ZurichOpenRepositoryand
Archive
UniversityofZurich
UniversityLibrary
Strickhofstrasse39
CH-8057Zurich
www.zora.uzh.ch
Year:2017
RiskAdjustmentinALPPSIsAssociatedWithaDramaticDecreasein
EarlyMortalityandMorbidity
Linecker,Michael;Björnsson,Bergthor;Stavrou,GregorA;Oldhafer,KarlJ;Lurje,Georg;
Neumann,Ulf;Adam,René;Pruvot,Francois-René;Topp,StefanA;Li,Jun;Capobianco,Ivan;
Nadalin,Silvio;Machado,MarcelAutran;Voskanyan,Sergey;Balci,Deniz;Hernandez-Alejandro,
Roberto;Alvarez,FernandoA;DeSantibañes,Eduardo;Robles-Campos,Ricardo;Malagó,Massimo
;deOliveira,MichelleL;Lesurtel,Mickael;Clavien,Pierre-Alain;Petrowsky,Henrik
DOI:https://doi.org/10.1097/SLA.0000000000002446
PostedattheZurichOpenRepositoryandArchive,UniversityofZurich
ZORAURL:https://doi.org/10.5167/uzh-139107
JournalArticle
PublishedVersion
Originallypublishedat:
Linecker,Michael;Björnsson,Bergthor;Stavrou,GregorA;Oldhafer,KarlJ;Lurje,Georg;Neumann,
Ulf;Adam,René;Pruvot,Francois-René;Topp,StefanA;Li,Jun;Capobianco,Ivan;Nadalin,Silvio;
Machado, Marcel Autran;Voskanyan, Sergey; Balci,Deniz;Hernandez-Alejandro, Roberto;Alvarez,
FernandoA;DeSantibañes,Eduardo;Robles-Campos,Ricardo;Malagó,Massimo;deOliveira,Michelle
L;Lesurtel,Mickael;Clavien,Pierre-Alain;Petrowsky,Henrik(2017).RiskAdjustmentinALPPSIs
AssociatedWithaDramaticDecreaseinEarlyMortalityandMorbidity.AnnalsofSurgery,266(5):779-
786.
DOI:https://doi.org/10.1097/SLA.0000000000002446
Risk Adjustment in ALPPS Is Associated With a Dramatic Decrease
in Early Mortality and Morbidity
Michael Linecker, MD,
Bergthor Bjo
¨
rnsson, MD, PhD,yz Gregor A. Stavrou, MD,§ô Karl J. Oldhafer, MD,§ô
Georg Lurje, MD,jj Ulf Neumann, MD,jj Rene
´
Adam, MD,
Francois-Rene
´
Pruvot, MD,yy
Stefan A. Topp, MD,zz Jun Li, MD,§§ Ivan Capobianco, MD,ôô Silvio Nadalin, MD,ôô
Marcel Autran Machado, MD,jjjj Sergey Voskanyan, MD,
Deniz Balci, MD,yyy
Roberto Hernandez-Alejandro, MD,zzz
Fernando A. Alvarez, MD,§§§
Eduardo De Santiban
˜
es, MD, PhD,§§§ Ricardo Robles-Campos, MD,ôôô Massimo Malago
´
, MD,jjjjjj
Michelle L. de Oliveira, MD,
Mickael Lesurtel, MD, PhD,
Pierre-Alain Clavien, MD, PhD,
and Henrik Petrowsky, MD
Objective: To longitudinally assess whether risk adjustment in Associating
Liver Partition and Portal Vein Ligation for Staged Hepatectomy (ALPPS)
occurred over time and is associated with postoperative outcome.
Background: ALPPS is a novel 2-stage hepatectomy enabling resection of
extensive hepatic tumors. ALPPS has been criticized for its high mortality,
which is reported beyond accepted standards in liver surgery. Therefore,
adjustments in patient selection and technique have been performed but have
not yet been studied over time in relation to outcome.
Methods: ALPPS centers of the International ALPPS Registry having per-
formed 10 cases over a period of 3 years were assessed for 90-day
mortality and major interstage complications (3b) of the longitudinal study
period from 2009 to 2015. The predicted prestage 1 and 2 mortality risks were
calculated for each patient. In addition, questionnaires were sent to all centers
exploring center-specific risk adjustment strategies.
Results: Among 437 patients from 16 centers, a shift in indications toward
colorectal liver metastases from 53% to 77% and a reverse trend in biliary
tumors from 24% to 9% were observed. Over time, 90-day mortality
decreased from initially 17% to 4% in 2015 (P ¼ 0.002). Similarly, major
interstage complications decreased from 10% to 3% (P ¼ 0.011). The
reduction of 90-day mortality was independently associated with a risk
adjustment in patient selection (P < 0.001; OR: 1.62; 95% CI: 1.36–1.93)
and using less invasive techniques in stage-1 surgery (P ¼ 0.019; OR: 0.39;
95% CI: 0.18–0.86). A survey indicated risk adjustment of patient selection in
all centers and ALPPS technique in the majority (80%) of centers.
Conclusions: Risk adjustment of patient selection and technique in ALPPS
resulted in a continuous drop of early mortality and major postoperative
morbidity, which has meanwhile reached standard outcome measures
accepted for major liver surgery.
Keywords: ALPPS, colorectal liver metastases, less invasive ALPPS
variants, outcome, risk adjustment, two-stage hepatectomy
(Ann Surg 2017;xx:xxx–xxx)
A
ssociating Liver Partition and Portal vein Ligation for Staged
Hepatectomy (ALPPS) is a novel 2-stage hepatectomy variant
that combines portal vein occlusion and parenchymal transection at
the first stage.
1,2
The major advantage of this procedure is an
exceptionally fast liver growth compared with ‘‘conventional’’
2-stage hepatectomies
3–6
enabling a higher resectability rate.
7
The
initial enthusiasm for the procedure has been consistently challenged
due to the high reported morbidity and mortality labeling this
procedure as a high-risk operation.
7
Some hepatobiliary centers have
therefore discontinued performing this operation.
Many surgical procedures including liver surgery have expe-
rienced unfavorable outcomes in the beginning, which was followed
by continuous improvement with more careful patient selection and
technical refinement. Previous reports on ALPPS, most of them with
a small number of patients, reveal a wide range of early postoperative
mortality ranging from 1% to 25%.
8–13
All of these studies are
mainly reporting morbidity and mortality rates of pooled populations
and outcome studies of ALPPS over time are lacking.
Therefore, the aim of the present study was to analyze whether
adjustments in patient selection, ALPPS technique, and interstage
management occurred and whether those were associated with
change of morbidity and mortality over time. For this purpose, we
From the
Swiss HPB and Transplantation Center, Department of Surgery,
University Hospital Zurich, Z urich, Switzerland; yDepartment of Surgery,
Linko
¨
ping University, Linko
¨
ping, Sweden; zDepartment of Clinical and
Experimental Medicine, Linko
¨
ping University, Linko
¨
ping, Sweden;
§Department of General and Abdominal Surgery, Asklepios Hospital Barm-
bek, Hamburg, Germany; ôSemmelweis University Budapest, Campus Ham-
burg, Germany; jjDepartment of General, Visceral and Transplantation
Surgery, University Hospital Aachen, RWTH Aachen, Aachen, Germany;
Centre He
´
pato-Biliaire, Ho
ˆ
pital Paul Brousse, Villejuif, France;
yyDepartment of Digestive Surgery and Transplantation, University Hospital,
Lille, France; zzDepartment of Surgery, University Hospital Du
¨
sseldorf,
Du
¨
sseldorf, Germany; §§Department of Hepatobiliary Surgery and Transplan-
tation University Medical Center Hamburg-Eppendorf Hamburg, Hamburg,
Germany; ôôDepartment of General, Visceral, and Transplant Surgery, Uni-
versity Hospital Tu
¨
bingen, Tu
¨
bingen, Germany; jjjj Department of Surgery,
University of Sa
˜
o Paulo, Sa
˜
o Paulo, Brazil;
Department of Surgery, A.I.
Burnazyan FMBC Russian State Scientific Center of FMBA, Moscow, Russia;
yyyDepartment of Surgery, Ankara University, Ankara, Turkey; zzzDivision of
HPB Surgery and Liver Transplantation, Department of Surgery, London
Health Sciences Centre, London, Ontario, Canada; §§§Liver Transplant Unit,
Division of HPB Surgery, Department of Surgery, Italian Hospital Buenos
Aires, Buenos Aires, Argentina; ôôôDepartment of Surgery and Liver
and Pancreas Transplantation, Virgen de la Arrixaca Clinic and University
Hospital, Murcia, Spain; jjjjjjDepartment of HPB- and Liver Transplantation
Surgery, University College London, Royal Free Hospitals, London, UK; and
Division of Transplantation, Hepatobiliary Surgery, University of
Rochester, Rochester, USA.
The study was supported by the Clinical Research Priority Program of the
University of Zurich as part of the project ‘‘Nonresectable liver tumors’’
and the Liver and Gastrointestinal Disease Foundation.
The authors report no conflicts of interest.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of
this article on the journal’s Web site (www.annalsofsurgery.com).
Reprints: Henrik Petrowsky, MD, FEBS, FACS, Swiss HPB and Transplantation
Center Zurich, Department of Surgery and Transplantation, University
Hospital Zu
¨
rich, Ra
¨
mistrasse 100, CH-8091 Zu
¨
rich, Switzerland. E-mail:
Henrik.Petrows ky@usz.ch.
Copyright
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0003-4932/16/XXXX-0001
DOI: 10.1097/SLA.0000000000002446
Annals of Surger y Volume XX, Number XX, Month 2017 www.annalsofsurgery.com | 1
ESA PAPER
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
designed a longitudinal cohort study based on identified centers of
the International ALPPS Registry meeting the inclusion criteria.
METHODS
Study Design
The present ALPPS cohort study is composed of data derived
from the International ALPPS Registry. The registry was set up in
2012 and is co-ordinately maintained by the Department of Surgery,
University of Zurich, Switzerland, approved by the Cantonal
Ethics Committee of Zurich (KEK 2013-0326) and is registered at
ClinicalTrials.gov (NCT01924741). The registry serves as a data
platform to prospectively collect the worldwide experience of this
procedure using a web-based data capture system secuTrial (Inter-
active System, Berlin, Germany). Using a longitudinal study design,
the objective of the study was to investigate whether risk adjustment
and technical modifications occurred over time and whether those
changes had an impact on early outcome including morbidity and
mortality. The study was approved by the Scientific Committee of the
ALPPS Registry on May 18, 2016 (http://www.alpps.net/?q=node/
82). Registry data were exported for the current analysis on
August 29, 2016. Questionnaires were sent to all centers exploring
center-specific risk adjustment strategies.
Definition of the Study Population
In an attempt to longitudinally monitor the effect of risk
adjustment over time, only ALPPS centers entering 10 cases over
a duration 3 years were considered for the study population.
Patients of eligible centers who were captured in the International
ALPPS Registry until December 31, 2015 were included in the
analysis. All centers fulfilling the inclusion criteria were approached
(i) to provide missing data commonly observed in registries, (ii) to
provide detailed information on procedure-related technique not
captured in the registry (eg, technical variants
14
), and (iii) to disclose
their individual strategy of risk adjustment for ALPPS over time
using a survey questionnaire (Table 1).
Risk Adjustment and Outcome Measures
Primary outcome measures were 90-day mortality after stage
1 surgery as well as major interstage and poststage 2 complications
(3b) as a global performance metrics of procedure-related morbid-
ity over time starting from the first registered cases in 2009 until the
end of 2015.
Secondary outcome measures included comorbidities (Charl-
son comorbidity index, cardiovascular disease, chronic obstructive
pulmonary disease, diabetes, and renal disease), size and growth of
the future liver remnant [standardized future liver remnant (sFLR),
future liver remnant/bodyweight, sFLR increase, D sFLR], concom-
itant resections, interstage interval defined as time period between
stage 1 and stage 2 surgery, liver failure using Model of End-stage
Liver Disease score and International Study Group of Liver Surgery
criteria, laboratory tests at poststage-1 day 5 and prestage-2, feasi-
bility of stage 2, and length of intensive care unit and hospital stay.
Risk adjustment was tested by calculating the predicted pre-
stage 1 and 2 mortality risks
15
for each patient incorporating age,
tumor entity, interstage complications 3b, and prestage 2 serum
bilirubin and creatinine
15
(Supplementary Table 1, http://links.
lww.com/SLA/B310). The analysis was performed for the periods
2011, 2012, 2013, 2014, and 2015. Furthermore, the development
of technical refinements of the ‘‘classical’’ ALPPS procedure
14
toward less invasive ALPPS variants (‘‘technically modified
ALPPS’’), including portal vein embolization (PVE)-ALPPS, Partial
ALPPS, Laparoscopic ALPPS, Tourniquet ALPPS, and Mini-ALPPS
was longitudinally analyzed (Supplementary Table 2, http://links.
lww.com/SLA/B310). The combination of at least 2 of the latter
variants was summarized as ‘‘combined modification.’’
Statistical Analysis
Descriptive Statistics and Univariate Analysis
Data were expressed using median and interquartile range for
continuous and absolute number (%) for categorical variables.
Longitudinal trends between the years of ALPPS performed were
calculated using Spearman’s r correlation for continuous variables
and Kendalls t for categorical variables. In preparation for multivar-
iate analysis on 90-day mortality and major interstage complications,
variables of interest were tested using Kruskal–Wallis test for
continuous and x
2
square test for categorical variables. Only clini-
cally useful parameters selected in a discussion between biostatis-
tician and clinician were included in regression analysis avoiding
automatic variable selection. P values 0.05 were considered
statistically significant.
Multivariate Logistic Regression Analysis
Variables, which showed a statistically significant change over
time, were considered for regression analysis. Parameters, which
were hypothesized to reduce 90-day mortality and major interstage
complications were subsequently split into 2 categories: (1) variables
representing risk adjustment (prestage-1 and prestage-2 risk) and (2)
variables representing the technical refinement of the procedure. The
influence of the respective variables was quantified using odds ratio
(OR) and 95% confidence interval (CI). Calibration of regression
models was assessed using the Hosmer–Lemeshow test.
All statistical analyses were performed using IBM SPSS
Statistics version 22 for Macintosh (IBM Corporation, Armonk, NY).
RESULTS
Study Population With Improved Data Quality
Of 836 registered patients from 123 centers, 16 centers (13%)
were identified meeting the inclusion criteria (at least 10 cases over a
period of 3 yr), providing a total of 437 patients for analysis.
Percent of data completeness of registry-captured variables has been
significantly increased from 86% (75%–94%) at the time of registry
data export to 97% (92%–99%) after approaching all centers. In
addition, new information on technical aspects of ALPPS variants,
which did not exist in the registry, was collected reaching 99% of
data completeness.
Participating centers were mainly located in Europe (n ¼ 11),
followed by Asia (n ¼ 2), South America (n ¼ 2), and North America
TABLE 1. Survey With Response of 16/16 (100%)
Survey Questions n %
Was patient selection adjusted? (yes/no) 16/16 100/0
Patient selection was adjusted for:
Age 10 63
Tumor entity 14 88
Risk assessment before surgery
850
Liver function testing 10 63
Timing of stage 2 11 69
Was ALPPS technique modified over time? (yes/no) 12/4 75/25
Did risk adjustment or technique improve
safety of ALPPS? (yes/no)
16/16 100/0
Refers to following parameters: ALPPS risk score (n ¼ 1), ASA (n ¼ 2), ECOG
(n ¼ 1), frailty (n ¼ 1), BMI (n ¼ 1), cardiopulmonary reserve (n ¼ 3).
ASA indicates American Society of Anesthesiologists Score; BMI, body mass
index; ECOG, Eastern Cooperative Oncology Group Score.
Linecker et al Annals of Surger y Volume XX, Number XX, Month 2017
2 | www.annalsofsurgery.com
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
(n ¼ 1). A median number of 26 (17–35) patients per center were
enrolled over a median period of 5 years (4–5 yr). Eleven of
16 (69%) centers started ALPPS in the early period (2011) and
5 centers (31%) in 2012. ALPPS procedures were annually distrib-
uted with 48 (11%), 103 (24%), 101 (23%), 105 (24%), and 79 (18%)
for the annual periods 2011, 2012, 2013, 2014, and 2015 (Supple-
mentary Table 3, http://links.lww.com/SLA/B310).
Survey Exploring Center-specific Risk Adjustment
Strategies
Before analysis, all centers participated in a survey exploring
center-specific risk adjustment strategies (Table 1). All centers
(n ¼ 16) stated that they adjusted patient selection over time for
age (63%), risk assessment before surgery (50%), liver function
testing (63%), and timing of stage-2 surgery (69%). The majority of
centers (75%) modified their surgical ALPPS technique.
Longitudinal Improvement of Procedure-related
Safety
Study population characteristics are presented in Table 2, and
operative characteristics in Table 3. Over time, there was a significant
decrease in the annual 90-day mortality rate (t ¼0.124;P¼ 0.002)
starting from 17% in the early period (2011) to 3.8% in 2015
(Fig. 1A). This development was parallelly accompanied by a steady
reduction of the annual overall (t ¼0.151;P< 0.001) and major
(t¼0.102;P¼ 0.011) interstage complication rates with 78% and
10% for the period 2011 versus 56% and 3% for 2015 (Fig. 1B,
Table 4).
In the same line, the prestage-2 ALPPS risk model as a
measure of interstage performance revealed a substantial decrease
over time from a mean of 11.6% to 3.1% (r¼0.260; P < 0.001;
Fig. 1A). Interstage serum bilirubin and international normalized
ratio levels, Model of End-stage Liver Disease, as well as intensive
care unit stay significantly decreased over time (Table 4).
Liver volumetric measures before stage-2 such as sFLR and
future liver remnant/bodyweight ratio changed toward smaller vol-
umes over time (Table 4). Liver mass gain, as measured by D sFLR,
also slightly decreased over time (Table 4). Interestingly, a prolonga-
tion of the interstage interval (10 d in the early period 2011 vs
14 d in 2015) was noted (Table 4). This development and the
increasingly use of functional liver tests indicate a more cautious
progression to stage-2 surgery. Failure to reach stage-2 surgery
slightly increased over time with a borderline significance (r ¼
0.051; P ¼ 0.467; Table 4). Of note, causes of mortality did not
change significantly over the years. For the periods 2011, 2012,
2013, 2014, and 2015 liver failure occurred in 3, 5, 7, 3, and 2 cases
(t ¼ 0.084, P ¼ 0.507), death of sepsis/infection in 8, 7, 6, 5, and
1 cases (t ¼0.219, P ¼ 0.070), cardiac death in 0, 2, 1, 0, and 0
cases ( t ¼0.064, P ¼ 0.420), and other causes of death in 0, 5, 2, 2,
and 0 cases (t ¼ 0.007, P ¼ 0.951). Due to the fact, that death of liver
failure or sepsis/infection is in some cases difficult to distinguish, in
8 patients cause of death was categorized for both.
TABLE 2. Study Population
Completeness
(%)
Parameter Before After 2011 2012 2013 2014 2015 Correlation P
Demographics
Age, yr 96 100 63 (54–71) 62 (53–70) 63 (56–70) 62 (52–71) 59 (49–65) 0.107 0.026
Sex; male, n, % 98 100 29 (60) 59 (57) 72 (71) 62 (59) 49 (62) 0.011 0.809
BMI, kg/m
2
97 100 25 (23–28) 26 (23–28) 27 (24–29) 25 (23–29) 25 (23–29) 0.017 0.726
Tumor
CRLM, n, % 94 99 27 (53) 62 (59) 75 (75) 76 (73) 61 (77) 0.127 0.003
Biliary tumor, n, % 94 100 12 (24) 21 (20) 14 (14) 17 (16) 7 (9) 0.107 0.013
HCC, n, % 90 99 1 (2) 10 (9) 7 (7) 5 (5) 3 (4) 0.040 0.311
Other, n, % 94 99 11 (21) 13 (12) 4 (4) 6 (6) 8 (10) 0.104 0.039
Comorbidities
Charlson comorbidity index — 100 6 (2–6) 6 (3–8) 6 (4–7) 6 (4–7) 6 (4–6) 0.021 0.665
Cardiovascular disease, n, % 98 100 17 (35) 42 (41) 44 (44) 32 (31) 20 (25) 0.093 0.026
COPD, n, % 98 100 2 (4) 1 (1) 3 (3) 4 (4) 2 (3) 0.015 0.728
Diabetes, n, % 98 100 4 (8) 11 (11) 8 (8) 13 (13) 4 (5) 0.022 0.594
Renal disease, n, % 98 100 3 (6) 2 (2) 1 (1) 1 (1) 0 (0) 0.100 0.048
Liver baseline characteristics
sFLR 88 95 0.23 (0.16–0.29) 0.24 (0.17–0.30) 0.22 (0.17–0.26) 0.23 (0.19 –0.28) 0.21 (0.16 – 0.26) 0.033 0.502
FLR/BW 88 96 0.37 (0.27–0.52) 0.37 (0.28–0.49) 0.37 (0.29–0.48) 0.36 (0.29 –0.45) 0.31 (0.25 – 0.48) 0.081 0.097
Bilobar tumor, n, % — 98 25 (52) 67 (65) 63 (64) 71 (71) 55 (71) 0.095 0.030
Chemotherapy
, n, % 98 100 20 (74) 56 (90) 69 (92) 69 (91) 58 (95) 0.112 0.043
Steatohepatitis, n, % 48 74 7 (22) 7 (10) 9 (12) 11 (14) 8 (12) 0.024 0.645
Fibrosis, n, % 46 76 7 (21) 12 (16) 14 (19) 18 (22) 10 (14) 0.015 0.759
Macrosteatosis, n, % 45 75 6 (18) 23 (31) 20 (28) 22 (28) 17 (25) 0.003 0.959
Serum Creatinine, mg/dL 82 92 0.84 (0.60– 0.98) 0.78 (0.65–0.90) 0.84 (0.70–0.92) 0.81 (0.70 –0.90) 0.81 (0.62 – 0.95) 0.000 0.996
Serum Bilirubin, mg/dL 91 98 0.60 (0.40 – 0.98) 0.64 (0.41 – 0.90) 0.50 (0.40–0.78) 0.50 (0.36–0.70) 0.50 (0.32 – 0.70) 0.219 <0.001
INR 86 91 1.0 (0.9–1.1) 1.0 (0.9–1.1) 1.0 (0.9–1.1) 1.0 (0.9–1.1) 1.0 (0.9–1.1) 0.018 0.719
MELD 76 84 7 (6– 8) 6 (6–8) 7 (6–8) 6 (6–7) 6 (6–7) 0.066 0.203
Continuous variables presented as median and interquartile range (IQR), correlation with Spearmans r; categorical variables presented as count and percent (%), correlation with
Kendalls t.
Refers to CRLM only.
BMI indicates body mass index; COPD, chronic obstructive pulmonary disease; FLR/BW, future liver remnant/bodyweight; INR, international normalized ratio; MELD, Model of
End-stage Liver Disease.
Annals of Surger y Volume XX, Number XX, Month 2017 ALPPS Risk Adjustment
ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.annalsofsurgery.com | 3
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Adjustment in Patient Selection Is Associated With
Reduced Early Mortality
The prestage-1 risk model of the ALPPS Risk Score
15
was
used as a standardized measure of patient selection for ALPPS
upfront. A significant shift toward younger patients (63 yr in
2011 to 59 yr in 2015) as well as toward colorectal liver metastases
(CRLM) (53% in 2011 to 77% in 2015) with concomitant decline
in biliary tumors (24% in 2011 to 9% in 2015) was observed
(Fig. 1C, Table 2). These changes in age and indications are also
reflected in a significant annually drop of the mean predicted pre-
stage-1 mortality risks (24% in 2011 vs 9% in 2015; r ¼0.168;
P < 0.001) (Fig. 1A).
Of note, CRLM patients in the later periods were more likely
to undergo chemotherapy than in the earlier periods suggesting a
more cautious management in terms of response to chemotherapy
(Table 2).
Multivariate logistic regression analysis identified the pre-
stage 1 risk model to be independently associated with a decline in
90-day mortality (P < 0.001; OR: 1.62; 95% CI: 1.36–1.93; Sup-
plementary Table 4, http://links.lww.com/SLA/B310). The Hosmer–
Lemeshow test revealed a stable model fit (P ¼ 0.429).
Less Invasive ALPPS Techniques Are Associated
With Improved Safety
Technically modified ALPPS procedures were independently
developed to reduce invasiveness of stage 1 surgery aiming to
improve safety. Over time, there was significantly increased use
of less invasive ALPPS variants (t ¼ 0.098;P¼ 0.027), including
PVE-ALPPS, Partial ALPPS, Laparoscopic ALPPS, Tourniquet
ALPPS, and Mini-ALPPS with a corresponding decline of ‘‘classic’’
ALPPS cases (t ¼0.102;P¼ 0.021) (Fig. 1D, Table 3). In 2015,
less invasive variants represented the half of all ALPPS cases (52%)
of which partial ALPPS was the most frequently performed variant
(68%).
Independently of patient selection using the prestage 1 risk
model, multivariate regression analysis showed that less invasive
ALPPS variants were associated with decreased 90-day mortality
rates (P ¼ 0.019; OR: 0.39; 95% CI: 0.18–0.86; Supplementary
Table 4, http://links.lww.com/SLA/B310). When the variable ‘‘year
of ALPPS performed’’ was included into the regression model, the
analysis revealed that the year of ALPPS performed did not signifi-
cantly affect the outcome (P ¼ 0.085; Supplementary Table 4, http://
links.lww.com/SLA/B310). This implies that improved outcome of
less invasive variants is not exclusively a result of a potential learning
curve bias of early cases.
Adjustment in Patient Selection and ALPPS
Technique Results in Improved Interstage
Complication Profile
To further investigate why patient selection and technical
refinement of stage 1 surgery translates into a dramatic decrease
of early mortality, we tested whether this finding was accompanied
by a parallel reduction of major interstage complications. Both
overall (t ¼0.151;P< 0.001) and major (t ¼0.102;P¼
0.011) interstage complications significantly declined over time
(Fig. 1B). Multivariate regression analysis demonstrated that a
TABLE 3. Operative Characteristics
Completeness
(%)
Parameter Before After 2011 2012 2013 2014 2015 Correlation P
‘‘Classic’’ ALPPS
,n,% — 100 33 (69) 69 (66) 75 (74) 71 (68) 38 (48) 0.102 0.021
Technically modified ALPPS, n, % — 100 15 (31) 35 (34) 26 (26) 34 (32) 41 (52) 0.098 0.027
One modification, n, % — 100 14 (29) 31 (30) 18 (18) 29 (28) 24 (30) 0.004 0.928
Combined modification
y
(>1), n, % — 100 1 (2) 4 (4) 8 (8) 5 (5) 17 (22) 0.173 <0.001
PVE-ALPPS, n, % — 99 6 (13) 4 (4) 11 (11) 7 (7) 8 (10) 0.015 0.741
Partial ALPPS, n, % — 99 5 (10) 10 (10) 13 (13) 20 (20) 28 (36) 0.180 <0.001
Tourniquet ALPPS, n, % — 99 5 (10) 19 (18) 9 (9) 8 (8) 5 (6) 0.096 0.024
Laparoscopic ALPPS stage 1, n, % 77 99 1 (2) 3 (3) 1 (1) 0 (0) 6 (8) 0.055 0.319
Laparoscopic ALPPS stage 2, n, % — 100 1 (2) 3 (3) 1 (1) 2 (2) 6 (8) 0.068 0.188
Mini-ALPPS, n, % — 99 0 (0) 0 (0) 0 (0) 2 (2) 8 (10) 0.180 0.001
Rescue ALPPS, n, % — 99 6 (13) 4 (4) 9 (9) 11 (11) 2 (3) 0.029 0.480
Stage 1
Concomitant resections, n, % — 99 7 (15) 14 (14) 8 (8) 8 (8) 5 (6) 0.087 0.048
Colorectal resections, n, % — 99 2 (4) 5 (5) 11 (11) 7 (7) 4 (5) 0.013 0.742
Gastric/bowel resections, n, % — 99 3 (6) 1 (1) 2 (2) 1 (1) 1 (1) 0.058 0.250
Pancreatic resections, n, % — 99 1 (2) 4 (4) 0 (0) 1 (1) 0 (0) 0.083 0.066
Cleaning of the FLR, n, % 40 99 19 (40) 58 (55) 53 (53) 57 (55) 52 (67) 0.101 0.018
Operation time, min 81 99 331 (268–480) 330 (240–330) 317 (241 – 420) 300 (216 –405) 320 (225–433) 0.060 0.211
Pringle, n, % 53 95 15 (34) 22 (22) 31 (32) 23 (23) 21 (30) 0.005 0.919
Transfusion, n, % 95 100 12 (25) 21 (20) 20 (20) 18 (17) 6 (8) 0.111 0.007
Stage 2
Concomitant resections, n, % — 100 1 (2) 2 (2) 2 (2) 2 (2) 3 (4) 0.030 0.527
Colorectal resections, n, % — 100 1 (2) 1 (1) 0 (0) 2 (2) 2 (3) 0.033 0.525
Gastric/bowel resections, n, % — 98 0 (0) 0 (0) 0 (0) 1 (1) 0 (0) 0.032 0.318
Pancreatic resections, n, % — 98 0 (0) 0 (0) 1 (1) 0 (0) 0 (0) 0.005 0.399
Operation time, min 66 95 156 (120 – 213) 150 (120 –200) 155 (110 – 218) 153 (115 – 213) 175 (116 – 259) 0.038 0.444
Continuous variables presented as median and interquartile range (IQR), correlation with Spearmans r; categorical variables presented as count and percent (%), correlation with
Kendalls t.
Refers to the initially described procedure (REF 1,2).
yCombination of at least 2 of the following 5 ALPPS variants.
Linecker et al Annals of Surger y Volume XX, Number XX, Month 2017
4 | www.annalsofsurgery.com
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
