Short- and Long-Term Outcomes of a Transdiaphragmatic Approach for Simultaneous Resection of Colorectal Liver and Lung Metastases

Abstract

Background.

Long-term outcomes for simultaneous resection of synchronous colorectal liver and lung metastases are unknown. To address this gap, we compared outcomes and costs of three strategies for such resection.

Methods.

Patients who underwent resection of synchronous colorectal liver and lung metastases during 2000–2018 were grouped by surgical strategy: simultaneous resection via a transdiaphragmatic approach (transdiaphragmatic) or separate abdominal and thoracic incisions (transthoracic) and nonsimultaneous staged resection (staged). Operative and postoperative outcomes, survival, cumulative lung recurrence, and surgical costs were evaluated.

Results.

The study included 63 patients, 29 with transdiaphragmatic, 14 with transthoracic, and 20 with staged resection. The groups had similar demographic and clinicopathologic characteristics. Lung resection-associated blood loss for the transdiaphragmatic group was similar to that for the transthoracic group (P = .165) but lower than that for the staged group (P = .006). Hospital stay was shorter for the simultaneous groups than for the staged group (P = .007). Median surgical costs were significantly higher in the staged group ($130,733, interquartile range [IQR] $91,109–$173,573) than in the transdiaphragmatic ($70,620, IQR $58,376–$86,203, P < .001) or transthoracic group ($62,991, IQR $57,405–$98,862, P < .001) but did not differ between the transdiaphragmatic and transthoracic groups (P = .786). Rates of postoperative complications, recurrence-free survival, overall survival, and cumulative lung recurrence were similar among the groups.

Conclusions.

Simultaneous resection of synchronous colorectal liver and lung metastases via a transdiaphragmatic approach is associated with lower blood loss, lower costs, and similar survival compared to staged resection.

INTRODUCTION

Colorectal cancer metastasizes most frequently to the liver (involved in 33% to 60% of patients with metastases) and lung (22%) [1,2]. A subset of patients develop synchronous or metachronous liver and lung metastases, and for these patients, resection of both liver and lung metastases has been shown to confer a survival benefit compared to nonresection [3,4]. Therefore, over the past two decades, resection of synchronous and metachronous colorectal liver and lung metastases has been advocated [5,6,7].

For synchronous colorectal liver and lung metastases, either simultaneous resection or staged resection (i.e., of liver metastases and then lung metastases or vice versa) can be performed. Although simultaneous resection traditionally requires both abdominal and thoracic incisions which present logistical, technical, and physiologic challenges, simultaneous resection may be feasible in selected patients. Nevertheless, some surgeons prefer staged resection to allow the patient to recover from the first operation before the second operation is performed [8,9,10].

In an attempt to reduce the invasiveness of synchronous abdominal and thoracic surgery, our group previously described a simple novel technique that creates an opening in the diaphragm on the side of the pulmonary metastasis(es) through which the lung metastasectomy can be performed, thus avoiding a thoracic incision. We have shown that simultaneous resection of colorectal liver and lung metastases via this transdiaphragmatic approach is safe and associated with lower blood loss and shorter length of hospital stay than staged resection [11].

However, whereas long-term outcomes for staged resection of synchronous and metachronous liver and lung metastases have previously been described [8,12], long-term outcomes for simultaneous resection of synchronous liver and lung metastases are currently unknown. To address this gap, we compared the long-term patient outcomes and cost-effectiveness of three strategies for resection of synchronous colorectal liver and lung metastases: simultaneous resection via a transdiaphragmatic approach, simultaneous resection via separate abdominal and thoracic incisions, and staged resection.

PATIENTS AND METHODS

The primary goal of our retrospective study was to compare short- and long-term outcomes among patients treated with the three different surgical strategies; the secondary goal was to evaluate the costs of each surgical strategy. The study was approved by the Institutional Review Board at MD Anderson Cancer Center.

Study population and general treatment approach

Patients who underwent combined resection of synchronous liver and lung metastases from colorectal cancer from January 2000 through December 2018 were identified from a prospectively maintained database. Synchronous liver and lung metastases were defined as liver metastases diagnosed within 3 months of the day of lung metastases diagnosis or vice versa. We excluded patients with R2 liver or lung resection, patients who required conversion of planned surgical strategy to other surgical approach, and patients who underwent segmentectomy or lobectomy of lung metastases because these types of lung resections are currently contraindications for the transdiaphragmatic approach in our practice. Patients were categorized into three groups according to the surgical strategy used to treat synchronous liver and lung metastases: 1) simultaneous resection using a transdiaphragmatic approach (transdiaphragmatic), 2) simultaneous resection using separate abdominal and thoracic incisions (transthoracic), or 3) nonsimultaneous staged resection (staged).

Liver resections were performed by hepatobiliary surgeons, and lung resections were performed by thoracic surgeons. The decision regarding adjuvant chemotherapy was made at the discretion of the treating oncologist and was based on number of cycles and tolerance of preoperative chemotherapy, presence of positive surgical margin, and pathological response to preoperative chemotherapy.

Analysis of complications and surgical costs

Surgical complications were defined as any deviation from the normal postoperative course within 90 days after hepatic and lung resection and graded according to the Clavien-Dindo classification [13].

Analysis of cost-effectiveness was limited to the day of the operation, and cost was calculated from all service charges, including charges for surgical equipment, use of the operating room, and anesthesia. For staged resection, the cost of surgery was calculated as the sum of the costs of the two separate operations.

Selection of approach and indications for a transdiaphragmatic approach

Selection of surgical strategy (simultaneous resection vs. staged resection) in patients with synchronous liver and lung metastases was discussed in a multidisciplinary meeting. For patients for whom simultaneous resection was recommended, the transdiaphragmatic approach was selected when lung metastases were located in the peripheral area of the lung, in the absence of hilar or mediastinal adenopathy, and could be removed with wedge resections. Otherwise, use of separate abdominal and thoracic incisions was selected. Large number, large size, and bilateral distribution of lung metastases were not generally considered contraindications for the transdiaphragmatic approach. Patients who had a history of previous thoracic surgery were excluded from the transdiaphragmatic approach.

Transdiaphragmatic approach for simultaneous resection of synchronous liver and lung metastases

General anesthesia was induced with placement of a double-lumen endotracheal tube to ventilate one lung separately from the other. Patients were placed in the supine position, and an inverse L or midline abdominal incision was generally performed for liver resections. Following resection of the liver metastasis(es), the lung bearing metastatic disease was fully deflated. The corresponding diaphragm was carefully cut open over the central tendon, with the surgeon making sure not to damage the phrenic nerve. The inferior pulmonary ligament was cut to mobilize the lung using electrocautery. Meticulous finger palpation was conducted to find lung lesions in each lung lobe. Lung resections were performed using endoscopic 45-mm or 60-mm staplers to ensure a sufficient surgical margin while the surgeon directly watched the surgical field through the diaphragm. A thoracoscope was also used to provide better surgical visualization if necessary. After the surgeon confirmed that air leakage and/or bleeding was not evident, usually a 24-Fr thoracic tube was placed through the intercostal space. Otherwise, a pulmonary drainage tube was placed through the diaphragm. The diaphragm was closed using a running 3–0 polypropylene suture.

Statistical analysis

Statistical analysis was performed with SPSS (version 20.0, SPSS Inc.) and EZR (version 1.37, Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R software (R Foundation for Statistical Computing, Vienna, Austria) [14]. Categorical variables were expressed in numbers and percentages and were compared among groups using Fisher’s exact test or the chi-square test, as appropriate. Continuous variables were expressed as median values with the interquartile range (IQR) and were compared using Kruskal–Wallis test. Holm method was used to adjust the P values in multiple testing [15]. Recurrence-free survival (RFS) and overall survival (OS) were calculated from the date of simultaneous resection of liver metastases and lung metastases or from the date of the first resection (either liver or lung) in patients who underwent staged resection. OS and RFS curves were constructed using the Kaplan-Meier method and compared using the log-rank test. P ≤ .05 was considered to indicate statistical significance.

RESULTS

Demographic and clinicopathologic characteristics

Of the 69 patients who underwent resection of synchronous liver and lung metastases in our study period, 63 patients met the inclusion criteria and were included in the analysis. Among these, 29 patients were treated with the transdiaphragmatic approach, 14 with the transthoracic approach, and 20 with the staged approach. No significant differences were found in demographic and clinicopathologic characteristics among the three groups (Table 1). Nearly all patients received preoperative chemotherapy. In this cohort of 63 patients, 10 (16%) had undergone prior liver resection for isolated colorectal liver metastases, including seven (24%) in the transdiaphragmatic group, two (14%) in the transthoracic group, and one (5%) in the staged group. Extrahepatic abdominal disease (hepatoduodenal lymph node or single peritoneal implant) was resected in two patients (7%) in the transdiaphragmatic group. The frequencies of somatic mutations were similar among the three groups (Table 1).

Table 1.

Demographic and clinicopathologic characteristics of patients who underwent synchronous liver and lung resection for colorectal cancer metastases

	Simultaneous resection
	Trans-diaphragmatic		Transtlioracic		Staged		P value
Number of patients	29		14		20
Age, yr	53	(41–56)	57	(47–63)	59	(43–66)	0.283
Sex, male	15	(52)	9	(64)	10	(50)	0.549
BMI, kg/m2	28	(24–32)	26	(24–28)	26	(22–28)	0.414
ASA score ≥ 3	23	(79)	12	(86)	15	(75)	0.854
Location, colon:rectum	20:9	(69:31)	9:5	(64:36)	8:12	(40:60)	0.103
T category ≥ 3*	26	(90)	12	(86)	15	(75)	0.395
Primary lymph node metastasis*	20	(69)	10	(71)	15	(75)	1.000
Synchronous liver metastasis†	19	(66)	8	(57)	13	(65)	0.889
Preoperative chemotherapy	27	(93)	14	(100)	18	(90)	0.809
Number of cycles	5	(4–10)	6	(4–8)	6	(4–8)	0.882
CEA (ng/mL)‡	4.6	(2.5–17.3)	3.7	(2–10)	4	(1.6–13.7)	0.681
Extra-hepatic disease	2	(7)	0	(0)	0	(0)	0.703
Somatic mutation§
RAS	11	(69)	7	(88)	8	(66)	0.364
BRAF	0	(0)	0	(0)	0	(0)	-
TP53	10	(63)	4	(50)	9	(75)	0.548
APC	9	(56)	6	(75)	9	(75)	0.594

Categorical variables are expressed in numbers and percentages. Continuous data are expressed as median values and the interquartile range.

Abbreviations: BMI. Body Mass Index: ASA. American Society of Anesthesiologists; CEA, carcino-embryomc antigen.

^*Data missing in 1 patient of the trans-diaphragmatic group, 1 patient of the trans-thoracic group, and 1 patient of the staged resection group.

^†Liver metastasis found within 12 months from the resection of primary colon cancer.

^‡Based on the data before simultaneous resection or first resection of the staged resection group.

^§Data missing on mutational status in 13 patients of the trans-diaphragmatic group, 6 patients of the trans-thoracic group, and 8 patients of the staged resection group.

Operative, histopathological, and postoperative outcomes

Table 2 shows operative and histopathological outcomes. The combined operative time for liver and lung resections did not differ significantly among the three groups. Estimated blood loss related to the lung resection was significantly greater in the staged group than in the transdiaphragmatic or transthoracic group, but estimated blood loss related to the liver surgery did not differ among the groups. Minimally invasive methods (laparoscopy, thoracoscopy, or both) were used for three patients (21%) in the transthoracic group and eight patients (40%) in the staged group. Conversely, the transdiaphragmatic approach always requires an abdominal incision. A total of 44, 25, and 46 wedge resections of the lung were performed in the transdiaphragmatic, transthoracic, and staged groups, respectively. In particular, the rate of upper lobes resections was similar between the three surgical strategies, namely 34%, 28%, and 35% in transdiaphragmatic, transthoracic, and staged approaches, respectively. The sites of wedge resection are detailed in Figure 1. A chest tube was placed in all patients except two patients (7%) in the transdiaphragmatic group. Only one patient required a conversion from a transdiaphragmatic to a transthoracic approach, because of technical difficulty in performing a metastasectomy of the lingula.

Table 2.

Operative and histopathological outcomes for patients who underwent synchronous liver and lung resection for colorectal cancer metastases.

	Simultaneous resection
	Trans-diaphragmatic		Transtlioracic		Staged		P value
Number of patients	29		14		20
Operative time, min *	285	(222–327)	283	(250–323)	361	(323–438)	0.810
Blood loss, ml
Liver surgery	300	(150–375)	163	(100–300)	250	(125–325)	0.383
Lung surgery	0	(0–10)	15	(0–45)	75	(38–125)	0.005±
Variables of liver resection
Number of liver metastases	2	(1–3)	1	(1–2)	3	(2–3)	0050
Largest diameter, mm	24	(15–35)	16	(14–23)	20	(15–31)	0.407
Positive surgical margins, n (%)	7	(24)	1	(7)	7	(35)	0.265
PVE, n (%)	5	(17)	2	(14)	4	(20)	1000
Two-stage hepatectomy, n (%)	5†	(17)	0	(0)	2	(10)	0.254
Concomitant RFA, n (%)	3	(10)	0	(0)	2	(10)	0.596
Major resections, n (%)	12	(41)	2	(14)	12	(60)	0.030‡
Variables of lung resection
Number of lung metastases	1	(1–1)	1	(1–1)	1	(1–2)	0.410
Largest diameter, mm	8	(6–15)	10	(7–18)	12	(10–19)	0.130
Surgical margin
Positive, n (%)	0	(0)	0	(0)	2	(10)	0.133
Margin size, mm	4	(2–10)	6	(3–10)	6	(3–9)	0.818
Number of wedge resections	1	(1–2)	1	(1–2)	1	(1–3)	0.535
Bilateral lung resection, n (%)	1	(3)	0	(0)	4	(20)	0081
Distance from pleura, mm¶	6	(0–18)	16	(2–24)	8	(0–13)	0.391
Newly detected lesions, n (%)§	5	(17)	1	(7)	4	(20)	0616
Malignant:benign, n (%)	4:2	(67:33)	0/1	(0:100)	7/2	(78:22)	0.322

Categorical variables are expressed in numbers and percentages. Continuous data are expressed as median values and the interquartile range.

Abbreviations: PVE. Portal Vein Embolization: RFA. Radio-Frequency Ablation.

^*Sum of liver resection and lung resection.

^±Multiple tests using Holm s method. Trans-diaphragmatic group vs staged resection group. P = 0.006

^†Lung resection was performed in the second stage of planned 2-stage hepatectomy.

^‡Multiple tests using Holm’s method. Trans-thoracic group vs staged resection group. P = 0.039

^¶Distance from the pleura refers to the depth of the most centrally located metastasis

^§Patients with other lesions detected intraoperatively by palpation and had not been visualized in preoperative modalities.

Figure 1.

Sites of wedge resections of the lung in patients with synchronous colorectal liver and lung metastases by surgical approach. A total of 44 wedge resections were performed in the transdiaphragmatic (TD) resection group, 25 were performed in the transthoracic (TT) resection group, and 46 were performed in the staged resection group.

Table 3 shows postoperative outcomes. Rates of overall complications and severe complications (Clavien-Dindo grade ≥ 3) were not significantly different among the three groups. Complications related to lung resection were experienced by two patients (7%) in the transdiaphragmatic group (pneumonia and pneumothorax), one patient (7%) in the transthoracic group (acute postthoracotomy pain), and three patients (15%) in the staged group (acute postthoracotomy pain in one patient and air leaks in two patients). Median length of hospital stay was greater in the staged group (8 days for the two hospital admissions combined) than in the transdiaphragmatic group (6 days) or transthoracic group (6 days) (P = .007). Adjuvant chemotherapy was delivered to 15 patients (52%) in the transdiaphragmatic group, nine (64%) in the transthoracic group, and 11 (55%) in the staged group.

Table 3.

Postoperative outcomes

	Simultaneous resection
	Trans-diaphragmatic		Transtlioracic		Staged		P value
Number of patients	29		14		20
Overall complications, n (%)	9	(31)	4	(29)	9*	(45)	1.000
Clavien-Dindo grade ≥ 3, n (%)	3	(10)	2	(14)	2*	(10)	1.000
Complications related to lung surgery, n (%)	2	(7)	1	(7)	3	(15)	0.772
Clavien-Dmdo grade ≥ 3, n (%)	2	(7)	0	(0)	0	(0)	0.703
Length of hospital stay, days	6	(5–8)	6	(4–7)	8*	(7–9)	0.007±
90-day mortality, n (%)	0	(0)	0	(0)	0†	(0)	-
Adjuvant chemotherapy, n (%)	15	(52)	9	(64)	11	(55)	0.721

Categorical variables are expressed in numbers and percentages. Continuous data are expressed as median values and the interquartile range.

^*Sum of two hospitalizations (liver surgery plus lung surgery).

^±Multiple tests using Holm s method. Trans-thoracic group vs staged resection group, P = 0.006

^†Assessed after the second surgery (either liver resection or lung resection).

Surgical costs

Median surgical costs were significantly higher in the staged group ($130,733, IQR $91,109–$173,573) than in the transdiaphragmatic group ($70,620, IQR $58,376–%86,203, P < .001) or transthoracic group ($62,991, IQR $57,405–$98,862, P < .001). Surgical costs did not differ significantly between the transdiaphragmatic and transthoracic groups (P = .786).

Recurrence-free survival, overall survival, and cumulative lung recurrence rate

The median follow-up period was 32 months (IQR, 17–53). Rates of RFS, OS, and cumulative lung relapse did not differ significantly among the three groups (Figures 2 and 3).

Figure 2.

Recurrence-free survival (A) and overall survival (B) in patients with synchronous colorectal liver and lung metastases by surgical approach.

Figure 3.

Lung relapse in patients with synchronous colorectal liver and lung metastases by surgical approach.

DISCUSSION

Our study shows that in patients with synchronous liver and lung metastases from colorectal cancer, the rates of survival and cumulative lung recurrence after simultaneous resection via a transdiaphragmatic approach were similar to those after simultaneous resection via a transthoracic approach or staged resection. Blood loss was lower in the transdiaphragmatic group than in the staged resection group, and postoperative complication rates did not differ among the three groups. Taken together, our results suggest that the transdiaphragmatic approach is a safe and oncologically acceptable option for resection of synchronous liver and lung metastases from colorectal cancer.

The optimal surgical approach for synchronous liver and lung metastases has not been well established. The staged resection approach is frequently used in clinical practice. Under this strategy, patients who present with both liver and lung metastases usually first undergo surgery of the organ with the lesion deemed more challenging to resect with curative intent. If a complete resection is achieved, patients then undergo resection of the other organ after they have recovered from the first operation. Simultaneous resections via a combined abdominal and thoracic approach with two separate incisions are occasionally performed in selected patients who are in good health and who require only limited liver and lung resections [8,9,10]. Our group previously developed a transdiaphragmatic approach for resecting synchronous liver and lung metastases at the same time without the need for a thoracic incision and demonstrated that this novel approach reduced blood loss and shortened length of stay compared to conventional staged surgery [11]. However, the long-term outcomes in patients undergoing the transdiaphragmatic approach were previously unknown.

The findings of our current study showed that OS, RFS, and cumulative lung recurrence rates were similar among the transdiaphragmatic, transthoracic, and staged resection groups. To the best of our knowledge, no previous study has reported the long-term outcomes after simultaneous resection of synchronous liver and lung metastases; instead, some previous studies have investigated long-term outcomes in patients treated with hepatectomy and eventually pulmonary resection or pulmonary resection and eventually hepatectomy. Results from these studies have shown wide variation in the 5-year OS rate, ranging from 11% to 74% [8,9,10,16, 17,18,19,20,21,22]. Moreover, these results are difficult to interpret because of the heterogeneity in patient selection, inclusion of both patients with synchronous and patients with metachronous liver and lung metastases, and different definitions established to assess survival outcomes. Indeed, while some studies calculated survival from the day of primary tumor resection, others calculated survival from the day of first metastasectomy or from the day of lung metastasectomy. Recently, Andres et al. [12] reported the largest series of long-term outcomes after staged resection only of synchronous liver and lung metastases in the era of new chemotherapy agents and showed that the 5-year OS rate was 45%. In our analysis, the 5-year OS rate was 48% after the transdiaphragmatic approach, 26% after the transthoracic approach, and 46% after the staged approach (P > .05).

In line with our previous report, our current study showed that short-term outcomes in the transdiaphragmatic group were comparable to or better than those in the staged resection group. We want to point out that lung metastases located in the upper lobes were not more difficult to resect trough the diaphragm incision compared to the other conventional approaches. In fact, despite the transdiaphragmatic group had the highest percentage of upper lobes resections, technical complexity and safety were comparable to those of conventional lung metastasectomy, as demonstrated by the equivalent operative time, lung-related morbidity and lung-relapse rate. The full deflation of the lung allows the upper lobes to be mobilized towards the diaphragm. Furthermore, the use of a thoracoscope for a better visualization of the surgical field and the proper inclination of the stapler may help to perform the lung wedge resection more easily and safely. Advantages of the transdiaphragmatic approach include avoiding two separate anesthesia episodes and two separate hospital admissions and eliminating the need for a thoracic incision to resect the lung metastasis(es). Another advantage of the transdiaphragmatic approach is that surgeons can palpate tiny lung metastases and localize them better than the video-assisted transthoracic approach which lacks palpation. A major hepatectomy is not a contraindication to simultaneous resection using the transdiaphragmatic approach. We have performed major hepatectomy with low rates of perihepatic fluid collection or bile leak using a 2-step air leak test after resection [23,24]. In the current series, the major hepatectomy rate in patients undergoing the transdiaphragmatic approach (41%) was significantly higher compared to patients undergoing the transthoracic approach (14%) and comparable to patients undergoing the stage resection approach (60%). Finally, a negative impact of a thoracotomy on postoperative physical quality of life has been reported in patients who underwent right-sided liver resection using a combined thoracoabdominal incision [25]. Thus, the transdiaphragmatic approach may reduce postoperative pain without approach-specific complications, although we did not set out to compare this outcome between the two simultaneous resection strategies. Still, in our series, we observed no cases of postoperative thoracic pain or transdiaphragmatic hernia in patients treated with the transdiaphragmatic approach.

Additionally, the transdiaphragmatic approach was more cost-effective than staged resection and cost about the same as the transthoracic approach even though we did not include costs related to the hospital length of stay in our calculations. The day-of-surgery cost-saving value of the transdiaphragmatic approach may be related to a reduction not only in the cost associated with a second hospitalization but also in the cost of surgical devices used only for the transthoracic part of the operation. Importantly, the transdiaphragmatic approach may also allow patients to start intended postoperative adjuvant chemotherapy sooner than the other approaches do by enhancing rapid recovery from surgery, although this was not studied in our current analysis.

A limitation of our study is that it was retrospective and covered a long study period during which the management of colorectal metastases may have changed. Another limitation is the selection bias that went into determining which approach to use for the patients in the study. However, despite these limitations, our data would support simultaneous resection of synchronous liver and peripheral lung metastases with a transdiaphragmatic approach, rather than the traditional approach (which clinicians may still be inclined to use) of using a separate transthoracic incision or performing staged resection. As our experience with this approach grows, we may be able to address anatomically more challenging lung metastases through the diaphragm incision.

CONCLUSION

Our study shows that simultaneous resection of synchronous liver and lung metastases using our transdiaphragmatic approach is associated with long-term outcomes similar to those after simultaneous resection using separate abdominal and thoracic incisions or staged resection. The transdiaphragmatic approach is associated with better short-term outcomes and lower costs than staged resection. These results suggest that the transdiaphragmatic approach is a safe, oncologically acceptable, and cost-effective option for the treatment of synchronous liver and lung metastases.