Castleman’s disease was described for the first time in 1954 by Benjamin Castleman, a pathologist at Massachusetts General Hospital, as an uncommon lymphoproliferative disorder and subsequently in 1956 as a benign, localized thymoma-like enlargement involving hyperplastic lymph nodes in the anterior mediastinum [
7,
8]. Earlier synonyms of CD included “angiofollicular lymph node hyperplasia”, “giant cell lymph node hyperplasia”, “follicular lymphoreticuloma”, “lymphoid hamartoma” and “benign lymphoma”.
Incidence and classification
The prevalence of CD has not been estimated, but it has been calculated that the number of cases in the United States ranges from 30.000 to 100.000 [
9]. Its incidence rate has not been reported in literature, although CD appears to be more common in the Asian population [
10].
A commonly used system to classify the heterogeneity of CD was proposed by McCarty et al. in 1955 [
11]. Based on clinical and radiological characteristics, CD can be classified as unicentric (unifocal) or multicentric (MCD) form, depending on the number of lymph nodes involved.
UCD represents the most common form (>90%) of CD and is asymptomatic in over half of cases. Sometimes, when the lesion is large enough, compressive or constitutional symptoms may be present. It tends to occur in the third and fourth decade of life with a slight female predominance with a median age of 35 years [
12,
13]. The age of the patient reported in this case is in line with the average age of all other patients with UCD reported in the literature.
Pathologic mechanism
The pathophysiological basis of Castleman’s disease is still unclear. However, chronic low-grade inflammation, immunodeficiency status and dysregulation autoimmunity have been proposed as likely pathogenic mechanisms. The critical role of inflammatory mediators such as interleukin 6 (IL-6) or interleukin 10 (IL-10) and human herpes virus 8 (only in Multicentric variant) has been well demonstrated in preclinical animal models [
14]. Dysregulation and overexpression of IL-6 stimulate hepatocytes to produce acute phase proteins which increase the levels of the hepcidin hormone, which correlates with anemia. IL-6 also stimulates B-cells and blood vessel proliferation promoting the overexpression of the vascular endothelial growth factor and the neoangiogenesis. Interestingly, a recent study has demonstrated that hyaline-vascular Castleman’s disease is often a monoclonal proliferation, consisting most likely of lymph node stromal cells [
15].
Histological features
CD can be classified into three histopathological patterns: a hyaline-vascular (HV) type, a plasma cell (PC) type and a mixed variant. Usually it is the HV type that represents 80–90% of cases and appears more frequently as unicentric localization (UCD) whereas the PC type is mostly multicentric (MCD) and accounts for only 10–20% of cases.
In the HV variant, lymph nodes involved in the disease, show increased numbers of lymphoid follicles that exhibit features of “regression”: a term referring to a predominance of dendritic cells relative to lymphocytes within germinal centers and consequent rearrangement of mantle zones, known as an “onion ring pattern”. Also, an increased number of small hyalinized vessels between and within follicles, named “lollipop follicles”, results in obliteration of medullary sinuses. In the unicentric localization the average size of lymph nodes is very wide, ranging from 1 to 12 cm. The lesion size reported in this case is consistent with those reported in literature.
Clinical manifestations
UCD usually is identified without symptoms at diagnosis and can be discovered incidentally in chest or abdominal-pelvic imaging performed for other reasons. The patients may present symptoms related to the compression of adjacent organs. Dyspnea, cough, hemoptysis, and chest pain can be present in thoracic disease whereas vomiting, postprandial discomfort and abdominal or lumbar pain in abdominal-retroperitoneal disease [
17‐
19]. Therefore, because there are no specific symptoms and clinical presentation can vary greatly, a diagnosis of UCD based only upon clinical features is difficult.
In our case the patient only complained about abdominal discomfort. Laboratory studies show normal levels of cytokines (C-reactive protein, IL-6) in the absence of anemia or thrombocytopenia with a normal T and B cell count.
Diagnosis
UCD diagnosis is based on clinical evaluation, which includes patient history, laboratory and radiological findings, but a final diagnosis can be achieved only by careful histological and immunohistochemical examination. Therefore, preoperative diagnosis is often not achievable. The laboratory evaluation of patients with UCD includes a complete cell blood count and metabolic panel, inflammatory markers, albumin and Human Immunodeficiency Virus (HIV) and Human Herpes Virus (HHV)- 8 tests. Plasmatic elevated levels of cytokines such as IL-6 and IL −10, can lead to its diagnosis but are not routinely recommended in clinical practice. As our patient had no abnormal laboratory tests, it was difficult to make the diagnosis of UCD on the basis of these analyses alone. Moreover, a circulating HHV-8 search resulted negative in our patient.
The UCD diagnosis of certainty is usually obtained by performing an excisional biopsy of the pathological lymph nodes. In the case of a less accessible disease core needle biopsy (FNAB) is preferred to fine needle aspiration (FNAC), which is commonly not diagnostic. In fact, a preoperative FNAB or FNAC is not recommended because of the difficulty of achieving an adequate amount of tissue, the possibility of spreading tumor cells and the risk of severe bleeding in hypervascular mass. In all cases reviewed in the literature, all authors performed a preoperative Computed Tomography (CT) scan, often proceeded by an Ultrasonography (US) and followed by Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET) or Fine Needle Aspiration Biopsy (FNAB). A preoperative diagnosis of CD was not suspected in any of the cases [
20,
21]. This is consistent with our clinical case in which a definitive preoperative diagnosis was not obtained. In particular, although endoscopic or ultrasound-TC guided fine needle biopsy is recommended by many authors, severe bleeding risk in hypervascular mass should be taken into account. Based on all these considerations, we do not perform a preoperative cytological diagnosis due to the risk of bleeding.
Although UCD is a not a malignant condition, different malignancies and other diseases have been associated with it.[
22]. Non-Hodgkin lymphoma and amyloidosis have been reported in approximately 18% of patients with MCD, as well as in patients with UCD [
23]. Paraneoplastic pemphigus is also associated with UCD in about 20% of cases and characterized by an increased risk of lymphoma [
24]. Lymphoma, lymph node metastasis, paraganglioma, gastrointestinal stromal tumor (GIST), ectopic pheochromocytoma, leiomyoma and leiomyosarcoma, liposarcoma, fibrosarcoma can be included in differential diagnosis with mesenteric UCD, especially for female patients [
25]. Moreover, the differential diagnosis must be performed with other causes of lymphadenopathy such as tuberculosis, luetic lymphadenitis, abscess, sarcoidosis, HIV and toxoplasmosis. Because the radiological findings for UCD are unspecific, the preoperative radiologic differential diagnosis of mesenteric disease most commonly includes hypervascular mesenchymal tumors such as GIST, neurogenic tumors such as ectopic pheochromocytoma, carcinoids or pancreatic cancer. Due to the face that many patients show very similar radiologic features, a differential diagnosis is very difficult to arrive at.
Unfortunately, the characteristics detectable for diagnostic tools (US: CT scan, MRI or PET) are not conclusive for CD even if Malara et al. described in detail the US and CT features of mesenteric UCD [
26]. Most cases of abdominal UCD cannot be visible on radiographs unless they are massive or have calcifications. Abdominal UCD usually presents as a homogeneus and hypoechoic solitary mass by US. In contrast, abdominal US of our patient showed heterogeneity of the mass, perhaps due to its large size. Homogeneity with intense contrast enhancement reflecting hypervascularity of the lesion is a characteristic finding at CT of abdominal UCD. Mesenteric UCD commonly appears at CT as a well-defined single mass of soft tissue without satellite nodules or surrounded by normal lymphadenopathy [
27,
28]. UCD usually results positive on fluorodeoxyglucose PET [
29].
Therapy
The standard treatment for UCD regardless of histological type (whether HV or PC), is a complete “en bloc” surgical resection, which is a curative approach in almost all cases without recurrence after 20 years of follow up [
30]. A subtotal resection presents a low recurrence rate and can be cured by re-excision. In Table
1 we report on all cases of abdominal UCD treated laparoscopically which have been published in the literature [
31‐
42]. In five cases the disease was localized in extralymphatic tissues such as pancreas, liver, spleen and the adrenal gland. In these cases, pancreatic cancer, splenic abscess, an accessory spleen, hepatocellular carcinoma and pheochromocytoma were suspected preoperatively. In contrast, lymphatic tissue localizations were defined preoperatively as adnexal mass, lymphoma or metastatic disease. Our preoperative diagnosis was consistent with that reported by Ohta et al. who performed a laparoscopic ileal resection suspecting a GIST localization [
35]. Our case was resolved without bowel resection because of the presence of an adequate dissection plane.
Table 1
Studies about laparoscopic treatment of abdominal castelman’s disease
1. | | Pelvic | F | None | CT, TVUS | Adenexal Mass | None | Single-Port Laparoscopic Mass Excision | 7-cm HV Type |
2. | | Liver VI Segment | F | Epigastric Pain | US, CT, MRI, PET, EGDS, Colonoscpy | HCC | None | Laparoscopic Assisted Right Lobectomy | 2-cm HV Type |
3. | | Hepatoduodenal Ligament | F | Right Quadrant Pain | CT, MRI, SA | Exophytic HCC | None | Totally Laparoscopic Resection | 3-cm HV Type |
4. | | Omentum | F | None | CT, MRI | | Fibrinogen | Totally Laparoscopic Resection | 3-cm PC and HV Type |
5. | | Jejunal Mesentery | F | None | US, CT, MRI | Duodenal Gist | None | Laparoscopic Assisted Resection | 7-cm HV Type |
6. | | Spleen | M | Abdominal Pain, Fever, Diarrhea | CT | Lymphoma, Splenic Hamartoma or Abscess | CRP, ESR | Laparoscopic Splenectomy | 7-cm HV Type |
7. | | Pancreas | F | Epigastric pain | CT, EUS, FNAB | Gastric GIST, Pancreatic Tumour | None | Laparoscopic Pancreatic Resection | 4-cm HV Type |
8. | | Right Retroperitoneal Mass | M | Nausea and Vomiting | EUS biopsy; CT-Pet | Lymphoma, Metastatic Disease, Extra-Adrenal Pheocromocytoma, Testicular Cancer. | None | Totally Laparoscopic Resection | 5,5 cm HV Type |
9. | | Posterior Surface of Abdominal Wall | M | Palpable Mass | CT | | None | Totally Laparoscopic Resection | 5 cm PC Type |
10. | | Lower Splenic Pole | M | Recurrent Palpitation and Vague Abdominal Pain | US, CT | Accessory spleen | None | Totally Laparoscopic Resection | 5-cm HV Type |
11. | | Right Adrenal Gland | M | None | US, CT | Adrenal Gland, Pheocromocytoma | None | Laparoscopic Adrenalectomy | 4,5 cm HV Type |
12. | | InterAorto-Caval Mass | F | Anemia | CT, FNAB | | | Converted Laparoscopy | 6.7-cm PC and HV Type |
As shown in Table
1, all the cases of abdominal UCD treated with LA were completed laparoscopically, with the exception of one [
42]
. In this case, the mass was adherent to the cava vein and so the authors converted the procedure to obtain safer vascular control. In the other cases, surgeons performed a mass removal laparoscopically or an "en bloc" resection of the organ in which was it contained. All procedures were bloodless. No other intraoperative or postoperative complications occurred and patients were discharged earlier (range 1–5 days). Based upon this positive experience all the Authors concluded that laparoscopy could be a safe and effective procedure for the treatment of UCD.
We opted for a LA to ensure the patient the typical benefits of the technique. Usually, we remove the specimen through a Pfannenstiel incision. In this case, both for the size of the lesion and for the presence of a previous umbilical incision, we opted for a xipho-umbilical incision as reported in the literature for more complex Gastric procedures.
The literature review suggests that radiotherapy can be a more favorable treatment to UCD than invasive surgical resection with a minimal complication rate and good prognosis [
43,
44]. Complete clinical and radiologic resolution of UCD is consistently documented in other articles. Intensity-modulated radiation therapy is better than three-dimensional conformal therapy due to its reduction of the dose gradient and toxicity to the surrounding normal tissue [
45]. De Vries et al. demonstrated that neoadiuvant radiotherapy used to downsize advanced unresectable UCD in order to achieve a radical excision could be a possible strategy of treatment [
26].
When surgical resection and radiotherapy are impossible, partial resection followed by clinical observation alone may be useful and can result in a lengthy remission.
In conclusion, although a rare disease, UCD should always be considered when a solid asymptomatic abdominal mass is incidentally found. The pelvis and retroperitoneum are USDs most frequent sites, and a correct pre-operative study and surgical timing can lead the patient to a full recovery. Moreover, based upon our experience we retain that a laparoscopic approach leads to better results than open surgery as it reduces postoperative pain and limits the length of hospital stay. In cases of an uncertain nature mass, LA must be considered as the last diagnostic tool and the first treatment one.