Download Systematic Pelvic and Aortic Lymphadenectomy in

Ann Surg Oncol (2012) 19:3522–3527
DOI 10.1245/s10434-012-2400-9
ORIGINAL ARTICLE – GYNECOLOGIC ONCOLOGY
Systematic Pelvic and Aortic Lymphadenectomy in Advanced
Ovarian Cancer Patients at the Time of Interval Debulking
Surgery: A Double-Institution Case–Control Study
Anna Fagotti, MD, PhD1, Pierandrea De Iaco, MD2, Francesco Fanfani, MD1, Giuseppe Vizzielli, MD1,
Federica Perelli, MD1, Federica Pozzati, MD2, Anna Myriam Perrone, MD2, Luigi Carlo Turco, MD1, and
Giovanni Scambia, MD1
Gynecologic Oncology Unit, Catholic University of the Sacred Heart, Rome, Italy; 2Gynecology and Reproductive
Medicine Unit, S. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy
1
ABSTRACT
Background. The prognostic role of systematic lymphadenectomy remains unclear in advanced ovarian cancer
(AOC). Only few retrospective case series have investigated the percentage of lymph node metastases after
neoadjuvant chemotherapy. This multi-institutional casecontrol study analyzed the prognostic role of systematic
lymphadenectomy in AOC patients at the time of interval
debulking surgery (IDS).
Methods. From January 2005 to December 2010, the
records of patients with AOC admitted to IDS at the
Catholic University of Rome (n = 101, controls) and at the
University of Bologna (n = 50, cases) were retrospectively
analyzed. The cases, routinely submitted to systematic
pelvic and aortic lymphadenectomy, were matched 1:2
with the controls, who did not routinely undergo lymphadenectomy. To correctly assess the prognostic role of
lymphadenectomy, only patients with optimally debulked
disease were included. Progression-free survival and
overall survival were analyzed by a log-rank test.
Results. After an overall mean follow-up of 36 months
(95 % confidence interval 33–39), 35 and 63 recurrences
(70.0 vs. 62.4 %; p = NS) and 15 and 24 deaths due to
disease (30 vs. 23.7 %; p = NS) were observed in the case
Electronic supplementary material The online version of this
article (doi:10.1245/s10434-012-2400-9) contains supplementary
material, which is available to authorized users.
Ó Society of Surgical Oncology 2012
First Received: 10 January 2012;
Published Online: 30 May 2012
A. Fagotti, MD, PhD
e-mail: [email protected]
and controls, respectively. The 2-year progression-free
survival rate was 36 versus 25 % (p = 0.834), and the
2-year overall survival rate was 69 versus 88 %
(p = 0.777), in the case and controls, respectively. The
median operating time was longer, and the percentage of
patients requiring blood transfusions was higher in the
cases than in the controls (225 vs. 210 min, p = 0.023, and
54 vs. 22.8 %, p = 0.0001, respectively).
Conclusions. Lymphadenectomy at the time of IDS could
be omitted, at least in high-risk patients.
Primary debulking surgery, including pelvic and aortic
lymphadenectomy, is the cornerstone of treatment of
advanced ovarian cancer (AOC).1,2 However, a percentage
of patients, which varies between 10 and 50 % according to
different selection criteria and surgical attitudes adopted in
the centers, still undergo neoadjuvant chemotherapy followed by interval debulking surgery (IDS).3–6 At the time
of surgery, these women are submitted to a large assortment of procedures as a result of different
clinicopathologic response to chemotherapy. In this context, the role of systematic lymphadenectomy may be
discussed.7 Indeed, lymphatic spread has been reported to
be a common feature in AOC. In particular, from 50 to
80 % of these patients have positive lymph node at diagnosis, mainly in the para-aortic and iliac areas (68 and
49 %, respectively).8 In patients with optimal intraabdominal debulking (residual tumor\1 cm) at the time of
primary cytoreduction, systematic lymphadenectomy has
been shown to improve progression-free survival (PFS)
without any impact on overall survival (OS).2 In this
context, the weight of postoperative complications and
consequently quality of life should be considered.
Lymphadenectomy at Interval Debulking Surgery
At present, an international prospective randomized trial in
AOC patients without intraperitoneal residual tumor at primary
debulking is ongoing, exploring whether the addition of complete resection of retroperitoneal microscopic disease may
theoretically achieve a truly absent microscopic disease status
and lead to a survival advantage.9 However, although there is
some evidence for the role of lymphadenectomy in up-front
surgery, only few retrospective case series have described the
rate of lymph node metastases (which varies from 37.5 to 75 %)
after neoadjuvant chemotherapy.10–12 Moreover, the prognostic impact of such treatment has not yet been clarified.12
The present study was a case-control analysis that aimed
to investigate the prognostic role of systematic lymphadenectomy in AOC patients at the time of IDS after
neoadjuvant chemotherapy. We also evaluated the rate of
surgical complications.
3523
TABLE 1 Patient characteristics
Characteristic
No. of all patients
Age (years), median (range)
Value,
cases
50
Value,
controls
101
63 (35–76) 62 (31–80) 0.607
Stage, n (%)
0.408
III
37 (74.0)
81 (80.2)
IV
13 (26.0)
20 (19.8)
Histotope, n (%)a
Type I
Type II
No. of cycles of neoadjuvant
chemotherapy, median (range)
0.194
3 (6.0)
47 (94.0)
2 (2.0)
99 (98.0)
6 (6–10)
4 (3–9)
41 (82.0)
79 (78.2)
9 (18.0)
22 (21.8)
49 (98.0)
96 (95.0)
1 (2.0)
5 (5.0)
Clinical response (RECIST), n (%)
Complete/partial
Stable
PATIENTS AND METHODS
Stable
From January 2005 to December 2010, the medical records
of patients with AOC (International Federation of Gynecology and Obstetrics stage IIIC–IV disease) admitted to IDS at
the Catholic University of Rome (control group) and the
University of Bologna (case group) were retrospectively
analyzed. All patients were submitted to three or more courses
of neoadjuvant platinum-based chemotherapy. The median
number of neoadjuvant cycles were different at the centers.
However, the preoperatively assigned clinical response
according to both radiologic Response Evaluation Criteria in
Solid Tumors and serologic Gynecological Cancer InterGroup (GCIG) criteria were homogenous in both groups
(Table 1).13,14 Only patients with complete, partial, or stable
response were submitted to IDS. The extent of debulking
surgery was classified as standard, radical, and supraradical,
except for lymphadenectomy.15 All cases were submitted to
systematic pelvic and aortic lymphadenectomy, whereas only
bulky nodes were removed in the control group. At the end of
surgery, the size and site of residual tumor were registered,
and two or more cycles of chemotherapy were administered.
In order to correctly assess the prognostic role of lymphadenectomy at IDS, only cases that experienced optimal residual
tumor (B1 cm) after surgery were included in the analysis.
Cases were considered those women submitted to systematic
pelvic and aortic lymphadenectomy, as routinely performed at
Bologna University. They were matched as closely as possible
with a nearly double number of controls; control subjects did
not routinely undergo lymphadenectomy at the Catholic
University of the Sacred Heart in Rome.
The whole count of pelvic and aortic lymph nodes removed
as well as the number of bulky and metastatic nodes were
registered. Operative time was calculated starting from the
skin incision to the end of all surgical procedures. The decision
0.0001
0.588
Clinical response (GCIG), n (%)
Complete/partial
p
0.382
Bold value indicates statistically significant (p \ 0.05)
RECIST Response Evaluation Criteria in Solid Tumors
GCIG Gynecological Cancer InterGroup
a
According to Kurman and Shih17
of performing an autologous blood transfusion was made
during or after surgery according to patient hemodynamic
conditions and hemoglobin levels. Postoperative recovery
was calculated starting from the first postoperative day to the
day of hospital discharge.
Classification of postoperative complications was made
according to Dindo et al.16 Briefly, this classification categorizes the complications into five grades, as follows:
grade 1 (minor events not requiring therapy), grade 2
(complications requiring medications only without invasive
procedures), grade 3 (complications leading to lasting
disability or organ resection), grade 4 (life-threatening
complication requiring intensive care or treatment in the
intensive care unit), and grade 5 (death due to complications). The incidence of lymphocele was also analyzed at
follow-up by computed tomographic scan.
Lymphadenectomy
Pelvic lymphadenectomy dissection began at the origin of
the external iliac vessels and continued caudally along the
medial border of the psoas muscle, with the lower limit of the
external iliac lymphadenectomy being represented by the
deep inferior epigastric vessels. The lateral boundaries of
lymphadenectomy were delineated superficially by the fascia covering the psoas muscle and deeply by the fascia
covering the internal obturator and levator ani muscles. The
medial margin of lymphadenectomy was represented by the
3524
A. Fagotti et al.
umbilical artery. The clearing of the obturator space began
with identification of obturator nerve and removal of the
superficial obturator nodes. Lymph nodes along the external
iliac vessels were removed en bloc with those adjacent to the
common iliac vessels. Aortic lymphadenectomy dissection
began at the aortic bifurcation by removing the superficial
intercavoaortic, precaval, and preaortic nodal groups.
Lymph nodes located lateral to the vena (i.e., paracaval nodal
group) were separated from the vena cava, then removed en
bloc. Lymph nodes behind the vena cava were removed if
enlarged. Removal of the lateroaortic nodes was performed
up to the level of the left renal vein.
Statistical Analysis
Univariate analysis was performed to verify any difference
between the two groups. Univariate analysis included chisquare analysis or Fisher’s exact test when appropriate for
categorical variables and the Student t test and Mann–Whitney test when appropriate for continuous variables. Survival
time was measured from the day of histologic diagnosis to the
date of recurrence (PFS) or death (OS). Mean and life tables
were computed by the product-limit estimate by the Kaplan–
Meier method and then analyzed by the log-rank test. All
statistical tests were two sided, and differences were considered significant at p \ 0.05. Statistical analysis was
performed by SPSS statistical software (SPSS, Chicago, IL).
RESULTS
Patient demographics are shown in Table 1. Preoperative clinical characteristics did not significantly differ
TABLE 2 Operative and
postoperative details
between the groups. Patients presented to surgery with
similar percentages of radiologic and serologic response in
both groups, independently from the number of cycles of
neoadjuvant chemotherapy.
All women underwent IDS, achieving an optimal cytoreduction with macroscopically absent residual tumor in
almost 80 % of the patients, at the price of supraradical
surgical procedures in about a third of the case and control
subjects (26 vs. 36.6 %; p = NS).
The median number of lymph nodes removed was 38 in
the cases and 4 in the controls (p = 0.0001). Rates of bulky
nodes did not differ between the groups (12 vs. 8.9 % for
case and controls, respectively, p = 0.550), as well as the
rates of histologic confirmed positive nodes (14 of 50,
28 %, for cases vs. 3 of 9, 33.3 %, for controls;
p = 0.708). The comparison of rates of positive bulky
nodes was not statistically different between the groups (3
of 6, 50.0 %, in the cases vs. 3 of 9, 33.3 %, in the controls,
p = 0.519). No statistical comparison can be performed for
micrometastases. Regarding operative parameters, patients
submitted to systematic pelvic and aortic lymphadenectomy showed a statistically significant median longer
operative time, more blood loss, and higher rates of blood
transfusions, but they had comparable lengths of stay with
respect to controls (Table 2).
The analysis of complications showed an overall low
grade of complications in about 85 % of patients (Table 3).
A statistically significant higher rates of blood transfusions
(grade II) were observed in cases than in controls
(p = 0.001), but equal percentages of grade III–IV complications occurred in both groups (18 vs. 11.9 %;
p = NS). In particular, grade III complications consisted of
Characteristic
Value, cases
No. of patients
Type of surgery, n (%)
50
Value, controls
101
a
Standard
10 (20.0)
29 (28.7)
Radical
27 (54.0)
35 (34.7)
Supraradical
13 (26.0)
37 (36.6)
Residual tumor (cm), n (%)
0
NA not applicable, DHb g/dl
largest variation in hemoglobin
levels between preoperative and
intra- or postoperative day 1
a
According to Sharma et al.15
0.075
0.791
41 (82.0)
81 (80.2)
9 (18.0)
20 (19.8)
Yes
6 (12.0)
9 (8.9)
No
44 (88.0)
92 (91.1)
11 (78.5)
NA
NA
38 (15–84)
\1
Bulky nodes, n (%)
Bold values indicate statistically
significant (p \ 0.05)
p
Microscopic positive nodes, n (%)
No. of lymph nodes removed, median (range)
0.550
4 (1–14)
0.0001
Operative time (min), median (range)
225 (125–390)
210 (90–400)
0.023
Blood loss (DHb g/dl), median (range)
3.25 (0.02–5.10)
1.40 (0.01–6.20)
0.0001
27 (54.0)
23 (22.8)
0.0001
8 (4–23)
9 (4–20)
0.232
Patients transfused, n (%)
Hospital stay (d), median (range)
Lymphadenectomy at Interval Debulking Surgery
3525
Cumulative
survival
TABLE 3 Postoperative complications
Characteristic
Value, cases
Value, controls
No. of all patients
50
101
p
Grade of surgical complications, n (%)a
I
20 (40.0)
*P = 0.777
1.0
Events Totals
Lymph
15
50
No lymph 24
101
0.8
73 (72.3)
0.0001
II
21 (42.0)
16 (15.8)
0.001
IIIa
0
1 (1.0)
0.480
IIIb
1 (2.0)
2 (2.0)
0.993
IVa
8 (16.0)
9 (8.9)
0.195
IVb
0
0
NA
0.6
0.4
0.2
Bold values indicate statistically significant (p \ 0.05)
NA not applicable
a
Classified according to Dindo et al.16
0
12
24
36
48
22
32
14
6
Months
Cumulative
survival
*P = 0.834
1.0
Lymph
No lymph
Events
35
63
Totals
50
101
0.8
Patients at risk
50
101
44
89
33
64
FIG. 2 Kaplan–Meier curves for ovarian cancer patients submitted
to IDS stratified by surgical performance: lymphadenectomy (lymph,
black line) vs. no lymphadenectomy (No lymph, gray line);
p = 0.777, log-rank test. OS is expressed in months on the x axis
0.6
0.4
0.2
0
12
24
36
48
Months
Patients at risk
50
101
48
95
22
38
to disease (30 vs. 23.7 %; p = NS) were observed in cases
and controls, respectively. No differences were observed in
terms of site of relapse according to the surgical procedure
adopted (Supplementary Table 1). Two-year PFS was 36
versus 25 % in cases and controls, respectively
(p = 0.834) (Fig. 1). Two-year OS was 69 versus 88 % in
cases and controls, respectively (p = 0.777) (Fig. 2).
DISCUSSION
13
6
6
2
FIG. 1 Kaplan–Meier curves for ovarian cancer patients submitted
to IDS stratified by surgical performance: lymphadenectomy (Lymph,
black line) vs. no lymphadenectomy (No lymph, gray line);
p = 0.834, log-rank test. PFS is expressed in months on the x axis
three reinterventions for bowel occlusion and one postoperative pleural drainage. The apparent high incidence of
grade IV complications has to be related to postoperative
admission to the intensive care unit, established at preoperative anesthesiologic evaluation. Follow-up computed
tomographic scan revealed a higher rate of postoperative
lymphocele in the cases (26 %) than in controls (0.9 %)
(p = NS).
Follow-up update was stopped after 48 months so we
would have comparable results for survival analysis. Mean
follow-up was 36 (95 % confidence interval 31–41) and 35
(95 % confidence interval 32–38) months in cases and
controls, respectively. During this time, 35 and 63 recurrences (70 vs. 62.4 %; p = NS) and 15 and 24 deaths due
As a result of the lack of standard staging procedures at
the time of IDS, the starting hypothesis of this study was
that AOC patients would behave similarly both at primary
cytoreduction and after neoadjuvant chemotherapy. In
particular, the persistence of microscopic lymph nodal
metastases would negatively influence the occurrence of
lymph node recurrence, thus affecting disease-free survival
but not OS.2 On the contrary, results from the present study
suggested that systematic pelvic and aortic lymphadenectomy performed in AOC patients at the time of IDS after
neoadjuvant chemotherapy had no value in terms of
improvement in PFS and OS. In other words, because the
rates of histologically confirmed disease-positive nodes
were equal in cases and controls (28 vs. 33.3 %; p = NS),
it seems that the removal of bulky nodes has only the same
prognostic value as systematic lymphadenectomy at the
time of IDS.
However, as during primary surgery, systematic lymphadenectomy is associated with a statistically significant
longer operative time and higher risk of blood loss,
3526
transfusion, and lymphocele occurrence. Although hemorrhage is a well-known complication during pelvic and
aortic lymphadenectomy, our rates of blood transfusion are
lower than those reported during primary surgery.2 This
difference is obviously related to a more aggressive surgery
during primary cytoreduction, although lymphadenectomy
after chemotherapy may be more difficult as a result of a
plastic reaction of the retroperitoneum.
Although it could be argued that the number of women
who could benefit from such a strategy would not be very
numerous if the maximal surgical effort is always pursued at
the time of primary diagnosis, results from clinical practice
and from recent data of the European Organisation for
Research and Treatment of Cancer 55971 phase III randomized trial suggest that approximately 30 % of AOC
patients referred to a tertiary-level center do not receive an
optimal cytoreduction at the time of primary surgery.18,19 In
these cases, maximal surgical effort is widely accepted as a
standard treatment for peritoneal disease and bulky lymph
nodes, whereas the therapeutic value of systematic lymphadenectomy remains controversial, although it has been
shown to be technically feasible and relatively safe.20–22
To our knowledge, this is the first case-control study
analyzing the impact of systematic lymphadenectomy in
this selected group of patients. Limitations of this study
include its retrospective nonrandomized design, although
the close matching between cases and controls should
theoretically assure the quality of data. In particular, as a
result of the prognostic objective of this study, only
patients with no residual disease or minor residual disease
with a maximum diameter 10 mm have been included.
Moreover, type of surgery has been stratified according to
an international classification, and no statistically significant differences were found between the groups except for
lymphadenectomy.15
Similar data were published several years ago by an
Italian group.21 This study analyzed patients submitted to
systematic pelvic and aortic lymphadenectomy at secondlook laparotomy for ovarian cancer. They observed a
comparable rate of positive nodes (25.8 %) and no differences in terms of 5-year survival in women with or without
lymphadenectomy. The higher percentages of positive
nodes after chemotherapy reported by other studies may be
influenced by the small number of patients and the type of
chemotherapy used.22,23
Finally, similar rates of positive nodes before and after
chemotherapy have been reported, thus implying that nodal
metastases are not as chemosensitive as peritoneal
lesions.11 However, the study did not report a benefit in
terms of survival associated with systematic lymphadenectomy in IDS patients.
In conclusion, data currently available in the literature
are confusing regarding the role of systematic
A. Fagotti et al.
lymphadenectomy in AOC women after neoadjuvant chemotherapy. However, as evaluated by an adjusted risk
model, we believe that the strict statistical methodology of
this study substantiates the hypothesis that this therapeutic
option can be omitted, at least in high-risk patients.24
Further prospective randomized studies are needed to
definitively resolve this question.
ACKNOWLEDGMENT
support.
This study is not supported by any grant
CONFLICT OF INTEREST
conflicts of interest.
The authors indicate nopotential
REFERENCES
1. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines). Ovarian cancer including fallopian tube cancer and primary
peritoneal cancer. Version 2.2012. NCCN.org. http://www.nccn.
org/professionals/physician_gls/pdf/ovarian.pdf. Accessed 1 Mar
2012.
2. Panici PB, Maggioni A, Hacker N, et al. Systematic aortic and
pelvic lymphadenectomy versus resection of bulky nodes only in
optimally debulked advanced ovarian cancer: a randomized
clinical trial. J Natl Cancer Inst. 2005;97:560–6.
3. Eisenkop SM, Freiedman RL, Wang HJ. Complete cytoreductive
surgery is feasible and maximizes survival in patients with
advanced epithelial ovarian cancer: a prospective study. Gynecol
Oncol. 1998;69:103–8.
4. Schwartz PE. Neoadjuvant chemotherapy for the management of
ovarian cancer. Best Pract Res Clin Obstet Gynaecol. 2002;16:
585–96.
5. Ozols RF. Management of advanced ovarian cancer consensus
summary. Advanced Ovarian Cancer Consensus Faculty. Semin
Oncol. 2000;27:47–9.
6. Van Der Burg ME, van Lent M, Buyse M, et al. The effect of
debulking surgery after induction chemotherapy on the prognosis
in advanced epithelial ovarian cancer. N Engl J Med.
1995;332:629–34.
7. Chambers SK. Systematic lymphadenectomy in advanced epithelial ovarian cancer: two decades of uncertainty resolved. J
Natl Cancer Inst. 2005;97:548–9.
8. Harter P, Gnauert K, Hils R, et al. Pattern and clinical predictors
of lymph node metastases in epithelial ovarian cancer. Int J
Gynecol Cancer. 2007;17:1238–44.
9. AGO Study Group (AGO-OVAR). Lymphadenectomy in ovarian
neoplasms (LION). An open randomized prospective multicenter
trial. AGO-OVAR OP.3. NCT00712218. ClinicalTrials.gov. http://
clinicaltrials.gov/ct2/show/NCT00712218. Accessed 1 Feb 2012.
10. Du Bois A, Reuss A, Harter P, Pujade-Lauraine E, Ray-Coquard
I, Pfisterer J. Potential role of lymphadenectomy in advanced
ovarian cancer: a combined exploratory analysis of three prospectively randomized phase III multicenter trials. J Clin Oncol.
2010;28:1733–9.
11. Rouzier R, Bergzoll C, Brun JL, et al. The role of lymph node
resection in ovarian cancer: analysis of the surveillance, epidemiology, and end results (SEER) database. BJOG. 2010;117:
1451–8.
12. Joulie F, Morice P, Rey A, et al. Are nodal metastases in ovarian
cancer chemoresistant lesions? Comparative study of initial
lymphadenectomy or after chemotherapy. Gynecol Obstet Fertil.
2004;32:502–7.
Lymphadenectomy at Interval Debulking Surgery
13. Eisenhauer EA, Therasseb P, Bogaertsc J, et al. New response
evaluation criteria in solid tumours: revised RECIST guideline
(version 1.1). Eur J Cancer. 2009;45:228–47.
14. Gronlund B, Hansen HH, Høgdall C, Høgdall EV, Engelholm
SA. Do CA125 response criteria overestimate tumour response in
second-line treatment of epithelial ovarian carcinoma? Br J
Cancer. 2004;90:377–82.
15. Sharma S, Driscoll D, Oduns K, Venkatadri A, Lele S. Safety and
efficacy of cytoreductive surgery for epithelial ovarian cancer in
elderly and high-risk surgical patients. Am J Obstet Gynecol.
2005;193:2077–82.
16. Dindo D, Demartines N, Clavien PA. Classification of surgical
complications: a new proposal with evaluation in a cohort of
6336 patients and results of a survey. Ann Surg. 2004;240:
205–13.
17. Kurman RJ, Shih IeM. The origin and pathogenesis of epithelial
ovarian cancer: a proposed unifying theory. Am J Surg Pathol.
2010;34:433–43.
18. Rose PG, Nerenstone S, Brady MF, et al; Gynecologic Oncology
Group. Secondary surgical cytoreduction for advanced ovarian
carcinoma. N Engl J Med. 2004;351:2489–97.
19. Vergote I, Tropé CG, Amant F, et al. Neoadjuvant chemotherapy
or primary surgery in stage IIIC or IV ovarian cancer. N Engl J
Med. 2010;363:943–53.
3527
20. Di Re F, Baiocchi G, Fontanelli R, et al. Systematic pelvic and
paraaortic lymphadenectomy for advanced ovarian cancer:
prognostic significance of node metastases. Gynecol Oncol.
1996;62:360–5.
21. Baiocchi G, Grosso G, di Re E, Fontanelli R, Raspagliesi F, di Re
F. Systematic pelvic and paraaortic lymphadenectomy at secondlook laparotomy for ovarian cancer. Gynecol Oncol. 1998;69:
151–6.
22. Scarabelli C, Gallo A, Zarrelli A, Visentin C, Campagnutta E.
Systematic pelvic and para-aortic lymphadenectomy during cytoreductive surgery in advanced ovarian cancer: potential benefit
on survival. Gynecol Oncol. 1995;56:328–37.
23. Burghardt E, Girardi F, Lahousen M, Tamussino K, Stettner H.
Patterns of pelvic and paraaortic lymph node involvement in
ovarian cancer. Gynecol Oncol. 1991;40:103–6.
24. Aletti GD, Santillan A, Eisenhauer EL, et al. A new frontier for
quality of care in gynecologic oncology surgery: multi-institutional assessment of short-term outcomes for ovarian cancer
using a risk-adjusted model. Gynecol Oncol. 2007;107:99–106.