Postoperative coagulopathy after live related donor hepatectomy: Incidence, predictors and implications for safety of thoracic epidural catheter.

BACKGROUND: Coagulopathy after living donor hepatectomy (LDH) may endanger donor safety during removal of thoracic epidural catheter (TEC). The present study was conducted to evaluate the extent and duration of immediate postoperative coagulopathy after LDH.
MATERIALS AND METHODS: A retrospective analysis of perioperative record of LDH over three years was conducted after IRB approval. Variables such as age, gender, BMI, ASA classification, liver volume on CT scan, preoperative and postoperative INR, platelet count (PC) and ALT of each donor for five days was noted. In addition, duration of surgery, remnant as percentage total liver volume (Remnant%), blood loss, day of peak in PC and INR were also noted. Coagulopathy was defined as being present if INR exceeded 1.5 or platelet count fell below 1 × 10 5 /mm 3 on any day. Data was analyzed using SPSS 20 for Windows. Between group comparison was made using the Student 't' test for continuous variables and chi square test for categorical variables. Univariate analysis was done. Multiple logistic regression analysis was used to find independent factor associated with coagulopathy.
RESULTS: Eighty four (84) donors had coagulopathy on second day (mean INR 1.9 ± 0.42). Low BMI, % of remnant liver and duration of surgery were independent predictors of coagulopathy. Right lobe hepatectomy had more coagulopathy than left lobe and low BMI was the only independent predictor. There was no correlation of coagulopathy with age, gender, blood loss, presence of epidural catheter, postoperative ALT or duration of hospital stay. High INR was the main contributor for coagulopathy.
CONCLUSIONS: Coagulopathy is seen after donor hepatectomy. We recommend removal of the epidural catheter after the fifth postoperative day when INR falls below 1.5.

Introduction

Living donor liver transplantation is a solution to shortage of deceased organ donation. Safety of the donor is vital as it influences the possible outlook of society towards voluntary organ donation.[1] Coagulopathy after major hepatic resection may compromise the donor's safety, especially at epidural catheter removal.[2] In the present study, we retrospectively analysed living donors who underwent hepatectomy (LDH) at our centre over a three year period (March 2010 to 2013). The primary objective was to study the extent, duration and predictors of coagulopathy in the immediate postoperative period. The secondary objective was to recommend the optimum postoperative day to remove epidural catheter to minimise the risk of epidural bleeding and its attendant complications.

Materials and Methods

Ethics

Approval was taken from the Institutional Review Board and consent waiver was obtained for analysis of data. Confidentiality was maintained using unique identifiers.

Study design, eligibility criteria and study procedure

Data of donor hepatectomy from March 2010 to 2013 including records of perioperative anaesthesia management, intraoperative events, postoperative recovery, nursing, physiotherapy, and computerized hospital data was retrospectively reviewed. We included all donors given general anaesthesia and those that had received thoracic epidural anaesthesia with consent. We excluded those donors in whom hepatectomy did not proceed or they received antiplatelet drugs postoperatively due to other medical reasons. From the anaesthesia records, we noted the age, gender, BMI, presence of co-morbidities, American Society of Anaesthesiologist (ASA) physical status classification, baseline preoperative haematocrit, INR, Platelet count and Alanine Aminotransferase (ALT) of each donor on the day of surgery (POD 0). From the preoperative CT scan of abdomen, we recorded the Total liver volume in cc (TLV). We also noted the anaesthesia technique, presence of epidural catheter, total duration of anaesthesia, type of donor hepatectomy (Right versus left lobe), graft weight (in grams), blood loss and blood transfused.

Remnant liver volume present in the donor was expressed as Remnant % or RmDWR.

Remnant % = (TLV — graft weight) × 100/TLV

RmDWR = (TLV — graft weight)/Body weight of the donor

From the postoperative record and hospital database, we collected the value of platelet count and peak value of INR of each day for first 5 postoperative days or till the epidural catheter was removed. We recorded the postoperative day, INR and platelet count on removal of epidural catheter and any catheter related complications as mentioned in records.

Donors were monitored with electrocardiography, invasive blood pressure by insertion of left radial artery catheter, central venous pressure by insertion of right internal jugular catheter, body temperature, end tidal carbon dioxide, arterial blood gas analysis and urine output. The mean blood pressure was maintained above 65-70 mmHg with a urine output more than 1 ml/kg/hour. Packed Red Blood Cells (PRBC) was transfused wherever the haemoglobin dropped below 8 gm%.

Cavitron Ultrasonic Surgical Aspirator (CUSA) was used for parenchymal transection. Heparin 50 units/kg was administered at the end of parenchymal transaction, before retrieval of graft to prevent graft thrombosis. However, no reversal with protamine was done in any case. After surgery, neuromuscular blockade was reversed and trachea was extubated in all donors on table. Thereafter, the patient was shifted to postanaesthesia care unit. For analgesia, either patient controlled epidural analgesia (PCEA) or intravenous patient controlled analgesia (IV PCA) was started with regular daily follow up by the Acute Pain Services Team (APST).

Case definition of coagulopathy

Coagulopathy was said to be present when INR >1.5 or platelet count less than 1 × 105/mm3 after surgery. The day when maximum numbers of subjects were found to be hypocoagulable was noted. Presence of coagulopathy on this day was then correlated with age, gender, BMI, presence of TEA, hepatectomy left or right, duration of anaesthesia, blood loss, remnant liver volume (expressed both as a percentage of remnant donor weight ratio RDWR with total liver volume (%TLV) as estimated by CT volumetry preoperatively).

Study groups for analysis

Donors were divided into two groups, Group 1-donors with coagulopathy and Group 2-donors without coagulopathy.

Statistical analysis

This was done using SPSS 20 for Windows (SPSS, Inc, Chicago, IL). Continuous data was expressed using measures of central tendency. Categorical data were presented as proportions. Between group comparison was made using the Student ‘t’ test for continuous variables and chi square test for categorical variables. Univariate analysis was done for presence of coagulopathy and included age, gender, BMI, duration of anaesthesia, blood loss, type of hepatectomy, presence of epidural, remnant liver volume and duration of hospital stay as the variables. Stepwise multivariate logistic regression analysis was done to evaluate independent factors associated with coagulopathy which was taken as the dependent variable. A P value of less than 0.05 was considered statistically significant.

Results

Demographics

A total of 100 donors who satisfied eligibility criteria were analyzed. All donors were healthy and (85/100) with 85% of donors were American Society of Anaesthesiologist (ASA) physical status class I and remaining (15/100) 15% ASA physical status class II. The mean age was 31.02 ± 8.87 years with a BMI of 23.2 ± 3.21. Fifty seven percent of the donors were females. General anaesthesia was administered in all cases with supplemental thoracic epidural analgesia in 72 donors. Right lobe hepatectomy was done in 86 individuals while left lobe was retrieved in remaining 14 donors. The mean duration of surgery was 12.2 ± 2.37 hours. The blood loss ranged from 350 to 750 ml and transfusion of PRBC was required in 7 patients. The remnant liver mass was observed to be 47.91 ± 11.4% of Total Liver Volume (TLV) with a remnant donor weight ratio of 0.988 ± 0.26. The median ALT peaked on first postoperative day (166 with IQ range 132-234) and decreased to 80 with IQ 61-109 on fifth postoperative day [Table 1]. The median duration of hospital stay was 13 days with an interquartile range from 9 to 16 days. Accidental removal of epidural catheter occurred in 3 donors (4.1%) in first three days while there was leakage of drug from catheter site in two donors (2.8%) requiring removal of the catheter. No symptoms suggestive of epidural site infection or sepsis were noted in any donor. A total of 36.1%, 18.1% and 5.6% epidural catheters were removed on postoperative day 5, 6, and 7 respectively followed by 33.3% on POD 4. No blood product was transfused before removal of catheter in any case.

Table 1

INR, platelet count, ALT — baseline (POD 0) and first five consecutive postoperative days (POD 1, 2, 3, 4, 5)

Investigation	POD 0	POD 1	POD 2	POD 3	POD 4	POD5
INR*	0.9±0.1	1.6±0.3	1.9±0.4	1.7±0.4	1.4±0.2	1.3±0.8
Platelet count*(105/mm3)	2.4±0.7	1.9±0.6	1.5±0.5	1.6±0.5	1.7±0.5	1.9±0.6
ALT (units/litre)#	23 (16-31)	166 (132-234)	130 (102-188)	101 (81-141)	93 (73-121)	80 (61-109)

*mean ± standard deviation, #median with interquartile range

INR and Platelet count changes

There was a rise in INR and fall in platelet count in all subjects after donor hepatectomy [Table 1]. Peak value of INR was found on second whereas the lowest platelet count was found on third postoperative day. On POD1, 62% of donors were categorized having coagulopathy compared to 84% on day 2, 61% on day 3 and 31% on day 4. Maximum incidence of donors with a hypocoaguable state was on the second postoperative day. Improvement in coagulation parameters started from the third day after surgery when a sharp decline in the number of donors fitting in coagulopathy criteria was noted. By the fifth postoperative day, only 14% donors remained hypocoaguable by INR/Platelet criteria [Table 2].

Table 2

Number of donors with coagulopathy — first five consecutive postoperative days (POD 1, 2, 3, 4, 5) after LDH

Number of donors	POD 1	POD 2	POD 3	POD 4	POD 5
INR >1.5	59	84	58	26	10
PC <100,000/mm3	5	12	12	8	6
Hypocoaguable by either of INR/PC	62	84	61	31	14

Between group comparison, univariate analysis

On comparison of the subjects with coagulopathy (Group 1: n = 84) with those without coagulopathy (Group 2: n = 16) on the second postoperative day, a lower BMI (22.83 ± 3.16 versus 25.42 ± 2.72), lower remnant liver volume (Remnant% 45.97 ± 9.4 versus 58 ± 15.32, RmDWR 0.956 ± 0.24 versus 1.16 ± 0.30), and longer duration of surgery (12.5 ± 2.3 versus 10.8 ± 2 hours) was found in group 1. Incidence of coagulopathy was significantly higher after right lobe (88%) than after left lobe LDH (P = 0.003). There was no correlation of development of coagulopathy in the postoperative period with age, gender, blood loss or presence of epidural catheter [Table 3]. The overall hospital stay was 12.5 (9-16) days in coagulopathic versus 14 (10-15.7) days in non-coagulopathic donors (P = 0.512).

Table 3

Comparison between two groups — demographic and perioperative variables

Variables	Postoperative coagulopathy (Day 2)		P value
	Present (Group 1; n - 84)	Absent (Group 2; n -16)
Male:Females	35:49	8:8	0.88
Age (mean±SD)	30.6±8.9	33.7± 7.5	0.156
BMI (Kg/m2)	22.8± 3.1	25.4 ± 2.7	0.003
Epidural±	62/22	10/6	0.35
Hepatectomy (right/left)	76 / 8	10/6	0.003
Duration of surgery (hours)	12.5±2.34	10.8±2.03	0.008
Blood loss (ml)	583±312	620±326	0.67
Remnant (% TLV)	45.9±9.4	58.1±15.3	0.007
Remnant donor weight ratio (RDWR)	0.95±0.24	1.1±0.3	0.019
Postoperative hospital stay (days)	12.5 (9-16)	14 (10-15.7)	0.51

Multivariate analysis

Low BMI, low remnant liver (expressed as a percentage of TLV), and long duration of surgery were independent predictors of immediate postoperative coagulopathy [Table 4].

Table 4

Independent predictors of coagulopathy on multivariate regression analysis

Variables	B coefficient	P value
BMI	0.274	0.002
Duration of surgery	−0.189	0.038
Remnant of liver (% TLV)	0.332	<0.001

Since incidence of a hypocoaguable state was more after right lobe donor hepatectomy, we calculated Remnant liver volume in donor (Remnant% and RmDWR) separately for subjects undergoing right lobe donor hepatectomy to avoid skewing of data. In this subgroup of patients, only low BMI was found to have statistical significance in prediction of coagulopathy in the immediate postoperative period with a low remnant % being a close second factor though failing to achieve a statistical significance [Table 5]. Persistent hypocoagulable state was present even on the fifth postoperative day in donors with BMI (20.6 ± 2.9) compared to those with BMI (23.7 ± 3.1) (P = 0.001).

Table 5

Predictors of coagulopathy after right lobe donor hepatectomy on multivariate regression analysis

Variables	Postoperative coagulopathy		P value
	Present (n - 76)	Absent (n - 10)
Remnant of liver (%TLV)	43.8±6.2	48.3±9.7	0.05
RmDWR*	0.92±0.21	1.01±0.27	0.22
BMI (Kg/m2)	22.9±3	26.2±2.9	0.002
Duration of surgery (hours)	12.6±2.4	11±2	0.06

*Remnant donor weight ratio

There was no correlation of presence of coagulopathy on each day with the respective ALT of that day.

Discussion

We observed significant coagulopathy with maximum severity and incidence on the second postoperative day. The rise in INR contributes more towards coagulopathy than fall in platelet count. On the fifth postoperative day, the mean INR decreased to 1.3 and only 14% patients had coagulopathy. We found that a low BMI was an independent predictor of development of a hypocoagulable state after right lobe LDH.

Removal of a considerable hepatic mass may lead to diminished hepatic synthesis of clotting factors causing a hypocoagulable state.[3456] Temporary alterations in the haemostatic balance have been documented after liver resection in healthy living donors as well as individuals with benign or malignant masses.[5678] Various studies have implicated blood loss, transfusion, temporary vascular flow interruption, fibrinolysis and decreased synthetic activity of the remnant liver as contributor of coagulopathy. There was no massive blood loss/transfusion in any of our donors. However, consumption and transient decrease in synthetic activity could explain the development of coagulopathy in our patients.

Postoperative coagulopathy was present in 84 % of our donors on the second day which persisted till the third postoperative day in 61% individuals. Kim et al. reported the presence of coagulopathy after donor hepatectomy however, they did not identify the course and severity of coagulopathy.[9] Our data is in agreement with Stamenkovic et al. who observed a peak in INR on the second day. However, we observed the nadir in platelet count on third postoperative day rather than the second.[10]

Patients who underwent right hepatectomy had a prolonged INR as compared to those who underwent left hepatectomy suggesting that parenchymal loss and extent of hepatectomy might have a significant effect on impairment of coagulation status after hepatectomy. The remnant volume as estimated by Remnantpercent (43.8 ± 6 versus 48.3 ± 9) and RDWR (0.9 ± 0.2 vs 1.0 ± 0.27), was lower in donors with coagulopathy compared to those without coagulopathy. Our mean remnant % in hypocoaguable donors were lower than that reported by Kim et al. (50 ± 12) after donor hepatectomy.[9] However, as a predictor of coagulopathy, the remnant % did not reach statistical significance in donors undergoing right lobe hepatectomy (P value >0.05). The vascular occlusion technique to reduce blood loss and minimize ischemia reperfusion injury like Pringle's manoeuvre have been shown to be an independent risk factor to predict postoperative coagulopathy.[11] Because of the retrospective nature of our study this aspect couldn’t be ascertained in our patients.

We observed a statistically significant association of a low BMI with development of a hypocoaguable state in the immediate postoperative period. Only one study has observed similar finding which is in contradiction with earlier studies which have shown that donors with high BMI may have higher incidence of steatosis and graft dysfunction leading to postoperative coagulopathy.[912] In donors with lower BMI, hypocoagulable state was observed on fifth postoperative day, suggesting that low BMI may have a significant association with persistence of hypocoagulable state. Gruttadauria et al. studied early liver regeneration of 70 living donors who underwent right hepatectomy. They found that a higher preoperative BMI was one of the predictors of greater liver regeneration.[13] Larger resections may lead to greater concentration of cytokines and promote growth.[14] Duration of surgery was a significant predictor of coagulopathy after LDH, though it did not achieve statistical significance when only donors with right lobe hepatectomy were analysed.

Coagulopathic state did not increase the postoperative hospital stay or was not associated with persistent rise in ALT. However, presence of coagulopathy may endanger the safety of the donor, especially if Thoracic Epidural Analgesia (TEA) is used for postoperative analgesia. Coagulation abnormalities post donor hepatectomy has been previously reviewed but no clear consensus on epidural catheter insertion and removal exists.[5] At an INR of 1.5, only 40% of the normal clotting factor activity may be present. Thus, INR should be less than 1.5 for normal haemostasis for safe removal of epidural catheter in patients on anticoagulants.[15] A platelet count of more than 100 × 103/mm3 is considered to be safe by many clinicians for the same purpose.[4] In view of the hypocoaguable state as seen in our study, it is unsafe to remove the catheter within the first four days of donor hepatectomy. In 86% donors, this coagulation derangement corrected spontaneously by fifth postoperative day. However, INR <1.5 must be confirmed before removal of TEC as 14% donors still remain hypocoagulable even on the fifth postoperative day. In three patients, we observed accidental removal of TEC in the first two days, but no epidural related complications were noted. We suggest that removal of the epidural catheter should be considered only when there is recovery in coagulopathic state and probably only fifth day after surgery, after confirming normal coagulation parameters.

In our study, no correction of coagulation was required for removal of epidural catheter. Our patients routinely received low molecular weight heparin (LMWH) from the second or third postoperative day. While routine PTT levels were not done but epidural catheter removal and subsequent dosing was done in accordance to current American Society of Regional Anesthesia guidelines.[15]

Though epidural analgesia seems safe, risk of bleeding and hematoma formation exists. Accidental removal of the epidural catheter within the first three days exposes the patient to these rare but serious complications of epidural hematoma or neurological compression. Because of the rarity of such complications the study was inadequately powered to confirm the safety of epidural analgesia in live liver donor hepatectomy.

The risk of infectious complications needs to critically reviewed because of inherent delay of epidural catheter removal. In most of our patients epidural catheter was removed on or after day 4 but no symptoms suggestive of infectious complication developed.

Several limitations exist in our study. The foremost being the retrospective nature of the study. Second, detection of a rare complication like epidural hematoma is very difficult in this small number of patients. Third, we followed the coagulation status for the first five postoperative days only and fourth, the efficacy of epidural for analgesia could not be compared to intravenous analgesia.

Conclusion

In conclusion, the incidence and severity of coagulopathy is maximum on the second postoperative day after donor hepatectomy. In donors with low BMI, planned for retrieval of right lobe of liver, it may be prudent to remove the TEC after the fifth postoperative day after confirming normal coagulation parameters, thereby decreasing the risk of complications associated with coagulopathy.