Marked Increase of Gamma-Glutamyltransferase as an Indicator of Drug-Induced Liver Injury in Patients without Conventional Diagnostic Criteria of Acute Liver Injury

Abstract

Introduction

Clinically significant drug-induced liver injury (DILI) is defined by elevations of alanine aminotransferase (ALT) ≥5 times the upper limit of normal (ULN), alkaline phosphatase (ALP) ≥2 × ULN, or ALT ≥3 × ULN and total bilirubin TBIL >2 × ULN. However, DILI might also occur in patients who do not reach those thresholds and still may benefit from discontinuation of medication.

Methods

Fifteen patients recruited for our prospective study on potentially hepatotoxic drugs were included. DILI diagnosis was based on RUCAM (Roussel Uclaf Causality Assessment Method) score and expert opinion and was supported by an in vitro test using monocyte-derived hepatocyte-like (MH) cells.

Results

Median RUCAM score was 6 (range 4–8), indicating that DILI was possible or probable in all cases. The predominant types of liver injury were mixed (60%) and cholestatic (40%). While no elevation above 2 × ULN of ALP and TBIL was observed, gamma-glutamyltransferase (GGT) above 2 × ULN was identified in 8 of the patients. Six of the 15 patients did not achieve full remission and showed persistent elevation of GGT, which was significantly associated with peak GGT elevation above 2 × ULN (p = 0.005).

Conclusion

Here we present a case series of patients with liver enzyme elevation below the conventional thresholds who developed DILI with a predominant GGT elevation leading to drug withdrawal and/or chronic elevation of liver parameters, in particular of GGT. Thus, we propose that DILI should be considered in particular in cases with marked increase of GGT even if conventional DILI threshold levels are not reached, resulting in discontinuation of the causative drug and/or close monitoring of the patients.

Introduction

The diagnostic criteria for drug-induced liver injury (DILI) were redefined by the DILI Expert Working Group (DEWG) in 2011 [1]. According to these consensus criteria, clinically significant DILI occurs if any of the following applies: Elevations of alanine aminotransferase (ALT) ≥5 times the upper limit of normal (ULN), alkaline phosphatase (ALP) ≥2 × ULN, or ALT ≥3 × ULN and total bilirubin TBIL >2 × ULN [1, 2]. A higher threshold for ALT in comparison to the criteria defined by the Council for International Organizations of Medical Sciences (CIOMS) was proposed [3], because ALT elevation below this threshold might merely represent mild liver injury of transient nature, that is, an adaption to the respective drug [2, 4], or might be due to underlying non-alcoholic fatty liver disease (NAFLD) [1]. However, it has also been recognized that in some instances drugs do not cause liver parameter elevation above the chosen thresholds but still might induce clinically significant liver injury. The DEWG therefore proposed to analyse and classify those cases individually and define them accordingly, that is, as “other forms of DILI” [1].

Here we present 15 cases of suspected drug-induced liver injury in which conventional DILI criteria were not fulfilled but clinically relevant liver injury nevertheless occurred.

Materials and Methods

Patient Cohort

The data of 385 patients who were recruited for our ongoing prospective study on the effects of potentially hepatotoxic drugs (ClinicalTrials.gov: NCT 02353455) between July 2012 and December 2020 were analysed. Written informed consent was obtained from each patient. The study protocol conformed to the ethical guidelines of the Declaration of Helsinki and was approved by the ethics committee of the Faculty of Medicine, LMU Munich (Project number 55-13). Out of the 385 patients who were enrolled in our prospective study, 15 patients with suspected DILI were identified who did not meet criteria for acute liver injury (ALI). Those 15 patients were selected for the current analysis. Out of the remaining 370 patients, 367 fulfilled conventional criteria for ALI; 223 were diagnosed with DILI while 145 were diagnosed with non-DILI-related ALI. Three patients who did not meet ALI criteria were excluded from the current analysis due to age below 2 years, missing follow-up information, and alcoholic steatohepatitis as the most likely diagnosis. Other causes for ALI were autoimmune hepatitis (AIH) in n = 42, biliary diseases in n = 29, viral hepatitis in n = 24, alcoholic liver disease in n = 12, ischemic hepatitis in n = 10, non-alcoholic steatohepatitis in n = 5, other autoimmune liver diseases in n = 5, infectious hepatitis other than viral hepatitis in n = 4, and others in n = 16 cases.

Definition, Diagnosis, and Classification of DILI

The diagnosis of DILI was based on clinical, laboratory, and histopathological findings, the Roussel Uclaf Causality Assessment Method (RUCAM) score [5], causality assessment by the physician in charge, and expert opinion in consideration of all clinical data available, the course of events, and data available in the LiverTox database [6, 7]. DILI diagnosis was supported by an in vitro test with monocyte-derived hepatocyte-like (MH) cells, which has been described in detail previously [8]. A hepatological work-up to exclude viral hepatitis and autoimmune or cholestatic liver diseases, including serological studies for hepatitis A, B, C, and E, cytomegalovirus, Epstein-Barr virus, and herpes simplex virus, and autoantibody screening for anti-nuclear, anti-mitochondrial, and anti-smooth muscle antibodies, as well as careful history of medication was performed in all patients recruited as described before [9, 10, 11]. A relevant alcohol consumption was defined as ≥30 g/day for men and ≥20 g/day for women according to the criteria of the European Association for the Study of the Liver [12]. The interpretation of the RUCAM score is as follows: ≤0, drug is excluded as the cause; 1–2, unlikely; 3–5, possible; 6–8, probable, and >8, highly probable. The type of liver injury was classified using the R ratio values, (ALT/ULN)/(ALP/ULN), with R ≥5 defining a hepatocellular, R ≤2 a cholestatic, and 2<R<5 a mixed type injury [1].

Results

Clinical Characteristics

We identified 15 patients enrolled in our prospective study on drugs with hepatotoxic potential who were diagnosed with DILI but did not fulfil the conventional criteria defined by DEWG [1]. Diagnosis of DILI was predominantly based on the clinical course of the disease, RUCAM score, and expert opinion. The clinical characteristics of the patients are summarized in Table 1. The main drug classes associated with liver injury in our cohort were antibiotics and immunosuppressants each in 3 cases (20%). Median RUCAM was 6 (range: 4–8) and 10 of the 15 patients (67%) had a RUCAM of 6 or higher indicating probable DILI. Patients presented with a mixed (60%) or cholestatic (40%) type liver injury. Two patients had underlying chronic liver disease, that is, AIH and NAFLD, respectively, and 2 patients reported a relevant alcohol consumption of 40 g/day (male, case 6) and 25 g/day (female, case 8), respectively. The AIH patient (case 9) received corticosteroids at the time of liver injury; nevertheless, increase in liver parameters was observed only after introduction of azathioprine. Considering all clinical data available, the evolution of liver parameters in those 4 patients was indicative for DILI. Both patients with underlying liver disease presented with a marked increase of gamma-glutamyltransferase (GGT) after drug intake, which decreased after drug withdrawal from 32.6 × ULN to 5.9 × ULN in the patient with NAFLD (case 3), and from 14.5 × ULN to 4.9 × ULN in the patient with AIH (case 9). Simultaneously, aminotransferases also declined markedly in both patients, rendering drug-induced liver injury the most likely diagnosis. Moreover, full remission was observed for one of the patients with relevant alcohol intake (case 6), while the other patient was lost to follow-up but had presented with normal liver enzymes at admission which only increased after intake of ibuprofen during hospitalization.

Table 1.

Clinical characteristics

Age, years	57 (18–76)
Body mass index, kg/m2	24.7 (20.8–33.2)
Male gender, n (%)	10 (67)
Underlying chronic liver disease, n (%)	2 (13)
Culprit drug classes, n (%)
Antibiotics	4 (25)
Immunosuppressive drugs	3 (19)
SNRI	2 (13)
NSAID	2 (13)
Others	5 (31)
RUCAM score	6 (4–8)
R ratio at onset1	2.1 (1.0–4.1)
Pattern of liver injury based on R value,1 n (%)
Hepatocellular	0 (0)
Mixed	9 (60)
Cholestatic	6 (40)
Time to remission, weeks	16 (6–267)
Total duration of follow-up, weeks	82 (1–277)

Categorical variables are presented as number and percentage, n (%). Continuous variables are presented as median (range). DILI, drug-induced liver injury; NSAID, non-steroidal anti-inflammatory drugs; RUCAM, Roussel Uclaf Causality Assessment Method; SNRI, serotonin-norepinephrine reuptake inhibitors.

¹The R ratio is defined as (ALT/ULN)/(ALP/ULN), with R ≥5 defining a hepatocellular, R ≤2 a cholestatic, and 2<R<5 a mixed type injury.

Table 2 gives an overview of the individual cases including the outcome. For most patients (n = 13; 87%) an abdominal ultrasound or alternative imaging was performed to exclude biliary or vascular pathologies. The monocyte-derived hepatocyte-like cell test was performed in every patient included in the current analysis and showed a positive test result for the suspected medication in 11 cases, while in 4 cases the test yielded negative results for the medications taken before the occurrence of liver injury (Table 2). The causative drugs were discontinued in every patient and drug re-exposure was only tried in 3 patients after the DILI episode: One patient showed recurrent liver injury upon treatment re-initiation with duloxetine (case 6), while 1 patient received isoniazid at lower dosages (case 1) and 1 patient was treated with methylphenidate (case 13) without recurrence of liver injury. Eight patients (57%) showed complete spontaneous resolution of liver injury, in 6 cases (43%), however, persistent GGT elevation was observed, and 1 patient was lost to follow-up. All the patients with chronic GGT elevation had shown reduction of GGT after drug withdrawal by 68%–86%; however, complete normalization of GGT levels was not observed until the end of follow-up.

Table 2.

Characteristics of individual cases including suspected medication, laboratory, radiological, and histopathological findings

Patient	Age	Gender	Suspected medication	RUCAM	Latency, days	Pattern of injury	Peak ALT	Peak AST	Peak ALP	Peak TBIL	Peak GGT	Outcome	Time to remission, weeks	Follow-up, weeks	MH cell test result
1	25	Male	Isoniazid	4	4	Mixed	4.5	2.0	1.0	0.5	1.0	Remission	6	68	Negative
2	63	Male	Tinzaparin	6	82	Cholestatic	0.9	0.4	1.0	0.8	6.6	Partial remission (reduction of GGT 53%) after 6 weeks. Full remission after 267 weeks	267	277	Positive: tinzaparin
3	50	Male	Ciprofloxacin	4	17	Cholestatic	3.9	4.2	2.0	0.6	32.6	Partial remission (reduction of GGT 82%, normalization of ALT/AST). Persistent elevation of GGT and ALP	13	82	Negative
4	65	Female	Ampicillin	6	4	Mixed	3.7	1.8	1.2	0.3	7.9	Partial remission (reduction of GGT 81%, normalization of ALT and AST). Persistent elevation of GGT and ALP	No follow-up for 2 years after ALI	178	Positive: ampicillin
5	76	Male	Pravastatin	6	20	Mixed	1.2	1.1	1.0	0.9	5.4	Remission	65	241	Positive: pravastatin
6	56	Male	Duloxetine	8	2	Cholestatic	1.0	2.7	1.7	0.9	27.4	Remission	36	49	Positive: duloxetine, Ibuprofen
7	62	Female	Azathioprine	6	80	Cholestatic	2.7	1.2	0.9	0.9	1.0	Remission	33	231	Positive: azathioprine
8	22	Female	Ibuprofen	6	3	Mixed	1.2	2.2	1.0	0.7	1.3	Lost to follow-up	na	1	Positive: Ibuprofen
9	37	Male	Azathioprine	8	49	Mixed	3.3	1.6	1.2	2.21	14.5	Partial remission (reduction GGT >66%). Persistent elevation of GGT	9	240	Positive: azathioprine
10	57	Female	Azathioprine	5	187	Cholestatic	1.5	1.6	1.9	0.4	9.2	Partial remission (reduction GGT 86%, normalization of other LFT). Persistent elevation of GGT	49	107	Positive: azathioprine
11	73	Male	Rituximab	6	14	Mixed	2.3	1.5	1.3	0.5	43.1	Partial remission (reduction GGT 84%, normalization of other LFT). Persistent elevation of GGT	15	201	Negative
12	67	Female	Duloxetine	5	2	Mixed	3.5	2.3	0.8	0.5	2.2	Partial remission (normalization of ALT after 67 weeks). Persistent elevation of GGT	67	68	Positive: duloxetine
13	18	Male	Methylphenidate	6	1,624	Cholestatic	1.8	0.9	1.0	1.6	0.8	Remission	16	33	Negative
14	45	Male	Metamizole	5	61	Mixed	2.7	1.5	1.0	1.9	0.6	Remission	6	8	Positive: metamizole
15	70	Male	Silymarin	7	na	Mixed	2.7	1.6	1.0	0.6	1.7	Remission	14	14	Positive: silibinin

Laboratory values are presented as times ULN. ALI, acute liver injury; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; DILI, drug-induced liver injury; GGT, gamma-glutamyltransferase; MH cell test, monocyte-derived hepatocyte-like cell test; LFT, liver function test; na, not applicable; RUCAM, Roussel Uclaf Causality Assessment Method; TBIL, total bilirubin; ULN, upper limit of normal.¹ Peak ALT and TBIL levels were measured on different days; the patient did therefore not fulfil DILI criteria at any time point.

Laboratory Parameters

Table 3 gives an overview of the laboratory parameters detected at baseline and respective peak levels. While ALP and TBIL were mostly normal at the time of liver injury diagnosis, an increase of baseline ALT and GGT was observed (median: 1.6 × ULN, respectively). Baseline GGT correlated significantly with baseline ALP (Pearson correlation: 0.61, p = 0.016) but not with baseline ALT (Pearson correlation: 0.08, p = 0.77) or with TBIL (Pearson correlation: −0.30, p = 0.30). Regarding laboratory parameters at respective peak levels, a marked increase of GGT was observed (median: 5.4 × ULN, range: 0.6–43.1 × ULN), while ALT, and in particular AST, ALP, and TBIL were only mildly elevated (Table 3). Peak GGT elevation above 2 × ULN was observed in 9 patients (60%) and in 8 of those even above 5 × ULN. Again, there was a significant correlation for peak GGT and ALP (Pearson correlation: 0.67, p = 0.007), but not for GGT and ALT (Pearson correlation: 0.25, p = 0.37) or GGT and TBIL (Pearson correlation: −0.15, p = 0.58).

Table 3.

Laboratory parameters

At baseline
ALT (X ULN)	1.6 (0.9–3.5)
AST (X ULN)	1.1 (0.4–3.1)
ALP (X ULN)	1.0 (0.5–1.9)
GGT (X ULN)	1.6 (0.3–23.8)
TBIL (X ULN)	0.6 (0.1–1.8)
At respective peak levels
ALT (X ULN)	3.0 (0.9–4.5)
AST (X ULN)	1.6 (0.4–3.1)
ALP (X ULN)	1.0 (0.8–2.0)
GGT (X ULN)	5.4 (0.6–43.1)
TBIL (X ULN)	0.7 (0.3–2.2)

Variables are presented as median (range). ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma-glutamyltransferase; TBIL, total bilirubin; ULN, upper limit of normal.

None of the patients fulfilled the DILI criteria by the DEWG at any stage of the injury, only 1 patient presented with peak ALP of 2 × ULN, however this was the patient with underlying NAFLD (case 3), and ALP increase was only 1.3 times when compared to baseline. ALT elevation above 2 × ULN, one of the diagnostic criteria defined by the CIOMS [3], was found in 6/15 patients (40%) at onset and 11/15 patients (73%) at peak levels. Median baseline GGT levels tended to be higher in patients with baseline ALT >2 ULN: GGT was 7.8 × ULN in patients with ALT >2 × ULN versus 1.3 × ULN in patients with ALT ≤2 × ULN; however, this did not reach statistical significance (p = 0.39).

Patients with baseline ALT or GGT elevation >2 ULN tended to develop persistent GGT elevation with a higher frequency: 67% of patients with initial ALT >2 × ULN versus 33% of patients with ALT ≤2 × ULN (p = 0.12) and 57% of patients with baseline GGT >2 × ULN versus 33% of patients with baseline GGT ≤2 × ULN (p = 0.28) presented with persistent GGT elevation at follow-up. Moreover, 6/9 patients (67%) presenting with peak GGT >2 × ULN developed chronic GGT elevation, while no patients with peak GGT ≤2 × ULN did (p = 0.016).

Discussion/Conclusion

Different thresholds are in use for the diagnosis of DILI [1, 3, 13, 14]. The most widely applied criteria established by the DEWG recommend the application of ALT levels of 5 × ULN or higher in order to exclude patients with NAFLD or clinically insignificant liver parameter abnormalities [1]. However, we hypothesized that clinically relevant liver injury can also occur at lower thresholds and here present 15 well-characterized cases of liver injury, which were classified as drug-induced despite not fulfilling the conventional DILI criteria. When RUCAM scores were calculated, all the patients scored 4 or above indicating that DILI was possible or probable in all cases. Differential diagnoses were excluded for all patients by a thorough hepatological work-up. While in the 3 patients receiving azathioprine, nodular regenerative hyperplasia could be considered as an alternative diagnosis [15]; imaging studies showed no signs of portal hypertension in any of these patients, making it a less likely differential diagnosis. Transient increase of liver enzymes, in particular of ALT and AST even in the case of drug continuation, has been described [16, 17], while such cases of “transaminitis” were not observed in our cohort. The causative drugs were discontinued in all patients and re-exposure was only conducted in 3 patients with recurrence of liver injury in one of them.

An interesting finding in our cohort was the marked increase of GGT in the majority of patients with 53% of the patients presenting with GGT above 5 × ULN. Elevation of GGT alone is usually not considered sufficient to indicate clinically relevant liver injury [1, 2, 18]. Although GGT is increased in up to 95% of patients with acute liver injury, other conditions such as hyperthyroidism or acute myocardial infarction can cause GGT increase leading to a low predictive value of GGT for liver diseases [19]. Cardiac or other diseases which cause GGT increases were excluded in our cohort. Thus, in our case analysis, GGT elevation was indeed representative of liver injury. Importantly, elevation of peak GGT of >2 times ULN was associated with chronic liver injury and persistence of GGT elevation, albeit at markedly lower levels after drug discontinuation. Since remission was not achieved in nearly half of the patients despite drug discontinuation in all cases, the observed subthreshold elevations of liver enzymes in our small cohort should be regarded as clinically significant.

Thus, although the DEWG criteria are a very good clinical test, which is also stressed by the fact that 367 out of the 385 patients with ALI enrolled in our study fulfilled those criteria, we show here by a thoroughly conducted analysis of a prospectively collected cohort that in some albeit rare instances clinically relevant liver injury can occur in patients who do not reach the conventional thresholds for the diagnosis of DILI. Importantly, these cases would have been missed by the application of standardized criteria putting the patients at risk of worsening or recurrence of liver injury if the culprit drug is not withdrawn.

These findings are of particular interest in the light of increasing usage of electronic healthcare records for pharmacovigilance studies [20, 21] since these “subthreshold” DILI cases are missed by automatic search algorithms [22]. Conversely, Udo et al. [23], who tested different algorithms comprising diagnostics codes and various thresholds of liver enzyme elevation showed that even mildly elevated ALT levels of only >2 × ULN, one of the threshold criteria defined by CIOMS [3], can be indicative of clinically relevant liver injury [23]. Moreover, since DILI cases not fulfilling the conventional criteria are not included in international DILI registries [13, 24], under-reporting and with this under-recognition is a likely possibility and will potentially increase in the future.

Our study has limitations, in particular the small number of patients included in the current analysis which limits the generalizability of our results and impedes the generation of a novel algorithm for the diagnosis of DILI. However, our results demonstrate that DILI might occur even if the conventional criteria for acute liver injury are not fulfilled and should therefore increase awareness of this rare presentation of DILI. Thus, in selected cases with unexplained GGT or mild ALT increase, a thorough manual analysis is warranted in order to identify atypical DILI cases and to avoid misclassification and possible prolongation of liver injury.

Statement of Ethics

The study protocol conforms to the ethical guidelines of the Declaration of Helsinki and was approved by the ethics committee of the Faculty of Medicine, LMU Munich (Project Number 55-13). Written informed consent was obtained from each participant.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

S.W. received funding from the Friedrich-Baur-Institution (grant number 43/20).

Author Contributions

S.W.: Conceptualization; Formal analysis; Data collection; Funding acquisition; Investigation; Methodology; Software; Visualization; Writing − original draft; Writing − review & editing. J.A. and G.D.: Data collection; Investigation; Visualization; Writing − review & editing. A.L.G.: Conceptualization; Funding acquisition; Methodology; Project administration; Resources; Supervision; Validation; Visualization; Writing − review & editing.

Data Availability Statement

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.