R HLH. It is important to note that this
study had limited treatment options. No patients underwent allogeneic
hematopoietic cell transplantation, and only two patients with R/R HLH
received salvage treatment with Ruxolitinib.
The limitations of this study include limited genetic testing within the
cohort, potentially leading to an underestimation of primary HLH cases
triggered by infections. In developing countries, such as Thailand,
access to comprehensive biomarker testing is limited. This includes
tests for markers of immune activation (such as soluble CD25 and CXCL9),
measurements of proteins affected in familial HLH, and functional
studies, as recommended by the North American Consortium for
Histiocytosis. 24 The diagnostic investigations, based
on the HLH-2004 criteria, were also restricted, with only 6 out of the 8
criteria assessable due to the lack of specialized laboratory
capabilities. 4 Another limitation was the range of
treatment options available for this cohort, which could have influenced
the outcomes. This retrospective study also had some missing data,
particularly in laboratory tests involving inflammatory markers and
coagulation profiles. However, the study identified clinical parameters,
specifically higher pSOFA scores and the younger ages of patients, as
useful tools for identifying individuals at risk for R/R HLH.