Detection of pulmonary actinomycosis by metagenomic next-generation sequencing in a cancer patient receiving maintenance olaparib: A review and case report

Abstract

A 52-year-old female patient receiving olaparib maintenance treatment was admitted to hospital with a low fever and chest tightness. A CT scan of the patient's chest showed diffuse ground glass shadow or miliary nodular shadow in the bilateral lungs. Her inflammatory biomarkers were almost normal, except the slightly elevated C-reactive protein. Moreover, lymphocytes count obviously decreased. Empirical treatment did not relieve her symptoms, while traditional testing developed negative results. The results of metagenomic next-generation sequencing (mNGS) revealed the presence of a potential pathogen, Actinomyces odontolyticus (A. odontolyticus), in bronchoalveolar lavage fluid (BLAF). Once large-dosed penicillin G was administered, the fever returned to normal and chest tightness disappeared. Reexamination of chest CT revealed that the pulmonary lesions was almost absorbed. Our case demonstrated that mNGS is a novel approach to identify pathogens sensitively and accurately, especially for uncommon and atypical infection.

1. Introduction

Pulmonary infection (PI) is prevalent worldwide with high morbidity and mortality, especially in elder and immunocompromised individuals [1]. With the improvement in survival time and the extensive use of novel targeted agents and immunotherapy drugs for malignant tumors, the prevalence of immunosuppression is increasing in cancer patients [2]. Concerning incidence rates based on specific types of immunosuppression, the rate of bacterial pneumonia after remission induction chemotherapy in acute leukemia was 21% [3]. The pneumonia incidence after anti-tumor therapy in lung cancer and in head and neck cancer were 5% [3] and 10% [4], respectively. Therefore, patients undergoing anti-cancer treatment are susceptible to PI.

Poly (ADP-ribose) polymerase (PARP) inhibitor, which interferes with the repair of DNA single‐strand breaks via the base excision repair pathway [5], has been shown to be an effective targeted drug for ovarian cancer (OC) patients. Previous studies revealed that olaparib could upregulate the expression of PD-L1 in OC cells and promote cancer-associated immunosuppression [6]. About 20% of OC patients experience a decrease in neutrophils or lymphocytes when undergoing olaparib maintenance therapy [7]. Impaired immune function facilitates cancer individuals more vulnerable to infections. Although there is no reported incidence of PI in OC patients available, opportunistic PIs were found in individuals. For example, a case of pneumocystis pneumonia (PCP) was reported in a 77-year-old woman with OC who was administered olaparib as maintenance therapy [8].

Actinomycosis is typically regarded as an uncommon disease and its current prevalence rates are unknown [9]. Actinomycosis is mainly caused by Actinomyces species, which are more likely to infect the immunocompromised hosts, such as cancer individuals [10]. The diagnosis of actinomycosis can be challenging due to the nonspecific clinical manifestations, including cough, fever, hemoptysis, and weight loss. In addition, a low culture rate interferes with the identification of a pathogen [11]. Moreover, the low specificity of radiological features poses a diagnostic challenge as various CT findings have been observed, such as alveolar or lobar consolidation, pleural thickening or effusions, ground-glass opacities, and necrotic mass [12]. Many cases might be confused with lung abscess, tuberculosis, lung malignancy, and fungal infection (e.g., Aspergillosis) [[13], [14], [15]]. However, the actinomycosis should be immediately diagnosed as the reported mortality rates vary between 0 and 28% [10]. Among the agent responsible for actinomycosis, Actinomyces turicensis and Actinomyces meyerii represented 22.2% each; Actinomyces israelii, Actinomyces neuii, and Actinomyces radingae represented 11.1% each; Actinomyces odontolyticus (A. odontolyticus) constituted 5.6% [16]. A number of antibiotics with little or no activity against Actinomyces spp. include penicillinase-resistant penicillins, aztreonam, co-trimoxazole, cephalexin, metronidazole, fluoroquinolone, and aminoglycosides [17]. The key to actinomycosis treatemnt is the long-term use of high-dose antibiotics [18]. Penicillin G is the preferred medication [19]. Additional drugs involve amoxicillin/clavulanic acid and ampicillin. Alternatives include imipenem, azithromycin, doxycycline, clindamycin, and ceftriaxone in case of allergy or nonresponse to penicillin. Prolonged treatment is often necessary to prevent recurrence [20].

Laboratory sputum culture for bacteria usually takes at least 48 h [21]. Colony morphology, gram staining, and biochemical tests were performed to identify bacteria [22]. Theoretically, difficulty potentially existed in the diagnosis 48 h after onset. In addition, the use of antibiotics before admission disturbed the sensitivity of traditional culture in detecting pathogens [23]. Metagenomic next-generation sequencing (mNGS) is an unbiased detection technique that can detect all sequences of pathogen genomes within 24–48h and is especially suitable for identifying rare, novel, and atypical pathogens [24]. In contrast with conventional testing methods, mNGS presents significant advantages in conditional or polymicrobial infection. This further facilitates precise diagnosis and optimal treatment.

Herein, we have presented a case of OC receiving olaparib maintenance treatment with A. odontolyticus pneumonia, in which conventional tests were negative but mNGS revealed a positive finding. Finally, high-dose penicillin G was administered for actinomycosis and then the patient recovered.

2. Case presentation

A 64-year-old female was admitted to the hospital with a low fever and chest tightness for two weeks. Before admission to our hospital, her highest temperature was 38.0 °C. She was empirically administered levofloxacin (0.5g qd ivgtt), meropenem (1.0g qd ivgtt), and fluconazole (0.2g qd ivgtt) for a week in her first admission to a local hospital; however, her symptoms were not significantly relieved, nor did her temperature return to normal.

As a patient with platinum-sensitive, relapsed OC [25], she had been administered olaparib tablets 300 mg twice daily for five months. Her treatment response was evaluated as stable disease. Hypertension was the only comorbidity and blood pressure was stably controlled by the administration of amlodipine besylate 5 mg once daily. No other drugs were taken.

Laboratory testing was performed. Routine blood testing showed that the white blood cell (WBC) count was 5.18 × 10⁹/L with a proportion of 73.5% being neutrophils. The C-reactive protein (CRP) reached 25.53 mg/L and procalcitonin (PCT) level remained normal. The count of T cells (0.31 × 10^9/L), B cells (0.04 × 10^9/L), and NK cells (0.13 × 10^9/L) in the peripheral blood were remarkably decreased. Routine urine testing, and liver and kidney function were normal. All detected auto-immune disease-related indicators were negative. In addition, complements C3 and C4, glucose 6-phosphate isomerase, and thyroid function-related indicators, including triiodothyronine, thyroxine, free triiodothyronine, free thyroxine, thyroid stimulating hormone, and parathyroid hormone, were normal. Echocardiography ruled out endocarditis. Based on clinical experience, the initial diagnosis was identified as an infection of an unknown pathogen.

An initial treatment with azithromycin (0.5g qd ivgtt) had been empirically administrated for 3 days. The patient's symptoms did not respond, but she still had a fever with a peak of 38.4 °C. Serological testing was performed, including antibodies to respiratory syncytial virus, adenovirus, influenza A and B viruses, parainfluenza virus, Coxiella burnetiid, Chlamydia pneumoniae, Mycoplasma pneumoniae, and Legionella pneumophila. Both her peripheral blood and sputum specimens were examined by bacterial culture. However, all cultures and serological tests were negative for micro-organisms. Furthermore, her whole chest computed tomography (CT) showed diffuse ground-glass shadow in bilateral lungs (Fig. 1A–D). An invasive lung biopsy was refused by the patient. After a multidisciplinary discussion, mNGS were performed on blood and bronchoalveolar lavage fluid (BLAF) samples to detect potential pathogens. The DNA was extracted using BALF and its concentration was measured at 0.792 ng/μL. The concentration of constructed DNA library using BALF was 30.4 ng/μL. The Q30 of raw reads after sequencing (BALF-mNGS) reached 92.13%. The results of mNGS revealed a good quality of sequencing data, negative control and positive control. A total of 14,463,053 reads were obtained by BALF-mNGS. A validated bioinformatics algorithm was applied to facilitate the pathogen diagnosis. A total of 1152 unique reads, which mapped to the genome reference of A. odontolyticus, were obtained. The relative abundance and coverage of A. odontolyticus were 15.48% and 4.05%, respectively (Fig. 2). No other microbes were considered pathogenic by mNGS according the mNGS positive criteria. The results showed that the cause of PI was most likely to be A. odontolyticus. Then, penicillin G (8MIU q8h ivgtt) was prescribed. Consequently, the patient's temperature returned to normal (Fig. 3) and inflammation indicators (WBC, and CRP) gradually dropped (Fig. 4A). Lymphocyte count increased (Fig. 4B). The result of blood gas analysis, including decreased arterial blood pCO2 (Fig. S1) and increased arterial blood pH (Fig. S2), manifested the recovery of respiratory function. Two weeks later, her chest CT scan revealed obvious absorption of pulmonary lesions (Fig. 1E–H) and she was discharged without any complications. There month later, the patient's lung lesions were nearly absorbed in the follow-up CT reexamination (Fig. 1I-L).

Fig. 1.

Chest computed tomography findings of the patient on the first day of admission (a–d), sixth day of admission (e–h), and one month after discharge (i–m).

Fig. 2.

The metagenomic next-generation sequencing of bronchoalveolar lavage fluid sample of ovarian cancer patient. The coverage of Actinomyces odontolyticus by mNGS (a). The detected specific read number and the percentage of Actinomyces odontolyticus (b).

Fig. 3.

The patient's temperature results.

Fig. 4.

The patient's routine blood test results.

The detailed method of mNGS is as follows: A standardized blood and BALF specimen collection was performed. The cells in samples were removed by centrifugation and the supernatant was collected for the subsequent cell-free DNA extraction using QIAamp DNA Micro Kit (QIAGEN, Hilden, Germany) based on its manual. DNA libraries were constructed using QIAseq™ Ultralow Input Library Kit (QIAGEN, Hilden, Germany) according to the instructions. The quality of constructed libraries was assessed by Qubit (Thermo Fisher Scientific, MA, USA) and Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, USA). All qualified libraries were sequenced on Nextseq 550 platform (Illumina, San Diego, USA), with 75 bp single-end reads of approximately 20 million per library after sequencing. After removing adapters and short (<35 bp), low-quality (Q < 30), and low-complexity reads using bowtie2, the data were then mapped to human reference database (hg38) to filter out human host DNA reads. The remaining reads were finally blasted against Microbial Genome Databases (fttp://ftp.ncbi.nlm.nih.gov/genomes/) using Burrows-Wheeler Aligner software, followed by species annotation analysis using Least Common Ancestor method. Negative controls (sterile deionized water) and positive controls (synthesize fragments with known quantities) were established for each batch of experiments using the same wet lab procedures and bioinformatics analysis as the clinical samples. The reads number and reads per million (RPM) of each detected microbe was calculated. For detected microbes, including bacteria (M. tuberculosis excluded), fungi (Cryptococcus excluded), and parasites, a positive mNGS result was given when its coverage ranked in the top 10 of similar microbial species (or genera) and was absent in the negative control ("No template" control, NTC) or when its ratio of RPM between sample and NTC (RPM_sample/RPM_NTC) > 10 if RPM_NTC≠0. For viruses, M. tuberculosis, and Cryptococcus, a positive mNGS result was considered when at least one unique read was mapped to species level and absent in NTC or RPM_sample/RPM_NTC > 5 when RPM_NTC≠0.

3. Discussion

Here, we presented a case report in which mNGS promoted clinicians quickly and accurately to diagnose pulmonary actinomycosis in an OC patient undergoing olaparib. This patient was hospitalized with a low fever and chest tightness. CT showed diffuse ground-glass shadow in bilateral lungs. Traditional culture and serological testing did not detect the causative agent, while mNGS identified A. odontolyticus in the BLAF specimen. Combined with her clinical characteristics, her diagnosis was regarded as an A. odontolyticus pneumonia. Similarly, a case of pulmonary actinomycosis caused by Actinomyces graevenitzii was reported, which was also diagnosed based on the positive criteria of mNGS on BLAF sample [26]. Considering the patient's drug-allergy history, a 6-month course of azithromycin was administered and the lung lesion was remarkably absorbed. As for histopathological examination, the transbronchial biopsy of a mass lesion in the right lung showed inflammation and mini necrosis of the bronchial mucosa [26]. In our study, an invasive lung biopsy was refused by the patient.

The gold standard for diagnosing pulmonary actinomycosis is histological examination (e.g., lung biopsy). However, mNGS provide a rapid and novel diagnostic method without invasive lung procedures, enabling patients to receive early treatment for this infectious disease. Compared with conventional diagnostic methods, mNGS has many advantages, but also some deficiencies. A prominent advantage is that mNGS, as a completely unbiased technology, detects a broader spectrum of pathogens (bacteria, fungi, viruses, and parasites) in the samples with a single assay that is not dependent on empirical clinical judgment [27]. Thus, mNGS can discover unknown pathogenic microorganisms that PCR cannot identify, especially for rare infectious pathogens. For the identification of viruses or parasites, the role of traditional test methods is really limited. Moreover, mNGS is suitable for various types of specimens, including tissue, sputum, bronchoalveolar lavage, pleural effusion, peripheral blood, and cerebrospinal fluid, and can be applied in the clinical practices of mixed infection, conditional infection, rare pathogen infection, and severe infection, especially in immunosuppressive hosts [[28], [29], [30]]. In addition, mNGS does have an advantage in decreasing the time of pathogen detection and improving diagnostic sensitivity [29]. Bacterial cultivation is generally time-consuming. mNGS has the capability of quickly and effectively identifying the different kinds of pathogens in the specimen within 24–48 h and improving the clinical diagnosis rate [24].

Plenty of studies have proved the successful implementation of mNGS in the diagnosis of infectious diseases. Most publications are case reports of pathogen infection, such as severe pneumonia and meningitis caused by Chlamydia psittaci [31], fascioliasis hepatica [32], endocarditis caused by Aspergillus flavus [33], encephalitis caused by Balamuthia mandrillaris [34], and pyogenic liver abscess caused by Klebsiella pneumoniae [34], promptly diagnosed by mNGS. Furthermore, mNGS has exhibited a higher diagnosis rate and sensitivity of pathogen identification compared to traditional microbiology techniques for infectious diseases, such as tuberculosis [35], Scrub typhus [36], meningitis and encephalitis [37], and post-liver transplantation infection [38] in some multi-sample research. However, it is generally recognized that mNGS can not likely replace conventional techniques in the short term for the following reasons: i) the high price of detection limits the clinical application of mNGS; ii) traditional serological tests will continue to play a key role in the diagnosis of emerging infectious diseases; ii) Polymerase chain reaction can be applied with rapid turnaround times; iii) functional assays such as culture and drug-sensitive test will likely always play an irreplaceable part in research [39]. In our case, the microbial cell-free DNA was detected by mNGS on blood samples and BLAF samples. A positive result was observed on BLAF. Our test result was consistent with previous studies, which revealed that BALF mNGS was more sensitive than blood mNGS in detecting PI [40]. Infection caused by Actinomyces species is usually recommended to be treated with penicillin injection for a long course. In fact, in the present case, the patient recovered from pulmonary actinomycosis with administration of high-dose penicillin G.

4. Conclusions

Clinicians should pay more attention to pulmonary actinomycosis and bring it as one of the differential diagnoses for refractory pulmonary diseases into the usual management, especially in immunocompromised patients. Our case illustrated the potential application of mNGS in detecting pathogenic microorganisms in samples that were not detected by traditional techniques. It is believed that in the near future, with the continuous reduction of sequencing costs, mNGS will be more and more widely used in clinical practice, benefiting more doctors and patients.

Consent

Written and verbal consent was obtained from the patient for this report.

Ethical approval

Ethical approval was obtained from the Ethics Committee of Hangzhou Cancer Hospital.

5. Authors'contributions

LWN and ZW collected the clinical data and drafted the manuscript. JH revised the manuscript. All authors read and approved the final version of the manuscript.

Funding

This case report was funded by grants from Hangzhou Medical and Health Science and Technology Project (A20200019).

6. Availability of data and materials

The dataset supporting the conclusions of this article is included within the article and its additional files.

Declaration of Competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: DR AC Amit Chopra

Appendix A. Supplementary data

The following are the Supplementary data to this article.

figs1.

figs2.