Diaphragmatic sulci and portal fissures

Abstract

Diaphragmatic sulci in the superior surface of the liver were found in 40% of cases at autopsy. All sulci were located in the right lobe and in 47% they were multiple. In order to evaluate possible predisposing factors for these accessory sulci, their topography and characteristics were observed in unfixed livers; moreover, intravenous injections of radio-opaque resins were performed in the portal and hepatic veins (HVs). After formalin fixation, the livers underwent CT and MR scans and a three-dimensional (3D) elaboration of the images was performed. Radiological examination revealed a correspondence between the topography of the sulci and the course of the right and middle HVs and their main tributaries in 67%. The corrosion casts showed the location of the sulci at the level of the boundaries between the ramifications of the terminal branches of the portal triad, where the HVs are located, in 73%. These findings suggest that, rather than the action of ‘special’ or hypertrophied muscle bundles, the pressure exerted by the diaphragm as a whole may be responsible for the production of sulci at the level of weak zones, represented by the portal fissures, where the watershed superficial hepatic parenchyma, owing to the absence of all but the smallest vascular branches, exhibits a lower resistance to external pressure.

Introduction

Accessory sulci in the superior surface of the liver have been attributed to anatomic variability and to the effects of pressure of the ribs and the diaphragm muscle (Zahn, 1882; Thompson, 1899). A high frequency of these sulci has been reported in autopsy studies (Ono et al. 2000), raising questions as to their vital origin with respect to post-mortem artefacts (Newell & Morgan-Jones, 1993). Radiological studies confirm the relative frequency of the diaphragmatic sulci also in vivo (Auh et al. 1984; Amstrong et al. 1990), although as pure morphological findings.

The possible clinical relevance of the diaphragmatic sulci as surface markings has never been suggested, since the functional anatomy (Bismuth et al. 1997) of the liver is based on the intrahepatic distribution of the vascular and biliary branches (Cantlie, 1898; McIndoe & Counseller, 1927; Hjortziö, 1951; Couinaud, 1957; Goldsmith & Woodburne, 1957; Tung, 1939; Ger, 1988), and the liver has been subdivided into sectors and segments by the so-called ‘avascular fissures’. This term (Hjortziö, 1951) refers to the corrosion casts where the fissures are due to the absence of all but the smallest branches of the portal triad, usually too fragile to preserve (Williams et al. 1995). They correspond to the terminal ramifications of the segmental branches and indicate the boundaries between adjacent hepatic segments. Moreover, Couinaud (1957) subdivided the liver into two lateral and two paramedian sectors, based on three main ‘portal fissures’ where the three hepatic veins (HVs) are located.

The purpose of this paper was to evaluate possible predisposing factors for the diaphragmatic sulci, through a radiological study of autopsied livers after injection of a radio-opaque resin in the venous systems and subsequent comparison of radiological findings with the vascular corrosion casts.

Materials and methods

During the anatomical dissecting courses for medical students, performed between 1999 and 2001 at Padova, Italy, 48 livers (25 male and 23 female cadavers; mean age: 52 years old; range: 30–75 years old) were sampled together with the diaphragm muscle (preserving the falciform, the coronary and the triangular ligaments) and the retroperitoneal structures, including the abdominal aorta and the inferior vena cava (IVC) (Macchi et al. 2001). Diaphragmatic sulci were found in 19 subjects. In all cases the anamnesis was negative for hepatic and respiratory diseases; however, in 12 cases a pulmonary emphysema was diagnosed at autopsy. In order to obtain vascular corrosion casts, simultaneous injections of an acrylic, radio-opaque resin (Beracryl; Troller, Fulenbach, Switzerland) in the portal and hepatic venous systems were performed. For each venous system 150 mL of resin were used. To avoid compression artefacts, the injections were carried out in the livers suspended in water. After a 1-week fixation period in 10% formalin solution, the livers underwent CT axial scans (General Electric PRO/SPEED, 120 kV, 130 mA, FOV − 30 cm, thickness 3 mm, rotation time 3, pitch 1, reconstruction interval 1 mm) and MR scans (General Electric Signa 0.5T; sequences: GE T1, SE DP and T2; 3D SP6R and FMPIR 90 STIR). An elaboration of the images was performed on a workstation SUN using Advantage Windows software to obtain 3D surface imaging, multiplanar reconstruction (MPR) and maximum intensity projection (MIP). The relationship between the diaphragmatic sulci and the portal and hepatic veins was analysed. On the radiological images, the segmental anatomy of the liver was delineated following the guidelines commonly used in radiological practice: the liver is subdivided into four sectors by the portal fissures, three vertical planes traced from the IVC through the right, middle and left HV, respectively (Couinaud, 1957; Bismuth, 1982). The sectors are further subdivided into segments by the transverse hepatic fissure, customarily drawn at the level of the right and left branches of the portal vein (Heiken, 1998). After digestion of the liver parenchyma with potassium hydroxide, on the corrosion casts the course of the portal and hepatic veins and that of diaphragmatic sulci was analysed. Comparison of the radiological images with the corrosion casts was performed to elucidate the relationship between the diaphragmatic sulci and the portal fissures with reference to the course of the HVs.

Results

Autopsy findings

In the survey of 48 cadavers, diaphragmatic sulci were found in the right lobe of 19 livers (11 females and eight males, mean age: 57 years) (40%). In 10 livers (53%) a single sulcus was present while in the other nine the sulci were multiple. A total of 31 sulci were examined. The sex, age, macroscopic appearance of the sulci with their measures and presence of emphysema at autopsy are reported in Table 1. The sulci had mean length of 5.7 cm (range 1.9–11.8 cm), mean width of 0.6 cm (range 0–2.8 cm) and mean depth of 1 cm (range 0.2–2.8 cm). All sulci were located in the superior surface of the right lobe, going from the right aspect of the IVC towards the right margin of the liver in 58% and from the left aspect of the IVC towards the anterior margin in 42%.

Table 1. Sex, age, macroscopic appearance of the diaphragmatic sulci, with measurements (cm) and presence or absence of emphysema at autospsy. N. sulci: number of sulci; RHV: right hepatic vein; MHV: middle hepatic vein.

						Relationship with Hepatic Veins
Case	Sex/Age	N. Sulci	Lenght	Width	Depth	Radiology	Corrosion cast	Emphysema
1	M/42	1	11.8	2.5	22.5	RHV	RHV	−
2	M/38	1	8.8	0.2	0.8	RHV	RHV	−
3	F/51	1	8.1	0.5	2.2	RHV	RHV	+
4	F/75	2	10.3	0.6	2.8	RHV	RHV	+
			3.5	0.7	0.5	−	−
5	F/40	2	5.0	0.4	0.3	Tributary of RHV	Tributary of RHV	−
			11.3	0.8	1.3	RHV	RHV
6	F/53	1	10.0	0.6	1.0	RHV	RHV	+
7	F/56	2	4.7	0.6	0.5	RHV	RHV	+
			4.8	0.2	0.3	−	−
8	M/65	2	6.0	2.8	2.5	RHV	RHV	+
			5.0	1.2	2.0	Tributary of RHV	Tributary of RHV
9	M/52	1	4.2	0.8	0.9	RHV	RHV	+
10	F/74	2	7.5	1.1	1.1	RHV	RHV	+
			4.5	0.8	0.7	−	−
11	F/44	1	9.0	0	2.1	−	−	−
12	M/38	1	5.0	0.6	1.1	−	−	−
13	F/69	4	4.0	0.3	0.2	−	−	+
			3.0	0.2	0.3	Tributary of RHV	Tributary of RHV
			4.2	0.3	0.4	−	Tributary of RHV
			5.0	0.6	0.2	MHV	MHV
14	M/75	2	5.0	0.4	0.3	−	−	+
			3.0	0.3	0.3	−	Tributary of RHV
15	M/74	3	3.0	0.4	0.3	Tributary of MHV	Tributary of MHV	−
			4.5	0.6	0.5	Tributary of MHV	Tributary of MHV
			6.0	0.3	2.0	MHV	MHV
16	F/56	1	6.1	0.4	0.8	Tributary of RHV	Tributary of RHV	+
17	F/72	2	1.9	0.2	0.2	Tributary of RHV	Tributary of RHV	+
			2.0	0.3	0.2	Tributary of RHV	Tributary of RHV
18	M/46	1	4.2	1.0	1.8	Tributary of RHV	Tributary of RHV	−
19	F/64	1	4.0	0.3	0.6	−	−	+
Mean	Value		5.7	0.6	1.0

Radiological findings

On the CT images the injected veins were radio-opaque while in MR they showed a low signal on T1- and T2-weighted sequences. On radiological images the course of the hepatic veins and of the portal branches was located to identify the boundaries of the segments. On axial images the diaphragmatic sulci appeared as incisures on the antero-superior surface of the liver. The relationship between the sulci and HVs was evaluated on the axial CT and MR images with reference to a line traced from the IVC to each of the incisures on the margin of the liver. A correspondence between the topography of the sulci and the course of the right and middle HVs and their tributaries was found in 67%. In particular the sulci corresponded to the right hepatic vein (RHV) in 10 instances (32%) and to its tributaries in seven instances (23%). Furthermore, the sulci showed a correspondence with the middle hepatic vein (MHV) in two instances (6%) and to its tributaries in two instances (6%) (Fig. 1). No correspondence between the sulci and vessels with significant calibre was found in 33%.

Fig. 1.

Case 15: a 74-year-old male subject. (A) Frontal view of the liver, after injection with red colour of the hepatic veins, showing three diaphragmatic sulci (asterisks). (B,C) The axial MR scan shows a low signal of the veins in T1-weighted sequence; the diaphragmatic sulci appear as incisures on the anterior margin of the liver; the trunk of the middle hepatic vein (M) and two tributaries are located in correspondence of the deep projection of the sulci (asterisks). (D) The corrosion cast of the hepatic veins confirms the correspondence between one sulcus and the middle hepatic vein (MHV) and between the other two sulci and its two tributaries (asterisks). IVC, inferior vena cava; RHV, right hepatic vein; LHV, left hepatic vein.

Corrosion casts

The main trunk of the HVs and the segmental branches of the portal vein were identified. In six cases the HVs emptied independently in the IVC, while in 13 cases a common trunk for the MHV and left hepatic vein was recognizable. A correspondence between the topography of the sulci and the course of the right and middle hepatic vein and their tributaries was found in 73%. In particular, the sulci corresponded to the RHV in 10 instances (32%) (Fig. 2) and to its tributaries in nine instances (29%). Furthermore, the sulci showed a correspondence with the MHV in two instances (6%) and to its tributaries in two instances (6%) (Fig. 1). No correspondence between the sulci and vessels with significant calibre was found in 27%.

Fig. 2.

Case 8: a 63-year-old male subject. (A) Cranial view of the surface-rendering image of a liver with two diaphragmatic sulci. (B) The three-dimensional reconstruction of the portal and hepatic venous systems shows that one sulcus (arrowhead) corresponds to the right hepatic vein and the other one (asterisk) to one of its tributaries. (C) Lateral view of the corrosion cast (hepatic veins were injected with blue stain and the portal vein with red) confirms the correspondence between one sulcus (curved line) and the right hepatic vein. (D) Cranial view of the corrosion cast shows the relationship between one sulcus (arrowhead) and the right hepatic vein (RHV) and between the other sulcus (asterisk) and one of the tributaries of the right hepatic vein. IVC, inferior vena cava.

Discussion

The presence of accessory sulci in the surface of the liver has been known for a long time (Zahn, 1882; Thompson, 1899) but their possible clinical relevance as surface markings has not previously been suggested. The sulci in the superior surface are considered acquired, owing to the costal pressure or to the relationship with the diaphragm muscle. The diaphragmatic sulci are localized in the superior surface of the right lobe and rarely on the left lobe (Zahn, 1882). They are narrow with variable depths ranging from 1 to 2 cm; sometimes they can be multiple (2–6 sulci). The corresponding serosa is usually normal (Buy, 1904). They are more frequent in females aged over 15 years (Yamagiwa, 1907).

According to some authors (Schafer & Symington, 1896; De Burlet, 1910) diaphragmatic sulci are derived from the uneven growth of the liver parenchyma caused by the variable resistance opposed by the different bundles of the diaphragm muscle. This hypothesis is in accordance with the morphogenesis of the embryonic liver that is influenced by the pressure exerted by the adjacent viscera and from their migration (descent of the heart, shortening of the inferior vena cava) (Ruge, 1902; Baxter, 1953; O'Rahilly & Muller, 1992). Moreover, the denomination of these accessory sulci derives from the description of folds of the diaphragm at their level (Zahn, 1882; Schafer & Symington, 1896) ascribed by some authors (Cosgrove, 1988; Sherlock & Dooley, 1993) to respiratory diseases (‘cough furrows’). But the presence of folds of the diaphragm in some cases does not prove causation (Newell & Morgan-Jones, 1993).

Given the frequency of the diaphragmatic sulci in our survey (40%), we investigated possible predisposing factors that might represent weak zones of the hepatic parenchyma particularly susceptible to the mechanical effect of the diaphragm. In radiological practice, the course of the HVs is used to identify the boundaries between the hepatic sectors. Therefore, the HVs constitute indirect landmarks for delineation of the portal fissures (Heiken, 1998). In our study, the radiological findings showed a correspondence between the topography of the sulci and the course of the right and middle hepatic vein and their tributaries in 67%. The corrosion casts showed the location of the sulci at the level of the boundaries between the ramifications of the terminal branches of the portal triad, where the HVs are located, in 73%. In particular, the sulci corresponded to the RHV in 32%, to its tributaries in 29%, and to the MHV in 6% and its tributaries in 6%. Thus our results indicate a relationship between the diaphragmatic sulci and the superficial extension of the portal fissures. The fissures are localized at the boundary between the territories of distribution of the segmental branches of the portal vein (Couinaud, 1957). At their level the superficial liver parenchyma is perfused by the terminal ramifications and since there are no anastomoses between the blood vessels of adjacent segments (Gupta et al. 1977), the only vessels with significant calibre are represented by the tributaries of the HVs, located in the depth of the fissures.

Moreover, in the corrosion casts at the level of the paramedian sectors, the ramifications of the segmental portal branches are disposed on parallel sagittal planes (Couinaud, 1957). The result is that the angio-architecture of segments IV and VIII consists of a succession of 3–4 parallel sagittal portal venous territories with interposed secondary portal fissures, where the roots of the HVs are located. Thus, the intervals between two adjacent portal venous territories represent potential ‘weak’ zones in the superficial hepatic parenchyma, and the frequency of sulci at the level of the paramedian sectors (42% of sulci in all livers studied), also documented by Ono et al. (2000), could be explained by the spatial course of the portal ramifications.

Although the role of the HVs as indirect landmarks for the correct delineation of the portal venous sectors and segments is under discussion owing to the fact that the territorial boundaries are not flat planes but are undulating (Downey, 1994), the major discrepancy between radiological and anatomical findings regards the central zones of the liver with respect to the marginal portions (Fasel et al. 1998) where the diaphragmatic sulci are located. However, this fact could explain the cases in our survey with lack of correspondence between the sulci and the HVs.

The difference between the results obtained with the two different methods can be ascribed to the limits of the spatial resolution of the radiological techniques with consequent lack of visualization of some minor tributaries of the HVs. However, both techniques were useful in the evaluation of the relationship between the diaphragmatic sulci and the HVs since radiology allowed a better visualization of the course of the main trunk of the HVs which are deeply located, while the corrosion casts also showed the minor tributaries of the HVs.

Therefore, the variability of course and the frequent multiplicity of the diaphragmatic sulci suggest that, rather than the action of ‘special’ or hypertrophied muscle bundles, the pressure exerted by the diaphragm as a whole can be responsible for the production of sulci at the level of predisposed weak zones of the superficial hepatic parenchyma. These zones correspond to the boundaries between the terminal ramifications of adjacent segmental portal branches and to the interval between adjacent secondary sagittal portal territories inside the segments. In the young subjects, the pressure exerted by the diaphragm might represent an obstacle to the homogeneous growth of the liver parenchyma at the level of these specific watershed districts. In the adults, pathologies that induce a chronic increase of the activity of the diaphragm cause an increase of the pressure that the muscle exerts on the adjacent liver which mainly acts at the level of preformed weak zones, corresponding to the portal fissures. At these levels the bundles of the diaphragm may undergo hypertrophy. This mechanism could also explain the inconsistent finding of hypertrophic bundles of the diaphragm muscle with respect to the high frequency of diaphragmatic sulci.

From a clinical point of view the diaphragmatic sulci can represent a useful landmark for the portal fissures and for the superficial projection of the deep course of the HVs and their tributaries, representing morphological evidence of the functional vascular anatomy of the liver.