Thirteen of the twenty-seven contacted institutions provided the requested data. During the defined 8-year period, these institutions had treated 1461 cases of malaria, and among those, 5 patients additionally met diagnostic criteria for HLH. One of these cases has been published previously [8].

In Germany, reporting malaria cases to the Robert Koch Institute (RKI), the German public health institute, is mandatory, and data are regularly published online [9]. According to RKI, a total of 6682 malaria cases had been reported in Germany during the selected period (2015: 1068; 2016: 970; 2017: 957; 2018: 896; 2019: 999; 2020: 366; 2021: 613; 2022: 813). We can, therefore, provide data for 21.9% (1461/6,682) of the reported malaria cases.

Our PubMed search yielded a total of 79 results with 45 case reports on 47 patients, including the above-mentioned patient whose case report we had already published [8, 10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53]. References of these articles comprised reports on five more cases which were included in the analysis [54,55,56,57,58].

Seven references were found that mentioned further twenty-four malaria-associated sHLH cases in a different context, mostly larger case series on infection-associated sHLH regardless of the causative agent. These publications did not contain specific clinical or diagnostic data on individual patients, but only the respective Plasmodium species and the outcome [59,60,61,62,63,64]. They are only considered in the following if they are explicitly mentioned. One of these cases has recently been published in detail and was included in our analysis [65, 66].

Table 1 gives an overview of the most important data from 57 patients (the 5 patients treated in our departments and 52 additional case reports from the literature). The complete dataset including the 23 cases from overview articles is available as Supplementary material.

In 55 of the 57 included cases, the patients’ age was recorded, ranging from 11 months to 73 years (median 23 years, IQR 8.5–34.5). Thirty-four patients were male and twenty-three were female. In 47 patients for whom data were available, the median delay between the onset of malaria symptoms and hospital admission (patient delay) was 7 days (IQR 5.5–12.5). Ethnicity of patients was available in only ten cases: five were Caucasian, four African, and one mixed African/Caucasian. Most patients have been treated in malaria-endemic countries in Asia (n = 48; 60%; including the overview articles), mostly India (26) and Thailand (6). Twenty-one malaria infections had been acquired in Sub-Saharan Africa, nineteen of the patients were travelers or migrants who were later diagnosed in their home countries.

Including our own cases and the cases from overview articles, most case reports describe infections with Plasmodium falciparum (55/80; 69%), followed by Plasmodium vivax (26/80; 33%) including co-infections with both Plasmodium species (4/80; 5%) and Plasmodium ovale wallikeri (1/80; 1%). In two cases, the Plasmodium species was unspecified. No HLH cases were attributed to Plasmodium malariae or Plasmodium knowlesi. Data on parasitaemia are available for 24 of the 38 cases with P. falciparum, with 11 (46%) exceeding a parasite load of 5%. In five cases, malaria was not the only infection diagnosed: co-infection with Dengue virus is mentioned in two cases [35, 48], and Mycobacterium tuberculosis, Mycoplasma pneumoniae or HIV plus HCV in one case each [26, 38, 40]. Two cases were diagnosed as airport malaria, and two cases were transfusion associated after bone marrow stem cell transplant [17, 44,45,46]. These four cases occurred outside endemic countries.

Overall, HLH-2004 criteria were met in 40 out of the collected 57 cases (70%), as far as documentation is available. Some of the cases have been published before the most recent revision of diagnostic criteria in 2007. In some other cases, current criteria were reportedly met, but not all relevant values were accessible.

Some cases were published before the introduction of HScore in 2014. HScore was mentioned in only one publication [48]. However, we retrospectively calculated HScore for all published cases. Unfortunately, only 18 of the 57 case reports (32%) included the complete data sets required for the calculation. In all other cases, we calculated the minimum and maximum possible value for the HScore, assuming that either no additional criteria or all criteria were met. Table 2 shows details on HLH-2004 criteria and additional HScore criteria. The available data show that 37 cases (65%) have an HScore of more than 169 points. Only two patients (4%) have a lower score, but interestingly, both met five HLH-2004 criteria. For the remaining 18 patients, values above or below 169 are possible depending on the missing data. The median HScore of cases with complete data sets is 244. In patients who received specific HLH therapy, the median HScore was 250, while patients without HLH therapy had a median HScore of 223. According to the Mann–Whitney U test, the difference was significant with a p value < 0.05. However, the individual diagnostic criteria (temperature, platelet count, hemoglobin, neutrophil count, triglycerides, fibrinogen, ferritin, sCD25, AST, spleen length) did not show any significant difference between the treated and the untreated group. Two of the patients treated in our departments fulfill only four HLH-2004 criteria but have an HLH probability of over 90% according to HScore. One of these patients did not have a bone marrow biopsy, while the other patient's bone marrow biopsy showed hemophagocytosis.

Severe complications were described in 20 out of 57 cases (35%), including acute kidney injury (AKI, n = 9), acute respiratory distress syndrome (ARDS, n = 7), somnolence/coma (n = 6), acute liver injury (ALI, n = 5), shock (n = 5), disseminated intravascular coagulation (DIC, n = 5), and unspecified multiple organ failure (MOF, n = 1). In three cases, fatal outcome was reported following multiple organ failure. One of these patients had suffered from co-infection with Dengue virus and Plasmodium vivax malaria [35], a second patient was diagnosed with airport malaria caused by P. falciparum late in the course of the disease [45]. The third patient suffered a fulminant course only 2 days after the onset of symptoms. The other 54 detailed reports relate full recovery of the patients, mostly without specific treatment for HLH. Only 24/57 (42%) patients received treatment for HLH, mostly corticosteroids (n = 18; 75%), intravenous immunoglobulins (IVIG) (n = 9; 38%), including both (n = 3; 13%). One of the patients received a salvage therapy comprising dexamethasone, etoposide, and ruxolitinib [8], another one was treated with dexamethasone and plasmapheresis.

Case series mentioning HLH because of malaria report 2 more patients successfully treated with IVIG and another 15 positive outcomes without treatment details. One patient with cerebral malaria died, although it is not clear whether sHLH was decisive in the fatal outcome. The other five cases contained no data on treatment or outcome either.

The Destatis query yielded three cases with ICD-10 codes corresponding to a combination of malaria and HLH. These cases match three of our patients with HLH diagnosed during hospital stay within the years 2016–2020. The fourth patient was diagnosed only retrospectively, and the corresponding ICD-10 code was not documented during his hospital stay. The fifth patient was diagnosed outside the period of available Destatis data. The register data make the existence of additional cases unlikely.