Malignancy in non-HIV infected patients with cryptococcal meningitis: a retrospective study

Background: Cryptococcal meningitis (CM) was found to be coexist with malignancy in non-human immunode�ciency virus (HIV) infected patients. The purpose of this study was to evaluate the clinical characteristics and therapeutic outcomes of CM in non-HIV infected patients with malignancy. Methods: A total of 320 CM patients were enrolled for analysis from January 2013-May 2019. One hundred and four patients underwent positron emission computed tomography (PET-CT) examination. The demographics, clinical characteristics, microbiological, radiological, therapeutic outcomes were analyzed in CM patients with and without malignancy. Results: Twelve patients with malignancy were found, of which 7 malignancy before CM (MBC), 5 malignancy after CM (MAC). CM patients with malignancy were older than non-malignancy ones. The prognosis of MBC patients was comparable to that of non-malignancy patients, but was extremely poor in MAC patients. Four out of 5 CM patients, who was suggested diagnosis as malignancy by PET-CT, were �nally con�rmed. Conclusions: This study found an increase rate of solid malignancies in CM patients. Screening malignancy in older CM patients was very important because it is closely related to prognosis and might affect treatment strategy. PET-CT might be a useful tool for early malignancy screening in CM.


Background
Cryptococcal meningitis (CM) is the one of the most common opportunistic infection and often occurs in the patients with human immunode ciency virus (HIV) infection [1,2].In non-HIV endemic areas, CM appears in a growing number of other forms of natural or iatrogenic immunosuppressive patients [3].In China, about one-third of non-HIV CM patients have predisposing factors, and the prognosis of these people does not seem to be signi cantly different from that of the immunocompetent population [4].
Malignancy is also considered to be immune-related diseases.HIV patients and genetic model animals with immunode ciency has shown an increase in the incidence of malignancy formation [5,6].Previous retrospective studies have shown that the incidence of cryptococcal infections is higher in the patients with malignancy, especially in hematological malignancy [7][8][9].Though the CM and malignancy may associate with immunode ciency at the same time, the coexistence of malignancy and CM in non-HIV patients is unclear now.In addition, the types of malignancies and their effect on CM prognosis need to be further studied.
To the best of our knowledge, no studies have been conducted to evaluate the clinical characteristics and therapeutic outcomes of CM in non-HIV infected patients with malignancy.The objective of this study is to analyze the epidemiological and clinical data of malignancy in non-HIV patients with CM, and assess the application of positron emission computed tomography (PET-CT) in the diagnosis of malignancy in the CM population.

Study populations
This is a retrospective study approved by the Medical Ethics Committee of the Third A liated Hospital of Sun Yat-sen University.All study participants have provided written consent for research and publication.
Records of all patients diagnosed with de nite CM from January 2013-May 2019 were examined.Figure 1 illustrates the owchart for the 357 recruited non-HIV infected CM patients.The enrolled CM patients were divided into three groups according to the discovery of malignancy: 1) no malignancy, 2) malignancy before CM (MBC), 3) malignancy after CM (MAC).De nite CM were identi ed by a positive result of India ink smear, microbiology culture in cerebrospinal uid (CSF) and histopathological nding in meninges or brain parenchyma (5 to 10 µm encapsulated yeasts).The diagnosis of malignancy before admission depended on the medical history.Patients newly diagnosed with malignancy after admission were con rmed by typical imaging ndings and tissue pathology (surgery or biopsy).Patients were excluded if they had a history of immunode ciency, such as HIV infection, rheumatic diseases, use of immunosuppressants and corticosteroids.
PET-CT examination Among these 320 CM patients, 104 of them underwent PET-CT examination.These patients were fasted for at least 4 hours with glucose levels less than 150 mg/dL before starting the scan.PET images of the whole body were obtained by using a PET-CT scanner (Discovery Elite, GE healthcare, Chicago, IL, USA) after intravenous injection of 5.2 MBq per kilogram body weight 18-uorodeoxyglucose ( 18 F-FDG) up to maximum of 370 MBq (10 mCi).The 3D PET data and CT images were reconstructed for diagnosis.

Treatment and outcomes
Antifungal treatment was based on amphotericin B, usually in combination with uconazole/voriconazole and 5ucytosine.Combination therapy will sterilize CSF within 2 weeks of treatment in most immunocompetent patients [10], so the patient's rst therapeutic evaluation was performed 2 weeks after using the standard antifungal treatment described above.With reference to previous study [4], "response" was de ned as survival within 2 weeks with relief of symptoms and clinical signs, improvement in CSF chemistry and cell count, reduce of cryptococcus counts in CSF.Our standard treatment composed by three stages according to the guidelines, including induction, consolidation and maintenance stages [11].If the patient has completed the standard treatment stages, with the clinical symptoms disappear, and negative CSF cryptococcus result, CM was considered to be cured, and patients were marked as "success".The patients were follow-up with regular rehospitalization, outpatient visits and telephone after discharge until got "success".The cause and time of death were recorded during follow-up.The cause of death within 6 months in these patients was analyzed.If the patient's death was due to an exacerbation of the central nerve system infection (assessed by radiography and CSF examination) and its associated complications, the cause of death was considered to be "CM-related".If the patient's death was due to malignancy, which caused organ dysfunction or associated complications, the death was considered to be "malignancy-related".The cause of death which cannot be attributed to the above two situations was considered as "other".

Statistical analysis
All statistical analyses were performed using SPSS (version 16.0, Chicago, IL, USA).Continuous data were analyzed using Student's t-test after con rming a normal distribution using the Shapiro-Wilk test.Mann-Whitney test was used if the data was not a normal distribution.Chi-square analysis was used to analyze categorical variables.Fisher's exact test was an alternative, if the data does not meet the chi-square requirements.Survival was de ned as the time from the diagnosis of CM to the death due to any cause or last follow-up.Survival curves were generated by the Kaplan-Meier product limit method.The difference in survival distribution was evaluated by the log rank test.

Demographic and clinical features
A total of 357 patients were included in this study.Thirty-seven patients were excluded according exclusion criteria mentioned above (Fig. 1).Among the enrolled patients, 232 (72.5%) were male and 88 (27.5%) were female, with an average age of 45.92 years.There were no signi cant differences in gender between malignancy and non-malignancy CM patients.CM patients with malignancy were older than non-malignancy patients (56.25 vs. 45.52,P 0.05) (Table 1).

Symptoms and time to admission
Headache and fever were the two most common initial symptoms, with incidence rates of 95.31% and 49.38%, respectively.The incidence rates of headache and fever were not signi cantly different between non-malignancy and malignancy patients (Table 1).Time to admission (time from initial symptoms to hospitalization and diagnosis of CM) were signi cantly different between non-malignancy and malignancy patients (Table 1).

CSF ndings
At the rst lumbar puncture, the CSF open pressure of patients with non-malignancy seemed higher than that of patients with malignancy but did not reach statistical difference (Table 1).Non-malignancy CM patients had lower cryptococcus counts than malignancy CM patients, but the difference was not statistically signi cant (Table 1).There were no signi cant differences in CSF white blood cell (WBC) count, protein, glucose, and chloride concentration between malignancy and non-malignancy CM patients (Table 1).

Outcomes
Most patients (302/308 and 12/12) completed the initial 2 weeks of antifungal treatment.Six patients died or were discharged from treatment within 2 weeks (4 death, 2 discharge).In the non-malignancy CM patients, after the initial 2 weeks of antifungal treatment, the clinical symptoms and signs of 97.68% patients improved, accompanied with decreasing CSF cryptococcus count, which represent the "response" to the treatment.
However, the response rate of malignancy CM patients was signi cantly lower than that of non-malignancy patients (66.67% vs. 97.68%).
Seven patients lost follow-up within 6 months, and 28 patients lost follow-up after 6 months.Mortality was nally evaluated in 305 patients at follow up for 6 months.Patients with malignancy had much higher 6-month mortality than non-malignancy patients (Table 1).At the end of this study, among the patients without malignancy, 35 patients died, 37 patients had not completed the therapy course yet, and 199 patients antifungal treatment acquired "success".Among the patients with malignancy, 5/12 (41.67%) died within 6 months, 6 patients antifungal treatment was "success" (one died of tumor recurrence 14 months after successful CM treatment), and one patient still took oral antifungal drugs for about 1 year and was in stable condition.Survival curves for non-malignancy CM and malignancy CM were shown in Fig. 2. Patients with malignancy had signi cant lower survival rate than non-malignancy ones (P < 0.001).

Characteristics of MBC and MAC patients
The clinical characteristics of MBC were shown in Table 2.The number of MBC cases was 7, age ranged from 31 to 68 years, with a median of 48.43 years and three of them were male.Six cases were epithelial-derived malignancy.Another case was diffuse large B lymphoma.The diagnosis time from malignancy to CM ranged from 2 to 60 months, with an average of 26.86 months.Most malignancies were treated by surgery and in stable condition.The treatment on CM in these patients was successful.The case with diffuse large B lymphoma has been undergoing antifungal treatment for approximately one year and has not yet cleared cryptococcus at the end of present study.

Discussion
Although CM is much more common in non-HIV patients with predisposing factors, including malignancy [2,12], there has been little information about the clinical features and prognosis of these patients.To our knowledge, the present study was the largest and most detailed series of CM cases in non-HIV patients with malignancy.In this study, we found that a longer time to diagnosis and worse prognosis in CM patients with malignancies and PET-CT might be a useful tool for screening malignancy in such patients.
Early researches have shown that malignancies, especially lymphomas [7][8][9], appear to be closely associated with cryptococcosis.Since then, there have been many case reports of cryptococcal infections as malignancy complications [13][14][15].These pioneering studies warned a close relationship between malignancy and CM, and this may be due to the malignancy-related immune alteration or lymphocyte-depleting chemotherapeutic regimens [9].Unlike previous studies that found CM in malignancy patients, our present study focused on CM patients, retrospectively analyzed and looked for potential malignant tumors.We found that most patients in our study had solid malignancies (except one lymphoma), and they were all epithelial origin.In immunocompromised genetically modi ed mice and immunode ciency populations, the incidence of malignancy has increased signi cantly [16][17][18].These phenomena were thought to be related to the decline in the ability of immunosurveillance.According to this perspective, CM and malignancy may share the same immune abnormality mechanism.
In current study, the incidence of CM with malignancy was 3.75%, which was much higher than that of the general population [19].Two recent retrospective studies have reported even higher rates of malignancy in CM patients [20,21].We found that malignancy was usually found in older CM patients, older than 50 years should be more vigilant about the possibility of malignancy.Up to 50% of CM patients with malignancy were hospitalized between 1 and 3 months after clinical symptoms appear, while 70.78% of non-malignancy CM patients were within one month of onset.This discrepancy suggested a signi cant difference in the severity of clinical symptoms between the two groups, which might be more prominent in patients without malignancy.
Comparing the intracranial pressure between the two groups did not reach a statistical difference, we considered the possible reason was that CM patients with malignancy were older.It is generally believed that the elderly are less sensitive to pain.Therefore, the clinical symptoms in CM patients with malignancy were atypical, which caused a delay in CM diagnosis.
Literature review analysis found that only less than 15% of patients with cryptococcus infection associated with malignancy were solid [9].However, recent studies have found that the incidence of CM patients with solid malignancies was roughly equivalent with haematologic [20,21].The present study described 12 malignancies in 320 CM patients, of which only one was haematologic.In the 7 MBC patients, the types of malignancy and the affected organs were varied, and they had good antifungal response and low mortality.In contrast, 5 MAC patients mainly involved the digestive and respiratory systems, with a mortality rate of 80%.We suggest that three main reasons might respond for the poor prognosis of MAC: 1) they were older, and senior has been con rmed to indicate a poor prognosis of CM [21,22], 2) when the malignancy was found, most of the patients lost the chance to treat the malignancy due to CM, 3) the patients and his family abandoned active therapy.
Early screening of malignancies is di cult [23].The sensitivity of different tests in screening of different types of malignancies is inconsistent.Of the 5 MAC patients in this study, only one patient found a lung cancer in routine chest CT.PET-CT is a non-invasive imaging test that has a unique advantage in the detection of early malignancies [24,25].In a variety of solid tumor and organ in ltrating leukemias, lesions of less than 5 mm can be found by PET-CT with 18 F-FDG radioactive probes [26,27].In the 104 patients who underwent PET-CT examinations, ve patients were prompted to have suspected malignancies, and 4 were eventually con rmed.Therefore, we recommend routine PET-CT examinations for CM patients older than 50 years in order to detect possible early malignancy in time.
Our study had limitation.In view of the economic cost, not all patients undergo detailed malignancy screening.
The possibility of missed malignancy diagnosis may exist.The clues of this possibility can be observed in the lower malignancy rate in without PET-CT patients than in PET-CT ones.

Conclusion
The results of present study suggested an increase in the incidence of solid malignancies in CM patients and those patients had a poor prognosis if the malignancies were newly diagnosed.These malignancies were relatively insidious, PET-CT might provide useful clue for nding the malignancies in CM patients older than 50 years.Flow chart displaying the enrollment of study participants.

Table 1
The clinical characteristics and laboratory data of the CM patients without or with malignancy.

Table 3
The clinical characteristics and prognosis of MAC patients.