Subdiaphragmatic phrenic nerve supply: A systematic review

Annals of Anatomy 254 (2024) 152269 Available online 29 April 2024 0940-9602/© 2024 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Subdiaphragmatic phrenic nerve supply: A systematic review María P´ erez-Montalb´ an a , Encarna García-Domínguez b , ´ Angel Oliva-Pascual-Vaca c , d , * a Universidad de Sevilla. Facultad de Enfermería, Fisioterapia y Podología, Departamento de Fisioterapia, Spain b UGC Aparato Digestivo, Hospital Universitario de Puerto Real, C´ adiz, Spain c Instituto de Biomedicina de Sevilla, IBiS, Departamento de Fisioterapia, Universidad de Sevilla, Spain d Escuela de Osteopatía de Madrid, Madrid, Spain ARTICLE INFO Keywords: Phrenic nerve Esophagus Stomach Liver Inferior vena cava Gallbladder Adrenal glands Neck pain Shoulder pain Hiccup ABSTRACT Objective: The aim of this systematic review is to study the subdiaphragmatic anatomy of the phrenic nerve. Materials and methods: A computerised systematic search of the Web of Science database was conducted. The key terms used were phrenic nerve, subdiaphragmat*, esophag*, liver, stomach, pancre*, duoden*, intestin*, bowel, gangli*, biliar*, Oddi, gallbladder, peritone*, spleen, splenic, hepat*, Glisson, falciform, coronary ligament, kidney, suprarenal, and adrenal. The ‘cited-by’ articles were also reviewed to ensure that all appropriate studies were included. Results: A total of one thousand three hundred and thirty articles were found, of which eighteen met the inclusion and exclusion criteria. The Quality Appraisal for Cadaveric Studies scale revealed substantial to excellent methodological quality of human studies, while a modified version of the Systematic Review Centre for Laboratory Animal Experimentation Risk of Bias Tool denoted poor methodological quality of animal studies. According to human studies, phrenic supply has been demonstrated for the gastro-esophageal junction, stomach, celiac ganglia, liver and its coronary ligament, inferior vena cava, gallbladder and adrenal glands, with half of the human samples studied presenting phrenic nerve connections with any subdiaphragmatic structure. Conclusions: This review provides the first systematic evidence of subdiaphragmatic phrenic nerve supply and connections. This is of interest to professionals who care for people suffering from neck and shoulder pain, as well as patients with peridiaphragmatic disorders or hiccups. However, there are controversies about the autonomic or sensory nature of this supply. 1. Introduction The classic pattern of the phrenic nerve (PN) is mainly made up of C4 fibers, with a lower proportion coming from C3 and C5 (Kessler et al., 2008; Ruiz and Latarjet, 2006; Andrade et al., 2021). It is made up mainly of motor fibers, but also contains sensory and autonomic fibers. In the neck, the nerve receives variable communicating fibers from the cervical sympathetic ganglia and their branches (Canella et al., 2010). The PN provides the sensitivity of the central portion of the diaphragm muscle and its only motor innervation, thus regulating its contraction (Hollinshead, 1985; Williams et al., 1995). In the thorax it also innervates the mediastinal and diaphragmatic pleura, the right atrium, the fibrous pericardium and the parietal layer of the serous pericardium (Canella et al., 2010; Hollinshead, 1985; Williams et al., 1995). The PN continues as it branches into the abdominal cavity. The right PN (RPN) crosses the diaphragm through the foramen for the vena cava, while the left PN (LPN) does so near the apex of the heart. Thus, most of the branches of the PN that supply the diaphragm are on its abdominal side (Moore et al., 2013; Drake et al., 2015), but the PN also innervates visceral and peritoneal structures in the abdominal cavity (Rouviere and Delmas, 2000). However, there still seem to be inconsistencies regarding its distribution. A better understanding of phrenic anatomy could contribute to the understanding of the different processes involving this nerve, such as neck-to-shoulder referred pain and hiccups. Therefore, the aim of this systematic review is to study the anatomy of the PN at the subdiaphragmatic level. * Correspondence to: Instituto de Biomedicina de Sevilla IBiS, Departamento de Fisioterapia, Facultad de Enfermería, Fisioterapia y Podología, Universidad de Sevilla, c/ Avicena s/n, Sevilla 41009, Espa˜ na. E-mail address: [email protected] (´ A. Oliva-Pascual-Vaca). Contents lists available at ScienceDirect Annals of Anatomy journal homepage: www.elsevier.com/locate/aanat https://doi.org/10.1016/j.aanat.2024.152269 Received 5 March 2024; Received in revised form 18 April 2024; Accepted 23 April 2024 Annals of Anatomy 254 (2024) 152269 2 2. Materials and methods The present systematic review has been performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). It was registered in the International Prospective Register of Systematic Reviews (PROSPERO), with registration number CRD42023405983. 2.1. Data source and search strategy Two authors independently conducted a systematic computerized search in the database Web of Science between January and September 2023. The key terms used in the search were phrenic nerve, subdiaphragmat*, esophag*, liver, stomach, pancre*, duoden*, intestin*, bowel, gangli*, biliar*, Oddi, gallbladder, peritone*, spleen, splenic, hepat*, Glisson, falciform, coronary ligament, kidney, suprarenal and adrenal. We also reviewed “cited-by” articles to ensure that all suitable studies were included. In case of a discrepancy between authors, a third author was in charge of reaching a consensus. 2.2. Study selection We included in this review any article published in scientific journals in English, Spanish, Italian, French or Portuguese, reporting original data about the phrenic supply of subdiaphragmatic tissues either in humans or animals. Articles were excluded if they do not study the subdiaphragmatic anatomy of the PN. Two researchers independently evaluated the studies in a blinded and standardized manner, considering the eligibility criteria previously established. In the event of disagreement, the issue was discussed together with a third member of the research team until a final consensus was reached. 2.3. Assessment of the methodological quality The QUality Appraisal for Cadaveric Studies (QUACS) scale was used to assess the methodological quality of the selected articles that involve humans. The QUACS scale has thirteen items which are scored as not stated (0) or present (1). Points are only awarded if there is no doubt. Those studies with 0–2 points are classified as having poor methodological quality, 3–5 as fair, 6–7 as moderate, 8–10 as substantial and higher than 11 as excellent (Wilke et al., 2015). A modified version of the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE) Risk of Bias Tool was used to assess the methodological quality of selected articles involving animals. The SYRCLE scale has ten items, which are scored as not stated (0) or present (1). These items are related to the assessment of risk of bias, which scores the internal and external validity of each study (Hooijmans et al., 2014). Two independent raters evaluated the studies with the scales. A consensus on the final score was reached, together with a third member of the research team whenever necessary. 2.4. Data extraction Once the studies were selected, two independent authors took the following information from each article: authors; year of publication; participants (humans and/or animals); sex and mean age; QUACS and/ or SYRCLE scale score; which PN is studied (right or left); and the subdiaphragmatic innervation of the PN. Data collected from the studies we included were pooled into tables. In case of a discrepancy between authors, a third author was in charge of reaching a consensus. 2.5. Anatomy textbooks To show what is generally accepted (and maybe taught) and in order to be used as a comparator, eight human anatomy textbooks were independently reviewed by two investigators, with another author solving discrepancies. A manual search was performed in those textbooks, searching for any data on subdiaphragmatic PN. Information about subdiaphragmatic anatomy was extracted. 2.6. Level of evidence The level of evidence for the PN supply of each abdominal structure was analyzed considering only data obtained in human samples (Table 7). It was considered not existent (no studies available), conflicting (inconsistent findings among multiple studies), limited (one lowquality study), moderate (consistent findings among multiple lowquality studies or one high quality study), or strong (consistent findings among multiple high-quality studies) (Van Tulder et al., 2003; Palomeque-Del-Cerro et al., 2017). 3. Results 3.1. Study selection The search strategy resulted in total of one thousand three hundred and thirty relevant articles that were retrieved through the Web of Science database. Then, seventy-three studies were excluded for not meeting the eligibility criteria. From the remaining one thousand two hundred and fifty-seven articles, fifteen of them were finally included. Following the literature review of the included studies, three more articles were added. All selected studies were written in English. Fig. 1 shows the flow diagram for the study selection process. 3.2. Study characteristics All the articles that were included were anatomy studies. Detailed descriptions of them are presented in Tables 1 through 4. Human and animal studies were selected. Two out of eighteen were case reports (Uzun et al., 2003; Natsis et al., 2004). Most of the included articles studied the RPN (eight out of eighteen), two studied the LPN (Haenssgen et al., 2020; Ogawa et al., 1958) and the others studied both branches. Seventy-seven percent of the studies included humans. Not all articles specified whether they were carried out in men or women. Only one study did not indicate the number of participants (Alexander, 1940). Three were conducted in children or fetuses (Mitchell, 1940; Pearson and Sauter, 1971; Hidayet et al., 1974). Forty-four percent of the studies included animals. Cats were the most numerous, appearing in four of the seven. Only one study included fourteen piglets (Haenssgen et al., 2020) and another study only included ten Suncus murinus (Yi et al., 2007). Four of the eighteen studies include both species (Haenssgen et al., 2020; Alexander, 1940; Hidayet et al., 1974; Yi et al., 2007). Only one of them differentiated between humans and animal data (Alexander, 1940). 3.2.1. Subdiaphragmatic supply of the PN Fifty-five percent of the studies involving only humans indicated that after the phrenic ganglion the nerve continues to the celiac ganglion and/or the adrenal nerve. Stomach innervation was also found in two studies (Mitchell, 1940; Gehwolf et al., 2020) and liver innervation in two (Natsis et al., 2004; Mitchell, 1940). The esophageal hiatus and crus were innervated in three other studies (Pearson and Sauter, 1971; An et al., 2012; Scott, 1965) and the inferior vena cava in one (Pearson and Sauter, 1971). Among the animal studies, two of them indicate innervation of the gallbladder by the PN (Hazarika et al., 1964; Iwahashi et al., 1991), one underlines the existence of phrenic afferents in the hepatic vein, the hepatic parenchyma and the inferior vena cava (Kostreva and Pontus, 1993). Innervation of the pancreas was also found in one study (Hazarika et al., 1964) and innervation of the esophageal hiatus was M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 3 suggested in another article (Ogawa et al., 1958). In the four mixed studies, the PN innervates the coronary ligament and the liver in one study (Alexander, 1940). Two of them indicate innervation of the esophagogastric junction only in humans (Haenssgen et al., 2020; Hidayet et al., 1974). In addition, one of them underscores the innervation of the inferior vena cava, the adrenal gland, the liver capsule in human species (Hidayet et al., 1974) and innervation of the gallbladder was found in another article (Yi et al., 2007). In total, half of the human specimens studied presented PN connections with any subdiaphragmatic structure. 3.3. Methodological quality of studies The evaluation of the methodological quality of the human studies using the QUACS scale (Wilke et al., 2015) revealed that the studies reported substantial to excellent methodological quality, scoring between eight to twelve out of a maximum of thirteen points, with only one study showed fair methodological quality (five points) (Table 1). The assessment of the methodological quality of the animal studies using the SYRCLE scale (Hooijmans et al., 2014) revealed that the studies denoted poor methodological quality. They scored four out of ten points (Table 1), which shows that all studies had a high risk of bias in items. 3.4. Anatomy textbooks Classic anatomy texts indicate that the PN pathways continue at this level (Moore and Agur, 2003; Lippert, 2006). It divides into several branches that supply the stomach (Lippert, 2006; Hochschild, 2017), the celiac ganglion (Loukas et al., 2013; Latarjet, 2019), the falciform and coronary ligaments of the liver (Netter, 2011), the Glisson capsule (Netter, 2011), the inferior vena cava (Rouviere and Delmas, 2000; Williams and Bannister, 1995), the liver (Lippert, 2006; Hochschild, 2017; Netter, 2011; Williams and Bannister, 1995), the gallbladder and the rest of the biliary tract – including the duodenal papila and Oddi sphincter (Netter, 2011). In addition to the adrenal gland (Rouviere and Delmas, 2000; Loukas et al., 2013; Williams and Bannister, 1995). Table 5 shows the information collected from books. Further, Table 6 synthetises the number of sources with data on PN supply for the subdiaphragmatic tissues. 3.5. Level of evidence Taking into account only human studies, there is strong evidence for the existence of PN innervation at the subdiaphragmatic level in the gastro-esophageal junction/esophageal hiatus, the celiac ganglion, the inferior vena cava, the liver and the adrenal glands/nerves. Moreover, Fig. 1. Flow chart diagram of the study selection process (Preferred Reporting Item for Systematic Reviews and Meta-Analyses, PRISMA, guidelines). M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 4 there is moderate evidence of this innervation in structures such as the stomach, the coronary ligament of the liver and the gallbladder (Table 7). Fig. 2 shows the percentage of documents which expose a phrenic connection with each structure. 4. Discussion The aim of this review was to study the subdiaphragmatic anatomy of the PN. According to our results, the PN has been shown to spread widely through the structures in the upper part of the abdominal cavity. The PN contains motor, sensory and autonomic fibers (Canella et al., 2010; Moore et al., 2013; Yano, 1928). In terms of the latter, the PN receives sympathetic fibers originating from the middle and lower cervical ganglia, from the sympathetic trunk, or from a sympathetic plexus on the pleural cupula (Canella et al., 2010; Moore et al., 2013; Yano, 1928). However, at the abdominal level, beyond the spread of the motor fibers on the lower surface of the diaphragm, there are controversies about the sympathetic or sensitive nature of the PN. 4.1. Autonomic supply Mitchell (1940) suspects that all the subdiaphragmatic fibers of the PN are autonomic. The sympathetic nature of the phrenic branch is supported by an analysis of its fiber composition, similar to that of the branches of the celiac plexus, which has shown to be connected to the PN (Uzun et al., 2003; Pearson and Sauter, 1971; Loukas et al., 2016; Table 1 Data collection selected studies. AUTHORS YEAR OF PUBLICATION PARTICIPANTS (HUMANS AND/ OR ANIMALS) SEX / MEAN AGE QUACS SCALE SYRCLE SCALE Mitchell, (1940) 15 infants - 5/13 - Alexander, (1940) Humans and cats - 8/13 4/10 Ogawa et al., (1958) 47 dogs - - 4/10 Hazarika et al., (1964) Cats - - 4/10 Scott, (1965) 21 humans - 8/13 - Pearson and Sauter, (1971) 10 human fetuses/ embryos - 8/13 - Hidayet et al., (1974) 30 humans (10 fetuses), 2 dogs and 5 cats. - 11/13 4/10 Iwahashi et al., (1991) 49 cats (4 groups) Female / male - 4/10 Kostreva and Pontus, (1993) 8 dogs - - 4/10 Uzun et al., (2003) 1 human Female 34 years old 8/13 - Natsis et al., (2004) 1 human Female 72 years old 9/13 - Rusu, (2006) 10 humans - 10/13 - Yi et al., (2007) 10 humans 10 animals (Suncus murinus) 5 females / 5 males Age range 50–94 years 10/13 4/10 An et al., (2012) 15 humans 6 females / 9 males Age range 46–75 years 9/13 - Loukas et al., (2016) 130 humans 88 females / 42 males Mean 70 years 12/13 - Verlinden et al., (2018) 35 humans 16 females / 19 males Between 58 and 101 years of age 12/13 - Haenssgen et al., (2020) 30 humans 14 piglets 16 females / 14 males Mean 84.4 years 11/13 4/10 Gehwolf et al., (2020) 53 humans 30 females / 23 males Age range 35–103 years 11/13 - Table 2 Subdiaphragmatic anatomy of the PN in human studies. PN, Phrenic Nerve; LPN, Left Phrenic Nerve; RPN, Right Phrenic Nerve. STUDIES PN SUBDIAPHRAGMATIC ANATOMY OF THE PN Mitchell, (1940) Both The LPN supplies several filaments to the cardiac end of the stomach. The RPN may supply the stomach and liver, an arrangement which has been seen so far in four cases. Scott, (1965) RPN The postero-medial branch supplies the right half of the esophageal hiatus, whereas the postero-lateral branches arch laterally within the muscle along the posterior edge of the central tendon. Pearson and Sauter, (1971) RPN Some of the posterior fibers pass to a small ganglion located near the inferior phrenic artery. These fibers pass beyond the ganglion in a bundle that communicates with the celiac plexus. The posterior division of both PN innervates the esophageal hiatus. In all but one of the embryos and fetuses studied, a communicating branch between the RPN and the LPN was found to pass anterior to the inferior vena cava just above the entrance of the hepatic veins. Uzun et al., (2003) RPN The phrenic ganglion was observed on the right nerve of the phrenic artery at the level of T10–11, in the right crus of diaphragm. It was located 40 mm above the celiac ganglion, where it is going. Natsis et al., (2004) RPN The RPN directing anterior to the pericardium penetrated the sternocostal part of the diaphragm, passed through the two layers of the falciform ligament and distributed to the upper surface of the serous layer of the liver. Rusu, (2006) RPN By the anastomosis between the RPN and the diaphragmatic nerve, morphologically variable phrenic ganglia have been regularly detected macroscopically. These ganglia were multiple in 50% of the samples studied. Two specimens presented two distinguishable phrenic ganglia each, located in the right inferior phrenic artery; these ganglia, serially connected, gave off adrenal nerves. An et al., (2012) Both The medial branches of the PN were subdivided and distributed to the muscular portions that form the esophageal hiatus. The lateral branches supplied the lumbar section posterior to the central tendon. Loukas et al., (2016) Both The LPN was present inferior to the diaphragm in 60% of cases. The LPG was present in 19% of the specimens and exhibited plexiform communication with multiple ganglia in 71,4%. It communicated only with the celiac ganglion in 21.4% and only with the adrenal gland in 7.1%. The RPN was present inferior to the diaphragm in 98% of cases. A RPG was present in 49.1% of specimens and showed plexiform communications with the celiac ganglion in 15.5% and the adrenal gland in 10.3%. Verlinden et al., (2018) RPN The abdominal course of the RPN was studied in five cadavers. The PN splits into two main myelinated branches that innervated the diaphragm. A TH- positive autonomic fascicle presented as a third, separate branch distal to this level and extended all the way to the celiac ganglion, this being the phrenic branch of the celiac plexus. Gehwolf et al., (2020) Both In ten cadavers in which a branch of the PN was identified, these nerves penetrated the left crus. In nine of theses ten cadavers, the target site was the fundus of the stomach. In one cadaver, the course of the nerve could not be accurately mapped. PN, Phrenic Nerve; LPN, Left Phrenic Nerve; LPG, Left Phrenic Ganglion; RPN, Right Phrenic Nerve; RPG, Right Phrenic Ganglion; RAPG, Right Accessory Phrenic Ganglion. M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 5 Verlinden et al., 2018). The areas of small-diameter myelinated fibers suggest the present of autonomic nerves (Kandel et al., 2000). Recent studies have shown that fibers are positive for tyrosine hydroxylase in both PN, indicating sympathetic activity (Verlinden et al., 2018). Interestingly, in Haenssgen’ study (2020), there was no tyrosine hydroxylase activity in the RPN at the diaphragm level in some of the specimens investigated. In this case, the PN could regain sympathetic fibers from the branches of the celiac plexus, which would use the PN as a morphological conduit (Verlinden et al., 2018). This idea is further supported by histological examination of the right phrenic and right accessory phrenic ganglia (Rusu, 2006; Tubbs et al., 2008). The presence of multipolar autonomic neurones within the phrenic ganglia suggests that these communications could provide autonomic innervation (Rusu, 2006; Tubbs et al., 2008). This autonomic innervation could have a vasoconstrictor function within the diaphragmatic muscle (Bałkowiec and Szulczyk, 1992). Moreover, the diversity and composition of the left branch of the PN at the gastro-esophageal junction suggest ancillary functions in the innervation of blood vessels and/or internal sphincter muscles (Haenssgen et al., 2020). The latter could be important during events leading to transient relaxation of the lower esophageal sphincter. Furthermore, electrical stimulating of the central cut end of the PN has also been observed to cause an increase in gastric movements. Increased movements of the small bowel were generally less noticeable or not always observable (Nakayama and Mori, 1968). In addition, when the phrenic innervation is interrupted, a considerable and consistent elevation of the pylorus and duodenal bulb to the gastric cardia follows. After discontinuation, the gallbladder also increased significantly (Jefferson et al., 1955). 4.2. Sensory supply Regarding the sensory innervation of the PN, it has been reported that, as it descends the thorax, it innervates the mediastinal pleura, the right atrium, the fibrous pericardium, the parietal layer of the serous pericardium, the diaphragmatic pleura and the central portion of the diaphragm (Canella et al., 2010; Hollinshead, 1985; Williams et al., 1995; Moore et al., 2013). Nakayama and Mori (1968) supposed that the PN contains the sensory nerves from the gallbladder, pancreas and diaphragmatic peritoneum. It is generally known that aggressions of several abdominal tissues can cause referred pain along the C3-C4-C5 dermatomes, at the neck, shoulder and arm (Goodman, 2010). Visceral nociception is perceived as secondary hyperalgesia in somatic tissues neuromerically connected with the affected tissue (Giamberardino and Vecchiet, 1995), generally known as Head´s zones (Head, 1893), possibly due to convergence of the afferent and a misinterpretation of spinal or supraspinal levels (Giamberardino and Vecchiet, 1995; Luz et al., 2015; Ruch and Patton, 1965). Therefore, since the PN is made up mainly of C3-C5 fibers and these levels are related to neck, shoulder and arm tissues (Moore et al., 2013; Netter, 2011; Williams and Bannister, 1995; Lee et al., 2008), any structure sensitively supplied by PN, including subdiaphragmatic, could be related to pain processes perceived in those areas. In fact, some authors even directly name it ‘phrenic pain’ (Flanagin et al., 2010). Sometimes neck-to-shoulder pain related to any abdominal disorder Table 3 Subdiaphragmatic anatomy of the PN in animal studies. PN, Phrenic Nerve; RPN, Right Phrenic Nerve; HPR, Horseradish Peroxidase. STUDIES PN SUBDIAPHRAGMATIC ANATOMY OF THE PN Ogawa et al., (1958) LPN In thirteen dogs, one or two branches of the LPN were observed from the posterior branch of the LPN, crossing to the posterior crus of the right hemidiaphragm and innervating the muscle bundles on the immediate right side of the esophageal hiatus. Hazarika et al., (1964) Both The PN innervated the pancreas and the gallbladder. In addition, it has been claimed that the PN proved to be the pathway involved in unexpected reflex responses when all other available pathways were thought to have been cut off. Iwahashi et al., (1991) Both After injection of HPR into the gallbladder wall. In two of four animals, cells labelled in the left cervical ganglia were found after the RPN section. Kostreva and Pontus, (1993) RPN The existence of mechanoreceptors of the hepatic vein, hepatic parenchymal, and inferior vena caval with phrenic afferents in the dog. Table 4 Subdiaphragmatic anatomy of the PN in mixed studies (humans and animals). PN, Phrenic Nerve; RPN, Right Phrenic Nerve; LPN, Left Phrenic Nerve. STUDIES PN SUBDIAPHRAGMATIC ANATOMY OF THE PN Alexander, (1940) RPN The branches of the RPN could be traced posteriorly to the region of the coronary ligament. Within the liver, myelinated fibers have been observed in some instances. Undoubtedly, the few myelinated fibers remaining are afferent phrenic components. Hidayet et al., (1974) Both In humans, a medial branch of the PN innervates the musculature at the margin of the esophageal hiatus. The posterior branch of the RPN is connected to a ganglion (phrenic ganglion) together with subdiaphragmatic nerve filaments of the celiac plexus in approximately half of the humans studied. This macroscopic plexus is seen only on the right side of humans, and connections may be traced from it to the diaphragm, inferior vena cava, the right adrenal gland and the liver capsule. Yi et al., (2007) Both In humans, innervation of the gallbladder involved three routes: via the anterior and posterior hepatic plexus and the PN. The PN sent offshoots forward to the hepatic portal, which ran along the sagittal sulcus of the liver. Offshoots were observed in eight out of ten cases, predominantly originating from the RPN in seven and from both the RPN and the LPN in one. Although there were no anterior and posterior hepatic plexuses, similar results were obtained for Suncus murinus. Haenssgen et al., (2020) LPN In twenty three of the thirty human specimens the LPN was involved in the innervation of the esophagogastric junction. Macroscopic dissection and nerve staining of the piglets revealed that none of them had phrenic branches to the esophagogastric junction. Table 5 Subdiaphragmatic anatomy of the PN as exposed in books. PN, Phrenic Nerve; RPN, Right Phrenic Nerve; LPN, Left Phrenic Nerve. BOOKS PN SUBDIAPHRAGMATIC ANATOMY OF THE PN William et al., 1995 Both The PN innervates the inferior vena cava, the liver and the adrenal gland. Rouviere and Delmas, (2000) Both The PN supplies the right adrenal gland, the subdiaphragmatic peritoneum of the hepatic region. It also intervenes in the supply of certain vasa, in particular the inferior vena cava. Moore and Agur, (2003) Both The PN sensitively innervates the diaphragmatic peritoneum, the gallbladder and the liver. Lippert, (2006) Both It receives afferent branches from the PN to the pleura, pericardium and peritoneum (especially visceral peritoneum of liver and cardia). Netter, (2011) Both The PN innervates the falciform and coronary ligaments of the liver, Glisson’s capsule, the liver, the gallbladder and the rest of the biliary tract – including the duodenal papila and the Oddi sphincter. Loukas et al., (2013) LPN The PN leaves the celiac ganglion to innervate the left adrenal gland. Hochschild, (2017) Both In the form of phrenic-abdominal branches the PN continues to the liver, stomach and kidneys, and provides part of its sensory innervation. Latarjet, (2019) RPN The RPN provides an inconstant fillet to the right celiac ganglion which approaches it at its superior border. M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 6 is exposed to diaphragmatic pressure or suffering. For instance, it can be read in Netter anatomy textbook that hepatic, gallbladder or duodenal harm can cause pain in the shoulder tip due to ‘diaphragmatic irritation’ (Netter, 2011). This has also been used as an explanation of shoulder pain in the case of hepatic tumors ablation (Head et al., 2007), donor right hepatectomy (Schumann et al., 2004), ventriculoperitoneal shunts (Tubbs et al., 2005), distended stomach or even spleen laceration (Ott et al., 2022). This can be true in several situations, but a direct implication of specific abdominal structures, independently of diaphragmatic status, must be considered according to our results if a sensory role is confirmed at this level. It is difficult to determine the exact origin of neck-to-shoulder pain in many cases of abdominal visceral pain because it may irritate not only the diaphragm, but also other adjacent tissues with phrenic supply. 4.2.1. Peritoneum We have not found any study performed in humans showing that the PN supplies the peritoneum. It might be because it is a generally accepted fact that does not needs more research. Animal experiments showed that the nociceptive sensitivity produced by stimulation of the central portion of the diaphragmatic peritoneum depends on the conduction of the afferent in the sensory fibers of the PN, independent of the afferent fibers in the vagus nerve and the efferent sympathetic pathways (Hinsey and Phillips, 1937). The peritoneum has been considered a source of shoulder pain in cases such as transient shoulder tip pain related to exercise (Morton and Callister, 2015), laparoscopic surgeries (Fernandez-Lopez et al., 2020), endometriosis and cyclic shoulder pain (Anderson et al., 2014; Redwine, 2002). Additionally, leakage from any damaged structure can finally irritate the peritoneum. 4.2.2. Hepatic and biliary system Regarding the supply of the PN to the liver and related structures, we have found different levels of evidence. Traction of the falciform and coronary ligaments of the liver has been considered a source of pain in the right shoulder in either laparoscopic surgery (Sao et al., 2019), or an increase in liver size (Netter, 2011), the latter also allowing for neck-to-shoulder pain due to the stretching of the Glisson capsule (Netter, 2011). Liver subcapsular hematoma and intraparenchymal hemorrhage can cause shoulder or neck pain (Barton and Sibai, 1996). The subcapsular hematoma may also press the peritoneum and diaphragm, but intraparenchymal hemorrhage that causes neck or shoulder pain could confirm the sensory supply of the PN from the liver parenchyma. According to Lewis, neck-to-shoulder pain is triggered by the gallbladder, the cystic duct and parenchymatous liver disease, or by inflammation or traction of the coronary and falciform ligaments, the Glisson capsule or the peritoneal cover of the subsurface of the diaphragm (Lewis, 1951). Almost 20% of patients complain of pain in the right shoulder area during electrocoagulation of the biopsy area of the liver, while these manipulations are not believed to cause sufficient traction of the hepatic ligaments to cause pain (Lewis, 1951). Lewis considers that, under the appropriate circumstances of congestion or inflammation, central hepatoma or abscess, the hepatic parenchyma triggers phrenic pain, without stretching the capsule, noticeable organ enlargement, or perihepatitis (Lewis, 1951). Twenty-three percent of patients suffer neck or right shoulder pain during attempts at insertion of the sheath into the right or middle hepatic vein (Khosa et al., 2003). Interestingly, we have not found evidence of hepatic veins supply in humans, but in animals (Kostreva and Pontus, 1993). In the case of radiofrequency ablation of hepatocellular carcinoma, visceral referred pain has been proposed to be caused by hepatic injury, hemoperitoneum or thermal injury of diaphragm (Kim et al., 2006). In the same way, liver trauma that causes shoulder pain could be due to the phrenic supply of the liver or peritoneal irritation (Navarro et al., 2000; Shilyansky et al., 1999). Table 6 Synthesis of the sources showing phrenic supply of subdiaphragmatic structures. ANATOMICAL STRUCTURES HUMANS STUDIES ANIMALS STUDIES MIXED STUDIES HUMAN ANATOMY TEXTBOOKS Peritoneum Rouviere and Delmas, (2000), Moore and Agur, (2003), Lippert, (2006) Gastroesophageal junction/ esophageal hiatus An et al. (2012), Scott, (1965), Pearson and Sauter, (1971) Ogawa et al. (1958) Haenssgen et al. (2020), Hidayet et al. (1974) Stomach Mitchell, (1940), Gehwolf et al. (2020) Hochschild, (2017) Celiac ganglion Uzun et al. (2003), Pearson and Sauter, (1971), Loukas et al. (2016), Verlinden et al. (2018) Latarjet, (2019) Falciform ligaments of the liver Netter, (2011) Coronary ligaments of the liver Alexander, (1940) Netter, (2011) Glisson´s capsule Natsis et al. (2004) Hidayet et al. (1974) Netter, (2011) Inferior vena cava Pearson and Sauter, (1971) Kostreva and Pontus, (1993) Hidayet et al. (1974) Rouviere and Delmas, (2000), William et al. 1995 Liver (parenchyma) Natsis et al. (2004), Mitchell, (1940) Kostreva and Pontus, (1993) Alexander, (1940) Moore and Agur, (2003), Lippert, (2006), Hochschild, (2017), Netter, (2011), William et al. 1995 Hepatic vein Kostreva and Pontus, (1993) Gallbladder Hazarika et al. (1964), Iwahashi et al. (1991) Yi et al. (2007) Moore and Agur, (2003), Netter, (2011) Biliary tract Netter, (2011) Adrenal glands/ nerves Rusu, (2006), Loukas et al. (2016) Hidayet et al. (1974) Rouviere and Delmas, (2000), Loukas et al. (2013), William et al. 1995 Pancreas Hazarika et al. (1964) Kidneys Hochschild, (2017) M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 7 Table 7 Level of evidence of PN supply for each structure according to human data. ANATOMICAL STRUCTURE TOTAL EXCELLENT SUBSTANTIAL MODERATE FAIR POOR LEVEL OF EVIDENCE Peritoneum Non-existent Gastroesophageal junction/esophageal hiatus 5 2 3 Strong Stomach 2 1 1 Moderate Celiac ganglion 4 2 2 Strong Falciform ligament of the liver Non-existent Coronary ligament of the liver 1 1 Moderate Glisson’s capsule 2 1 1 Strong Inferior vena cava 2 1 1 Strong Liver (parenchymal) 3 2 1 Limited Hepatic vein Non-existent Gallbladder 1 1 Moderate Biliary tract Non-existent Adrenal glands/nerves 4 2 1 1 Strong Pancreas Non-existent Kidneys Non-existent Fig. 2. Subdiaphragmatic structures connected to the phrenic nerves. Percentages of documents about each structure. M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 8 However, some authors do not consider any of the hepatic tissues but the peritoneum to be the source of shoulder pain in cases such as hepatic cysts (Yamada et al., 1994) and hepatic colic crisis (Rouviere and Delmas, 2000). Similarly, shoulder pain in living liver donors after the surgical procedure has been considered to be due to malposition of the donor body during long surgery (Akpek et al., 2003). It is interesting to note that, in the normal population, shoulder pain has not been shown to be related to sleep posture, but to organic suffering (Holdaway et al., 2018). In fact, preventing overstretching of the right arm by laterally attaching it to the body did not prevent shoulder pain in these hepatectomized patients (Ozkardesler et al., 2008). Furthermore, the PN infiltration has been shown to reduce shoulder pain after laparoscopic procedures for cholelithiasis (Yi et al., 2017), hepatectomy (Bak et al., 2021) and liver radiofrequency ablation (Liu et al., 2021). Finally, hepatic cysts drainage has been shown to eliminate shoulder pain (Gerwin, 2002), so cholecystectomy has been able to resolve chronic neck pain (Draper, 2008). In this sense, Moore et al. (2013) indicates that the PN supply of the gallbladder is sensitive in nature. 4.2.3. Esophagus and stomach We have found moderate level of evidence that the PN supplies the stomach, with some authors exposing that it is sensory innervation (Hochschild, 2017; Sis¸u et al., 2012). Additionally, we found a strong level of evidence for the gastroesophageal junction and hiatus. Furthermore, the supply of the PN to the esophagus is widely accepted. Thus, neck and shoulder pain following gastric perforation and bleeding ulcers could be related not only to the phrenic supply of the stomach but also to peritoneal irritation (Lax and Singh, 2020; Pappano and Bass, 2006). Hiatal hernia and gastroesophageal reflux trigger pain in the neck, left shoulder, arm and jaw (Flanagin et al., 2010). Related to this viscerosomatic reflex, Sandifer syndrome has been described, which is characterized by torticollis due to gastroesophageal reflux. Patients suffering from this syndrome have spasms and pain in the neck muscles (Shahnawaz et al., 2001; Rana et al., 2013). It has been proposed to be a vagal reflex (Cerimagic et al., 2008). However, phrenic afferences might contribute to the syndrome. Postpandrial shoulder pain has also been reported due to a distended stomach (Ott et al., 2022). Interestingly, experimental research on gastric pain is carried out by stomach distension, which triggers a spasm in the upper trapezius (Ozaki et al., 2002; Zhang et al., 2013). Furthermore, gastric electrical stimulation decreases the electromyographic activity of the neck muscles, as a sign of visceral pain inhibition (Sun et al., 2014). Regarding phrenic pain referred to the jaw due to hiatal hernia (Flanagin et al., 2010), afferent convergence has been reported for the jaw and the PN in rats (Matsumoto et al., 1999), just as convergence of trigeminal, visceral and phrenic input has been shown in primates (Chandler et al., 1999). Furthermore, the trigeminocervical nucleus in humans descends until C4 (Brazis et al., 1985; Barr and Kiernan, 1983), so a trigeminophrenic convergence could occur. Finally, gastric volvulus has also shown to produce pain in the neck and left arm (Saghir et al., 2019). 4.2.4. Adrenal glands Regarding the connection of the PN to the adrenal glands, we have found strong evidence of its existence. It has been shown to constitute a sensitive supply (Barral and Croibier, 2009). Some cases of adrenal injury and shoulder pain have been reported, such as adrenal cancer (Branham et al., 2022; Wrightson et al., 2002) or hematoma (Garber and Renzi, 1988; Decina et al., 1992). However, considering the size of the masses, other abdominal structures such as the splenic artery, stomach (Garber and Renzi, 1988) and liver (Branham et al., 2022) were involved. In addition, referred pain after laparoscopic procedures for adrenal disorders might be related to the usual peritoneal irritation of these procedures (Kozłowski et al., 2019; Suzuki, 2001). Therefore, we do not have data to support a sensitive phrenic supply of the adrenal glands. 4.2.5. Inferior vena cava Strong evidence has been found in humans for the supply of the PN to the inferior vena cava. Clinically, a patient with shoulder pain as a consequence of acute adrenal hemorrhage has been reported to show restricted blood flow in the inferior vena cava due to the pressure effect of the hematoma. This may be the source of the referred pain to the shoulder, but extravasation was also found, so peritoneal irritation cannot be avoided (Acar et al., 2015). 4.2.6. Pancreas Regarding the pancreas, it has been shown to have the PN supply in animal studies (Hazarika et al., 1964), and it has been assumed that the PN contains sensory nerves for the pancreas in humans (Nakayama and Mori, 1968). However, we have not found evidence of this PN supply in humans. Diabetes and metabolic syndrome have frequently been associated with neck and shoulder pain (M¨ antyselk¨ a et al., 2010; Vehmas et al., 2013; Rechardt et al., 2010; Kim et al., 2023). The low-intensity chronic inflammatory process that affects not only internal organs but also the musculoskeletal tissues has been considered a key factor (Csonka et al., 2023). Pancreatitis has also been shown to produce shoulder pain (Campbell, 1983), but the referred pain could also be due to peritoneal irritation. Furthermore, diabetes has previously demonstrated to cause visceral hypersensitivity in structures such as the esophagus, stomach and duodenum (Zhang et al., 2013; Frøkjær et al., 2007), supporting a possible link between visceral and metabolic disorder and shoulder and neck suffering. It should be noted that this gastric hypersensitivity in diabetes increases the reflex spasm of the neck muscles, showing enhanced viscerosomatic consequences (Zhang et al., 2013). Additionally, the ability of long-standing visceral disorders to induce central sensitization, increase musculoskeletal pain and generate myofascial trigger points might connect pancreatic, gastric and/or esophageal suffering with neck and shoulder pain in patients with metabolic disorders (Gerwin, 2002; Pacheco-Carroza, 2021; Cervero, 2000, 2009; Han and Lee, 2009). 4.2.7. Other structures The PN achieves the celiac plexus and ganglia with a strong level of evidence according to our results. This can be related to the fact that celiac plexus block may trigger shoulder pain (Marcy et al., 2001). Regarding the spleen, intrasplenic contrast injection triggers left shoulder pain, but it appears to be related to contrast extravasation and leakage into the perisplenic space and peritoneum and not into the splenic tissues (Jain et al., 1999). This is also considered the cause of the PN stimulation in cases showing left shoulder pain after splenic rupture (Saad and Rex, 2008; Sachdev et al., 2012). However, despite the fact that we have not found evidence of the PN splenic supply, a case of left shoulder pain due to a foreign body in the spleen has been reported (Isbir et al., 2012). In relation to the kidney, sensitive innervation by the PN has been exposed (Hochschild, 2017), but we have not found any study on it. Thus, shoulder pain after laparoscopic nephrectomy and renal transplantation has been considered secondary to diaphragmatic or peritoneal irritation (Ismail et al., 2005). 4.2.8. Function of PN sensory supply in peridiaphragmatic tissues The PN sensitive supply of visceral structures seems to have a protective role. In the case of gastroesophageal reflux, PN afferences can contribute to the sensation of cervical metameres to trigger cervicoscapular spasms with the aim of avoiding aspiration and voiding the esophagus, as seen in the Sandifer Syndrome (Lehwald et al., 2007). Furthermore, the fact that the same nerve that provides the motor supply to the diaphragm also provides a sensitive supply to those M. P´ erez-Montalb´ an et al. Annals of Anatomy 254 (2024) 152269 9 structures surrounding it might have a protective role because it allows a better control of respiratory contractions. Since diaphragmatic excursion modifies abdominal and thoracic pressures, in the event of injury to the peridiaphragmatic structures, the PN blocks the diaphragm, avoiding any further injury. For example, subjects with gallstones avoid diaphragmatic breathing (Moore et al., 2013). This mechanism could be the basis for the respiratory pattern of obese subjects, with swollen peridiaphragmatic structures and/or increased hepatic pressure. In this sense, some phrenic afferents appear to modulate respiratory rate and are speculated to be involved in the perception of dyspnea (Pappano and Bass, 2006; Frankshteĭn et al., 1978). 4.3. Hiccups Our results might also contribute to explaining the ability of certain abdominal structures to cause hiccups when irritated (Adu-Gyamfi and Kudaravalli, 2022; Eren Durmus¸ et al., 2022; Amoruso et al., 2013; Grifoni et al., 2013). The implication of the PN in hiccups is even highlighted, since it sometimes occurs along with shoulder pain (Lee and Chang, 2020). Interestingly, it has been reported that the most common cause of hiccups is distension of the stomach, which would be the afference that triggers the reflex (Steger et al., 2015). 4.4. Limitations We searched only in one database. However, the initial number of studies to be evaluated was high. Although many abdominal structures have been shown to have phrenic connections, the number of studies reporting that the connection for some structures is scarce. Furthermore, the variability between subjects and species is unclear. So, more research is needed about every subdiaphragmatic structure, with special focus on variability and differentiation between autonomic and sensory functions. It should also be noted that due to the lack of adequate scales to assess the methodological quality of anatomical studies in animals, we used the SYRCLE scale (Hooijmans et al., 2014), which has several items that fit intervention studies, but not anatomic studies. That is the reason why our results indicate that the reviewed studies have low quality. The anatomy textbooks were selected simply by availability. Other textbooks might show different information. However, we have used them simply as a comparator, and to show what nowadays might be accepted and taught. We have not considered the information obtained in those texts to be evidence. Furthermore, the textbooks we have used included the most recognized ones. 5. Conclusion This review provides the first systematic evidence for the supply and connections of the subdiaphragmatic PN. It has been demonstrated for the gastroesophageal junction, the stomach, the celiac ganglia, the liver and its coronary ligament, the inferior vena cava, the gallbladder and the adrenal glands. This is of interest for professionals who deal with people suffering from neck and shoulder pain, as well as patients with peridiaphragmatic disorders or hiccups. Since the number of studies for many structures is low, more research is needed, with special focus in determining between-subjects variability and type of fibers, either autonomic or sensitive. Ethical Statement The authors declare that the work presented in this article has been carried out in accordance with all ethical considerations. CRediT authorship contribution statement Encarna García-Domínguez: Writing – review & editing, Writing – original draft, Investigation, Conceptualization. María P´ erez-Mon- talb´ an: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. ´ Angel Oliva-Pascual-Vaca: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References Acar, D., Gülpembe, M., Akilli, N.B., et al., 2015. Acute adrenal hemorrhage in an elderly patient. J. Acad. Emerg. Med Case Rep. 6, 10–12. https://doi.org/10.5152/ jaemcr.2015.742. Adu-Gyamfi, K.O., Kudaravalli, P., 2022. The ominous hiccup: pyogenic liver accesses as a complication of acute sigmoid diverticulitis. Am. J. Gastroenterol. 117 (10). Akpek, E.A., Arslan, G., Erkaya, C., et al., 2003. Anesthetic risks for donors in livingrelated liver transplantation: analysis of 30 cases. Transpl. Int 16 (8). https://doi. org/10.1007/s00147-003-0614-2. Alexander, W.F., 1940. The innervation of the biliary system. J. Comp. Neurol. 72, 357–370. https://doi.org/10.1002/cne.900720205. Amoruso, L., Kaga, M., Wang, L., 2013. Gastric ulcer causing recurrent hiccups: a case report. Am. J. Gastroenterol. 108, S221–S222. An, X., Yue, B., Lee, J.H., Lee, M.S., Lin, C., Han, S.H., 2012. Intramuscular distribution of the phrenic nerve in human diaphragm as shown by Sihler staining. Muscle Nerve 45 (4), 522–526. https://doi.org/10.1002/mus.22141. Anderson, T.L., Aguirre, F., Ayuso, A., 2014. Diaphragmatic endometriosis in a patient with cyclic shoulder pain. J. Minim. Invasive Gynecol. 21 (1), 23–24. https://doi. org/10.1016/j.jmig.2013.05.012. Andrade, L.S., Jaison, J., B, J., Lewis, C.R., Prasanna, L.C., 2021. A study on the variants of phrenic nerve roots with histological correlation. Transl. Res Anat. 23, 100123. Bak, T.S., Bøgevig, S., Christensen, A.P., Tollund, C., Hillingsø, J., Aasvang, E.K., 2021. Phrenic nerve block on severe post-hepatectomy shoulder pain: A randomized, double-blind, placebo-controlled, pilot study. Acta Anaesthesiol. Scand. 65 (9), 1320–1328. https://doi.org/10.1111/aas.13928. Bałkowiec, A., Szulczyk, P., 1992. Properties of postganglionic sympathetic neurons with axons in phrenic nerve. Respir. Physiol. 88 (3), 323–331. https://doi.org/10.1016/ 0034-5687(92)90006-i. Barr, M.L., Kiernan, J.A., 1983. The human nervous system, 4th ed. Harper & Row, New York, pp. 126–129. Barral, J.P., Croibier, A., 2009. Manipulaciones de los nervios perif´ ericos, 1ª edici´ on. Espa˜ na: Elsevier Masson, Barcelona. Barton, J.R., Sibai, B.M., 1996. Hepatic imaging in HELLP syndrome (hemolysis, elevated liver enzymes, and low platelet count). Am. J. Obstet. Gynecol. 174 (6), 1820–1825. https://doi.org/10.1016/s0002-9378(96)70216-8. Branham, Z., Fox, A.D., Ullah, A., Patel, N.G., Terris, M., Parikh, J., 2022. Insights into Clinical Features and Outcomes of Adrenal Cortical Carcinosarcoma. Diagn. (Basel) 12 (10), 2419. https://doi.org/10.3390/diagnostics12102419. Brazis, P.W., Masdeu, J.C., Biller, J., 1985. Localization in clinical neurology. Little, Brown, Boston, pp. 159–170. Campbell, S.M., 1983. Referred shoulder pain. An elusive diagnosis. Post. Med 73 (5), 193–203. https://doi.org/10.1080/00325481.1983.11697838. Canella, C., Demondion, X., Delebarre, A., Moraux, A., Cotten, H., Cotten, A., 2010. Anatomical study of phrenic nerve using ultrasound. Eur. Radio. 20 (3), 659–665. https://doi.org/10.1007/s00330-009-1579-z. Cerimagic, D., Ivkic, G., Bilic, E., 2008. Neuroanatomical basis of Sandifer’s syndrome: a new vagal reflex? Med Hypotheses 70 (5), 957–961. https://doi.org/10.1016/j. mehy.2007.09.011. Cervero, F., 2000. Visceral pain—central sensitization. Gut 47 (Suppl IV), iv56–iv57. Cervero, F., 2009. Visceral versus somatic pain: similarities and differences. Dig. Dis. 27 (Suppl 1), 3–10. https://doi.org/10.1159/000268115. Chandler, M.J., Qin, C., Yuan, Y., Foreman, R.D., 1999. Convergence of trigeminal input with visceral and phrenic inputs on primate C1-C2 spinothalamic tract neurons. Brain Res 829 (1-2), 204–208. https://doi.org/10.1016/s0006-8993(99)01348-7. Csonka, V., Varjú, C., Lendvay, M., 2023. Diabetes mellitus-related musculoskeletal disorders: Unveiling the cluster of diseases. Prim. Care Diabetes S1751-9918 (23), 00141–00149. https://doi.org/10.1016/j.pcd.2023.08.003. Decina, P.A., Vallee, D., Mierau, D., 1992. Acute pancreatitis presenting as back pain: a case report. J. Can. Chiropr. Assoc. 36 (2), 75–83. Drake, R.L., Vogl, A.W., Mitchell, A.W.M., 2015. Gray’s anatomy for students. 3ª ed. Elsevier, Barcelona. Draper, B., 2008. Chronic cervicothoracic dysfunction alleviated by cholecystectomy. Cl. íN. Chirop. 11, 77–82. Eren Durmus¸, M., ¨ Ocal, S., Duman, A., ¨ Ocal, R., K¨ oker, G., 2022. An unexpected cause of hiccups: esophageal variceal band ligation. J. Gastrointestin Liver Dis. 31 (4), 477–478. https://doi.org/10.15403/jgld-4588. M. P´ erez-Montalb´ an et al.