Tumori, 94: 779-786, 2008

Primary duodenal adenocarcinoma

Mario Solej1, Silvia DAmico1, Gabriele Brondino2, Marco Ferronato1, and Mario Nano1
1 University of Turin, School of Medicine, Section of General Surgery, San Luigi Gonzaga Hospital, Orbassano, and 2University of Turin, Department of Housing and Town, Polytechnic, Turin, Italy

ABSTRACT

Aims and background. Primary duodenal adenocarcinoma is a rare tumor with a poorly defined natural history and prognostic factors. It presents with nonspecific symptoms, and for this reason the diagnosis is often delayed. It is a serious problem for the surgeon because of the difficulty in obtaining an early diagnosis and standardizing basic tenets for an appropriate surgical approach. The aim of this work was to conduct a review of the literature analyzing the points most frequently debated about this pathology. Methods and study design. A bibliographic search was carried out on the main search engines to find studies regarding duodenal adenocarcinoma, published in English, from January 1992 to January 2007. Results. A total of 19 articles was selected. Results concerning symptoms, location of the tumor, diagnostic examinations, surgical treatment, histopathology of the tumor, survival and follow-up were obtained and discussed. Conclusions. All patients who are medically fit to undergo surgery should be given the option of aggressive resection regardless of tumor size, tumor invasion or appearance of positive lymph nodes. Hopefully, an early diagnosis will correlate with improved long-term survival.

Introduction Primary duodenal adenocarcinoma is a rare tumor with a poorly defined natural history and prognostic factors. It represents 0.3-1% of all gastrointestinal tumors and 25-35% of malignant tumors of the small intestine1-6. It presents with nonspecific symptoms, and for this reason the diagnosis is often delayed. Little is known about its biology, and fundamental questions regarding the pattern and significance of metastatic spread remain unanswered4. Primary duodenal adenocarcinoma is a serious problem for the surgeon because of the difficulty in obtaining an early diagnosis and standardizing basic tenets for a suitable surgical approach5. In the literature, there is heterogeneous data about the clinical presentation, surgical treatment and outcome of the patients. Moreover, almost all the articles are retrospective, and the case studies reported in most of these have been collected over several decades. There are only a few articles that deal with duodenal adenocarcinoma as a pathology in itself. It is often dealt with in the ambit of periampullary tumors, which tends to lead to even more dispersion of the data. Even recent articles have collected few cases with retrospective analysis. Notwithstanding the development of diagnostic techniques and the improvement of surgical techniques, it was not possible to determine with statistical significance which factors can influence the resectability of the tumor4. The aim of our work was to conduct a review of the literature and analyze the points of this pathology most frequently debated, in order to update the state-of-the-art of duodenal adenocarcinoma.

Key words: duodenum, primary duodenal adenocarcinoma, surgery, survival. Acknowledgements: The authors thank Doctor Ileana Baldi, MD, Department of Epidemiology, University of Turin, for the contribution to the study design. Correspondence to: Mario Nano MD, FACS, Universitary Section of General Surgery, San Luigi Gonzaga Hospital, Regione Gonzole 10, 10043 Orbassano (Turin), Italy. Tel +39-011-9026224; fax +39-011-9026887; e-mail mario.nano@unito.it Received March 10, 2008; accepted March 14, 2008.

780

M SOLEJ, S DAMICO, G BRONDINO ET AL

Material and methods Data search and study selection A bibliographic search was carried out on the main search engines (Pubmed, Embase, Ovid and Cochrane databases) to find studies regarding duodenal adenocarcinoma published in English from January 1992 to January 2007. Any study design was accepted (retrospective, perspective, randomized, systematic review and meta-analysis). The key words were adenocarcinoma, duodenum, primary duodenal adenocarcinoma, survival and surgery. A search was carried out on Pubmed using Mesh terms (duodenal neoplasm, survival, surgery); on selected articles research was carried out using related links. The related articles function was used to broaden the search, and all abstracts and citations scanned were reviewed. The bibliographic research and the selection of the articles was carried out independently by two of the authors (MS and SDA). All the articles not specifically regarding duodenal adenocarcinoma, tumors of the ampullary and the small intestine, articles on conditions with a predisposition for duodenal adenocarcinoma (familial adenomatous polyposis, duodenal polyps, Gardners syndrome, etc.), and the case reports and reviews without personal case studies were excluded. When two articles were published by the same author or the same institute, the most recent article was chosen. Articles that reported case studies of cadavers were excluded. All the articles chosen were found on Pubmed and were retrospective, with the exception of one perspective article7. Both authors (MS and SDA) identified the same articles. The following data was taken from each article: the year the collection of data for the study started and the year it finished, sample size, mean age, sex, symptoms (type and duration), tumor site, diagnosis, type and intent of intervention (radical or palliative), postoperative morbidity and mortality, main prognostic factors (tumor diameter, grading, staging parietal invasion, pancreatic infiltration, lymph node involvement, resection margins), survival and mean follow-up. Initially, we set up the work with the intention of carrying out a systematic review, accepting the types of study (carried out with retrospective data) and the choice of only articles in English as limits. The criteria which we initially chose to include (which represent the most debated and controversial themes in the literature, i.e. type of surgical intervention based on the site, survival based on the site, survival based on type of surgical intervention, survival based on lymph node involvement) did not produce an adequate sample size with a consequent low level of statistical evidence. Moreover, the selected articles had other limitations due to the heterogeneity of classifications used for the site of the tumor (topographic classification versus anatomic

based on the portions of the duodenum) and different grouping of some categories (clinical symptoms and signs). Therefore, the work was set up according to the rules of a classic review. Data regarding survival were quoted only when expressed in percentages. Global survival data after radical and palliative surgery, after pancreaticoduodenectomy or duodenal segmental resection, in patients with positive and negative lymph nodes and based on the site were considered. Other data regarding survival were not considered because of the limited number of articles reporting them. Some authors considered survival at 2 years, which was not shown in the tables; others only that at 5 years. Only 10 studies reported the mean follow-up.

Results Nineteen articles which contained observational data were selected. The samples exceeded 100 only in 26,7 of the 19 studies chosen. The total number of cases collected was 1021. The duration of the data collection varied from a minimum of 8 to a maximum of 41 years (median, 15). Signs and symptoms The most frequent symptoms found are jaundice (11.1-55.8%), nausea/vomiting (21-51.8%), pain (11.159.1%), weight loss (10.1-57.6%) and anemia (22.5-60.6%). Two articles did not report information regarding the symptomatology2,8. The interpretation of the term obstruction, used in four articles9-12 is not univocal because it does not allow distinction between biliary and intestinal obstruction. However, Hu et al.10 and Barnes et al.11 also reported the percentage relative to jaundice: probably they were referring to the gastroduodenal transit. Only 7 studies reported the average duration of the symptomatology, which varied from 1.4 to 8 months4,5,9-13 (Table 1)1-19. Site of the tumor The most frequent location of duodenal adenocarcinoma is the second portion of the duodenum (42.8-81.8%). Four articles used the topographic classification11,14-16: the most frequent site with this classification is peripapillary (50-62.9%). Two articles did not report data indicating the site of the tumor5,17. In some articles, the percentages obtained from the sum of the frequencies of the tumor in the proximal portion of the duodenum (D1-D2) and/or of the distal portion (D3 D4) were considered2,3,7,13,18. The percentages were respectively 58.8-81.8% and 22-41.2% (Table 1)1-19. Diagnosis The most frequently used examinations are esophagogastroduodenoscopy (31.8-100%) and duodenogra-

PRIMARY DUODENAL ADENOCARCINOMA

Table 1 - Number of patients, symptoms, site of tumor, examinations used for diagnosis and percentage of success Authors No. cases Most frequent symptom (%) Other symptoms Site of tumor % Diagnosis %

781

EGDS EGDS + Transit Sohn et al.1 Ryder et al.2 Bucher et al.3 Stell et al.4 Rose et al.5 Bakaeen et al.6 Sarela et al.7 Kaklamanos et al.8 Tocchi et al.9 Hu et al.10 Barnes et al.11 Delcore et al.12 Rotman et al.13 Santoro et al.14 Ohigashi et al.15 Cangemi et al.16 Hurtuk et al.17 Lowell et al.18 Scott-Coombes Williamson19 55 49 9 45 79 101 137 63 47 43 67 35 66 89 31 14 52 17 22 Pain (46) Hemorrhage (56) Anemia (31.1) Weight loss (59) Pain (54) Pain (44) Pain, obstruction (53.2) Jaundice (55.8) Pain (61) Anemia, jaundice D3-D4 D2 D1-D2 D2 29 61 66.7 55.5 73.5 77.8 42.8 78.7 44.1 55.9 51.4 48.5 62.9 IP 50 50

Transit+

CT

CT+

Jaundice Vomiting, jaundice Pain, jaundice Weight loss, jaundice D2 Weight loss, jaundice D1-D2 D2 Weight loss D3 D2 PP D2 D2 PP SP 50 PP

100 100 83

100 44.4 92

33.3

100

100

77.7

60

90

52

79

100 88.3 89 57.1 98.4 81.3

63.8 86.8 82 95 93.8 88 38.5

100 72.1 88 82.8 100 93.2

80.9 29 62.7 93 81.8 81.9 61.5

70.2 34.8 50

57.6 79.5 62.7

Pain, anemia Obstruction, weight loss Pain, obstruction (60) Hemorrhage Pain (59.1) Weight loss, vomiting Anemia (60.6) Vomiting, pain Pain (41.9) Anemia, jaundice Mass (23.1) Anemia, hemorrhage Pain, vomiting (46.1) Pain (58.8) Anemia, vomiting Pain (59) Jaundice, vomiting

43 67 51.5 100 33.7 38.5

D1-D2 D2

58.8 81.8

58.8 31.8

90 100

41.1 9.1

85.7 100

PP, peripapillary; SP, suprapapillary; IP, infrapapillary; EGDS, esophagogastroduodenoscopy; +, diagnosis performed with that type test; Transit, contrast radiography; CT, computerized tomography.

phy (33.3-100%). Other methods (computerized tomography [CT], magnetic resonance imaging [MRI], endoscopic retrograde cholangiopancreatography and angiography) are used to identify the stage and to decide the surgical strategy. Seven articles did not deal with this topic1,2,5,7,8,15,17. Data collected by CT and MRI are only indicative because most of the articles included cases collected over a period of time which also included periods in which these instruments did not yet exist (Table 1)1-19. Resectability and surgery The articles analyzed showed a resectability of 42.8%16 to 87%1. As regards curative surgery, pancreaticoduodenectomy was performed in 19.1%9 to 100%15 of cases. For some authors, 15%6 to 87.3%19 of pancreaticoduodenectomy interventions were pylorus preserving1,3,6,18,19. Duodenal segmentectomy was performed in 5%5 to 38%18 of cases. Some authors preferred to perform pancreaticoduodenectomy for tumors in the first and second portion of the duodenum, reserving duodenal segmentectomy for distal tumors1,7,11-13. Among the authors who reported palliative interventions, digestive bypass was performed the most often, with a percentage varying from 4.2%14 to 86.6%8 (Table 2)1-19.

Histopathology of the tumor The diameter of the tumor varied between 3 and 5 cm1,2,5,6,12,13,15,17,19. As regards the degree of differentiation, most of the authors preferred to group the different grades, and the grouping was not identical for all the authors. With these limitations, G1-2 was the most frequent grading (22.6-100%). Results regarding staging showed lymph node involvement was present in 25.8%13 to 76.2%14, pancreatic invasion in 22%2,13 to 81.8%10 and parietal infiltration in 38%6 to 81.4%10. Four articles reported positive resection margins (8.3-12.9%)1,2,4,19 (Table 3)1-19. Survival and follow-up Global survival at 5 years varied from 23%9 to 57%15. At 3 and 5 years for patients who underwent radical surgery, it varied from 47%2 to 61%6 and from 25%14 to 75%3, respectively. Three articles considered survival at 5 years based on the type of surgical intervention performed: for pancreaticoduodenectomy it varied from 30-69% and for duodenal segmentectomy from 0-60%1,8,12. Only two articles reported survival at 5 years based on the site: for D1 D2 and D3-D4 it varied from 32-67% and from 13-58%, respectively1,8. Survival at 5 years in patients with the ab-

782

M SOLEJ, S DAMICO, G BRONDINO ET AL

Table 2 - Resectability, main types of surgical intervention with radical and palliative intent Authors Resectability (%) PD Sohn et al.1 Ryder et al.2 Bucher et al.3 Stell et al.4 Rose et al.5 Bakaeen et al.6 Sarela et al.7 Kaklamanos et al.8 Tocchi et al.9 Hu et al.10 Barnes et al.11 Delcore et al.12 Rotman et al.13 Santoro et al.14 Ohigashi et al.15 Cangemi et al.16 Hurtuk et al.17 Lowell et al.18 Scott-Coombes Williamson19 87 63 83 56 53 67 72 58.7 53.2 65.1 61 82 69.7 73 77 42.8 67 76 77.3 73 87 33.3 75 48 30 78 67.6 19.1 25.6 75 60 82.6 41.7 100 28.5 77.1 17.6 Curative surgery % PPPD 77 22.2 DS 27 13 22.2 16.7 5 15 22 29.7 24 37.2 25 22.8 13 16.8 14.3 22.8 38.4 GJb 13 37 Palliative procedure % BDb GJb-BDb Bps/Lap

15

25 33 86.6 46.8 50 13.4 20 30 39 14.2 8.3

55 66.6 23 42.9 35.2 60

4.2

15 20.8 14.3 28.4 20

11.7

11.7 20

53.8 82.3

PD, pancreatoduodenectomy; PPPD, pylorus-preserving pancreatoduodenectomy; DS, duodenal segmentectomy; GJb, gastro-jejunal bypass; BDb, biliodigestive bypass; Bps/Lap, biopsy/exploratory laparoscopy/laparotomy.

Table 3 - Histopathology of tumor Authors Tumor (cm) Grading % Positive margins % 8 12.9 T2, 12.1; T3, 62.5; T4, 24.2 8.33 4 4 G2, 45; G3, 42 T1, 10; T2, 20; T3, 38; T4, 28 T% N+ % M % PI % PAI %

Sohn et al.1 Ryder et al.2 Bucher et al.3 Stell et al.4 Rose et al.5 Bakaeen et al.6 Sarela et al.7 Kaklamanos et al.8 Tocchi et al.9 Hu et al.10 Barnes et al.11 Delcore et al.12 Rotman et al.13 Santoro et al.14 Ohigashi et al.15 Cangemi et al.16 Hurtuk et al.17 Lowell et al.18 Scott-Coombes Williamson19

4.6 4.9

G1-2, 70; G3, 30 G2-3, 37

G,1 25.5; G2, 40.4 ; G3, 34 G1-2, 74.4; G3, 25.6 5 4.9 4.5 61.2%>3.5 G1, 25.7; G2, 54.3; G3, 20 G1-2, 86.4; G3, 1.5 G1, 62.5; G2-3, 37.5 G1-2, 100 G1, 22.6; G2, 61.3; G3, 16.1 G1-2, 82.35; G3, 17.64 3

T2, 29.8; T3, 40.4; T4, 29.8

T4, 89

56 44 63 58.33 38 32 43 48 21.3 50 36.1 63 25.8 76.2 58.3 48.4 29.41 59.1

14

22

64 41.66 71 38

21.3 81.8 31.43 22.7 54.1 81.4 69.2

12.9

T1, 19.4; T2, 16.1; T3, 32.3; T4, 32.3 T2, 5.9; T3, 47.1; T4, 47.1

T, tumor depth; N+, lymph node metastasis; M, metastasis; PI, pancreatic invasion; PAI, parietal invasion.

sence or the presence of lymph node involvement was 38%10 to 81%7 and 22%6 to 72%12, respectively. The mean follow-up reported by 10 authors varied from 24.2-104 months1-3,6,7,9,10,13,17,19 (Table 4)1-19.

Discussion Adenocarcinoma of the duodenum is the least common of the periampullary tumors1. The rarity of the dis-

PRIMARY DUODENAL ADENOCARCINOMA

Table 4 - Survival and follow-up Authors OS (%) 1 yr 3 yr Sohn et al.1 81 64 OS RS (%) OS PS (%) OS PD/DS (%) 3 yr 5 yr 1yr 69/0 OS N+ (%) 3yr 5yr 1yr 43 OS N(%) 3yr 5yr 66

783

OS site OS, Follow-up, (%) m m 5 yr D1-2, 67; D3-4, 13 49 36 68 104

5 yr 1 yr 3 yr 5 yr 1 yr 3 yr 5 yr 1 yr 47 90 47 53 43 75 60 54 71 37 51 54 54 59 50 45 25 50 20 0 13 91/76

24 Ryder et al.2 Bucher et al.3 Stell et al.4 Rose et al.5 Bakaeen et al.6 Sarela et al.7 Kaklamanos et al.8 Tocchi et al.9 Hu et al.10 Barnes et al.11 Delcore et al.12 Rotman et al.13 Santoro et al.14 Ohigashi et al.15 Cangemi et al.16 Hurtuk et al.17 Lowell et al.18 Scott-Coombes Williamson19

50 0 35 0 30/60 38 22 56 15 62 68 81 60 20

43

31 37 48 25 23 29 62 33 57

61

73

48 36 D1-2, 32; D3-4, 58 23.9 24.2 72

0 52/0

54 47 72

38 59

23 6.4 54

45 69

44 0 25 59

76 52 34

OS, overall survival; RS, radical surgery; PS, palliative surgery; PD/DS, pancreaticoduodenectomy vs duodenal segmentectomy; N+, lymph node metastasis; N-, negative lymph node metastasis.

ease makes the determination of factors affecting survival5 and the determination of the optimal form of management difficult4. Most of the articles found in the literature are composed of small personal series or retrospective multicenter studies with data collected over several decades in order to obtain sufficient numbers so as to be able to perform meaningful analyses4. The heterogeneity of the symptomatology found in the selected articles did not allow us to determine which and whether one symptom was prevalent with respect to the others. Within these limits, weight loss, abdominal pain, anemia and jaundice were the most frequent symptoms found. The incidence of jaundice varied considerably, up to 60%19. Hu et al.10 coined the term painful jaundice to contradistinguish it from painless jaundice observed in pancreatic cancer and carcinoma of the ampulla of Vater. The same author elaborated a classification of duodenal adenocarcinoma based on the symptomatology (1, icteric; 2, obstructive; 3, insidious), hypothesizing that type 1 tumors are in an advanced phase, as confirmed by other authors14. In spite of this, surgical treatment is not contraindicated. Some authors have shown how the nature of presenting symptoms did not appear to influence the resectability of the lesion4,10. Hurtuk et al.17 reported how the presenting symptoms can be indicative of advanced disease and invasion into nearby structures, explaining

why those patients with at least four symptoms had a worse outcome despite aggressive surgical resection. In the literature, differences in the clinical symptomatology depending on the site have not been reported. The only sign appears to be jaundice, more frequently seen in the periampullary forms, even though its frequency is not directly proportional to the incidence in this location. Rotman et al.13 stressed that the survival curve was better for patients with than for those without anemia, even though the difference was not statistically significant. We concluded that patients with anemia come to the attention of the doctor sooner. Bakaeen et al.6 found that weight loss was predictive of a worse survival. The nonspecific clinical presentation, which often mimics benign conditions, is the main pitfall for an early diagnosis of duodenal cancer9. The index of suspicion for the tumor remains very low, and the lack of specific symptoms or physical signs substantially delays diagnosis and treatment12. In fact, all the authors pointed out that on the average symptoms were present before the diagnosis for a minimum of 3 months, but in the most important case studies the duration of the symptomatology was around 6-8 months. Delcore et al.12 found that patients who had symptoms for 4 months or longer had a worse 2-year survival rate those who were symptomatic for less than 4 months.

784

M SOLEJ, S DAMICO, G BRONDINO ET AL

Notwithstanding the technological development of diagnostic instruments, endoscopy and duodenography remain the principal tests carried out and, if well integrated with other examinations, allow diagnosis and staging in 80-100% of the cases3,6,14. For predominantly anatomic reasons, these diagnostic techniques present limitations for the diagnosis of tumors located in the third and fourth portions of the duodenum. When there is clinical suspicion of adenocarcinoma in these sites, second-level examinations become useful (CT, MRI, ultrasonography). These examinations together with others of the third level (echoendoscopy, celiac axis arteriography, endoscopic retrograde cholangiopancreaticography, laparoscopy) allow in most cases preoperative staging to be established and the therapeutic strategy to be defined, whatever the site. Finally, the diagnostic study in the case of primary duodenal carcinoma can be subdivided into three phases: first phase, diagnosis of site and nature; second phase, staging; third phase, definition of treatment14. The largest areas of controversy with regard to survival for patients with primary duodenal adenocarcinoma are location of tumor, type of resection and some prognostic factors of which the most important is lymphatic spread (N+) (Table 5)1-19. The most frequent location of primary duodenal adenocarcinoma is the second portion. The significance of the primary location in regard to prognosis remains controversial (Table 5)1-19. Lowell et al.18 reported longer survival among patients with distal tumors, whereas Sohn et al.1 and Stell et al.4

reported longer survival in patients with proximal tumors. Other authors did not find statistically significant differences in survival based on site (Table 5), even though it is evident that there is a more favorable prognosis for distal duodenal adenocarcinoma. Patient selection and stage probably account for such differences. The hypothesis expressed for the difference in survival between proximal and distal duodenal adenocarcinoma also includes embryogenetic factors. The different embryological origins of the infrapapillary duodenum (midgut) and the suprapapillary and peripapillary duodenum (foregut) could be partly responsible for the aggressive behavior of tumors of the duodenum originating in the anterior intestine. Barnes et al.11 found that adenocarcinoma of the duodenum behaves similarly to colon cancer, whereas Rose et al.5 and Sarela et al.7 reported a similarlity to gastric cancer, in particular to gastric antral cancer. As regards therapy, the only treatment for duodenal adenocarcinoma which offers the possibility of a cure is excision by radical surgery14. There is general agreement that patients with resectable duodenal cancer have a longer survival than patients with unresectable tumors. Even today, the choice of the type of intervention is a topic for discussion. It is correlated above all to the possibility of carrying out an adequate lymphadenectomy and seems to be closely related to the location of the primary tumor. The most frequently performed interventions with a curative intent are pancreaticoduodenectomy and its variants (of these, the pylorus-preserving

Table 5 - Prognostic factors and statistical significance Authors Age Sex Symptom l symp Site ToR RvsNR CRvsPR PSvsNS T stage Nod G stat Margins Tumor P M size invasion SS NS SS Adjuv T

Sohn et al.1 Ryder et al.2 Bucher et al.3 Stell et al.4 Rose et al.5 Bakaeen et al.6 Sarela et al.7 Kaklamanos et al.8 Tocchi et al.9 Hu et al.10 Barnes et al.11 Delcore et al.12 Rotman et al.13 Santoro et al.14 Ohigashi et al.15 Cangemi et al.16 Hurtuk et al.17 Lowell et al.18 Scott-Coombes Williamson19

NS

NS

SS NS SS NS NS

SS NS

SS SS

SS SS

NS NS

NS SS

NS

SS SS NS NS NS

NS NS NS NS NS

SS (wt loss)

NS NS NS

SS SS SS SS NS

NS NS SS (gib) NS NS SS (jaundice) SS

NS NS

NS NS NS

NS NS NS

NS NS

NS NS NS NS

NS

SS (T4) SS NS NS SS NS NS NS NS NS

NS SS SS SS SS NS NS SS NS NS

NS NS NS NS NS NS SS NS NS NS

NS NS

NS NS NS

NS

SS NS NS NS SS SS SS

NS

NS

NS

NS

NS NS NS SS

SS SS

NS

SS (T4) NS

SS, statistically significant; NS, not statistically significant; symp, symptoms; l symp, duration of symptoms; ToR, type of resection; RvsNR, resection vs non resection; CRvsPR, curative resection vs palliative resection; PSvsNS, palliative surgery vs no surgery; nod stat, lymph node involvement; G, grading; P invasion, pancreatic invasion; Adjuv T, adjuvant therapy; gib, gastrointestinal bleeding.

PRIMARY DUODENAL ADENOCARCINOMA

785

pancreaticoduodenectomy is performed most often) and the duodenal segmentectomy. It has been suggested that duodenal segmentectomy represents a less morbid option9. However, the use of this procedure remains a debated subject. Analysis of the studies has highlighted a marked heterogeneity in the percentage of pancreaticoduodenectomies performed. Also these data are limited by the period of time in which the gathering of information occurred. As things stand at present, we believe that pancreaticoduodenectomy is the intervention of choice for tumors of the first, second and third portions of the duodenum in the absence of distant metastasis, even though our analysis does not allow us to make definite conclusions. Notwithstanding that in the past it was associated with elevated morbidity and mortality, pancreaticoduodenectomy has been considered the procedure that satisfies the principles of an adequate curative cancer operation1,3,7,15,17,19 because it also offers the theoretical advantage of en bloc resection with lymphadenectomy. According to the authors of the aforementioned papers, duodenal segmentectomy is inadequate for anatomical reasons. In fact, the mesentery of the duodenum appears short because it is fused in the retroperitoneum and dissection must be carried out close to the bowel. Barnes et al.11 also stated that lymphatic drainage of the duodenum is carried out not only by the superior mesenteric system but also by the inferior and celiac mesenteric districts, the last two being districts which are not removed during duodenal segmentectomy. Moreover, some primary operative steps such as the assessment of tumor resectability, search for an adequate cancer-free proximal margin and attainment of a tension-free anatomosis remain inadequate. Notwithstanding these criticisms, only Sohn et al.1 found a statistically significant difference in survival with pancreaticoduodenectomy with free margins and survival with duodenal segmentectomy (Table 5). Instead, supporters of duodenal segmentectomy9,10 state that it is an intervention with a curative intent when performed after an accurate selection of patients and they advocate its use for distal duodenal tumors. These authors affirm that duodenal segmentectomy allows complete removal of the tributary mesenteric lymph nodes of the respective segment of the duodenum to be resected. Kaklamanos et al.8 reported that the number of lymph nodes removed by duodenal segmentectomy did not differ from the number removed by pancreaticoduodenectomy. Tocchi et al.9 described a technique of duodenal segmentectomy associated with intestinal derotation for tumors of the third and fourth portions of the duodenum with satisfactory results in terms of survival. Rotman et al.13 instead justified the use of duodenal segmentectomy for tumors of the first and fourth portions of the duodenum. For tumors of the first portion, Santoro et al.14 advised a partial pancreaticoduo-

denectomy, because lymphadenectomy must include both tributary districts (celiac and mesenteric). As regards procedures with a palliative aim, biliodigestive bypasses are the most commonly performed surgical interventions. In a percentage which varies from 6%5 to 28%17, only one exploratory laparoscopy or laparotomy with biopsy was performed. Palliative procedures are not compatible with survival at 5 years. In the articles analyzed, lymph node involvement was present in 21.3%9 to 76.2%14. However, only 6 authors69,11,12 found a statistically significant difference in survival with respect to patients with negative lymph nodes. Of these, only 4 authors6-8,11 found a statistical significance in case series larger than 50 patients. Sarela et al.7 studied and compared the nodal stage-specific survival with that associated with gastric cancer, showing how the examination of >15 lymph nodes improved prognostic discrimination by the pN category. Although it is a well-known prognostic factor for other cancers of the gastrointestinal tract, with duodenal adenocarcinoma it is not possible to really demonstrate whether lymph node involvement is or is not a negative prognostic factor. The debate remains open also for the other prognostic factors (Table 5), and in the literature there is no unanimity in the results obtained. The lack of agreement about all of these issues probably reflects the limitation inherent in retrospective reviews and, because the disease is uncommon, the small number of patients in most case series2. Some authors1,2,5,6,9,11,17 have demonstrated that the T stage is correlated, with statistical significance, with a worse survival. In particular, Hurtuk et al.17 and Rose et al.5 observed a negative impact on survival for stage T4. In their series, Kaklamanos et al.8 reported a shorter survival in patients with transmural lesions, but T stage was not shown to be a significant predictor. Other authors7,10,12,13 did not report differences in survival based on parietal infiltration. Some authors analyzed pancreatic invasion as a factor in itself: only Ohigashi et al.15 and Hurtuk et al.17 found that invasion into the pancreatic parenchyma, even only at a macroscopic level, was a significant prognostic factor. Only two studies8,17 analyzed the influence of M stage on survival, with evidence of a shorter survival in M1 patients. Santoro et al.14 considered the presence of peritoneal carcinosis, distant metastasis and vascular invasion as contraindications to the exeresis of duodenal adenocarcinoma. Only two studies2,17 found that tumor size is an important predictor of survival. Ryder et al.2 demonstrated a correlation between the diameter of the tumor and survival: the greater the diameter, the shorter the survival. They explained this correlation by hypothesizing that large tumors have been present longer and have had more time to metastasize or invade adjacent structures. Interestingly, Hurtuk et al.17 found that large tu-

786

M SOLEJ, S DAMICO, G BRONDINO ET AL

mors (>3.5 cm) were less likely to be invasive and on the contrary small tumors (<3.5 cm) tended to be more aggressive than large ones. Ryder et al.2 found a statistically significant correlation between grading and survival. Such data were not confirmed in any other article. Among the authors1,5,6,10,17 who analyzed the positivity of the resection margins as prognostic factors, only Sohn et al.1 found a statistical significance. The analysis of survival at 3 and 5 years (Table 4) showed that the global survival for this type of tumor is better than those of other tumors of the peripapillary region. The mean follow-up was not long enough in some case series. It is evident that the patients treated with a curative surgical intervention had better survival, which in some case studies was as high as 75%3. In our analysis, it is evident that some factors such as the presence of lymph node metastasis reduce survival even if there was not always a corresponding statistical significance for such data. Finally, few articles1,2,6,7,11 dealt with the usefulness of chemotherapy and radiotherapy. Information in the literature concerning this approach is scarce and needs further evaluation. In conclusion, the absence of randomized clinical trials (not feasible at present because of the rarity of the tumor), the scantiness of cases in some articles, and the retrospective analysis (not reliable because the elevated possibility of overestimation) preclude a consensus process and the compilation of guidelines for the diagnostic-therapeutic procedure. In the literature, duodenal adenocarcinoma as a pathology in itself and not as a subgroup of periampullary tumors is beginning to be considered. As things stand at present, all patients who are medically fit to undergo surgery should be given the option of aggressive resection regardless of tumor size, tumor invasion or the appearance of positive lymph nodes. We believe that pancreaticoduodenectomy is at present the intervention of choice for tumors of the proximal portion of the duodenum in the absence distant metastasis. We reserve duodenal segmentectomy for tumors of the fourth portion and for selected patients. Hopefully, early diagnosis will correlate with improved long-term survival. Major advances in understanding the natural history of this disease and in developing new treatment strategies will only be made by large collaborative studies4.

References
1. Sohn TA, Lillemoe KD, Cameron JL, Pitt HA, Kaufman HS, Hruban RH, Yeo CJ: Adenocarcinoma of the duodenum: factors influencing long-term survival. J Gastrointest Surg, 2: 79-87, 1998.

2. Ryder NM, Ko CY, Hines OJ, Gloor B, Reber HA: Primary duodenal adenocarcinoma: a 40-year experience. Arch Surg, 135: 1070-1075, 2000. 3. Bucher P, Gervaz P, Morel P: Long-term results of radical resection for locally advanced duodenal adenocarcinoma. Hepatogastroenterology, 52: 1727-1729, 2005. 4. Stell D, Mayer D, Mirza D, Buckels J: Delayed diagnosis and lower resection rate of adenocarcinoma of the distal duodenum. Dig Surg, 21: 434-439, 2004. 5. Rose DM, Hochwald SN, Klimstra DS, Brennan MF: Primary duodenal adenocarcinoma: a ten-year experience with 79 patients. J Am Coll Surg, 183: 89-96, 1996. 6. Bakaeen FG, Murr MM, Sarr MG, Thompson GB, Farnell MB, Nagorney DM, Farley DR, van Heerden JA, Wiersema LM, Schleck CD, Donohue JH: What prognostic factors are important in duodenal adenocarcinoma? Arch Surg, 135: 635-642, 2000. 7. Sarela AI, Brennan MF, Karpeh MS, Klimstra D, Conlon KC: Adenocarcinoma of the duodenum: importance of accurate lymph node staging and similarity in outcome to gastric cancer. Ann Surg Oncol, 11: 380-386, 2004. 8. Kaklamanos IG, Bathe OF, Franceschi D, Camarda C, Levi J, Livingstone AS: Extent of resection in the management of duodenal adenocarcinoma. Am J Surg, 179: 37-41; 2000. 9. Tocchi A, Mazzoni G, Puma F, Miccini M, Cassini D, Bettelli E, Tagliacozzo S: Adenocarcinoma of the third and fourth portions of the duodenum: results of surgical treatment. Arch Surg, 138: 80-85, 2003. 10. Hu JX, Miao XY, Zhong DW, Dai WD, Liu W, Hu W: Surgical treatment of primary duodenal adenocarcinoma. Hepatogastroenterology, 53: 858-862, 2006. 11. Barnes G Jr, Romero L, Hess KR, Curley SA: Primary adenocarcinoma of the duodenum: management and survival in 67 patients. Ann Surg Oncol, 1: 73-78, 1994. 12. Delcore R, Thomas JH, Forster J, Hermreck AS: Improving resectability and survival in patients with primary duodenal carcinoma. Am J Surg, 166: 626-630, 1993. 13. Rotman N, Pezet D, Fagniez PL, Cherqui D, Celicout B, Lointier P: Adenocarcinoma of the duodenum: factors influencing survival. French Association for Surgical Research. Br J Surg, 81: 83-85, 1994. 14. Santoro E, Sacchi M, Scutari F, Carboni F, Graziano F: Primary adenocarcinoma of the duodenum: treatment and survival in 89 patients. Hepatogastroenterology, 44: 11571163, 1997. 15. Ohigashi H, Ishikawa O, Tamura S, Imaoka S, Sasaki Y, Kameyama M, Kabuto T, Furukawa H, Hiratsuka M, Fujita M, Hashimoto T, Hosomi N, Kuroda C: Pancreatic invasion as the prognostic indicator of duodenal adenocarcinoma treated by pancreatoduodenectomy plus extended lymphadenectomy. Surgery, 124: 510-515, 1998. 16. Cangemi V, Volpino P, Gentili S, Giuliani A, Di Martino M, Cangemi R, Clementi M, Piat GP: Carcinoma of the duodenum: management and survival in 14 cases. Panminerva Med, 39: 24-29, 1997. 17. Hurtuk MG, Devata S, Brown KM, Oshima K, Aranha GV, Pickleman J, Shoup M: Should all patients with duodenal adenocarcinoma be considered for aggressive surgical resection? Am J Surg, 193: 319-325, 2007. 18. Lowell JA, Rossi RL, Munson JL, Braasch JW: Primary adenocarcinoma of third and fourth portions of duodenum. Favorable prognosis after resection. Arch Surg, 127: 557560, 1992. 19. Scott-Coombes DM, Williamson RC: Surgical treatment of primary duodenal carcinoma: a personal series. Br J Surg, 81: 1472-1474, 1994.