818 860
4.
Discussion
In our series of 68 patients undergoing midline EP-RPLND,
we saw a return of bowel function on postoperative day
2 and median LOS of 3 d. There were no cases of ileus.
Notably, 41 patients (60%) were PC, for which 19 (46%) had a
mass
>
10 cm, yet no significant differences in outcomes
were seen between the primary and PC groups. We
postulate that by avoiding entry into the peritoneal cavity,
the risk of paralytic ileus and small bowel obstruction (SBO)
can be minimized, as can risk of injury to intraperitoneal
structures. Preserving the peritoneal sac also minimizes
insensible fluid loss, which is particularly valuable in the PC
bleomycin-treated population, in whom perioperative fluid
management is critical. Overall, our PC LOS represents a
significant difference to that found in a 2016 population-
based analysis of PC-RPLND morbidity, in which 25% of
patients had a hospital stay of
>
7 d (median 9 d)
[6].
Because this approach offers excellent exposure, even for
a full bilateral dissection in the setting of large masses, our
lymph node yield was not compromised, with a median of
36 nodes resected. This is in line with contemporary large
series reporting a median of 28–38 nodes resected; yields in
this range have been associated with a higher percentage of
positive nodes found and a lower risk of relapse
[7,8]. In
addition, nerve sparing was possible in all primary RPLND
patients and 78% of PC patients, with antegrade ejaculation
rates of 91.6% and 97.8%, respectively. This is consistent
with rates reported for recent series
[9].
Although RPLND has undergone significant refinements
since the original descriptions, it remains a fairly morbid
procedure for an otherwise healthy, young population
[10,11]. Even in the primary RPLND setting, Baniel et al
[12]described an 11% overall postoperative complication rate in
a series of more than 478 patients, the largest such series
reported, including 11 patients (2%) with SBO, of which
eight required emergent laparotomy and lysis of adhesions.
In 2014, Subramanian et al
[13]described a 24% complica-
tion rate in the primary setting and 32% in the PC setting in
their series of 204 patients at a tertiary center. This included
41 patients with ileus, ranging from a requirement for
bowel rest for at least 6 d to TPN (
n
= 16) and SBO requiring
reoperation (
n
= 1), with an additional two patients requir-
ing this in the long term. Two smaller contemporary cohorts
showed a decline in postoperative complications and
hospital stay over time, but still had LOS of 6 and 4 d,
respectively
[14,15]. Therefore, there should be a focus on
strategies that minimize perioperative and potentially long-
term complications, and a midline EP approach is an
example with demonstrated results.
Oncologic outcomes were not compromised in our
series. Four patients experienced relapse in the retro-
peritoneum. Three of these had clinical stage IIA embryonal
predominant NSGCT and underwent primary RPLND. The
first, who was a dialysis-dependent diabetic, developed an
in-field recurrence seen on surveillance computed tomog-
raphy; the patient underwent salvage chemotherapy with
resolution of this lesion, then developed lymphadenopathy
in the pelvis, with subsequent pelvic lymphadenectomy
revealing no disease. He was subsequently lost to follow up
but died of unknown causes 8 mo later. The second patient
developed in-field and pulmonary relapses 2 mo after
RPLND, was successfully salvaged with chemotherapy, and
remains disease free. The third patient had a persistent
1-cm mass in the retroperitoneum with negative markers,
but opted for salvage chemotherapy and is disease-free. Our
results are consistent with reported series showing that
30–40% of patients with stage II NSGCT will relapse after
primary RPLND, and that essentially all can be salvaged
successfully with chemotherapy
[16] .The last patient
underwent RPLND after salvage chemotherapy and was
found to have high-volume teratoma and 2% viable disease
at RPLND. He then had a marker and retroperitoneal relapse
and underwent high-dose chemotherapy with stem cell
transplant and repeat RPLND, and subsequently died of
disseminated disease. This is also unfortunately consistent
with the poorer outcomes associated with RPLND in the
salvage setting
[17].
To the best of our knowledge, a midline EP approach for
RPLND has not been described by other groups. It has been
described in the literature on general surgery, particularly to
gain access to the aorta for open abdominal aortic aneurysm
repair
[18–20] ,with series demonstrating a trend towards
lower rates of ileus and shorter LOS, including one random-
ized controlled trial comparing midline TP versus EP
approaches, although the results in the latter did not reach
statistical significance
[21] .Almost 20 yr ago, Christmas et al
[22]reported 80 cases of PC RPLND performed extraper-
itoneally via a thoracoabdominal (
n
= 71) or flank (
n
= 9)
incisionat the level of the12th rib; they achieved full access to
the retroperitoneum and reported no cases of postoperative
ileus. In our experience these incisions are more morbid than
a midline one and are unnecessary for adequate exposure.
Minimally invasive techniques also have the potential to
minimize the surgical morbidity of RPLND and warrant
discussion here. Steiner et al
[23]described the largest
laparoscopic series to date of 100 PC cases of RPLND for stage
II disease. All but one case was completed laparoscopically
Table 3 – Ejaculatory outcomes for retroperitoneal lymph node
dissection (RPLND)
Primary RPLND PC RPLND
Total cases (
n
)
27
41
Nerve-sparing procedure,
n
(%)
27 (100)
32 (78)
Antegrade ejaculation,
n
(%)
22/24 (91.6)
30/31 (96.8)
Retrograde ejaculation,
n
(%)
2/24 (8.3)
1/31 (3.2)
Missing data (
n
)
3
1
Table 2 – Clavien-Dindo complications at 90 d
Clavien grade
Frequency,
n
(%)
Total (in 11 patients)
12 (17.6)
Grade I
6 (55)
Grade II
5 (45)
Grade IIIa
0
Grade IIIb
1 (1.5)
Grade IV
0
Grade V
0
E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 8 1 4 – 8 2 0
818