Other
Gastrointestinal Imaging
July 2000

CT of Primary Cystic Pancreatic Neoplasms: Can CT be Used for Patient Triage and Treatment?

Abstract

OBJECTIVE. The purpose of this study was to determine whether CT can be used to distinguish serous cystadenomas from mucinous cystadenomas or cystadenocarcinomas of the pancreas and play an enhanced role in patient triage and treatment.
MATERIALS AND METHODS. A blinded retrospective analysis of CT scans from 50 patients with pathologically proven primary cystic pancreatic neoplasms was performed independently by three radiologists. Using classic CT criteria as reported in the literature, each tumor was categorized as definitely serous, mucinous, or indeterminate. Tumor location, size, presence of calcification, and size of largest cyst were recorded. Data for each reviewer were analyzed independently. Consensus data were then subjected to multivariate logistic regression analysis.
RESULTS. The ability of our reviewers to diagnose serous neoplasms ranged from 23% to 41%. Eight mucinous neoplasms were mistaken for serous tumors by two of the three reviewers. When consensus between at least two of the three reviewers was used for diagnosis, 27% of serous neoplasms were correctly diagnosed and all of the mucinous tumors were correctly identified as uncertain or mucinous, yielding the same clinical end point. For multivariate logistic regression analysis, a cyst smaller than 2 cm had a statistically significant association (p=0.005) with serous tumors, and the presence of peripheral tumoral calcification had a statistically significant association (p=0.01) with mucinous tumors.
CONCLUSION. There is significant variability in the CT appearance of serous and mucinous neoplasms of the pancreas, making CT an insensitive tool for differentiating these tumors. All tumors with peripheral calcifications were identified as mucinous neoplasms.

Introduction

With the advent of 1-sec and sub-second helical CT, the role of CT as the primary imaging study for a wide range of clinical symptoms and presentations has continued to expand. With its increased use it is not surprising that incidental, serendipitous, or iatrogenic conditions will be detected [1]. Numerous publications have addressed the incidental adrenal mass or renal cell carcinoma. Likewise, cystic pancreatic lesions are not infrequently found in asymptomatic patients who lack a history of current or prior pancreatic pathology [2,3].
By far, the most common cystic lesion of the pancreas is a pseudocyst. Although cystic pancreatic neoplasms can emulate pancreatic pseudocysts, a careful clinical history or correlation with laboratory values (amylase or lipase) will often enable a specific diagnosis. When the diagnosis of pancreatitis can be excluded, or is clinically unlikely, a primary cystic neoplasm must be considered. Traditionally, these tumors have been classified as cystadenomas or cystadenocarcinomas of the serous (microcystic) and mucinous (macrocystic) subtypes [4]. It is well documented that serous or microcystic neoplasms have no malignant potential and that the mucinous or macrocystic type may coexist with or become a cystadenocarcinoma [4, 5]. Early differentiation of primary cystic neoplasms from adenocarcinomas and other pancreatic malignancies is important because asymptomatic microcystic adenomas do not require surgical excision and macrocystic cystadenomas or cystadenocarcinomas require surgery [4,5,6,7,8]. In addition, even the overtly malignant mucinous cystadenocarcinomas carry a far better prognosis than solid ductal adenocarcinomas [4, 5, 9, 10].
The radiographic appearance of microcystadenomas and macrocystic adenomas or adenocarcinomas has been described in several small series [4, 6,7,8, 11,12,13]. The largest single-institution study to date, reported by Johnson et al. [8], includes CT findings in 35 patients with primary cystic pancreatic neoplasms. We present an analysis of the CT findings in 50 patients with cystic pancreatic masses examined at our institution to determine whether specific CT criteria permit the definitive diagnosis of these tumors and whether CT can play a role in patient triage to preclude unnecessary invasive procedures in patients with serous cystadenomas of the pancreas.

Materials and Methods

Using surgicopathologic records at our institution spanning 1984-1998, 250 cases of proven primary cystic pancreatic neoplasms were identified. Neoplasms such as mucin-secreting duct-ectatic adenocarcinomas and solid and papillary epithelial tumors (Hamoudi tumor) that have cystic characteristics but are not included in the classic description of primary cystic tumors were excluded. From these cases, 50 patients with serous or mucinous cystadenomas or cystadenocarcinomas had preoperative CT scans available for analysis. CT scans were performed on a variety of scanners, although most were helical. Forty-four CT scans were obtained with and six without IV contrast material. Collimation ranged from 3 to 10 mm and interscan spacing ranged from 3 to 8 mm.
Sixteen mucinous cystadenomas and 12 mucinous cystadenocarcinomas were present in 28 patients (16 women, 12 men), ranging in age from 24 to 86 years with a median age of 64 years. Twenty-two serous cystadenomas were present in 22 patients (17 women, five men), ranging in age from 28 to 84 years with a median age of 64 years.
Blinded retrospective analysis was performed independently by three experienced radiologists who analyzed each tumor using established CT criteria (Table 1) and then identified the tumor as a definite serous cystadenoma, a definite mucinous cystadenoma or cystadenocarcinoma, or indeterminate [4]. The reviewers were instructed to make a diagnosis as if this were a true clinical situation with the knowledge that all tumors identified as serous would not undergo biopsy or surgical excision. In addition, independent observations regarding the presence or absence of calcification, size of the largest cyst and overall mass, and location were recorded. Results for individual reviewers and a consensus of their observations requiring agreement of at least two of the three reviewers were compiled.
TABLE 1 CT Criteria for Distinguishing Serous from Mucinous Neoplasms
CT CriteriaSerousMucinous
Dimension of largest cyst<2 cm≥ 2 cm
Calcifications± central± peripheral
Metastatic diseasenoyes or no
Adenopathy
no
yes or no
Adapted from [4].

Results

Tumor Characteristics

Mucinous neoplasms.—Consensus data show tumoral calcification was present in eight mucinous neoplasms (29%), of which seven (88%) were peripheral in location (Table 2). Mucinous tumors ranged in greatest dimension from 1.5 to 15 cm. The largest cyst in each tumor was larger than 2 cm in 24 patients (86%). Tumors were septated in four patients (15%), and mural excrescences were seen in two patients (7%). One mucinous cystadenoma had aggressive characteristics with tumor extending into the adjacent psoas muscle and tracking into the patient's proximal thigh (Fig. 1A,1B).
TABLE 2 Number and Percentage of Cases Under Each Tumor Type for Each Characteristic Evaluated
CharacteristicSerous (n = 22)Mucinous (n = 28)
Calcium present4 (18)8 (29)
Central41
Peripheral07
Largest cyst present  
>2 cm8 (36)24 (86)
<2 cm
14 (64)
4 (14)
Note.—Numbers in parentheses are percentages.
Fig. 1A. —34-year-old woman who presented to emergency department with abdominal pain. Contrast-enhanced CT scans show aggressive, atypical mucinous cystadenoma extending into iliopsoas muscle (arrows). All reviewers identified this tumor as either uncertain or definitely mucinous.
Fig. 1B. —34-year-old woman who presented to emergency department with abdominal pain. Contrast-enhanced CT scans show aggressive, atypical mucinous cystadenoma extending into iliopsoas muscle (arrows). All reviewers identified this tumor as either uncertain or definitely mucinous.
Serous neoplasms.—Calcification was present in four serous tumors (18%), all of which were centrally located. Tumors ranged in greatest dimension from 1.4 to 27 cm. The largest cyst in each tumor was smaller than 2 cm in 14 patients (64%) (Table 2). One tumor had a stellate appearance, but no tumors showed papillary excrescences or internal debris.

Reviewer Analysis

Reviewer 1 correctly identified 32%, reviewer 2 identified 23%, and reviewer 3 identified 41% of the serous cystadenomas (Table 3). Although the ability to definitively diagnose serous neoplasms was low for reviewers 1 and 2, only one mucinous neoplasm was mistaken for a serous tumor by reviewer 2, and reviewer 1 correctly identified all mucinous lesions as mucinous or uncertain, yielding the same clinical endpoint. Reviewer 3 correctly identified more serous neoplasms than the other two reviewers but erroneously diagnosed seven mucinous tumors as serous cystadenomas.
TABLE 3 Number and Percentage of Each Tumor Type Classified by Each Reviewer
True Diagnosis: Serous (n = 22)True Diagnosis: Mucinous (n = 28)
ReviewerInterpretation
SerousMucinous or UncertainSerousMucinous or Uncertain
17 (32)15 (68)127 (93)
25 (23)17 (77)028 (100)
3
9 (41)
13 (59)
7
21 (75)
Note.—Numbers in parentheses are percentages.
To determine whether the opinions of multiple radiologists has an impact on diagnosis, a consensus reading was held in which a diagnosis for each tumor was determined by agreement between at least two of the three reviewers (Table 4). Of the 22 serous tumors, only six were correctly diagnosed by at least two of the three reviewers and only one was identified as unanimously serous. For eight of the serous tumors, all three reviewers were unable to exclude the diagnosis of a mucinous neoplasm and answered mucinous or uncertain (Fig. 2A,2B). All mucinous neoplasms were correctly categorized as mucinous or uncertain (yielding the same clinical end point of biopsy or surgery).
TABLE 4 Results Based on Consensus Data
ConsensusTrue Diagnosis
Serous (n = 22)Mucinous (n = 28)
Mucinous07 (25)
Uncertain16 (73)21 (75)
Serous
6 (27)
0
Note.—Numbers in parentheses are percentages.
Fig. 2A. —Two typical serous cystadenomas, one correctly and one incorrectly diagnosed by reviewers. 74-year-old woman whose primary complaint was early satiety and abdominal bloating. Contrast-enhanced CT scan shows serous cystadenoma composed of many cysts smaller than 2 cm and central tumoral calcifications correctly diagnosed by all reviewers.
Fig. 2B. —Two typical serous cystadenomas, one correctly and one incorrectly diagnosed by reviewers. 71-year-old woman with incidental finding on CT performed for unrelated symptoms. Contrast-enhanced helical CT scan shows serous cystadenoma of head of pancreas (arrow) measuring 2.5 cm and composed of multiple subcentimeter cysts. Compare this benign lesion with mucinous neoplasm shown in Figure 6 and note similarities. All three reviewers mistook this lesion for mucinous neoplasm.
Multivariate logistic regression was performed with true diagnosis (serous or mucinous) as the dependent variable and with the following independent variables: peripheral calcification (present or absent), cyst size (<2 cm or ≥2 cm), age, gender, overall mass size, and mass location (pancreatic head, body, or tail). In this analysis, there was a statistically significant association between peripheral calcification and mucinous tumors (p = 0.01). We also noted a statistically significant association between a cyst larger than 2 cm and mucinous tumors (p = 0.05). No other statistically significant associations were found in the multivariate analysis.

Discussion

Pancreatic cystadenomas and cystadenocarcinomas are often radiographically similar to other pancreatic processes such as benign pseudocysts or pancreatic tumors that have cystic changes (i.e., solid and papillary epithelial neoplasms, necrotic islet cell tumors, intraductal papillary mucinous tumors, and, occasionally, pancreatic ductal adenocarcinomas) [9, 14,15,16,17,18,19,20,21]. Early differentiation of cystic neoplasms from the previously mentioned entities and accurate classification as microcystic or macrocystic tumors are important because asymptomatic microcystic adenomas do not require surgical excision and macrocystic cystadenomas or cystadenocarcinomas do [4, 22, 23].
Preoperative attempts to definitively diagnose cystic tumors as serous or mucinous using invasive and noninvasive techniques have had limited success. Although CT has remained the primary imaging technique for evaluation of the pancreas, MR imaging and sonography are often useful in revealing and characterizing pancreatic lesions of all types but have not been successful at differentiating subtypes of cystic pancreatic masses [12, 13].
Cyst fluid aspiration with cytologic evaluation, measurement of viscosity, and analysis for tumor markers and enzymes has shown promise as an ancillary diagnostic tool [24,25,26]. A series of 18 patients with primary cystic pancreatic neoplasms who underwent a total of 23 fine needle aspiration biopsies using CT (n = 16), sonography (n = 3), or intraoperative (n = 4) guidance resulted in a definitive diagnosis in 50% of serous lesions and 75% of mucinous neoplasms [25]. However, complications such as inadvertent spillage of potentially malignant fluid into the peritoneum may occur. Additionally, aspiration of nonpancreatic cysts (i.e., mesenteric duplication cysts) or dilution of cyst fluid as a result of communication with the pancreatic duct or other structures may lead to false-positive or false-negative diagnoses. The retrospective study by Johnson et al. [8] evaluated 35 CT scans and 23 sonograms of 45 pathologically proven serous and mucinous pancreatic tumors. Unlike the study by Johnson et al., our reviewers were not as successful at differentiating a serous cystadenoma from a mucinous tumor in most cases. We were able to definitively diagnose serous tumors in 23-41% of serous cystadenomas by independent readings and 27% by consensus reading compared with 93% of serous tumors (14/15 tumors) seen on CT by Johnson's reviewers. This difference is, in part, a result of differences in study design. Although both studies used established CT characteristics to define serous and mucinous tumors, the study by Johnson et al. did not allow reviewers to defer a diagnosis if they believed they could not confidently exclude a mucinous tumor. The only deviation from serous cystadenoma or mucinous cystadenoma or cystadenocarcinoma allowed was to judge a mass “atypical” serous or “atypical” mucinous. By including an “uncertain or indeterminate” category, we believe this more closely simulates a true clinical situation in which the reviewers were asked to combine established criteria with their impression on the basis of their experiences and make a decision that would have a strong impact on patient treatment. The importance rests on whether one can preoperatively determine that the patient needs no further intervention, that is, to definitively diagnose a serous neoplasm with a high level of certainty. Despite more than 30 years of combined experience interpreting CT examinations, our reviewers were uncertain of the diagnosis in an average of 57% of patients.
It is important to identify when the radiographic findings are not specific enough to confidently exclude a mucinous lesion. Classically, microcystic adenomas are described as multilocular well-circumscribed masses that often contain a central stellate scar and calcifications. Innumerable small cysts ranging in size from 1 to 20 mm are filled with clear fluid and have a honeycombed appearance as a result of intervening connective tissue septa (Fig. 3A,3B). Mucinous cystadenomas and cystadenocarcinomas can be unilocular or multilocular masses filled with mucin, hemorrhage, and debris [4,5,6]. Cysts vary in size but are usually larger than 2 cm and frequently have internal papillary excrescences (Fig. 4). Based on our data, classic criteria were met by only a few primary cystic pancreatic neoplasms. Specifically, only three serous neoplasms had all of the typical characteristics that raise the possibility of a bias in case selection because a disproportionate number of “atypical” cystic neoplasms may have been referred for biopsy or surgical excision. However, other institutions have published similar experiences with “atypical” imaging features. In the study by Johnson et al. [8], 47% of serous lesions and 50% of mucinous cystadenocarcinomas were judged “atypical” on the basis of CT findings. More recently, a case series from the same institution described 10 of 18 primary cystic pancreatic neoplasms as atypical based on radiographic characteristics [25].
Fig. 3A. —Classic microcystic adenomas in two patients. Contrast-enhanced helical CT scan in 50-year-old woman shows microcystic adenoma (arrowheads). Note multiple small (<1 cm) cysts with tumoral calcification (arrows) in center of mass.
Fig. 3B. —Classic microcystic adenomas in two patients. 76-year-old woman with serous cystadenoma (arrow) incidentally noted on helical CT scan performed for cholecystitis. Mass is well defined and composed of many small (<5 mm) cysts.
Fig. 4. —Contrast-enhanced CT scan in 86-year-old man reveals classic appearance of mucinous cystad-enocarcinoma. Note multiple large cysts (>2 cm) with peripheral tumoral calcification (short arrow), solid components, and mural nodules (long arrow).
Our reviewers had greatest difficulty identifying serous tumors that contained cysts larger than 2 cm. Of the eight serous tumors with cysts larger than 2 cm, the consensus of the reviewers provided the correct diagnosis for only one. All reviewers answered uncertain or mucinous (Fig. 5) for five of the eight tumors. In addition, CT scans depicting tumors smaller than 3 cm were difficult to interpret. Twelve tumors (24%) (six mucinous and six serous) were 3 cm or smaller (Fig. 6). Two of three reviewers answered uncertain for the diagnosis of 10 of these tumors.
Fig. 5. —Contrast-enhanced CT scan in 38-year-old woman shows atypical serous cystadenoma with single large cyst measuring 9 cm. Note that there are no internal septations or peripheral calcifications to suggest mucinous neoplasm. All three reviewers identified this tumor as mucinous or uncertain.
Fig. 6. —Contrast-enhanced CT scan in 65-year-old woman shows atypical mucinous cystadenoma in head of pancreas. Mass (arrow) measures 1.8 cm and contains some small septations. Note similarities between this mucinous neoplasm and serous cystadenoma in Figure 2B. Two reviewers were uncertain of diagnosis and one reviewer mistakenly identified this neoplasm as serous.
The presence and location of tumor calcification is an important characteristic that can be helpful in arriving at an accurate diagnosis. None of the serous neoplasms in our study had peripheral calcifications, but 25% of the mucinous tumors did. Additionally, central calcifications were present in 10% of all tumors in this study (5/50), of which 80% were serous neoplasms. A potential pitfall is misinterpretation of vascular or ductal calcifications as tumoral in origin. One reviewer mistook a serous cystadenoma with heavy splenic artery calcifications as a mucinous neoplasm with peripheral tumoral calcification (Fig. 7A,7B).
Fig. 7A. —CT scans with IV contrast material in 77-year-old woman. Serous cystadenoma of tail of pancreas shows heavy splenic artery calcifications (arrow).
Fig. 7B. —CT scans with IV contrast material in 77-year-old woman. Splenic artery calcifications may be confused with tumoral calcifications (arrow) at periphery of lesion.
It is clear that CT provides limited assistance in the differentiation between serous and mucinous neoplasms. Even when pseudocysts and other pancreatic pathology easily confused with serous and mucinous neoplasms were eliminated and reviewers were provided with this information, they had great difficulty distinguishing between the two groups. Because of the variability in the radiographic appearance of serous cystadenomas and overlap in CT characteristics with mucinous neoplasms, most serous neoplasms still require ancillary testing such as biopsy to reach a definitive diagnosis.
However, we have also verified the finding of Johnson et al. [8] that the combined opinion of three radiologists improves accuracy. Despite our low sensitivity for identifying serous tumors, no mucinous neoplasm would be left untreated as a result of the consensus interpretation, and invasive procedures may have been avoided in 27% of patients. Although more data is needed on a greater number of patients, this case series shows significant variability in the imaging characteristics of serous cystadenomas and mucinous neoplasms of the pancreas, showing CT to be an insensitive tool for differentiating these tumors.

Footnote

Address correspondence to E. K. Fishman.

References

1.
Smith PA, Fray FM, Fishman EK. Spiral computed tomography evaluation of the kidneys: state of the art. Urology 1998; 51:3-11
2.
Bosniak M.A. Observation of small incidentally detected renal masses. Semin Urol Oncol 1995; 13:267-272
3.
Newell-Price J, Grossman A. Adrenal incidentaloma. Postgrad Med J 1996; 72:207-210
4.
Compagno J, Oertel JE. Mucinous cystic neoplasms of the pancreas with overt and latent malignancy (cystadenocarcinoma and cystadenoma): a clinicopathologic study of 41 cases. Am J Clin Pathol 1978; 69:573-580
5.
Buetow P, Rao P, Thompson L. Mucinous cystic neoplasms of the pancreas: radiologic-pathologic correlation. RadioGraphics 1998; 18:433-449
6.
Friedman AC, Lichtenstein JE, Dachman AH. Cystic neoplasms of the pancreas. Radiology 1983; 149:45-50
7.
Itai Y, Moss AA, Ohtomo K. Computed tomography of cystadenoma and cystadenocarcinoma of the pancreas. Radiology 1982; 145:419-425
8.
Johnson CD, Stephens DH, Charboneau JW, Carpenter HA, Welch TJ. Cystic pancreatic tumors: CT and sonographic assessment. AJR 1988; 151:1133-1138
9.
Loftus EV, Olivares-Pakzad BA, Batts KP, et al. Intraductal papillary-mucinous tumors of the pancreas: clinicopathologic features, outcome and nomenclature. Gastroenterology 1996; 110:1909-1918
10.
Madura JA, Wiebke EA, Howard TJ, et al. Mucin-hypersecreting intraductal neoplasms of the pancreas: a precursor to cystic pancreatic malignancies. Surgery 1997; 122:786-793
11.
Wolfman NT, Ramquist NA, Karstaedt N, Hopkins MB. Cystic neoplasms of the pancreas: CT and sonography. AJR 1982; 138:37-41
12.
Mergo P, Helmberger T, Buetow P, Helmberger R, Ros P. Pancreatic neoplasms: MR imaging and pathologic correlation. RadioGraphics 1997; 17:281-301
13.
Minami M, Itai Y, Ohtomo K, Yoshida H, Yoshikawa K, Iio M. Cystic neoplasms of the pancreas: comparison of MR imaging with CT. Radiology 1989; 171:53-56
14.
Buetow PC, Parrino TV, Buck JL, Pantongrag-Brown L, et al. Islet cell tumors of the pancreas: pathologic-imaging correlation among size, necrosis and cysts, calcification, malignant behavior and functional status. AJR 1995; 165:1175-1179
15.
Hertzanu Y, Bar-Ziv J, Freund U. Computed tomography of unusual calcified pancreatic tumors. J Comput Assist Tomogr 1989; 13:75-76
16.
McDowell RK, Gazelle GS, Murphy BL, et al. Mucinous ductal ectasia of the pancreas. J Comput Assist Tomogr 1997; 21:383-388
17.
Nino-Murcia M, Olcott EW, Jeffrey RB. Dualphase helical CT of locally invasive pancreatic adenocarcinoma. J Comput Assist Tomogr 1998; 22:282-287
18.
Ohtomo K, Furui S, Onoue M, et al. Solid and papillary epithelial neoplasm of the pancreas: MR imaging and pathologic correlation. Radiology 1992; 184:567-570
19.
Procacci C, Graziani R, Bicego E, et al. Intraductal mucin-producing tumors of the pancreas: imaging findings. Radiology 1996; 198:249-257
20.
Stafford-Johnson DB, Francis IR, Eckhauser FE, Knol JA, Chang AE. Dual-phase helical CT of nonfunctioning islet cell tumors. J Comput Assist Tomogr 1998; 22:335-339
21.
Van Hoe L, Baert AL. Pancreatic carcinoma: applications for helical computed tomography. Endoscopy 1997; 29:539-560
22.
Hodgkinson DJ, Remine WH, Weiland LH. Pancreatic cystadenoma: a clinico-pathologic study of 45 cases. Arch Surg 1978; 113:512-519
23.
ReMine SG, Frey D, Rossi RL, Munson JL, Braasch JW. Cystic neoplasms of the pancreas. Arch Surg 1987:122:443-446
24.
Lewandrowski K, Lee J, Southern J, Centeno B, Warshaw A. Cyst fluid analysis in the differential diagnosis of pancreatic cysts: a new approach to the preoperative assessment of pancreatic cystic lesions. AJR 1995; 164:815-819
25.
Carlson SK, Johnson CD, Brandt KR, Batts KP, Salomao DR. Pancreatic cystic neoplasms: the role and sensitivity of needle aspiration and biopsy. Abdom Imaging 1998; 23:387-393
26.
Ros P, Hamrick-Turner J, Chiechi M, Ros L, Gallego P, Burton S. Cystic masses of the pancreas. RadioGraphics 1992; 12:673-686

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 99 - 103
PubMed: 10882255

History

Submitted: March 5, 1999
Accepted: December 15, 1999

Authors

Affiliations

Charlene A. Curry
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
John Eng
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Karen M. Horton
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Bruce Urban
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Stanley Siegelman
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Brian S. Kuszyk
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.
Elliot K. Fishman
All authors: Department of Radiology, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media