Haemophagocytic syndromes are rare disorders characterized by high fever, lymphadenopathy, hepatosplenomegaly, cytopenias, and hyperferritinaemia. They can be categorized into primary and secondary forms. Primary, or familial, haemophagocytic syndrome likely comprises a group of distinct genetic diseases resulting in impaired cytotoxicity, and most often affects children under the age of 2 years. This form is usually called haemophagocytic lymphohistiocytosis. Secondary, or reactive, haemophagocytic syndrome (RHS), may occur at any age, most commonly in the setting of malignancy (particularly lymphoma), infection or autoimmune disease [1,2].
Herpes viruses are the most common infectious aetiology involved in RHS [3,4] but the syndrome has been associated with a variety of other viral, bacterial, and parasitic pathogens [4,5]. A preexisting immune deficiency is found in about 30–60% of cases, which suggests an important pathogenetic role of the underlying immune status [3,6–9].
Since 1992, HIV infection, alone or in concert with other opportunistic infections, has been reported to be a cause of RHS [7,10,11]. The hypercytokinaemia triggered by haemopathy/malignancy or opportunistic infections and the generalized defects in natural killer (NK) and T-cell cytotoxicity associated to HIV infection could explain, at least in part, the predisposition of HIV-infected patients to develop RHS. However, whereas haemophagocytosis pictures seem not rare among HIV-infected patients , most of the published data on RHS are case reports or small series, published in the preHAART era. Data on RHS in large series of HIV-infected patients in HAART era are lacking.
In this context, we reviewed all cases of RHS occurring in HIV-infected patients in three French Departments of Infectious Diseases/Internal Medicine during a 2 year-period.
Patients and methods
In December 2008, we conducted a retrospective study in three departments of Infectious Diseases/Internal Medicine in three French tertiary centres. HIV-1-infected adult patients for whom the diagnosis of RHS was suspected or confirmed in these departments during a 2-year period (1 January 2006–31 December 2007) were included. Two sources were used to assess identification of patients eligible for this study. First, the medical records of patients coded during the study period in each of these departments as D76.1 (Haemophagocytic lymphohistiocytosis), D76.2 (Haemophagocytic syndrome, infection-associated) or D76.3 (Other histiocytosis syndromes) and B20 (HIV disease resulting in infectious and parasitic diseases) or B21 (HIV disease resulting in malignant neoplasms), B22 (HIV disease resulting in other specified diseases), B23 (HIV disease resulting in other conditions) or B24 (unspecified HIV disease) following the International Classification of Diseases (ICD-10)  were reviewed. Second, we reviewed the results of all bone marrow needle aspiration prescribed by physicians of these three departments during the study period. Medical records of HIV-1-infected patients with haemophagocytosis features on bone marrow needle aspiration were examined. In the case of a patient with multiple episodes of RHS during the period study, only the first one was taken into account.
All data reported in this retrospective study were based on hospital records. Baseline demographic (age, sex), clinical (year of diagnosis of HIV infection, past history of opportunistic infections, type of antiviral therapy at time of RHS, highest recorded temperature, presence or not of organomegaly and lymphadenopathy) and laboratory (plasma HIV-1 RNA, CD4/CD8 lymphocyte counts, leucocytes and platelets counts, haemoglobin, C-reactive protein, fibrinogen, ferritin, triglycerides, liver enzymes, lactate dehydrogenase (LDH), blood urea nitrogen, creatinine and serum sodium levels, prothrombin time) data were recorded. Moreover, when available, viral loads of Epstein–Barr virus (EBV), cytomegalovirus (CMV) and human herpes virus-8 (HHV-8) were collected. For all biological parameters except viral loads and lymphocyte counts, we recorded data available the day of bone marrow aspiration or during the 3 previous days. For patients for whom bone marrow aspiration was not performed, data available the day of RHS treatment initiation or during the 3 previous days were recorded. For viral loads and CD4/CD8 lymphocyte counts, we recorded data available during a period of 20 days around the bone marrow cytology (10 days before or 10 days after). Lastly, data about the aetiology of RHS and the outcome of included patients were collected.
Criteria definition for the diagnosis of haemophagocytic syndrome
In the absence of any specific marker, it is suggested that the diagnosis of haemophagocytic syndrome can be established if five out of eight diagnostic criteria are fulfilled . These eight criteria are fever [seven or more days of a temperature as high as 38.5°C (101.3°F)], splenomegaly (a palpable spleen greater than 3 cm below the costal margin), cytopenia (two or more of the three lineages haemoglobin <90 g/l, platelets <100 × 109/l, neutrophils <1.0 × 109/l), low fibrinogen (<1.5 g/l) and/or hypertriglyceridaemia (>3.0 mmol/l), haemophagocytosis without evidence of malignancy, low/absent NK cell activity, hyperferritinaemia (>500 μg/l) and increased soluble CD25 levels (>2400 U/ml). In our retrospective study, CD25 levels and NK cell activity were not available. Thus, we considered the diagnosis of haemophagocytic syndrome as certain if at least five out of the six other criteria were fulfilled. If only four criteria were fulfilled, we considered the diagnosis as possible but not certain.
Proportion was used as descriptive statistic for categorical variables. Continuous variables were described by median (range). The Fisher exact test for categorical variables and the Wilcoxon test for continuous variables were used to compare groups. The overall survival univariate analysis was conducted using the Kaplan–Meier method and groups were compared using the Log-Rank test. Because of multiple comparisons, statistical analyses were tested at the significance threshold of α = 0.01. All statistical analyses were two-tailed. Analyses were performed using SAS software (version 8.2, SAS Institute, Cary, North Carolina, USA).
The medical charts of 72 patients were reviewed. Following our classification criteria, 43 had a certain diagnosis of RHS and 15 a possible diagnosis (ferritin, fibrinogen or triglycerides levels were not available for four of these 15 patients). Platelet count tended to be lower and features of haemophagocytosis were more frequently observed in patients with a certain diagnosis of RHS (Table 1). These two groups did not significantly differ for other demographic (age, sex, origin), clinical (hepatomegaly, lymphadenopathy) or biological (leucocytes, haemoglobin, C-reactive protein, liver enzymes, LDH, blood urea nitrogen, creatinine and serum sodium levels, prothrombin time) parameters. Respectively 16 out of 43 (37%) and two out of 15 (13%) patients died within 3 months after the diagnosis of RHS (P = 0.11).
The 14 others patients had three or less criteria excluding the diagnosis of RHS.
The median age of the 58 patients with certain or possible RHS was 42 [23–85] years. Forty-four of them (76%) were men and 33 (57%) originated from Africa. Hepatomegaly and/or splenomegaly were observed in all but five patients. All patients had fever (median duration of fever at time of RHS: 17 (5–180) days].
A bicytopenia or pancytopenia was observed in 49 (84%) patients. Low fibrinogen was observed in only 12 (21%) out of the 56 patients for whom it was available. Ferritin level was available in 56 patients and was high more than 500 μg/l in all but one of them (98%). Triglyceridaemia more than 2.0 mmol/l was observed in 37 out of 55 (67%) patients for whom it was available and was more than 3.0 mmol/l in 27 of them (49%). Haemophagocytosis was observed on bone marrow aspirate for 36 out of 57 (63%) patients (for one patient, bone marrow aspirate was not performed) and on bone marrow (n = 3), lymph nodes (n = 2) or liver (n = 1) biopsies in six patients with no haemophagocytosis on bone marrow aspirate.
Thirty-seven (64%) patients received RHS treatment whereas the 21 (36%) others only received the RHS trigger's treatment (i.e., antiinfectious therapy, chemotherapy). Specific RHS therapies were etoposide 120–200 mg intravenously, alone or with corticosteroids, (n = 24) or corticosteroids alone (1 mg/kg orally to 500 mg intravenously, n = 12). One patient received only intravenous immunoglobulins (1 g/kg per day for 2 days).
At time of RHS, the median duration of HIV infection was 4 [0–22] years (Table 2). CD4/CD8 lymphocyte counts and plasma HIV-1 RNA were available for 50 and 52 patients, respectively. At time of RHS episode, the median CD4 lymphocyte count was 91 [2–387]/μl. Eighteen out of 52 patients (35%) had plasma HIV-1 RNA less than 50 copies/ml. Thirty-three patients (57%) received HAART at time of RHS with a median duration of 5 [0–132] months. Lastly, 29 patients (50%) had a past history of opportunistic infections.
Aetiology of reactive haemophagocytic syndrome
All patients had a CT scan and blood cultures including direct examination and culture for mycobacteria. EBV, CMV and HHV-8 viral loads were available for 53, 54 and 49 patients, respectively. Aetiology of RHS was found in 56 out of 58 patients (Table 2). For 31 of them (55%), a haemopathy/malignancy was the cause of RHS and an underlying infection was found in 23 patients (41%) (one of them had an acute HIV infection). Lastly two women were diagnosed with systemic lupus erythematosus.
Patients with haemopathy/malignancy presented more frequently with splenomegaly and had lower aspartate aminotransferase (AST) and LDH levels and CD8 cell count than patients with infection (Table 3).
At the time of RHS, median plasma EBV viral load was 20 000 [0–650 000] copies/ml and an EBV viral load above 100 000 copies/ml was evidenced in seven (13%) patients (disseminated toxoplasmosis n = 1, CMV infection n = 1, acute HIV infection n = 1, disseminated herpes simplex virus-2 infection n = 1, multicentric Castleman disease n = 2, Hodgkin lymphoma n = 1). Median plasma CMV viral load was 5000 [0–105 000] copies/ml and a CMV viral load above 10 000 copies/ml was found in eight (15%) patients. Patients with Kaposi sarcoma and/or multicentric Castleman disease had a median HHV-8 viral load of 64 000 [1000–1 830 000] copies/ml.
Five patients developed RHS within the first 2 months of the HAART. For all but one, aetiology was evidenced (disseminated toxoplasmosis = 1, non tuberculous mycobacteria n = 1, tuberculosis n = 1, Hodgkin lymphoma n = 1). For the last one, no aetiology was found despite a large screening. The diagnosis of immune reconstitution inflammatory syndrome (IRIS) was hypothesized.
Twenty-four patients (41%) required intensive care during the course of RHS and 18 (31%) died, 13 of them within 3 months after the diagnosis of RHS. Respectively, seven (33%) out of the 21 patients who did not receive specific RHS treatment and 11 (30%) out of the 37 patients who received specific RHS treatment died.
Among the 31 and 23 patients with haemopathy/malignancy-associated RHS or infection-associated RHS, nine (29%) and eight (35%) died during follow-up, respectively. In univariate analysis, the overall survival was not statistically different between the two groups (P = 0.68) (Fig. 1).
This study of RHS in HIV-1-infected patients is the largest reported to date and is the only one available in HAART era. These 58 HIV-1-infected patients were diagnosed with RHS during a 2-year period in three HIV/AIDS reference centres. Haemopathy/malignancy, and in particular Hodgkin lymphoma, was retained as the cause of RHS in 55% of patients. Infectious diseases, mainly herpesviruses infections and tuberculosis, were found in 41% of cases. The prognosis remains poor with 31% of death which is however better than in the pre-HAART era.
Systemic infections were the triggering events firstly associated with RHS [15,16] and almost all infectious agents have been described as potential triggers [4,5,10]. RHS has been reported in HIV-infected patient, especially in the pre-HAART era [6–8,11]. In the series by Grateau et al. 0.6% of the HIV-infected patients followed in the institution were diagnosed with RHS . During this pre-HAART period, RHS was mainly reported in patients with HIV end-stage disease and opportunistic infections were the most frequently reported trigger event . Except our previous study about HHV-8-related RHS , this study is the only one available in the HAART era and RHS seems now more frequently triggered by haemopathy/malignancy, especially lymphoma. These data probably reflect the increasing prevalence of haemopathy/malignancy in HIV-infected patients, corresponding to increased survival and decrease of opportunistic infections related to the use of HAART [18,19]. In our study, about one-third of patients died during follow-up which is much higher than in the global HIV-infected population, at least during the first 5 years following infection . This is however lower than the 50–100% mortality rates reported in RHS-patients in the pre-HAART era (Table 4). In patients without HIV infection, mortality rates of RHS vary from 25 to 45% [8,16,21].
In our study, patients with haemopathy/malignancy presented more frequently with splenomegaly and had lower ASAT and LDH levels and CD8 cell count. If it is not surprising that splenomegaly is more frequent in these patients, especially those with lymphoma, the results regarding LDH and AST level are more intriguing. Higher levels of LDH and AST have been evidenced in patients with poorer prognosis [22,23] but this point was not evidenced in our study. Clifford et al.  recently showed that Hodgkin lymphoma diagnosed in HIV-infected patients was strongly associated with lower total CD4 and CD8 cell counts. This may partially explain our results regarding the lower CD8 cell count.
In our study, 33 out of 58 patients (57%) originated from Africa, which is much higher than the 20% for the global population infected by HIV in France in 2005 . Patients originated from Africa are likely to access to medical care later than the others and to present with opportunistic infections or haemopathy that could trigger RHS. However, we cannot totally rule out a recruitment bias in the three study centres. If genetic loci related to the activity of perforin and granzyme granules have been associated with the familial forms of haemophagocytic syndrome [26–29], to our knowledge, no genetic defects have been linked to the reactive forms of haemophagocytic syndrome.
Our study has two main limitations. First, because of its retrospective nature, we cannot be certain that all patients have been identified even if we have used two sources for patient identification. Moreover, about 20% of patients had missing data regarding CD4/CD8 lymphocyte count and/or HIV/EBV/CMV/HHV-8 viral loads. However, we think that these two points could not have biased our results in a major way. Second, the diagnosis of RHS was based on six of the eight usual diagnostic criteria. By choosing to establish the diagnosis of haemophagocytic syndrome using these criteria we intended to do better than defining the cases only on the basis of haemophagocytosis. Indeed, we think that evidence of active haemophagocytosis in bone marrow is not sufficient to establish the diagnosis of haemophagocytic syndrome because it is probable that HIV itself can cause various degrees of histiocytic reaction without any associated disease [12,30,31]. However, the validity of such an adaptation of the original criteria cannot be defined.
In our study, five patients developed RHS within the first 2 months of HAART and the diagnosis of IRIS was hypothesized [32–34]. In HIV-infected patients, IRIS occurs weeks to months after HAART initiation and could result from a disproportionate inflammatory response of the immune system to various antigens with a proinflammatory Th-1 cytokines storm [35,36]. Increased production of these cytokines such as INF-γ has also been involved in the pathophysiology of haemophagocytic syndrome [37–39], but it is unclear whether these are epiphenomena, or cause the hallmark of the syndrome, that is, haemophagocytosis. Indeed, cytokines are not immediately causal in the familial form of haemophagocytic syndrome, which results from molecular defects, impairing the CD8T/NK cells granule-dependent cytotoxic activity [27–29]. Thus, further works are needed to assess whether RHS is a direct result from a cytokines storm associated to IRIS or is due to a transient reduction of CD8 cells cytotoxicity caused, for example, by infectious agents and/or unknown somatic mutations [40,41].
L.F. study design, medical evaluation, analysis of results, statistical analysis and writing of the first draft of the paper.
O.L. study design, medical evaluation, recruitment of study participants, analysis of results, critical reading of the manuscript.
J-L.M. medical evaluation, recruitment of study participants, critical reading of the manuscript.
W.K. medical evaluation, recruitment of study participants, processing of biological samples obtained.
L.G. medical evaluation.
C.M. processing of biological samples obtained, critical reading of the manuscript.
A.D.L. processing of biological samples obtained.
J.C. medical evaluation, recruitment of study participants, critical reading of the manuscript.
C.L. medical evaluation, recruitment of study participants.
P.C. medical evaluation, recruitment of study participants, critical reading of the manuscript.
J-M.M. medical evaluation, recruitment of study participants.
V.M. study design, medical evaluation, recruitment of study participants, analysis of results, critical reading of the manuscript.
1. Janka G, Imashuku S, Elinder G, Schneider M, Henter JI. Infection- and malignancy-associated hemophagocytic syndromes. Secondary hemophagocytic lymphohistiocytosis. Hematol Oncol Clin North Am 1998; 12:435–444.
2. Emmenegger U, Schaer DJ, Larroche C, Neftel KA. Haemophagocytic syndromes in adults: current concepts and challenges ahead. Swiss Med Wkly 2005; 135:299–314.
3. Reiner AP, Spivak JL. Hematophagic histiocytosis. A report of 23 new patients and a review of the literature. Medicine (Baltimore) 1988; 67:369–388.
4. Tiab M, Mechinaud F, Harousseau JL. Haemophagocytic syndrome associated with infections. Baillieres Best Pract Res Clin Haematol 2000; 13:163–178.
5. Fisman DN. Hemophagocytic syndromes and infection. Emerg Infect Dis 2000; 6:601–608.
6. Tiab M, Mechinaud F, Hamidou M, Gaillard F, Raffi F, Harousseau JL. Hemophagocytic syndromes. A series of 23 cases. Ann Med Interne (Paris) 1996; 147:138–144.
7. Grateau G, Bachmeyer C, Blanche P, Jouanne M, Tulliez M, Galland C, et al
. Hemophagocytic syndrome in patients infected with the human immunodeficiency virus: nine cases and a review. J Infect 1997; 34:219–225.
8. Sailler L, Duchayne E, Marchou B, Brousset P, Pris J, Massip P, et al
. Etiological aspects of reactive hemophagocytoses: retrospective study in 99 patients. Rev Med Interne 1997; 18:855–864.
9. Wong KF, Chan JK. Reactive hemophagocytic syndrome: a clinicopathologic study of 40 patients in an Oriental population. Am J Med 1992; 93:177–180.
10. Doyle T, Bhagani S, Cwynarski K. Haemophagocytic syndrome and HIV. Curr Opin Infect Dis 2009; 22:1–6.
11. Bourquelot P, Oksenhendler E, Wolff M, Fegueux S, Piketty C, D'Agay MF, et al
. Hemophagocytic syndrome in HIV infection. Presse Med 1993; 22:1217–1220.
12. Niedt GW, Schinella RA. Acquired immunodeficiency syndrome. Clinicopathologic study of 56 autopsies. Arch Pathol Lab Med 1985; 109:727–734.
14. Henter JI, Tondini C, Pritchard J. Histiocyte disorders. Crit Rev Oncol Hematol 2004; 50:157–174.
15. Chandra P, Chaudhery SA, Rosner F, Kagen M. Transient histiocytosis with striking phagocytosis of platelets, leukocytes, and erythrocytes. Arch Intern Med 1975; 135:989–991.
16. Risdall RJ, McKenna RW, Nesbit ME, Krivit W, Balfour HH Jr, Simmons RL, et al
. Virus-associated hemophagocytic syndrome: a benign histiocytic proliferation distinct from malignant histiocytosis. Cancer 1979; 44:993–1002.
17. Fardet L, Blum L, Kerob D, Agbalika F, Galicier L, Dupuy A, et al
. Human herpesvirus 8-associated hemophagocytic lymphohistiocytosis in human immunodeficiency virus-infected patients. Clin Infect Dis 2003; 37:285–291.
18. Engels EA, Pfeiffer RM, Goedert JJ, Virgo P, McNeel TS, Scoppa SM, et al
, for the HIV/AIDS Cancer Match Study. Trends in cancer risk among people with AIDS in the United States 1980–2002. AIDS 2006; 20:1645–1654.
19. Bonnet F, Burty C, Lewden C, Costagliola D, May T, Bouteloup V, et al
, Agence Nationale de Recherches sur le Sida et les Hépatites Virales EN19 Mortalité Study Group; Mortavic Study Group. Changes in cancer mortality among HIV-infected patients: the Mortalité 2005 Survey. Clin Infect Dis 2009; 48:633–639.
20. Bhaskaran K, Hamouda O, Sannes M, Boufassa F, Johnson AM, Lambert PC, et al
, CASCADE Collaboration. Changes in the risk of death after HIV seroconversion compared with mortality in the general population. JAMA 2008; 300:51–59.
21. Fardet L, Coppo P, Kettaneh A, Dehoux M, Cabane J, Lambotte O. Low glycosylated ferritin, a good marker for the diagnosis of hemophagocytic syndrome. Arthritis Rheum 2008; 58:1521–1527.
22. Guo X, Li Q, Zhou CY. Retrospective analysis of 41 childhood hemophagocytic syndrome. Zhonghua Xue Ye Xue Za Zhi 2007; 28:449–453.
23. Dhote R, Simon J, Papo T, Detournay B, Sailler L, Andre MH, et al
. Reactive hemophagocytic syndrome in adult systemic disease: report of twenty-six cases and literature review. Arthritis Rheum 2003; 49:633–639.
24. Clifford GM, Rickenbach M, Lise M, Dal Maso L, Battegay M, Bohlius J, et al
, Swiss HIV Cohort Study. Hodgkin lymphoma in the Swiss HIV Cohort Study. Blood 2009; 113:5737–5742.
26. Ménasché G, Feldmann J, Fischer A, de Saint Basile G. Primary hemophagocytic syndromes point to a direct link between lymphocyte cytotoxicity and homeostasis. Immunol Rev 2005; 203:165–179.
27. Stepp SE, Dufourcq-Lagelouse R, Le Deist F, Bhawan S, Certain S, Mathew PA, et al
. Perforin gene defects in familial hemophagocytic lymphohistiocytosis. Science 1999; 286:1957–1959.
28. Feldmann J, Callebaut I, Raposo G, Certain S, Bacq D, Dumont C, et al
. Munc13-4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic lymphohistiocytosis (FHL3). Cell 2003; 115:461–473.
29. Feldmann J, Le Deist F, Ouachee-Chardin M, Certain S, Alexander S, Quartier P, et al
. Functional consequences of perforin gene mutations in 22 patients with familial haemophagocytic lymphohistiocytosis. Br J Haematol 2002; 117:965–972.
30. Khalil SH, Nounou RM, Frayha H, Halim MA, Ellis M, Black FT. Bone marrow morphologic findings in patients with human immunodeficiency virus (HIV) infection. Ann Saudi Med 1996; 16:16–19.
31. Marche C, Tabbara W, Michon C, Clair B, Bricaire F, Matthiessen L. Bone marrow findings in HIV infection: a pathological study. Prog AIDS Pathol 1990; 2:51–60.
32. De Lavaissière M, Manceron V, Bourée P, Garçon L, Bisaro F, Delfraissy JF, et al
. Reconstitution inflammatory syndrome related to histoplasmosis, with a hemophagocytic syndrome in HIV infection. J Infect 2009; 58:245–247.
33. Cuttelod M, Pascual A, Baur Chaubert AS, Cometta A, Osih R, Duchosal MA, et al
. Hemophagocytic syndrome after highly active antiretroviral therapy initiation: a life-threatening event related to immune restoration inflammatory syndrome? AIDS 2008; 22:549–551.
34. Huang DB, Wu JJ, Hamill RJ. Reactive hemophagocytosis associated with the initiation of highly active antiretroviral therapy (HAART) in a patient with AIDS. Scand J Infect Dis 2004; 36:516–519.
35. Shelburne SA 3rd, Hamill RJ, Rodriguez-Barradas MC, Greenberg SB, Atmar RL, Musher DW, et al
. Immune reconstitution inflammatory syndrome: emergence of a unique syndrome during highly active antiretroviral therapy. Medicine (Baltimore) 2002; 81:213–227.
36. Bourgarit A, Carcelain G, Martinez V, Lascoux C, Delcey V, Gicquel B, et al
. Explosion of tuberculin-specific Th1-responses induces immune restoration syndrome in tuberculosis and HIV co-infected patients. AIDS 2006; 20:F1–7.
37. Mazodier K, Marin V, Novick D, Farnarier C, Robitail S, Schleinitz N, et al
. Severe imbalance of IL-18/IL-18BP in patients with secondary hemophagocytic syndrome. Blood 2005; 106:3483–3489.
38. Osugi Y, Hara J, Tagawa S, Takai K, Hosoi G, Matsuda Y, et al
. Cytokine production regulating Th1 and Th2 cytokines in hemophagocytic lymphohistiocytosis. Blood 1997; 89:4100–4103.
39. Jordan MB, Hildeman D, Kappler J, Marrack P. An animal model of hemophagocytic lymphohistiocytosis (HLH): CD8+ T cells and interferon gamma are essential for the disorder. Blood 2004; 104:735–743.
40. Chuang HC, Lay JD, Hsieh WC, Wang HC, Chang Y, Chuang SE, Su IJ. Epstein-Barr virus LMP1inhibits the expression of SAP gene and upregulates Th1 cytokines in the pathogenis of hemophagocytic syndrome
41. Holzelova E, Vonarbourg C, Stolzenberg MC, Arkwright PD, Selz F, Prieur AM, et al
. Autoimmune lymphoproliferative syndrome with somatic Fas mutations. N Engl J Med 2004; 351:1409–1418.