Vancomycin-induced thrombocytopenia in endocarditis: A case report and review of literature

Abstract

BACKGROUND

Thrombocytopenia is a serious complication in the medical practice of numerous drugs. Vancomycin is frequently used for the prophylaxis and treatment of suspected or identified methicillin-resistant positive infections. Several cases with vancomycin-induced thrombocytopenia (VIT) have been reported. However, these have rarely been extensively reviewed. The present report describes a case of VIT in endocarditis, and reviews all VIT cases reported in the literature.

CASE SUMMARY

A 26-year-old male diagnosed with infective endocarditis was admitted. The patient was treated with multiple drugs, including vancomycin, which was initially intravenously given at 1000 mg every 12 h and subsequently at 500 mg every 8 h on day 3. On day 11, the platelet count decreased to 51 × 10⁹/L, vancomycin was switched to 500 mg every 12 h, and platelet transfusion was given. On day 17, the platelet count dropped to 27 × 10⁹/L, and platelet transfusion was administered again. On day 23, vancomycin was adjusted to 500 mg every 8 h as the trough concentration dropped to the minimum effective concentration. On day 33, the platelet count declined to approximately 40 × 10⁹/L. After platelet transfusion, the platelet count rebounded to 90 × 10⁹/L on day 35 but dropped again to 42 × 10⁹/L on day 43. Based on the time-to-platelet count curve and Naranjo’s Adverse Drug Reaction Probability Scale score, VIT was suspected. After vancomycin discontinuation and platelet transfusion, the platelet count gradually normalized.

CONCLUSION

The diagnosis of VIT can be achieved through the time-to-platelet count curve and Naranjo’s Adverse Drug Reaction Probability Scale score. The platelet count cannot be normalized simply by platelet transfusion alone, and vancomycin discontinuation is essential.

Key Words: Case report; Vancomycin; Thrombocytopenia; Endocarditis; Platelet transfusion; Literature review

Core Tip: The present report describes a case of vancomycin-induced thrombocytopenia (VIT) in endocarditis and reviews all VIT cases reported in the literature, in terms of indications, diagnosis, management, and potential molecular mechanisms. The present case highlights the importance of accurate diagnosis of VIT in endocarditis, which can be achieved through the time-to-platelet count curve and the Naranjo’s Adverse Drug Reaction Probability Scale score. The platelet count cannot be normalized simply by platelet transfusion alone, and the discontinuation of vancomycin is essential. Due to the anamnestic response to vancomycin re-exposure and the formation of vancomycin-dependent antiplatelet antibodies, the re-administration of vancomycin should be avoided.

INTRODUCTION

Thrombocytopenia, which is generally defined as a platelet count of less than 100 × 10⁹/L of blood or a decrease in the platelet count of 20%-50% from baseline[1,2], has been considered as an adverse reaction correlated to drugs used in the treatment of various diseases[3]. Drug-induced thrombocytopenia is mostly caused by immune-mediated platelet degradation and is always drug dependent, indicating that the platelet count can return to baseline levels after the discontinuation of medications[4]. Although it has been reported that approximately 10 million persons per year are suspected to suffer from drug-induced thrombocytopenia[5], its incidence has not been well-defined[4].

Antibiotic-induced thrombocytopenia has been documented[3]. The commonly reported antibiotics that induce thrombocytopenia are β-lactam antibiotics agents (such as penicillin, nafcillin, ticarcillin, cefazolin, cefuroxime, ceftriaxone, and piperacillin) and linezolid[6]. Vancomycin is a first-generation glycopeptide antibiotic that is often used for the prophylaxis and treatment of suspected or identified methicillin-resistant infections, such as Staphylococcus aureus infection[7], and has been considered as an uncommon cause of thrombocytopenia[6]. Several clinical observational studies have suggested that the incidence of vancomycin-induced thrombocytopenia (VIT) might be higher than that induced by linezolid[1]. However, the incidence of VIT may have been overestimated, because the definition for thrombocytopenia used among these studies vary. A platelet count of less than 150 × 10⁹/L was used in two studies[8,9], and a decrease in platelet count of at least 50% from baseline was used in another study[1]. In addition, the clinical manifestations, diagnosis, and management of VIT have not been well-established, and the underlying molecular mechanisms by which vancomycin induces thrombocytopenia needs to be further elucidated[10]. Over the past two decades, cases with VIT have been continuously reported[11]. However, these cases have rarely been extensively reviewed.

In June 2019, a young male patient with endocarditis was admitted to our hospital, and he developed VIT after vancomycin therapy during the treatment for endocarditis. The present report describes this case and extensively reviews all VIT cases reported in the literature, in terms of indications, diagnosis, management, and potential molecular mechanisms.

CASE PRESENTATION

Chief complaints

A 26-year-old Chinese male presented with dyspnea, fatigue, arrhythmias, fever, and cough.

History of present illness

The patient was admitted to the hospital on June 6, 2019 and diagnosed with infective endocarditis, heart failure, and ventricular septal defect, with a series of manifestations of inflammatory response syndrome.

History of past illness

He had a medical history of uncontrolled hypotension and rheumatic heart disease.

Personal and family history

He denied any family history.

Physical examination

He had a body temperature of 39.1 °C and a heart rate of 100 beats per min.

Laboratory examinations

Initial laboratory testing showed no abnormality.

Imaging examinations

The electrocardiosignal data revealed a high echo of the tricuspid valve, suggesting a neoplasm.

FINAL DIAGNOSIS

Infective endocarditis, heart failure, and ventricular septal defect.

TREATMENT

During hospitalization, the patient was treated with multiple courses of drug therapy, which included vancomycin, omeprazole (40 mg q.d. for 10 d), ceftazidime (2 g t.i.d. for 9 d) and metoprolol (50 mg q.d. for 11 d). Based on the bacterial culture and drug susceptibility test, methicillin-resistant Staphylococcus aureus was identified with the minimum inhibitory concentration of less than 2 mg/L. Therefore, according to The Clinical Practice Guidelines by the Infectious Diseases Society of America for the Treatment of Methicillin-Resistant Staphylococcus aureus Infections in Adults and Children[12], vancomycin was prescribed for the patient with infective endocarditis. Initially, the patient was intravenously treated with 1000 mg of vancomycin every 12 h and 960 mg of benzylpenicillin every 8 h on day 2 for suspected gram-positive infections. The blood cultures were negative for two consecutive tests. Then, the intravenous administration of vancomycin was changed to 500 mg for every 8 h on day 3, for a trough concentration of 25.87 µg/mL beyond the upper limit of 20 µg/mL, and the platelet count was maintained within 100-110 × 10⁹/L for 7 d. On day 11, the administration of vancomycin was switched to 500 mg every 12 h, because the platelet count decreased to 51 × 10⁹/L and the peak concentration of vancomycin reached 57.2 µg/mL. Platelet transfusion (1 U) was given on day 11, and the platelet count slightly increased to 64 × 10⁹/L on day 14, but dramatically dropped afterwards to 27 × 10⁹/L on day 17. Then, platelet transfusion (1 U) was given again on day 17, and the platelet count steadily increased up to 67 × 10⁹/L on day 31. The dose of vancomycin was adjusted to 500 mg every 8 h on day 23 when the trough concentration of vancomycin dropped to 8.86 µg/mL in order to ensure the minimal effective treatment concentration (10 µg/mL). However, on day 33, the hemoglobin level sharply dropped from 101 g/L to 54 g/L, the platelet count continuously declined to approximately 40 × 10⁹/L, and platelet transfusion (1 U) was given. The platelet count rebounded to 90 × 10⁹/L on day 35 but dropped to 42 × 10⁹/L again on day 43. Based on these above observations, and along with the time-to-platelet count curve that illustrated the decline in platelet level after the administration of vancomycin (Figure 1), VIT was suspected, vancomycin was discontinued and replaced with daptomycin, and platelet transfusion (1 U) was given on the same day. The platelet count level increased and reached up to 120 × 10⁹/L on day 50 and remained stable thereafter.

Figure 1 Time-to-platelet count curve during hospitalization. Vancomycin was initiated on day 2 and continued on day 43, as indicated by the downward arrows. Platelet transfusion was given on day 11, 17, 33, and 43, as indicated by the upward arrows.

The concurrent antibiotic, benzylpenicillin, was continuously used during the period with little effect on the platelet count. The respiratory rate, white blood cell count, and hemodynamic variables remained stable, and the microbiological cultures from the sputum and urine revealed negative results, implying that no infection was correlated with thrombocytopenia for the present patient. The other causes of thrombocytopenia, including thrombotic thrombocytopenic purpura, hemolytic anemia, disseminated intravascular coagulation, and platelet-clumping pseudothrombocytopenia, were also excluded, because the results of the blood smear, bilirubin, and hemoglobin were normal. There have been no reports on the occurrence of drug-induced thrombocytopenia caused by other medications during treatment, including omeprazole, ceftazidime, and metoprolol. Therefore, a diagnosis of VIT was initially established using the time-to-platelet count curve. In addition, the Naranjo’s Adverse Drug Reaction Probability Scale score (Naranjo score) for vancomycin was 8, indicating the probable correlation between thrombocytopenia and vancomycin.

OUTCOME AND FOLLOW-UP

The patient was discharged from the hospital.

DISCUSSION

A literature search for all case reports on VIT published in English from April 1985 to June 2020 was performed on PMC/PubMed, Scopus, and Web of Science. Keywords, which included vancomycin and thrombocytopenia, were used to search the titles and abstracts. References of relevant case reports, according to the brief research, were also carefully checked. The demographic data (e.g., age and gender), platelet count, diagnosis, and treatment were recorded and analyzed.

A total of 36 articles that reported 37 cases with VIT were identified in the literature. Thus, the data of 38 cases, which included the case described in the present report, were reviewed and are summarized in Table 1. Among these cases, 14 cases were females and 24 cases were males. The mean age of these cases was 51.19 ± 26.98 years-old (range: 3 mo to 81-years-old). Furthermore, four children were under 2-years-old, and 36 cases were adults 26-81 years-old. The most common indications for the treatment with vancomycin were endocarditis (n = 6), sepsis (n = 6), pneumonia (n = 5), and cellulitis (n = 2). These were followed by other various indications, such as prosthetic infection, hemodialysis, bone graft, acute pancreatitis, etc. The diagnosis of VIT mainly relied on the platelet count, with or without the Naranjo score[13]. The test results for the vancomycin-dependent antibodies were available for 19 patients, and all except three were positive. Among the 27 patients with available bleeding information, 21 patients exhibited bleeding signs. The treatments for these cases comprised of platelet transfusion, prednisone, dexamethasone, immunoglobulin, and daptomycin, and the discontinuation of vancomycin. For the six cases with endocarditis, VIT was diagnosed as probable by the Naranjo score. The information on bleeding signs was available for four cases, and three of these cases presented with bleeding signs. The presence of vancomycin-dependent antibodies was examined only in two patients, and positive results were obtained in one of them. After vancomycin discontinuation, VIT was generally resolved for all patients within 4-15 d. Despite clinical significance revealed by the present literature review, it must be clearly stated that this is only a narrative review, which possesses all the limitations of the selection of papers. Thus, a methodologically acceptable systematic review should be carried out when more data with high quality are available in order to draw a firm conclusion.

Table 1 Summary of case reports of vancomycin-induced thrombocytopenia.

Author (publication year)	Age (yr)	Gender	Indication	Platelet count (109/L)1	Naranjo Score	Treatment	Time for resolution (d)2	Vancomycin-dependentantibodies	Bleeding sign
Present case	26	M	Endocarditis	27	Probable	Platelet transfusion and daptomycin	12	NA	+
MacDougall et al[14] (2020)	81	M	Prosthetic infection	2	NA	Prednisone, immunoglobulin, and daptomycin	10	+	-
Ajit et al[27] (2019)	61	M	Hemodialysis	< 10	NA	Prednisone, immunoglobulin, eltrombopag and plasma exchange	15	+	+
Getz et al[18] (2019)	71	M	Transverse myelitis	8.6	Definite	Dexamethasone and immunoglobulin	8	+	+
Chen et al[16] (2018)	0	F	Purulent meningitis	8	Probable	Methylprednisolone	4	+	+
Danieletto et al[17] (2017)	57	F	Bone graft	9	Probable	Tigecycline	7	-	-
Kalra et al[28] (2016)	0.7	M	Sepsis	30	NA	Platelet transfusion	3	NA	-
Schueler et al[15] (2016)	55	M	Purpura	1	NA	Prednisone and immunoglobulin	3	NA	+
Yamanouchi et al[29] (2016)	72	F	Sepsis	2	NA	Steroid therapy and carbapenem	7	+	+
Lobo et al[13] (2015)	67	M	Pneumonia	2	Probable	Platelet transfusion, methylprednisolone and immunoglobulin	3	+	+
Ahmed et al[30] (2015)	63	M	Diabetic foot	2	Probable	Daptomycin	10	+	-
Candemir et al[10] (2013)	54	F	Hematoma	42	Probable	Platelet transfusion and daptomycin	4	NA	NA
Wetzel et al[31] (2013)	64	F	Sepsis	7	Probable	Prednisone and immunoglobulin	8	NA	+
Ruggero et al[32] (2012)	41	M	Pneumonia	15	Definite	Daptomycin	5	NA	+
Arnold et al[23] (2013)	66	F	Endocarditis	4	Probable	Platelet transfusion and immunoglobulin	5	+	+
Rowland et al[33] (2013)	51	M	Acute pancreatitis	9	Probable	Only vancomycin discontinuation	5	NA	+
Anand et al[34] (2011)	54	M	Cellulitis	1	Possible	Corticosteroids and daptomycin	2	+	+
Ganly et al[22] (2011)	67	M	Sepsis	2	NA	Platelet transfusion and immunoglobulin	8	NA	+
Shah et al[11] (2009)	60	M	Shoulder infection	9	Possible	Platelet transfusion	3	+	+
Lee et al[35] (2009)	76	M	Diabetic foot	< 15	Possible	Platelet transfusion and teicoplanin.	5	+	-
Apiwattanakul et al[36] (2008)	16	F	Endocarditis	123	Possible	Prednisone and cefotaxime	5	NA	-
Kenney et al[20] (2008)	61	M	Gangrene and bacteremia	3	Probable	Platelet transfusion and immunoglobulin	4	+	+
Pauldine et al[37] (2008)	60	M	Pneumonia	10	Possible	Platelet transfusion	12	+	+
Dilli et al[38] (2008)	0.6	M	Neonatal respiratory distress	41	NA	Only vancomycin discontinuation	3	-	NA
O’Donnell et al[21] (2007)	56	F	Prosthetic infection	2	Possible	Immunoglobulin	7	+	+
Bay et al[39] (2006)	2	M	Pneumonia	11	NA	Only vancomycin discontinuation	4	NA	NA
Winteroll et al[24] (2004)	72	M	Sepsis	3	Probable	Platelet transfusion	10	+	+
Peel et al[40] (2003)	45	M	Peritonitis	5	Probable	Platelet transfusion and prednisone	4	NA	+
Marraffa et al[41] (2003)	50	M	Endocarditis	10	Probable	Clindamycin and methylprednisolone	5	NA	+
Rocha et al[42] (2002)	38	F	Prosthetic infection	68	Possible	Only vancomycin discontinuation	4	NA	NA
Shahar et al[43] (2000)	43	F	Surgery-site infection	118	Probable	Only vancomycin discontinuation	3	NA	NA
Govindarajan et al[44] (1999)	72	M	Epidural abscess	13	Probable	Trimethoprim/sulfmethoxazole	10	NA	NA
Kuruppu et al[45] (1999)	72	F	Endocarditis	14	Possible	Only vancomycin discontinuation	5	-	NA
Mizon et al[46] (1997)	71	F	Sepsis	< 10	Probable	Antibiotics withdrawn	4	+	+
Zenon et al[47] (1991)	54	M	Cellulitis	17	Probable	Antibiotics withdrawn	7	NA	NA
Christie et al[48] (1990)	73	F	Pneumonia	14	Probable	Platelet transfusion	2	+	NA
Christie et al[48] (1990)	31	M	Fever	< 10	Possible	Platelet transfusion	11 (not continued)	+	NA
Carmichael et al[49] (1986)	42	F	Endocarditis	82	Possible	NA	9	NA	NA

¹The lowest value during the vancomycin treatment.

²The time from vancomycin discontinuation to the resolution of vancomycin-induced thrombocytopenia. +: A positive result; -: A negative result. F: Female; M: Male; NA: Not available or not tested.

The present patient with infective endocarditis was diagnosed with VIT, which rapidly progressed. The early diagnosis of VIT was performed using the time-to-platelet count curve and Naranjo score. During treatment, the repeated platelet transfusions failed to increase the platelet levels. Subsequently, simply switching vancomycin to daptomycin returned the platelet count close to the baseline level.

The early diagnosis of VIT is quite difficult due to many suspected causes, such as severe infection, disseminated intravascular coagulation, heparin, and other medications. The correct diagnosis can be masked by the simultaneous administration of medications, which can cause drug-induced immune thrombocytopenia[14] or a complicating disease, such as primary idiopathic thrombocytopenic purpura or chronic hepatitis C[15]. Therefore, the early recognition of VIT was of key importance for the following treatment. According to previous studies[6], the investigators used the platelet count at different time points as the initial diagnosis of VIT, which can reveal a definite time-independent relationship with drug administration. All other causes of thrombocytopenia were excluded. Furthermore, the Naranjo score, which has been applied to help with the diagnosis of VIT[16-18], also indicated a probable correlation between thrombocytopenia and vancomycin. Noticeably, endocarditis represents the most common indication that is associated with VIT. The probable correlation between thrombocytopenia and vancomycin can be made through the Naranjo score for all cases, indicating that the Naranjo score is a useful tool to initially diagnose VIT in patients with endocarditis. A bleeding sign is also a useful clinical manifestation to which attention should be given in order to make a definite diagnosis, since most cases would present bleeding signs when VIT occurs. This is particularly true in patents with endocarditis, based on the analyzed data. However, the value of the detection of vancomycin-dependent antibodies in the diagnosis of VIT in patients with endocarditis remains to be elucidated due to the very limited data available.

The critical management after suspicions of VIT is to determine whether to continue or discontinue the vancomycin therapy. The seriousness of the bleeding and the dropping trend of the platelet count should be carefully considered and closely monitored when continuing the vancomycin therapy. However, once VIT is diagnosed, the vancomycin must be stopped, and the platelet transfusion should be taken into consideration as a supplemental treatment for some patients with various success rates, although there are still some transfusion-resistant patients[10,13,19,20]. If vancomycin is not stopped, the survival time of the transfused platelets would be obviously reduced, and the platelet transfusion will not always expectedly increase the platelet count of affected patients[3]. In any of these circumstances, platelet transfusion should be considered for severe thrombocytopenia with a platelet count of below 20 × 10⁹/L and bleeding[19]. In addition to platelet transfusion, therapies with intravenous corticosteroids, intravenous immunoglobulins, rituximab, and plasma exchange have been shown to be beneficial in some cases after vancomycin discontinuation[3,21-23]. However, these medical approaches may not always be effective. In the present case, VIT was effectively resolved by platelet transfusion with the switching of vancomycin to daptomycin.

The mechanisms of VIT remains unclear. The formation of different drug-dependent platelet antibodies, including hapten-dependent, quinine-type, fiban-type and drug-specific antibodies, and other platelet destruction antibodies, are the widely documented mechanisms for VIT[11,21,24]. In contrast, naturally occurring antibodies may not contribute to VIT[25]. It has been postulated that vancomycin can bind to platelet glycoproteins and induce the generation of antibodies, which can attach to the drug-platelet complex, resulting in cell lysis[26]. In the present case, the platelet count gradually dropped after the administration of vancomycin (1000 mg per day) for 7 d. Although the vancomycin was changed to a low-dose (500 mg per day), the platelet count continued to drop. It was postulated that the persistent thrombocytopenia might be due to the anamnestic response for vancomycin re-exposure, and vancomycin-dependent antiplatelet antibodies were formed once vancomycin was re-used.

CONCLUSION

The present case highlights the importance of the accurate diagnosis of VIT in patients with endocarditis. The time-to-platelet count curve and Naranjo score are useful tools for the diagnosis of VIT. The platelet count cannot be normalized simply by platelet transfusion alone. Instead, the discontinuation of vancomycin or switching vancomycin to other antibiotics, such as daptomycin, is essential for effectively treating VIT. Due to the anamnestic response to vancomycin re-exposure and the formation of vancomycin-dependent antiplatelet antibodies, the re-administration of vancomycin should be avoided.