What this study adds
The overall prevalence of tigecycline-induced pancreatitis (TIP) was relatively low in a real-world setting. Serum amylase and lipase levels should be closely monitored in patients with comorbid renal insufficiency during tigecycline treatment.
Comorbid renal insufficiency was identified as an independent risk factor for TIP.
Tigecycline significantly decreased the infection indices in non-TIP patients without worsening pancreatitis. Tigecycline is safe and efficient for treatment of pancreatitis with intra-abdominal infection.
1 | INTRODUCTION
Tigecycline is the first member of the glycylcycline class of antimicrobials to be a derivative of minocycline, having a 9-t-butylglycylamido group added the first carbon of the minocycline D ring.1 This new antibiotic inhibits bacterial protein synthesis by binding to the 30S ribosomal subunit,2and overcomes the two main mechanisms of bacterial resistance: ribosomal protection and efflux pump.3 The increasing incidence of multidrug-resistant (MDR) and extensively drug-resistant (XDR) bacterial pathogens is a major public health concern. Owing to its broad-spectrum antibacterial activity, tigecycline is widely used for treatment of complicated skin and soft tissue infections, intra-abdominal infections, and community-acquired pneumonia caused by MDR pathogens, especially carbapenem-resistant bacteria.4,5
Tigecycline was approved by the US Food and Drug Administration in June 2005 and by Chinese government in 2011. With the increasing clinical use of tigecycline, adverse effects, such as nausea, vomiting, and diarrhea, have become inevitable. Acute pancreatitis (AP) has been reported in a limited number of case reports and can be fatal if unrecognized.6,7 In 2006, AP was added to the list of side effects for tigecycline after post-marketing surveillance. Currently, there are no predictive factors for AP occurrence and no sensitive identification methods for AP prevention during tigecycline treatment. The purpose of this study was to analyze the risk factors and clinical characteristics of tigecycline-induced pancreatitis (TIP) and to evaluate the safety and efficiency of tigecycline use in patients with previously diagnosed pancreatitis before tigecycline use.
2 | METHODS
2.1 | Study design
This single-center retrospective case-control study was conducted at Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, a 4600-bed tertiary hospital. The study was approved by the Ethics Committee of Tongji Hospital. The need for informed consent from the patients was waived because of the retrospective nature of the study.
2.2 | Patients
Consecutive patients treated with intravenous tigecycline for >3 days in our hospital were screened from January 1, 2017, to December 31, 2020. The patients were divided to juveniles (<18 years of age) and adults (≥18 years of age). Data for the screened patients were exported from the Department of Computer Center and evaluated by two independent researchers. The inclusion criteria were: patients with duration of tigecycline therapy for at least 3 days. The exclusion criteria were: (1) patients who lacked complete medical records and laboratory examination results; (2) patients who were readmitted to our hospital several times and treated with tigecycline (only first admission was included); (3) patients with history of AP episode, chronic pancreatitis, autoimmune pancreatitis, or pancreatic cancers; and (4) patients administered definite or probable drugs associated with pancreatitis (such as acetaminophen, azathioprine)8 during tigecycline treatment, with exception of octreotide because most patients with pancreatitis were administered octreotide. The selection of the patients is depicted in Figure 1. Patients were evaluated for the prevalences of comorbid diseases including hypertension, coronary heart disease, diabetes mellitus, chronic obstructive pulmonary disease, bronchial asthma, renal insufficiency (including acute and chronic renal insufficiency), liver insufficiency (including hepatitis, cirrhosis, and autoimmune liver disease), solid organ tumor, and autoimmune disease based on the diagnoses in their discharge records.
2.3 | AP evaluation
Patients with a discharge diagnosis of AP were evaluated for the following common causes: biliary stones, alcohol exposure, endoscopic retrograde cholangiopancreatography (ERCP), trauma, hypertriglyceridemia, hypercalcemia, tumor, and autoimmune disease. AP patients who had one of the common causes and developed AP before tigecycline treatment were classified as non-tigecycline-induced pancreatitis (non-TIP). For these patients, we recorded length of hospital stay, cause of AP, severity of AP,9 any surgery performed to treat AP, modified computed tomography (CT) severity index score,10 amylase and lipase levels, and infection indices before and after tigecycline treatment. These data were used to examine the safety and efficiency of tigecycline treatment in non-TIP patients.
Patients who did not have the above causes were evaluated for the presence of TIP using a probability assessment scale for drug-induced pancreatitis based on the Naranjo scale that was developed by Weissman et al11 in 2020. According to the requirements for the scale, we completed a form involving 10 questions for each patient without the common causes and obtained a summative score (>9, highly probable; 6–8, probable; 3–5, possible; and ≤2 doubtful). Patients with TIP were evaluated for the Naranjo score, clinical symptoms, severity of AP,9 modified CT severity index score,10 average time from tigecycline therapy to symptom onset, serum amylase and lipase levels, and times for symptoms and amylase and lipase levels to return to their normal ranges.
Patients without a discharge diagnosis of AP were classified as non-pancreatitis.
2.4 | Data collection
The following data were extracted from the electronic medical records: sex, age, admission department, main and secondary diagnoses, baseline routine blood tests (white blood cell count, percentage of neutrophils, red blood cell count, hemoglobin, platelet count), liver function tests (alanine aminotransferase, aspartate transaminase, albumin, globulin, total bilirubin, direct bilirubin, alkaline phosphatase [ALP], γ-glutamyl transpeptidase), renal function tests (blood urea nitrogen, serum creatinine, uric acid, bicarbonate radical, estimated glomerular filtration rate), serum lipid parameters (total cholesterol, triglyceride, low-density lipoprotein), electrolyte indices (potassium, sodium, calcium, corrected calcium), infection indices (high-sensitivity C-reactive protein, erythrocyte sedimentation rate, procalcitonin), coagulation markers (prothrombin time [PT], fibrinogen, activated partial thromboplastin time [APTT], d-dimer), amylase, and lipase. Tigecycline treatment data (loading dose, maintenance dose, treatment duration) and combination therapy information were recorded. The total dose of tigecycline was calculated for each patient. Based on the discharge diagnoses and medical records, sites of infection or indications for tigecycline therapy were classified as follows: lung, chest, cranium, endocarditis, skin or soft tissue, abdomen, intestine, biliary tract, liver abscess, urogenital tract, mixed infections, sepsis, septicemia, septic shock, agranulocytosis, hemophagocytic syndrome, multiple organ failure, trauma, and major surgery-related infection.
2.5 | Statistical analysis
All statistical analyses were performed using SPSS version 22.0 (IBM Corporation, Somers, NY, USA). Presence of a parametric distribution was evaluated by the Kolmogorov–Smirnov test. Parametric and categorical data were presented as mean ± standard deviation (SD) and rate, respectively. Non-parametric data were presented as median and interquartile range (IQR). Comparisons between two groups were made by Student’s t-test for parametric data and the Mann–Whitney test or Wilcoxon test for non-parametric data. The chi-square test was used for categorical variables. Univariate and multivariate analyses were performed using logistic regression models to identify independent risk factors for TIP. All biologically plausible variables that showed significance at P < .10 in univariate analyses were entered into a multivariate forward logistic regression analysis. A value of P < .05 was considered to indicate statistical significance.
3 | RESULTS
3.1 | Demographic data
After the screening process, a total of 3910 patients were included in the study. The enrolled patients comprised 3823 adult patients aged ≥18 years and 87 juvenile patients aged <18 years. Among the adult patients, there were 103 patients with a discharge diagnosis of AP. Based on the medical records, 82 patients were diagnosed with AP before tigecycline therapy and classified as non-TIP, 21 patients were classified as TIP, and 3720 patients were classified as non-pancreatitis.
In the adult patients, tigecycline was most commonly used in the hematology department followed by the intensive care unit and the organ transplantation department (Supplementary Figure 1). The demographic data for TIP and non-pancreatitis are shown in Table 1. The median length of hospital stay was longer and the prevalence of comorbid renal insufficiency was higher in the TIP group compared with the non-pancreatitis group. There were no significant differences in the sex distribution and prevalences of other comorbid diseases between the two groups.
The juvenile patients comprised 49 boys (56.3%) and 38 girls (43.7%) with a mean age of 11.74 ± 6.0 years.
3.2 | Indications for tigecycline
In the adult patients, the most common indication for tigecycline therapy was pneumonia, followed by agranulocytosis and major surgery-related infection. There were no significant differences in the classifications of the indications for tigecycline therapy between the TIP group and the non-pancreatitis group (Supplementary Table 1).
For the juvenile patients, the most common indication for tigecycline therapy was agranulocytosis (36.8%), followed by pneumonia (11.5%) and sepsis (10.3%).
3.3 | Prevalence of TIP
In the adult patients, the prevalence of TIP was 21/3741 (0.56%). The prevalences of TIP in the hematology department, intensive care unit, and organ transplantation department were 7/879 (0.80%), 2/501 (0.43%), and 5/392 (1.28%), respectively. Compared with the total patients, there were no significant differences in the prevalences of TIP in the hematology department (0.80% vs. 0.56%, P = .419) or intensive care unit (0.43% vs. 0.56%, P = .722), while the prevalence in the organ transplantation department showed a tendency to be higher (1.28% vs. 0.56%, P = .087).
For the juvenile patients, the prevalence of TIP was 1/87 (1.15%).
3.4 | Clinical characteristics of TIP
The detailed characteristics of the patients with TIP are summarized in Table 2. Of the 21 patients with TIP, 57.1% received a loading dose and the majority had a maintenance dose of 50 mg q12h. Most patients presented with overt abdominal symptoms, while five patients had no clinical symptoms. The mean time from tigecycline use to symptom onset was 7.2 days. All patients had mild pancreatitis, and the median modified CT severity score was 2. No patients were re-administered tigecycline. Abdominal symptoms were relieved soon after tigecycline withdrawal (mean time: 3.6 days). The amylase and lipase levels returned to their normal ranges at a mean time of 7.1 days after tigecycline withdrawal.
3.5 | Risk factors for TIP
We compared the laboratory data and therapeutic regimens between the TIP group and the non-pancreatitis group (Table 3). The ALP level in the TIP group was significantly higher than that in the non-pancreatitis group, while the PT and APTT levels in the TIP group tended to be lower. There were no significant differences in the data for blood routine tests, other liver function tests, serum lipid level, electrolyte indices, renal function tests, or infection indices between the two groups. The therapeutic regimens, including loading dose, maintenance dose, treatment duration, and total dose of tigecycline, were similar between the two groups.
The following variables with significant differences (P< .1) between the TIP group and the non-pancreatitis group were included in the multivariate analysis: age, length of hospital stay, percentage of comorbid renal insufficiency, multiple organ failure as indication for tigecycline, ALP, PT, and APTT. In the multivariate analysis, comorbid renal insufficiency was identified as an independent risk factor for TIP with an odds ratio (OR) of 3.032 (Table 4).
3.6 | Safety and efficiency of tigecycline therapy in non-TIP
In the study cohort, 82 patients were diagnosed with pancreatitis before tigecycline therapy and classified as non-TIP. The detailed information for these patients is shown in Table 5. The most common cause of pancreatitis was biliary stones, followed by hypertriglyceridemia. There were two cases of pancreatitis in pregnancy in the hypertriglyceridemia group, 81.7% of patients had severe pancreatitis, and 22% of patients underwent surgery for removal of necrotic tissue. Acute necrotizing pancreatitis was noted in 39 (47.6%) patients. Finally, 63.4% of patients showed improvement. The modified CT score after tigecycline use did not differ from that before tigecycline use (5.3 ± 1.8 vs. 5.3 ± 2.0, P = 1.0). The amylase and lipase levels after tigecycline use did not increase compared with those before tigecycline use, but significantly decreased as the pancreatitis improved (Figure 2A, B). After a median 8 days of tigecycline treatment, infection indices including white blood cell count (Figure 2C), percentage of neutrophils (Figure 2D), high-sensitivity C-reactive protein (Figure 2E), and procalcitonin (Figure 2F) were significantly decreased compared with the indices before drug use.
4 | DISCUSSION
In the present study, we retrospectively analyzed patients administered tigecycline for at least 3 days in our hospital during a 4-year period. We enrolled both adult patients and juvenile patients with the aim of analyzing the prevalences of TIP in this real-world setting and identifying risk factors for TIP. We also created a detailed summary for information related to tigecycline use in 82 non-TIP patients.
Tigecycline is a vital antibiotic treatment option for infections caused by MDR and XDR bacteria, especially in the intensive care unit. Our findings showed tigecycline was most commonly used in the hematology department, followed by the intensive care unit. Tigecycline was also commonly used for immunosuppressed recipients in the organ transplantation department. Regarding the indications for tigecycline, the most common indication was pneumonia, consistent with the finding in a previous study (>60%).12
The prevalences of TIP in the adult and juvenile patients were 0.56% and 1.15%, respectively. The prevalence for patients in the organ transplantation patients was slightly higher compared with those in the other patients. A study involving phase 3 and 4 tigecycline trials revealed that only 0.24% of 3,788 tigecycline-treated patients developed pancreatitis.13 In the second Periodic Safety Update Report for tigecycline, the incidence of pancreatitis was estimated at 1‰ and 1%, with 3 cases of necrotizing pancreatitis and 2 cases of fatal pancreatitis.14 However, all 21 patients with TIP in the present study had mild pancreatitis and their main CT feature was edema of the pancreas. Our results were similar to those of a secondary analysis involving 19 studies on tigecycline-induced AP, in which edematous infiltrate was the main imaging feature in patients and most cases had mild AP.15 We further found that the mean time from tigecycline use to symptom onset was 7.2 days, while previous studies reported 8.5 days14 and 12.5 days.16In our patients, abdominal symptoms were relieved at a mean of 3.6 days after tigecycline withdrawal and the enzyme levels returned to their normal ranges at a mean of 7.1 days, similar to the findings in a previous study (4 days and 5 days, respectively).15 We also noted that five patients had no obvious symptoms, consistent with the previous study.15 The possible reasons for TIP may be the high elimination rate of tigecycline by the biliary tract17 and the similar mechanism to tetracycline-induced pancreatitis (reaction with 30S ribosomal units and blockade of protein synthesis leading to triglyceride accumulation in the pancreas).18
The mean Naranjo score in our patients was 7.2. Possible reasons why the mean Naranjo score did not reach 9 (highly probable for drug-induced pancreatitis) may be that none of the patients were re-administered tigecycline (relevant question: if a drug re-challenge was performed, did AP recur?) and that all patients had infections (relevant question: does the patient have or was the patient recently diagnosed with an infection [bacterial, fungi, or viral] which could cause AP?).
Our multivariate analyses identified comorbid renal insufficiency as an independent risk factor for TIP (OR = 3.032). However, the renal function indices did not differ significantly between the TIP group and the non-pancreatitis group. This inconsistency may be explained by the small number of TIP patients (n = 21). Nevertheless, our results indicated that TIP is an uncommon event and that serum amylase and lipase levels should be closely monitored in patients with comorbid renal insufficiency during tigecycline treatment.
Acute necrotizing pancreatitis is a severe and life-threatening disease, and infection occurs in about 30% of cases as the most important prognostic factor.19 In AP patients, the major cause of death, after early organ failure, was secondary infection of pancreatic or peripancreatic necrotic tissue, leading to sepsis and multiple organ failure.20 In the present study, we evaluated the safety of tigecycline use in 82 AP patients. We found that the modified CT score after tigecycline use was similar to that before tigecycline use and that the amylase and lipase levels significantly decreased as the pancreatitis improved. The infection indices were also significantly decreased, indicating that tigecycline was efficient in controlling infection in patients with pancreatitis. Previous studies demonstrated that tigecycline was good at penetrating necrotic pancreatic tissue21 and was both safe and effective for treatment of infected necrotizing pancreatic fluid collection and sepsis caused by infected pancreatic necrosis.4 There is also evidence for efficacy of tigecycline use in treatment of complex intra-abdominal infection or abscess after AP.4 Our results further confirmed that tigecycline was suitable and safe for treatment of AP, especially severe necrotizing pancreatitis, with abdominal infection.
There were some limitations to the present study. First, it was a retrospective case-control study and may have a selection bias. Second, the dose of tigecycline used in each patient was not adjusted by the body weight because many patients had severe disease and their body weight could not be obtained.
In summary, the overall prevalence of TIP was low, most of the AP cases were mild, and withdrawal of tigecycline as early as possible usually permitted quick recovery and avoidance of severe complications. Comorbid renal insufficiency was identified as an independent risk factor for TIP. Tigecycline is safe and efficient for treatment of AP, especially necrotizing pancreatitis, with intra-abdominal infection.