Liver regeneration after adult living donor and deceased donor split-liver transplants



As the number of living donor (LD) and deceased donor (DD) split-liver transplants (SLTs) have increased over the last 5 years, so too has the interest in liver regeneration after such partial-liver transplants. We looked at liver regeneration, as measured by computed tomography (CT) volumetrics, to see if there were significant differences among LDs, right-lobe LD recipients, and SLT recipients. We measured liver volume at 3 months postoperatively by using CT, and we compared the result to the patient's ideal liver volume (ILV), which was calculated using a standard equation. The study group consisted of 70 adult patients who either had donated their right lobe for LD transplants (n = 24) or had undergone a partial-liver transplant (right-lobe LD transplants, n = 24; right-lobe SLTs, n = 11; left-lobe SLTs, n = 11). DD (vs. LDs) were younger (P < 0.01), were heavier (P = 0.06), and had longer ischemic times (P < 0.01). At 3 months postoperatively, LDs had attained 78.6% of their ILV, less than the percentage for right-lobe LD recipients (103.9%; P = 0.0002), right-lobe SLT recipients (113.6%; P = 0.01), and left-lobe SLT recipients (119.7%; P = 0.0006). When liver size at the third postoperative month was compared with the liver size immediately postoperatively, LDs had a 1.85-fold increase. This was smaller than the increase seen in right-lobe LD recipients (2.08-fold), right-lobe SLT recipients (2.17-fold), and left-lobe SLT recipients (2.52-fold). In conclusion, liver regeneration, as measured by CT volume, seems to be greatest in SLT recipients. LD recipients seem to have greater liver growth than their donors. The reason for this remains unclear. (Liver Transpl 2004;10:374–378.)

The liver differs from most other organs in its remarkable ability to regenerate after injury or partial resection. This ability has been well documented and studied in animal models of liver resection as well as in human models, the latter usually involving liver resection for an underlying disease process.1–3 Recently, the tremendous increase in the number of partial-liver transplants has renewed interest in studies of liver regeneration. Such transplants, which include split-liver transplants (SLTs) from deceased donors (DD) as well as living donor (LD) liver transplants, are now beginning to account for a significant proportion of all liver transplants performed. Recipients of such transplants, as well as the LDs undergoing partial hepatectomy, constitute an ideal group in which to develop clinical models to study liver regeneration.

Liver growth after such transplants occurs at a rapid pace. Most recipients and LDs have near-complete regeneration of their liver volumes within a matter of a few weeks.4 Both the inability to achieve this degree of regeneration and the ability to perhaps increase this pace would have important clinical implications. Incomplete regeneration may have an impact on long-term liver function for both LDs and recipients. Increasing the rate of regeneration may have benefits for both LDs and recipients. For example, allowing for smaller pieces of the liver to be successfully transplanted might make surgery safer for LDs and more effective for recipients.

The purposes of this study were to measure hepatic regeneration after partial transplants, to compare results in donors and recipients, and to determine the impact of factors such as donor source and graft type on measured liver volumes.


SLT, split-liver transplant; DD, deceased donor; LD, living donor; CT, computed tomography; ILV, ideal liver volume; GW/RW, graft weight/recipient weight; Post-tx, post-transplant; INR, international normalized ratio.


Between January 1, 1999, and May 31, 2003, we performed 46 partial liver transplants in 46 adult recipients at the University of Minnesota (24 right-lobe LD transplants and 22 adult SLTs [11 right-lobe, 11 left-lobe]). These 46 recipients, plus the 24 LDs who underwent right hepatectomy, constituted a total of 70 patients for our study. Pediatric patients (defined as <16 years of age) were not included in this analysis.

Living Donor Transplants

All LD transplants used the right hepatic lobe, with preservation of the middle hepatic vein in the donor. No intermittent vascular occlusion to the liver was carried out during the donor hepatectomy. The recipient operation was done in a standard fashion, with reimplantation of accessory hepatic veins and of veins draining segments 5 and 8 if deemed necessary because of the size of the vessels. Biliary reconstruction was with a hepaticojejunostomy in 22 of the recipients, duct to duct in 2.

Split Liver Transplants

All SLT graphs were prepared in situ, with transection of the liver through its midplane. Doing so generated a right-lobe graft (used in a normal-sized adult recipient, i.e. ≥70 kg) and a left-lobe graft (used in a smaller adult recipient, i.e. <70 kg). Details of the surgical procedure have been previously described.5 The middle hepatic vein and main hilar structures were preserved with the left-lobe graft.

Liver Volume

For LDs and all recipients, liver volume was measured at 3 months postoperatively using helical computed tomography (CT) studies. Spiral CT images of the abdomen were obtained using 1.0 mm slice thickness at 0.75 mm slice collimation. Source images were reviewed and postprocessed on a standard workstation. Multiplanar reconstructions and 3-dimensional volume-rendered images were obtained. The liver was sculpted every 5 mm, excluding the inferior vena cava but not the intrahepatic vessels, and a volume was calculated. The 3-month time point was used for the follow-up study because it represents a time when all donors and recipients are consistently seen at our center.

The patient's ideal liver volume (ILV) was calculated with the following equation6:

equation image

For all patients, we compared their measured actual liver volume at 3 months postoperatively with their calculated ILV to obtain the degree of liver regeneration. This value was expressed as a percentage. For LD as well as SLT recipients, we also compared liver volume at 3 months postoperatively with starting liver volume (measured intraoperatively) to determine the degree of liver graft growth. The volume of the graft at the time of surgery was assumed to have a volume corresponding to its measured weight, with a specific gravity of 1.0.7 This degree of growth was expressed as a ratio. For LDs, the volume of their residual liver mass (i.e., their left lobe), as measured by their preoperative CT scan, was used for comparison.

Statistical Analysis

Categorical variables were analyzed using the chi-square test and, when applicable, the Fisher exact test. Continuous variables were analyzed using the student's T-test.


We divided patients into 4 groups for analysis: right-lobe LDs (n = 24), right-lobe LD recipients (n = 24), right-lobe SLT recipients (n = 11), and left-lobe SLT recipients (n = 11). Patient characteristics for the 4 groups are shown in Table 1. LDs were older (P < 0.01) and smaller (P = 0.06) than deceased donors. Right-lobe grafts from the LD groups were significantly smaller (P = 0.04) than from deceased donors; right-lobe grafts from both LDs and deceased donors were significantly larger than left-lobe grafts from deceased donors (P = 0.01). However, the ratio of graft weight to recipient weight ratio was similar in all 3 recipient groups (P = not significant [NS]), close to 1.0%.

Table 1. Patient Characteristics By Group
 LDsRight-Lobe LD RecipientsRight-Lobe SLT RecipientsLeft-Lobe SLT RecipientsP Value (DD vs. LD)
  1. Abbreviations: GW/RW, graft weight/recipient weight; LD, living donor; SLT, split-liver transplant; DD, deceased donor.

Mean Donor Age (yrs)36362424<0.01
Mean Donor Weight (kg)72.472.486.586.50.06
Mean Graft Weight (gm)7957958755520.04
Mean GW/RW Ratio (%)0.991.020.93ns
Mean Recipient Age (yr)494840ns
Mean Cold Ischemia Time (hr)09.16.9<0.01

Liver Volume

The liver volumes for all living donors (n = 24) and for all recipients who were alive at 3 months postoperatively with their original graft (right-lobe LD recipients [n = 23], right-lobe SLT recipients [n = 10], and left-lobe SLT recipients [n = 9]) were measured and analyzed. Liver growth was expressed as a ratio of the volume at 3 months postoperatively to the immediate postoperative liver volume (Table 2; Fig. 1). For all 3 recipient groups, liver volume increased by a mean of about 2.2-fold by 3 months; we found no significant difference among these 3 groups (P = NS). The residual liver volume of LDs increased by a mean of about 1.8-fold, a somewhat lower growth than in the 3 recipient groups (P = 0.06).

Table 2. Liver Regeneration at 3 Months Postoperatively, As Measured by Change in Liver Volume and % Achievement of ILV at That Time
RecipientsChange in Liver Volume From Immediately Postoperatively% Achievement of ILV
MeanRangeStandard DeviationMeanRangeStandard Deviation
  1. Abbreviations: ILV, ideal liver volume; LD, living donor; SLT, split-liver transplant.

Living donors1.91.4 - 2.20.2878.662 - 10211.3
Right-lobe LD2.11.0 - 3.20.7103.977 - 16521.7
Right-Lobe SLT2.21.0 - 2.80.61113.680 - 15329
Left-Lobe SLT2.61.6 - 3.30.61121.497 - 17027.7
Figure 1.

Change in liver volume at 3 months postoperatively (vs. immediately postoperatively), by patient group.

Liver volume at 3 months postoperatively was expressed as a percentage of the calculated ILV (Table 2; Fig. 2). At this time, LDs had attained 78.6% of their ILV, less than the percentage for right-lobe LD recipients (103.9%; P = 0.0002), right-lobe SLT recipients (113.6%; P = 0.01), and left-lobe SLT recipients (119.7%; P = 0.0006).

Figure 2.

Percent achievement of ILV at 3 months postoperatively, by patient group. Dotted line represents ILV.

Figures 1 and 2 represent snapshots of liver volumes at 3 months posttransplant. It is unclear how much liver volumes changed before or after that. Very few patients had CT scans performed prior to 3 months. A number of the transplant recipients, however, underwent repeat CT imaging for various reasons (most commonly for follow-up due to an initial diagnosis of hepatocellular carcinoma) after 3 months posttransplant (Table 3). Of 7 recipients who underwent repeat imaging at a mean of 13 months (range, 8 to 30 months) posttransplant, liver volume as compared with the 3-month volume decreased by a mean of only 3.6% (range, −10.7% to +11%). None of the LDs underwent repeat CT imaging after 3 months postoperatively, as it was not felt to be clinically indicated. There were no CT scans available for assessment of liver volume prior to the 3-month mark.

Table 3. Change in Graft Size After 3 Months Postoperatively, in Recipients
RecipientGraft TypeTime Post-tx to 2nd CT (month)Change in Graft Size (percent)
  1. Abbreviations: Post-tx, post-transplant; LD, living donor; SLT, split-liver transplant; CT, computed tomography.

1Right-lobe LD16+11
2Right-lobe LD14+0.5
3Right-lobe LD6−6.0
4Right-lobe SLT15−13.2
5Right-lobe SLT18+0.1
6Left-lobe SLT30−7.0
7Left-lobe SLT8−10.7


Patients who undergo a partial-liver transplant or a partial hepatectomy offer ideal clinical models for the in vivo study of liver regeneration. Partial-liver transplants are becoming increasingly popular, given the growing organ shortage. To optimally use these partial grafts, to expand the number of such transplants, and to decrease the risks associated with them (both for LDs and for recipients) we need a better understanding of liver regeneration in the clinical situation is required.

In an animal model, the regenerating liver can be removed and its size accurately measured to quantify liver growth.8 Of course, doing so is not possible in the clinical situation. Other methods are needed to assess liver regeneration in human patients. Liver growth is generally associated with a normalization of synthetic liver function test results (e.g., serum bilirubin and international normalized ratio [INR]); therefore, serial measurements may serve as a crude indicator of liver regeneration. Radiologic imaging with CT scans to obtain volume measurements of the liver has been shown to correlate well with liver size9 and probably represents the best method at present to measure liver size.

Our analysis of CT scans for measuring liver volume yielded several observations regarding liver growth after surgery. The ability of the liver to rapidly regenerate was well illustrated in all 4 patient groups. By 3 months postoperatively, most patients had almost doubled their liver size, as compared with their residual liver volume immediately postoperatively. This rapid rate of liver growth has been well documented in numerous animal and clinical models.4, 8 Interestingly, in our study, recipients tended to have a greater increase in liver volume as compared with LDs. By 3 months posttransplant, liver volume increased by a mean of approximately 2.2-fold for all recipients (SLT and LD, right-lobe and left-lobe). For LDs, the mean increase was about 1.8-fold. This difference in growth may be partially due to the different methods used to measure the initial starting volume (i.e., the volume of the liver immediately postoperatively). For donors this was measured from a CT; for recipients this was measured using a scale in the operating room. However, there is generally a good correlation between these two methods for measurement of liver volume.9 Moreover, other studies of liver regeneration in living donors have reported similar values for liver volume changes after donation.10

A similar trend was noted when the liver volume at 3 months postoperatively was expressed as a percentage of the calculated ILV. Again, recipients seemed to have achieved a greater percentage of their ILV (approximately 110%) than did LDs (about 80%). Another interesting finding was that SLT recipients seemed to have somewhat more liver growth than LD recipients, with a higher percentage of ILV achieved. This was, however, not a statistically significant difference.

Admittedly, our liver volume data offer a snapshot only—specifically, at 3 months postoperatively. The majority of patients had only one CT scan after the procedure, so it is not possible to make estimates on the rate of liver regeneration and evaluate how it may differ among the 4 groups. To do so would require multiple CT scans at various times postoperatively, which we felt were not clinically indicated. It is possible that liver volume may continue to change after the 3-month time point, perhaps increasing in LDs as regeneration continues and decreasing in recipients as initial congestion and edema resolve. However, as shown in Table 3, liver volume changed only to a small degree after the first 3 months, at least in the recipients.

In summary, our study demonstrated that liver regeneration occurs at a rapid pace in recipients of partial-liver grafts and in LDs after partial hepatectomy. The liver of LDs seemed to regenerate at a less vigorous pace, as compared with recipients, and may not be as complete. However, it is important to keep in mind that these radiologic findings may not have any clinical impact, at least for LDs. Synthetic liver function was usually completely normal in LDs by 1 week postoperatively, so the fact that they achieve only 80% of their ILV by 3 months may not be of any clinical significance. Nonetheless, studies such as this are important to help solve the mystery of liver regeneration and to ultimately make partial-liver transplants safer for donors and more effective for recipients.


The authors thank Deann Ronning and Mary Knatterud for their help in the preparation of this manuscript.