Outcome and Prognostic Factors of Hemophagocytic Lymphohistiocytosis in Children: Experience From a Low- and Middle-Income Country

Objective Hemophagocytic lymphohistiocytosis (HLH) is a life-threatening condition especially in low- and middle-income countries (LMICs). This study was done to evaluate the outcome and prognostic factors of HLH in patients presenting to our center. Methods The study was carried out at the Paediatric Oncology Department of Combined Military Hospital (CMH) in Rawalpindi, Pakistan. All cases of HLH, from one month to 15 years of age enrolled between January 1, 2013 to June 30, 2023, were included. IBM SPSS Statistics for Windows, version 25.0 (released 2017, IBM Corp., Armonk, NY) was used for statistical analysis, and t-test and chi-square tests were used for comparison between continuous and categorical variables. Frequencies and percentages were calculated for categorical variables. Results Out of 115 patients, seven (6%) abandoned the treatment. The data of 108 cases, including 58 males (53.7%), were analyzed. The mean age at diagnosis was 31.5 ± 39.03 months. The mean time to reach a pediatric oncologist was 30.20 ± 22.15 days. Fever and pallor were common symptoms occurring in 107 (99.1%) and 98 (90.7%) cases, respectively. Jaundice was present in 44 (40.7%), visceromegaly in 64 (59.3%), and bruising/bleeding in 16 cases (14.8%). Twenty-six (24.1%) patients underwent hematopoietic stem cell transplant (HSCT), out of which 17 (65.4%) children were cured. Overall survival at two years, five years, and 10 years was 38%, 37%, and 36.1%, respectively. Disease-free survival at two years, five years, and 10 years was 33.3%, 32.4%, and 31.5%, respectively. Conclusion HLH leads to high mortality due to delayed or misdiagnosis in LMICs. Early diagnosis and early referral to a pediatric oncologist is the detrimental factor in survival for HLH. HSCT is the treatment of choice for primary, refractory, or relapse cases.

The Histiocytic Society in 1991 published the initial diagnostic guidelines.In 2004, these were modified by considering genetic analysis.Lately, a new analytical score, the H-score, has been used for diagnosis.The Hscore significantly correlates with the HLH-2004 criteria, allowing for an effectual estimation of an individual's risk of having HLH [6][7][8].In centers where genetic testing is not obtainable, the H-score is a IBM SPSS Statistics for Windows, version 25.0 (released 2017, IBM Corp., Armonk, NY) was used for statistical analysis, and t-test and chi-square tests were used for comparison between continuous and categorical variables.Frequencies and percentages were calculated for categorical variables.The time from the date of complete response (CR) until relapse or death was defined as disease-free survival (DFS).The time from the day of diagnosis to the day of the last follow-up or death was defined as overall survival (OS).
Censoring was done at the date of the last contact (31st August 2023).
The median follow-up time was 40.90 + 45.19 months (interquartile range, 10.40-90.57).Kaplan-Meier survival curves estimated the DFS and OS and were compared using the log-rank test.The Cox proportionalhazard regression model was utilized for univariate and multivariate analysis of prognostic factors with 95% confidence intervals (95% CIs).P-values of <0.05 were considered significant.

Results
During the study period, a total of 115 patients with pediatric HLH were registered at the Pediatric Oncology Department of CMH, Rawalpindi.Among them, seven (6%) patients left the treatment after enrolment and were excluded.Data from 108 cases, including 58 males (53.7%) and 50 females (46.3%), were analyzed.The mean age at diagnosis was 31.59 ± 39.03 months (ranging from one month to 14 years).The mean time to reach the pediatric oncologist was 30.20 ± 22.15 days (ranging from two to 92 days).

FIGURE 1: Mortality distribution of cases (n = 69)
After the initial response and considering financial implications, 26 (24.1%) cases were planned for HSCT, out of which 17 (65.4%)children were cured, graft failure occurred in five (19.2%) children, and four (15.4%) children succumbed to treatment-related mortality.

OS and DFS
The median follow-up duration for the cohort was 49.13 ± 33.10 months.At the 10-year mark, the OS rate was observed to be 36.1%,while the DFS rate was 31.5%.A detailed analysis of the survival rates reveals that the OS rates at two, five, and 10 years were 38%, 37%, and 36.1%,respectively.This indicates a relatively stable survival rate after the initial two years, suggesting that early survival is a favorable factor for these patients.Similarly, the DFS rates at two, five, and 10 years were 33.3%, 32.4%, and 31.5%,respectively, indicating that patients who remain disease-free at the two-year mark have a relatively high likelihood of maintaining disease-free status in subsequent years.Figure 2 and Figure 3 visually represent these survival outcomes offering a graphical depiction of the Kaplan-Meier survival curves for OS and DFS.These findings emphasize the importance of early therapeutic interventions to improve long-term survival and disease-free outcomes in children with HLH.

FIGURE 2: Disease-free survival (DFS) curve for children with hemophagocytic lymphohistiocytosis (HLH) FIGURE 3: Overall survival (OS) curve for children with hemophagocytic lymphohistiocytosis (HLH) in our study
The OS of patients who underwent HSCT was considerably higher compared to those who did not undergo HSCT (69.2% vs. 25.6%,respectively), as illustrated in Figure 4.This significant difference highlights the importance of HSCT in improving long-term survival outcomes for pediatric HLH patients.The male-to-female ratio in our study was 1.16:1, which is in congruence with previous studies quoting ratios of 1.8:1 and 1.30 [8,14].Other studies by Janka et al. and Mottaghipisheh et al. have also documented a male preponderance, although no clear association has been verified [8,10] [15,16].By contrast, another study from Iran by Mottaghipisheh et al. reported 11.5 months as mean age of presentation [8].FHLH was observed in only 5.5% of cases as all patients were not screened keeping in view the limitation of resources and huge financial burden of genetic testing.This number is quite higher in previous literature with Alsohime et al. reporting 43% of cases of FHLH and 23.5% cases reported by Paul et al. [3,17].
Seventeen children (15.7%) were diagnosed cases of Griscelli syndrome, which is a known association observed in previous literature.Sasan et al. observed Griscelli syndrome in 12% of cases [14].Twenty (33.3%) cases in our study had visceral leishmaniasis (VL-HLH)-associated HLH.An Indian study by Paul et al. reported 2.1% cases of VL-HLH, whereas Gamero et al. reported 41% cases, which reflects endemicity of visceral leishmaniasis in these specific regions [3,18].Fever and pallor were observed in 90% of children in our study, which is a well-reported phenomenon in previous literature.Mottaghipisheh et al, and Shazia et al, reported fever in nearly all of their patients, while Yong-hai Zhou et al. reported a figure of 98.7% [8,15,19].However, Alsohime et al. reported fever in only 27.6% of patients, which was in contrast to our results [17].Bruising was present in 14.8% of children in our study, which is in accordance with previous literature as narrated by Alsohime et al. who reported bleeding in 13.8% [17].Forty-four (40.7%) children in our study had jaundice, and 64 (59.2%) children had visceromegaly on initial presentation.Zhou et al. also reported hepatomegaly in 95.6%, splenomegaly in 92.1%, and jaundice in 34.5% of patients [15].Similarly, Mottaghipisheh et al. also reported splenomegaly as the most constant presenting sign detected in nearly all their patients [8].
Hyperferritinemia was the hallmark biochemical abnormality observed in 90.7% of cases in our study, which has been observed in many previous studies as well.Allen CE et al. revealed that ferritin level >10,000 stands at 90% sensitivity and 96% specificity for HLH [20].Similarly, a study by Shazia et al. from Pakistan also found increased serum ferritin levels in 82% of their patients [20].
Contrary to previous literature, we found children who were cured to have higher mean ferritin levels as compared to those who expired.However, this relationship was not significant (p = 0.5).No other study in our knowledge has observed this peculiar association.Mottaghipisheh et al. and Paul et al. did not observe hyperferritinemia correlation with disease severity or adverse outcomes, such as relapse or death [3,8].
Mean bilirubin levels in our study were 16.7 (SD = 12.4) in treated patients as compared to 41.3 (SD = 58.3) in children who expired (p < 0.05).This effect has been studied in the past and reported in the literature by several studies that total bilirubin twice the upper limit of normal is associated with a higher mortality rate [17,21].
Rungrojjananon et al. reported that platelet count <50,000 cells/mm 3 , hyperbilirubinemia, and treatment response at weeks two and eight after treatment initiation were associated with higher mortality although we did not observe these associations in our study [22].However, we found that a higher triglyceride level was associated with inferior outcomes, with survivors having a triglyceride level of 4.04 (1.78%) as compared to nonsurvivors with a level of 7.00 (5.91%; p < 0.05).Mottaghipisheh et al. also reported this similar association, contrary to the study reported by Alsohime et al. who reported high mean triglyceride levels in survivors of HLH [8,17].
We did not observe any prognostic association between age, gender, cytopenias, albumin levels, and fibrinogen levels and the outcome of the child, as shown in Table 1.Ramchandar et al. also did not observe any association similar to our results [23].However, Abbasi et al. reported a significant association between these parameters and the outcome of the child [7].

TABLE 1: Comparison of prognostic factors among survivors and non-survivors in HLH
A p-value of <0.05 was considered significant.
ANC: absolute neutrophil count, WBC: white blood cell, PT: prothrombin time, PTTK: partial thromboplastin time kaolin, ALT: alanine transaminase HLH is a life-threatening disease for children with studies reporting poor prognosis without treatment and has a median survival of one to two months [24].A study conducted by Japanese scholars found that outcomes in children with HLH who were treated with the same protocol were different among HLH subtypes [15,25].Unfortunately, 69 (63.9%) children eventually expired in our study, which is considerably higher than data from the developed countries.Previous studies have stated a lower mortality rate ranging from 13.8% to 28.1% of pediatric HLH [15,26].Allen et al. and Abbasi et al. also observed a higher mortality percentage of 32% and 44%, respectively [7,20].This is mainly due to the delayed presentation of patients to oncologists and limited resources in LMICs.In addition, malnutrition is also a common feature in children from LMIC further increasing mortality risk.The differences observed in survival and prognostic factors in various studies may be due to geographical differences in patient population, different protocols used in various centers, and the lack of high-end medical care in LMICs.
Thirty-two (29.6%) children in our study expired within 30 days of disease onset, whereas 41 (37.9%) children died during the initial therapy induction phase, which is similar to previous literature by Zhang et al. and Zhou et al. who observed 30-day mortality as 30.9% and 29.52%, respectively [15,27].Tan et al. reported that the pooled mortality rate was 32.6% (95% CI: 23.4-42.4)and was higher in primary in comparison to secondary HLH [28].
We found a peculiar observation that children presenting with secondary HLH due to visceral leishmaniasis had a better survival rate than the rest of the cases.Mottaghipisheh et al. also reported a higher OS of three years in VL-associated HLH (p < 0.001) [8].
The OS rate in our study was 37.91% at two years, 37.1% at five years, and 36.1% at 10 years, whereas the DFS was 33.3% at two years, 32.4% at five years, and 31.5% at 10 years.Mottaghipisheh et al. and Tanya et al. also reported high mortality of the primary HLH diseases with OS of three years at 35.9% and 21.89%, respectively [8,24].Messina et al. narrated the five-year probabilities of OS and event-free survival (EFS) as 71% and 60%, respectively [29].Moreover, one study by Ponnett et al. reported that the survival rate of HLH improved from a one-year survival rate of less than 5% to a five-year survival rate of 21% [24].
HLH has a high risk of relapse in the pediatric population.We documented a high relapse rate of 56.5% of the total cohort.Half of the children presenting with relapse eventually expired due to disease-related complications or refractory disease.

Conclusions
HLH is a rapidly progressive disease that leads to a high mortality rate due to delayed or misdiagnosis in resource-limited regions where it can masquerade initially with varied clinical presentations.This study provides valuable insights into the long-term outcomes of pediatric patients with HLH in an LMIC.Delayed presentation to a pediatric oncologist and higher triglyceride levels at the time of presentation were associated with poor outcomes.The 10-year OS and DFS rates were 36.1% and 31.5%,respectively, with OS and DFS showing significant stability after the initial two years.
Another important observation was the markedly higher OS in patients who underwent HSCT, highlighting the importance of incorporating HSCT into treatment protocols where feasible.Awareness of primary care pediatricians for an early referral to a pediatric oncologist is required as it will improve the outcome and chances of survival.Overall, this study emphasizes the importance of HSCT in the management of HLH and calls for enhanced genetic diagnostic facilities in LMICs to better understand and treat this complex condition.

Human subjects:
Consent was obtained or waived by all participants in this study.Institutional Review Board/Ethical Committee of Combined Military Hospital, Rawalpindi issued approval 545.The article has been scrutinized thoroughly, and no violation of patient safety and privacy, breach of data, and violations of human rights and institutional rights have been observed.Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work.Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work.Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.
. The mean age at diagnosis was 31.59 ± 39.03 months in our study, which coincides with other studies by Zhou et al. and Bayram et al. who observed a mean age of 25 and 30 months, respectively [29]nother cross-border study byPonnatt et al., a 13% rate of relapse/refractory HLH was reported[24].Twenty-six children in our study underwent HSCT, out of which four (15.4%) children expired due to transplant-related mortality and 17 (65.4%)childrenwerecured.Graft failure was seen in five (19.2%) children.Allen et al. reported a post-HSCT mortality rate of 35% in his study[20].Similarly, in a study by Messina et al., 26 cases (24%) died due to transplant-related causes, whereas 14 (13%) and 10 (9%) patients experienced graft rejection and/or relapse, respectively.Twelve out of 14 children were given a second HSCT after graft failure/relapse and were reported alive and disease-free[29].One significant limitation of this study on HLH conducted in an LMIC is the lack of comprehensive genetic studies, particularly for FHLH.Due to constrained resources and limited access to advanced diagnostic technologies, genetic testing to identify specific mutations associated with FHLH was only performed in six cases.