Management of hemophagocytic lymphohistiocytosis in pregnancy

Background: The diagnosis of hemophagocytic lymphohistiocytosis (HLH) in pregnancy is challenging due to its rarity. There is currently no consensus on the treatment of HLH during pregnancy. We aim to analyze and summarize the clinical characteristics of HLH in pregnancy, and to discuss effective diagnostic and treatment options. Methods: Thirteen patients with HLH during pregnancy who were diagnosed and treated at the Peking Union Medical College Hospital of the Chinese Academy of Medical Sciences from January 2000 to December 2019 were studied retrospectively. We collected data on treatment regimens and on maternal and pregnancy outcomes. Results: All patients had a singleton pregnancy, with a median age of 28 years (range, 22–33 years) and a median gestational age of 23 weeks (7–36 weeks). There were underlying associated diagnoses in six patients; of the patients, 12 (92.3%, 12/13) were treated with corticosteroids, and a good eciency was achieved in 5 (41.7%, 5/12) of them. Two patients who were treated with dexamethasone and etoposide after termination of pregnancy achieved CR. Two patients attained remission after termination of pregnancy. Four pregnant women died, and the mortality rate was 30.77% (4/13). Fetal or neonatal death up to 1 week after delivery occurred in eight (61.54%) pregnancies, and there were four cases of miscarriage, two of stillbirth, and two of neonatal death. Complications included premature birth (57.14% of neonates), small for gestational age (SGA, 7.70%), premature rupture of membranes (15.38%), and fetal stress (15.38%). Conclusion: Early diagnosis and treatment are important for maternal survival, and corticosteroids are the rst choice for most patients with HLH during pregnancy. For patients who do not respond to corticosteroids, etoposide, and termination of pregnancy may be life- saving.


Background
Hemophagocytic lymphohistiocytosis (HLH), also known as hemophagocytic syndrome, is a type of hyperin ammatory response caused by primary or secondary immune disorders. The principal clinical features of HLH are persistent fever, hepatosplenomegaly, and a decline in the number of blood cells [1]. The disease is divided into familial (primary) HLH and acquired (secondary) HLH. Primary HLH usually presents in childhood. It is caused by genetic mutation.
Secondary HLH is typical in most adult HLH cases, which is secondary to infection, autoimmune diseases, and malignant tumors. As HLH becomes rapidly fatal, with mortality rates ranging between 26.5% and 74.8% [2], timely identi cation of suspected HLH cases and a correct diagnosis are important.
It is rare for HLH to manifest during pregnancy and most of the relevant literature involves case reports [3][4][5][6]. Given the pregnant patients' age, secondary HLH was most likely. HLH has symptoms similar to those of obstetric complications, including hemolysis, elevated liver enzymes, low platelet count (HELLP), and acute fatty liver. Additionally, because there are other causes and related factors that cooperatively induce HLH [7], diagnosing HLH in pregnancy is di cult. Clinical management of HLH also appears inconsistent across the published cases, and the effect of medications during pregnancy on the fetus needs to be considered. Thus, there is currently no consensus on the treatment of HLH during pregnancy. In the present study, we retrospectively analyzed the clinical data from 13 cases of HLH during pregnancy at our hospital and emphasized the importance of rapid diagnosis and treatment.

Study population
We included 13 patients with HLH during pregnancy in this study. All the patients had singleton pregnancies, with a median gestational age of 23 weeks (range, 7-36 weeks). There was one case of HLH in the rst trimester of pregnancy (7.70%), seven in the second trimester (12-28 weeks; 53.85%), and ve in the third trimester (38.46%). Four (30.77%) patients were primiparas and nine (69.23%) were multiparas. There were other causes/related factors of HLH in six patients, including one with Still's disease complicated by cytomegalovirus (CMV) infection, three with systemic lupus erythematosus (SLE), 1 with SLE complicated by CMV infection, one with parvovirus B19 infection, seven with unclear causes. The patients' baseline characteristics are shown in Table 1.

Clinical symptoms and laboratory features
Fever and elevated ferritin levels were the most common symptom (13/13), followed by splenomegaly (12/13), increased brin or triglyceride levels (10/13), cytopenia, a diminution in two or three types of cells in peripheral blood (9/13), increased sCD25 levels (7/13), and attenuated natural killer (NK)-cell viability (3/13). The results of other tests showed that the incidence of abnormal liver function was high, with elevated liver enzymes (13/13), lactic dehydrogenase (13/13), and bilirubin levels (5/13). The copy number of CMV-DNA in two patients exceeded the normal range, and one patient was positive for parvovirus B19 (Table 1).

Therapy and outcomes
As shown in Table 1, 6 of the 13 patients (cases 1-6) were diagnosed with HLH and initiated treatment during pregnancy. They all received corticosteroids as rst-line treatment, with four patients (cases 2-5) also receiving intravenous immunoglobulin [IVIG]) and one (case 6) combined with etoposide. Three patients (cases 1-3) achieved partial remission (PR) and required termination of pregnancy because of the disease or stillbirth and ultimately achieved complete remission (CR). One patient (case 4) received methylprednisolone and IVIG and did not experience remission, and she continued the same treatment after vaginal delivery, achieved a PR, and was discharged from the hospital 35 days after delivery. One patient (case 6) was treated with corticosteroids/etoposide without remission. After cesarean section, she received a regimen of dexamethasone, IVIG, etoposide, and cyclosporine A (CsA) and achieved a PR 10 days after the operation and was then discharged from the hospital. The condition of one patient (case 5) continued to deteriorate progressively despite the use of dexamethasone and IVIG, and she died of multiple organ failure the day after spontaneous abortion.
Seven patients (cases 7-13) were diagnosed with HLH after termination of pregnancy and started speci c treatment thereafter. Two (cases 7-8) were treated with single corticosteroids, one patient (case 9) was treated with corticosteroids/CsA, 3 (cases 10-12) were treated with corticosteroids/etoposide, and 1 (case 13) was treated with extracorporeal membrane oxygenation (ECMO) immediately after delivery. The condition of ve patients deteriorated after delivery, and three (42.9%, 3/7) died. Case 8 was treated with corticosteroids (without IVIG) 1 day after cesarean section, and the patient's condition became aggravated and she died 15 days after the operation due to multiple organ failure. The condition of case 11 worsened after cesarean section, and the patient was started on methylprednisolone treatment on the 8th day after the operation. She received dexamethasone (IVIG) and etoposide treatment on the 14th day after the operation and died 22 days after the operation due to multiple organ failure. Case 13 was admitted to the hospital with multiple organ failure, received ECMO immediately after complete curettage of the uterine cavity, and died the day after delivery. Case 7 received corticosteroids (without IVIG) treatment after cesarean section and then achieved a PR. Case 9 began treatment with hydrocortisone (with IVIG) and CsA the day after induced labor, and then achieved PR. The condition of case 10 after cesarean section deteriorated and the patient then received dexamethasone (with IVIG) and etoposide treatment, achieving a CR. Case 12 received intravenous dexamethasone on the 2nd day after spontaneous abortion and achieved a PR on the 5th day after delivery, developing a fever and showing increased triglyceride levels again on the 23rd day after delivery. This patient was then treated with a combination of etoposide and achieved a CR.

Obstetric and neonatal events
Obstetric and neonatal events are shown in , respectively, and vaginal delivery was then performed; with these newborns dying on the 3rd day and the same day after birth, respectively. The most common obstetric complication was premature delivery (57.14% of neonates), followed by SGA (7.70%, 1/13).

Discussion
In our study, the most common time of onset for HLH in pregnancy was in the second trimester of pregnancy, followed by the third trimester, which is similar to the previous studies (Table 3). Our hypothesis for these phenomena is that pregnancy may be a regulatory immune state, immunologic alterations with advancing pregnancy impair the clearance of pathogens, resulting in an increased frequency of disease caused by some pathogens [8,9].
There is currently no guideline for HLH during pregnancy. Treatments are mostly selected on clinical experience and clinical manifestations. In general, the treatment of HLH includes inhibition of life-threatening in ammatory responses using immunosuppressive agents and cytotoxic drugs, treatment of the underlying cause, or both. As for the causes of HLH in pregnancy, among the 13 patients in this study, 5 had autoimmune diseases, 1 with parvovirus B19 infection, and 7 with unclear causes. The prognosis is variable with different underlying diseases. We found that patients with autoimmune diseases and viral infections have good outcomes and the patients for whom causes were not known had worse outcomes with respect to maternal health. The Topo II inhibitor etoposide is one of the essential drugs in HLH-94 and HLH-04 regimens. Prognostic factors of adult hemophagocytic syndrome indicated that the use of etoposide as the rst-line treatment tended to be associated with a better outcome [33]. Etoposide is a cell cycle-speci c antitumor drug that is classi ed as category D by the FDA. Etoposide is commonly used in the treatment of ovarian cancers, and it is considered safe for the fetus if given during the second or third trimester. No neonatal malformations have been reported [34]. Song et al. [27] reported the use of etoposide in a pregnant patient with HLH. No congenital malformations were found in the fetus. However, in a study performed in mice, etoposide had adverse effects on fetal ovarian development.
Exposure of pre-follicular ovaries to etoposide resulted in a near-complete elimination of germ cells prior to follicle formation [35]. In the current study, four patients were treated with etoposide, of whom one was treated during pregnancy, and we observed no abnormalities in the neonates. Our other three patients were treated after delivery, and two achieved CR, while one died of multiple organ failure 22 days after delivery. The number of cases of etoposide application during pregnancy was small; the timing, dose and frequency of the drug-as well as the effect of the drug on the fetus-still require further investigation.
CsA may be an effective treatment for patients with HLH who do not show a response to corticosteroids [31]. In our study, two patients were treated with CsA, including one with the combination of hydrocortisone and CsA, and one with the combination of dexamethasone, etoposide, and CsA. The FDA classi es CsA as class C drug for pregnancy. CsA can inactively cross the placenta and enter the fetal circulation [36]. A systematic review suggested that use of CsA during pregnancy is associated with premature delivery and low birth weight, but it is di cult to determine if any risks associated with CsA therapy during pregnancy are due to drug exposure alone or to pre-existing maternal comorbidities [37].
Patients with HLH in pregnancy can go into remission after termination of pregnancy. Teng et al. [3] hypothesized that the pathogenesis of HLH during pregnancy was similar to preeclampsia, where the immature placenta releases genetically foreign material into the maternal circulation. Maternal Tlymphocytes (which are unable to recognize unfamiliar human lymphocyte antigens), may then trigger a systemic in ammatory response and cytokine storm.
Termination of pregnancy may thus prevent the maternal condition from continuing to deteriorate and allow for timely chemotherapy. In previous studies, six patients (40%, 6/15 from a total of 40 cases) attained remission after termination of pregnancy [3-5, 11, 15, 27], including four with unclear causes, one complicated by SLE, and one complicated by autoimmune hemolytic anemia. In our study, termination of pregnancy was effective in two patients. However, the condition of six cases was exacerbated or did not improve after termination of pregnancy. The overall effect of termination of pregnancy is still controversial in HLH. The relationship between pregnancy and HLH requires further elucidation. If corticosteroid treatment is ineffective, termination of pregnancy may be an effective method of treatment.
There are inherent biases to our study because it was a retrospective study conducted in a referral center. The majority of patients worsen at the local hospitals. Therefore, this may have generated bias in the evaluation of treatment effects. Additionally, the details of most neonatal outcomes were relatively unclear. Therefore, it is important to perform more investigations to develop a standard treatment protocol for HLH in pregnancy.

Conclusions
In summary, the speci c mechanisms underlying HLH during pregnancy are unclear. Although several associated factors have been investigated, the etiology of many cases remained unclear. Corticosteroids are the rst choice for most patients with HLH during pregnancy, causes and related factors need to be identi ed and treated accordingly. Etoposide and termination of pregnancy may then be effective for patients with ineffective corticosteroid treatment. For patients after delivery-especially for severe patients-etoposide may be used as soon as possible to improve the prognosis. Our conclusions, however, still need to be further con rmed using a larger sample size.

Methods
This study method was approved by the Peking Union Medical College Hospital Review Board (reference number: S-K1161). The need for written informed consent was waived because of the retrospective nature of the study, and the data set was deidentified in order to protect patient privacy. Using a Etiologies of and factors related to HLH including rheumatologic, infectious, and oncologic workups were recorded for diagnostic and prognostic purposes. HLH-related laboratory indices that were recorded included routine blood results, liver function, serum ferritin levels, brinogen levels, triglyceride levels, hemophagocytosis, NK-cell viability, soluble CD25 levels. The presenting signs and symptoms, treatment, and outcome of HLH during pregnancy were also recorded.
We collected maternal characteristics including age, gravidity, parity, gestational age at disease onset, maternal outcomes, gynecological and obstetric history, major medical history, and major family history. The perinatal outcomes included preterm labor (delivery after 24 and before 37 full weeks of gestation), SGA (birth weight <the 10th percentile), preeclampsia, eclampsia, HELLP, PROM, gestational age, method of terminating pregnancy, birthweight, Apgar score, miscarriage (spontaneous fetal loss before 20 weeks of gestation), stillbirth, and neonatal death. The term "stillbirth" was used to describe fetal deaths at 20 weeks of gestation or later. Neonatal death was de ned as the death of an infant between 0 and 7 days after birth.
Descriptive statistics-such as frequency, percentage, and range-were used for presentation of variables. The distribution of age and gestation is shown as medians and interquartile ranges. Differences between groups were assessed using the Student's t-test. Categorical variables, including clinical characteristics and complications, are expressed as proportions and were compared using the Chi-square test or Fisher's exact-probability test. All statistical analyses were This study method was approved by the Peking Union Medical College Hospital Review Board (reference number: S-K1161). The need for written informed consent was waived because of the retrospective nature of the study, and the data set was deidentified in order to protect patient privacy. Our study was done in compliance with the Declaration of Helsinki.

Consent for publication
Not applicable.

Availability of data and materials
The datasets generated and/or analyzed during the current study are not publicly available due to patient data safety restrictions but are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.