Assessment of hepatic function: Implications for perioperative outcome and recovery

Techniques of volumetry

Formal measurement of liver volumes is most commonly accomplished by using computed tomography (CT) or magnetic resonance imaging (MRI). Other imaging modalities may also be used, but CT and MRI are commonly obtained in patient care for characterization of lesions and operative planning, and therefore additional tests are typically not needed. Cross-sectional images obtained from either of these modalities are sequentially marked with the planned resection line, following which the surface area is derived and multiplied by the slice thickness ( Fig. 4.1 ). Excellent correlation has been demonstrated between the planned FLR and the actual FLR radiologically, as well as between the calculated resected liver volume and the surgical specimen. ^,

Because of the variability in total liver size based on patient body habitus, the FLR volume is typically expressed as a ratio of FLR to TLV. Although the measurement of the FLR is fairly standard, there are several variations to calculate the TLV. The simplest and most intuitive technique involves manually tracing the borders of the liver in a variety of planes and using software to calculate the total volume in the same manner as the FLR calculation. There are several limitations to this technique. Most notably, because resection is usually considered on the basis of hepatic tumors, the volume of the tumors is implicitly included in the measurement of the TLV. This is problematic because the tumor volume does not contribute to hepatic function and so provides a falsely elevated value of the TLV and hence a falsely diminished anticipated FLR ratio. Manually measuring the volume of each tumor and subtracting it from the TLV to yield the total functioning liver volume can correct this but can be labor intensive and prone to measurement error. Some software packages can perform automated subtraction of the tumor volume. The direct measurement technique of TLV is further limited by the fact that the parenchyma beyond tumors may be abnormal because of biliary or vascular obstruction. These limitations typically do not apply to the assessment of the FLR, which usually does not contain tumors.

An alternative method referred to as the total estimated liver volume (TELV) was first proposed by Urata and colleagues in Japan for use in liver transplantation. Rather than measuring the TLV directly and subtracting the volume of liver tumors, this technique estimates the TLV based on body surface area (BSA). The formula was subsequently modified to apply to Western patients, based on the observation that Urata’s formula underestimated TELV by an average of 323 cm ³ . ^, The resulting equation (TELV = −794 + 1267 × BSA) has been extensively studied and found to yield a precise estimate of TLV across institutions with different CT scanners and three-dimensional reconstruction techniques. When the TELV is used as the denominator to calculate the FLR ratio (i.e., FLR/TELV), the resultant ratio is referred to as the standardized FLR (sFLR).

The measured TLV was compared with the TELV in a study of 243 patients who underwent major liver resection (three or more segments). There was a strong correlation between the two measures across the population; however, in overweight patients (body mass index [BMI] > 25), TELV was significantly higher, yielding a lower sFLR in these patients. Based on the surgeons’ thresholds, 47 patients were deemed to have insufficient liver volume for resection using TLV compared with 73 patients using TELV. According to institutional practices at the time, patients who had sufficient liver volume based on TLV underwent resection. The subset of patients who had insufficient volume based on TELV had significantly higher rates of post-hepatectomy liver failure (PHLF) and mortality than did the patients who had sufficient volume based on both calculations. Therefore the authors concluded that TELV (i.e., sFLR) is a better measure of postoperative hepatic insufficiency risk.

Increasingly sophisticated software packages are being developed that incorporate semi-automated and fully automated segmentation for both CT and MRI. A number of studies have shown these to be very accurate and time-efficient when compared with the gold standard of manual volumetry for TLV ^, and individual liver segment volumes. These automated software packages have also been shown to be accurate and time-saving for living donor liver transplant patients and for planning a standard right trisectionectomy. Measuring the FLR for a resection not following a standard anatomic plane still requires manual volumetry.

Volumetric thresholds

Despite the refinement in methods to measure the FLR, the clinical application of the information gathered remains controversial. It has long been clear that patients with lower FLR are at increased risk for hepatic dysfunction, but the exact threshold below which resection should not be performed is debated. Several studies have attempted to address this fundamental question, yielding different conclusions. The variable results may be attributable to the heterogeneity of included patients (some having background liver disease and others healthy livers), methods used to calculate the FLR (TLV vs. TELV), indications for PVE, and definitions of hepatic dysfunction. Furthermore, only two studies analyzed their results using a formal receiving operator characteristic (ROC) curve to determine the optimal FLR threshold, and both studies were limited by small sample sizes. ^, Allowing for these admittedly crucial differences, the optimal cutoff for patients with a normal background liver appears to be between 20% and 30%. A 2006 expert consensus statement recommended a minimum of 20% FLR for major hepatic resection in a patient with a healthy liver and to consider PVE for any FLR less than that.

Patients who have received preoperative chemotherapy are at risk for background liver injury that impairs regeneration after partial hepatectomy (see Chapters 69 , 98 , and 102 ). There is general consensus that patients treated with extensive preoperative chemotherapy or who have evidence of background liver injury require a larger FLR to allow safe hepatectomy, although the exact threshold is again controversial. Two studies examined this question and performed formal ROC curve analyses, reporting optimal thresholds of 31% and 48.5%, respectively. ^, The largest study includes 194 patients undergoing extended hepatectomy on the right side, stratified by extent of preoperative chemotherapy, with long-duration chemotherapy defined as greater than 12 weeks (86 patients). Using a minimum P -value approach, the authors concluded that the optimal cutoff value of FLR for preventing postoperative liver insufficiency in these patients was 30%. Patients who have received extensive chemotherapy and have an sFLR between 30% and 40% should be investigated closely for any suggestion of underlying liver dysfunction and could be considered for PVE.

The optimal FLR threshold in patients with documented underlying liver disease is even less certain, given the additional variability of defining the extent of background liver injury. Some authors advocate for PVE in all patients with chronic liver disease before right side hepatectomy, and others apply a conservative threshold as high as 40%. ^, Given the importance of background liver function, additional functional tests to assess the liver remnant should be considered before embarking on major hepatectomy in the setting of significant background liver disease.

Volumetry after hypertrophy

In patients at increased risk for PHLF, hypertrophy of the FLR may be induced by preoperative ipsilateral PVE (see Chapter 102C ). Other techniques to achieve hypertrophy, such as the ALPPS procedure (see Chapter 102D ) or radioembolization with yttrium-90 (see Chapter 94 ), are discussed in other chapters. Cross-sectional imaging is typically repeated 2 to 6 weeks after PVE, and the FLR (or sFLR) may be recalculated. The post-PVE FLR can be interpreted with the same thresholds previously discussed, although the degree of hypertrophy (DH), defined as absolute difference between FLR before and after PVE, appears to be more informative. The authors of this study recommended that patients without cirrhosis undergoing PVE should have both an sFLR of at least 20% and a DH of at least 5% to undergo safe hepatectomy. In addition to its therapeutic intent, PVE functions as a diagnostic test analogous to a cardiac stress test; patients who do not experience substantial growth in the FLR after PVE should be suspected of harboring background liver disease and approached with caution.

Recognizing that the DH is contingent on the duration from PVE to reimaging, surgeons have proposed incorporating some measure of growth rate into consideration. The kinetic growth rate (KGR) may be calculated by dividing the DH by the number of weeks elapsed since PVE. In one study of 107 patients who underwent liver resection for colorectal liver metastases with an sFLR volume of greater than 20%, KGR was a more accurate predictor of postoperative hepatic insufficiency than absolute sFLR or DH (area under the curve [AUC] 0.830). In this study, patients with a KGR less than 2% per week suffered a 21.6% hepatic insufficiency rate and an 8.1% 90-day mortality rate compared with no hepatic insufficiency or 90-day mortality in patients with a KGR greater than 2% per week. In a similar study of 153 patients who underwent major hepatectomy after PVE, post-PVE absolute FLR correlated poorly with liver failure. Both DH and KGR were good predictors of liver failure (AUC 0.80 and 0.79, respectively). Notably, posthepatectomy liver failure did not develop in any patients with a KGR greater than 2.66% per week.

In summary, for patients with insufficient FLR (or sFLR) to safely undergo hepatectomy, response to PVE provides a good measure of the remnant liver’s ability to hypertrophy. Post-PVE FLR should be interpreted in combination with some measure of extent of hypertrophy (either DH or KGR) to optimally predict a patient’s risk for post-hepatectomy liver insufficiency.