Peritoneal Metastases from Appendiceal Cancer

INTRODUCTION

Appendiceal neoplasms are diagnosed in approximately 1% of all appendectomy specimens.¹ The early symptoms of appendiceal cancer may be nonspecific, or they may mimic the clinical picture of appendicitis. Not surprisingly, most patients are diagnosed incidentally during surgical exploration or late when peritoneal or systemic dissemination has already occurred, as colonoscopy rarely will diagnose an appendiceal cancer.

The common pathway of all appendiceal tumors regardless of grade and cell of origin involves invasion of the appendiceal wall, luminal obstruction, and perforation with subsequent dissemination of malignant epithelial cells throughout the peritoneal cavity (Figs. 1 and 2). Systemic/extraperitoneal metastases are distinctly unusual for appendiceal mucinous lesions. The subsequent course of disease for a mucinous lesion depends on the grade of appendiceal primary as defined by Bradley and colleagues.²

Fig. 1.

Ruptured appendiceal mucocele.

Fig. 2.

Eviscerated omental metastatic deposits form a low-grade appendiceal primary.

Patients with low-grade mucinous tumors (LGA) will typically progress slowly over years and can possibly develop pseudomyxoma peritonei, which describes accumulation of mucinous ascites within the peritoneal cavity. Approximately 7% of the low-grade lesions have lymph node involvement, and up to 16% will dedifferentiate into higher-grade lesions during the course of the disease.^3–5

High-grade lesions (HGA) are more likely to metastasize systemically, resulting in a poorer prognosis. The variability of HGA clinical presentation stems from the variability in biologic behavior and grade of the primary lesions: moderately differentiated, poorly differentiated, as well as signet ring cell histologies that are all included in the HGA group.²

Pseudomyxoma peritonei (PMP) is a descriptive term, referring to the presence of muscinous ascites. Mucinous ascites is by for most commonly associated with mucinous appendiceal neoplasms. However, pseudomyxoma can be produced by several primary tumor types, including appendiceal, colon, ovarian, mucinous pancreatic, and low-grade urachal primaries among others. The prognosis of patients with PMP of nonappendiceal origin depends on the primary tumor type, while the surgical selection for cytoreductive surgery (CRS)/hyperthermic intraperitoneal chemotherapy (HIPEC) for non-LGA PMP patients follows different selection criteria.

Not every appendiceal primary is associated with mucin production or ascites. Patients often present with solid peritoneal disease that has no phenotypic difference from any other gastrointestinal malignancy with peritoneal dissemination. In addition, not every PMP or appendiceal cancer is associated with long-term survival. Such factors as histologic grade, tumor biology of the primary lesion, age, functional status, and extent of disease at the time of diagnosis determine the disease-free survival and overall survival of these patients.

The authors’ group’s approach to peritoneal dissemination from appendiceal tumors has been optimal CRS with the goal of removal of all gross disease if feasible. This typically entails selective peritonectomy and multivisceral resection followed by HIPEC. LGA primaries will be treated with CRS/HIPEC without systemic chemotherapy, while HGA primaries usually are treated with upfront systemic chemotherapy prior to CRS/HIPEC. Selected patients (approximately 10%) who recur may again be candidates to undergo repeat CRS with HIPEC as dictated by their performance status, clinical staging, and symptoms.⁶

Preoperative Patient Evaluation

Selection criteria

Appropriate patient selection is of paramount importance in the management of patients with PSD. All patients presenting to the authors’ multidisciplinary clinic have a complete history and physical examination followed by imaging with either MRI of the abdomen and pelvis or computed tomography (CT) of the chest, abdomen, and pelvis (with oral and intravenous contrast). The authors also will obtain baseline blood counts renal and hepatic functions as well as tumor markers including CEA, CA19–9, and CA125. All patients undergo pathologic review of previous biopsy or resected tissue.

In general, the authors use the following eligibility criteria for any patient presenting with documented peritoneal surface malignancy:

1. Patients should be medically with electrocorticography (ECOG) performance status of no more than 2

2. Absence of metastatic disease outside the abdomen

3. The primary lesion is resectable (or has been previously resected)

4. The peritoneal disease is resectable

5. Parenchymal hepatic metastases if present are easily resectable

6. There is no bulky retroperitoneal disease

In cases of appendiceal cancer specifically, the authors’ selection criteria are modified based on the grade of the appendiceal primary.

For low-grade appendiceal cancer a cytoreduction is attempted regardless of the volume of disease. LGA primaries progress as sclerotic metastases and as space-occupying lesions within the peritoneal cavity, which are uniformly fatal if left untreated, typically resulting in death from bowel obstruction. LGA tumor biology is typically indolent, and even in cases of incomplete cytoreduction patients receive the benefit of symptomatic control and improved overall survival that can often be measured in years.⁷ In addition a multi-institutional retrospective review of 2298 patients with LGA with peritoneal carcinomatosis index (PCI) from 31 to 39 had a 10-year survival of 68% when a complete cytoreduction was achieved.⁸ However, even if not completely debulkable, there is frequently benefit to cytoreduction of low-grade lesions. The decision to proceed with heated intraperitoneal chemotherapy after incomplete cytoreduction depends on the volume of residual disease and the amount of ascites. In general, patients with voluminous ascites are perfused, because HIPEC with CRS will control the production of ascites in approximately 90% of the patients.⁹ In patients without symptomatic ascites, but with excessive post-CRS residual disease, the perfusion is aborted.

In high-grade appendiceal primaries, the PCI along with the specific distribution of the peritoneal disease is taken into consideration before proceeding with cytoreduction. Patients with imaging showing disease not amenable to complete cytoreduction are not taken into the operating room for CRS. These patients are treated with systemic chemotherapy followed by restaging imaging, potentially supplemented with a laparoscopic exploration, to evaluate resectability. Generally, HGA patients are preferably treated with upfront chemotherapy with the possible exception of the patient who presents with limited metastatic disease (PCI <10) that is amenable to nonmorbid resection. In case of postchemotherapy presence of ascites, these patients will be evaluated with diagnostic laparoscopy to determine resectability. Ascites, hydronephrosis, and bowel obstruction are regarded as signs of inability to achieve complete CRS due to volume and distribution of disease.^5,9,10 Such patients may be candidates for palliative procedures (such as stoma and/or gastrostomy tube placement), followed by second-line systemic therapy.

Preoperative imaging

All patients have a thin-cut contrast-enhanced CT of the thorax, abdomen, and pelvis or MRI, ideally within 30 days of the scheduled operation.

CT is the authors’ modality of choice for low-grade appendiceal malignancies to obtain a rough estimate of the distribution of disease and avoid surprises of possible extra-abdominal involvement. Even though MRI with gadolinium has increased sensitivity in identifying smaller peritoneal implants, the authors believe that this additional information has no impact on the clinical decision-making algorithm for low-grade appendiceal patients. In addition, it increases the cost of the preoperative evaluation and patient discomfort. The authors use MRI more frequently for the long term follow-up of CT-negative patients who have elected not to proceed with CRS/HIPEC or surveillance of completely cytoreduced patients. MRI with dilute oral barium and delayed enhanced intravenous gadolinium for mucinous appendiceal lesions when compared with intraoperative findings has been shown to be superior to CT scan in detecting peritoneal metastasis, with a sensitivity of 82% to 89%.¹¹

The strength of the CT is its fundamental ability to detect anatomic details and differences in tissue density. Unfortunately, in peritoneal carcinomatosis, one often encounters subcentimeter lesions spread in a carpet- or plaque-like fashion. Comparison of intraoperative findings with CT findings showed a CT scan sensitivity between 25% and 37%, with a negative predictive value that ranged between 47% and 51%, while the CT sensitivity for lesions less than 1 cm was between 9% and 24%.¹² In a similar study, the false-negative rate for the CT to detect small bowel lesions was 60%.¹³ Therefore, the authors explain to their patients that what is seen in the preoperative imaging is rarely what one gets in the operating room. There is also significant variability among radiologists in the interpretation of the extent of peritoneal carcinomatosis, making it advantageous for the surgeon who treats PSD to develop expertise in the interpretation of abdominal imaging.⁸ Despite thorough preoperative imaging, approximately 5% to 10% of patients are deemed not to be operative candidates on exploration.

Positron emission tomography (PET) has approximately 10% sensitivity in low-volume peritoneal carcinomatosis.¹⁴

For LGA primaries specifically, PET offers no additional information given the low proliferation index of these lesions and the almost universally false-negative or indeterminate reading of the final examination. For HGA primaries, PET has a place in cases of debatable disease outside the peritoneal cavity or in evaluating specific peritoneal lesions for local recurrence. PET with or without CT has serious limitations in predicting the extent of carcinomatosis, and is rarely if ever obtained in the authors’ institution.¹⁵

Colonoscopy for Appendiceal Tumors

Colonoscopy itself for identification of appendiceal cancer is diagnostic less than 5% of the time, with most endoscopists describing a smooth, submucosal cecal mass at the appendiceal orifice with or without free-flowing intraluminal mucin. The authors do offer colonoscopy in patients with low-grade appendiceal primaries who have not had one in the previous 5 years, because close to 44% of patients with appendiceal primaries have synchronous colonic polyps. Given the potential of these patients for long survival, the authors prefer to address a colonic resection at the time of CRS-HIPEC. For high-grade nonmucinous appendiceal tumors, a colonoscopic examination does not commonly alter the course of the disease and is requested in selected patients based on age and volume of disease.¹⁶

Tumor Markers

All patients have preoperatively drawn CEA, CA125, and CA19–9. The likelihood of having increased tumor markers has been observed to be equivalent in low- and high-grade lesions. It has been shown that normal preoperative levels of all three are associated with increased likelihood to obtain a complete cytoreduction, probably functioning as a marker of low-volume disease. In addition, normal preoperative CA125 (in male and female patients) has been shown to be associated with prolonged overall survival. Postoperatively, tumor markers are obtained in 3- to 6-month intervals. Levels are taken into consideration along with diagnostic imaging findings and the presence or not of symptoms in evaluating patients for possible recurrence. Patients with voluminous disease and normal preoperative levels rarely benefit from further testing for those markers after HIPEC. A decision to offer a repeat cytoreduction is never based exclusively on the laboratory tumor marker values.¹⁷

Clinical Outcomes

Clinical outcomes of appendiceal peritoneal surface malignancies after optimal cytoreduction depend primarily on histology, the extent of peritoneal disease at the time of diagnosis, the completeness of resection, and functional status of the patient.^18–20 Appendiceal primaries can be grouped based on their ability to produce mucin. In general, mucin-producing primaries have a better biologic behavior than their nonmucin-producing counterparts.

Ronnett and Sugarbaker had classified pseudomyxoma patients in 3 groups in terms of survival: (1) disseminated peritoneal adenomucinosis (DPAM) with 5- and 10-year survival of 75% and 68%; (2) mucinous carcinomatosis with intermediate or discordant features (peritoneal mucinous carcinomatosis [PMCA] I/D) with 5- and 10-year survival of 50% and 21%; and (3) PMCA with 5- and 10-year survival of 14% and 3%, respectively.²¹ In the authors’ clinical experience, DPAM was prognostically indistinguishable from what Ronnett had defined as PMCA I/D. Therefore, the authors concluded that categorizing mucinous appendiceal primaries into low- and high-grade lesions is more predictive of overall survival and response to CRS-HIPEC, with 5-year survival for the low- and high-grade mucinous cohorts of 62.5% and 37.7%, respectively.² DPAM and intermediated primaries also had similar outcomes in a review of 2298 patients with PMP (P<.001).⁸ In the Wake Forest classification, low-grade mucinous carcinoma peritonei includes all cases formerly classified as DPAM, well-differentiated mucinous carcinomatosis, PMCA I/D, and well-differentiated variants of mucinous adenocarcinoma or low-grade appendiceal mucinous neoplasms. High-grade mucinous carcinoma peritonei applies to cases histologically recognized as either moderately or poorly differentiated adenocarcinoma, PMCA, and cases with signet-ring cell component.² Within the HGA cases and the cohort with signet features, invasion of signet cell into tissue is associated with worse survival than floating signet cells in mucin, with median survival of 0.5 years versus 2.4 years (P = .03), respectively.²²

Patients who develop PMP from appendiceal primaries have been considered the best candidates for CRS-HIPEC, with 5-year survival ranging between 60% and 97% and 15-year overall survival up to 59%.^{19,20,23–25} Similarly, 5-year survival of 45% for the high-grade group has been reported for patients with PCI greater than 20 who had a complete cytoreduction. High-grade mucinous patients with PCI less than 20 who had a complete cytoreduction achieved a 5-year survival of 66%.²⁶ However, high-grade nonmucinous appendiceal primaries that include appendiceal adenocarcinoma, goblet cell, and carcinoid tumors derive significantly less benefit from a CRS-HIPEC procedure, with a 3-year survival of approximately 15%. In the authors’ experience, goblet cell patients with low-volume disease (PCI <11) and without nodal involvement experience a median survival of 29 months after complete CRS/HIPEC.²⁷

The completion of CRS is probably the most dominant driver of survival outcomes.¹⁸ What is defined as complete CRS is different for LGA and HGA primaries. CCO and CC1 (or R1 and R2a) resection have similar but not identical outcomes for LGA primaries, while CC1 resection in HGA lesions have similar survival with CC2 resections. In the authors’ series of 481 CRS/HIPEC cases for LGA and HGA appendiceal primaries, a complete CRS had superior outcomes compared with those who underwent incomplete CRS (respective medians of 175, 73, 29, and 17 months for R0/R1, R2a, R2b, and R2c resections P<.001).⁵ This finding confirms data from the authors’ institution and others that demonstrate a significant survival advantage for patients undergoing R0/R1 resection compared with those with R2 resections.^19,20 The extent of disease, as described by the PCI at the time of CRS/HIPEC, seems to be a significant factor in overall survival even in cases where a complete cytoreduction was achieved. Elias and colleagues¹⁹ reported that in 206 patients with pseudomyxoma treated with complete cytoreduction, the 5-year survival was 57% for patients with a PCI greater than 19% and 83% for patients with PCI less than 19 (P 5 .004). These data underline the importance of earlier diagnosis and treatment. At the same time, increased PCI by itself should not function as an exclusion criterion for operative treatment, given that increased PCI is not incompatible with prolonged survival, especially in low-grade lesions.⁸

For HGA lesions, PCI is predictive in a linear fashion of the ability to achieve complete macrosocopic CRS. Unlike colon cancer, where a complete CRS for PCI greater than 17 is not associated with a survival benefit, it is unknown if a similar PCI threshold exists for HGA primaries.²⁸ Colon cancer and HGAs, however, are distinct entities with different sequencing profiles.²⁹ Currently, the authors will proceed with a CRS/HIPEC, regardless of the recorded PCI at exploration, as long a complete CRS is feasible and safe. On the other hand, an incomplete CRS for HGA should not be attempted, as it offers no survival benefit, while it is associated with operative morbidity and delays or inability to receive systemic chemotherapy.^5,9

Low-grade mucinous lesions, without excluding DPAM, may also present with positive lymph nodes in less than 10% of cases.⁵ The authors do not perform a right hemicolectomy routinely in LGA patients unless the right colon cannot be stripped from peritoneal disease. Lymph node involvement, besides being a predictor of surgical morbidity, is also a negative prognostic feature for both LGA and HGA primaries. Thus nodal metastases in this setting seem to function as a surrogate of aggressive tumor biology. In the authors’ hands, node-positive patients with a complete cytoreduction exhibit median survival that is less than the median survival of their node-negative counterparts for both LGA (85 months vs not reached [82% alive at 90 months]) and HGA primaries (30 vs 153 months) (P<.001). In addition, in the authors’ experience, node positivity was a more important than grade predictor of survival, as the node-positive LGA patients had worse long-term survival than the node-negative HGA subjects, even after an R0/R1 complete cytoreduction.⁵

Systemic chemotherapy for PSD of low-grade appendiceal neoplasms is considered largely ineffective. This has been related to the inability of systemically delivered drugs to reach effective intraperitoneal concentrations and the slow growth kinetics of the low-grade malignant cells. A phase 2 study in advanced unresectable low-grade appendiceal primaries with concurrent mitomycin C and capecitabine showed a response in 38% of the patients in the form of either stabilization or radiologic reduction in the volume of disease.³⁰ Lack of progression of disease of LGA primaries while on chemotherapy should be interpreted cautiously, and is likely a solid treatment goal, because lack of progression is common for these lesions regardless of chemotherapy treatment. Further, in determining chemotherapy responses, it should be kept in mind that the mucin does not respond to chemotherapy, only the cells producing it. In the authors’ experience, administration of postoperative chemotherapy had no effect on OS for LGA primary lesions.

For high-grade lesions, it seems that systemic chemotherapy might improve progression-free survival, with the overall survival benefit being derived from the ability to achieve a complete cytoreduction.³¹ Chemotherapy with folinic acid, fluorouracil and oxaliplatin (FOLFOX) in the neoadjuvant setting for high-grade appendiceal mucinous lesions was related to progression of disease in 50% of patients who had surgical exploration. Tumor response was observed in 29% of the examined specimens.³² More recently, the authors have also utilized FOLFOXIRI based on the superior response rates seen in treating marginally resectable hepatic lesions in colorectal cancer.

Preoperative ECOG performance status is a significant prognosticator of survival. In the authors’ experience in patients with 1000 HIPEC procedures from a variety of primaries, ECOG was predictive of survival in univariate and multivariate analysis (P<.0001 hazard ratio [HR] 2.8 for ECOG 2 – HR 4.3 for ECOG 3 or 4).¹⁸ This is also true for appendiceal primaries, specifically where for each stepwise increase in ECOG score, the risk of death is increased 8.8-fold.¹⁰ The authors use the performance status routinely as a selection criterion for CRS/HIPEC.

Age itself has not been found to be significant in predicting survival.^19,20 The authors do not use age as an exclusion criterion for CRS-HIPEC, but rather as another factor to be considered before the procedure. In the authors’ series with patients older than 70, complete CRS was associated with a median survival of 33 months for appendiceal primaries. It is important to mention that in elderly population, the completeness of cytoreduction was significant predictor of survival (P = .007) only when surgical complications were not included as a covariate. With complications included in modeling only, the institutional experience (P = .001) and the absence of complication itself (P = .001) were predictive of survival, suggesting challenges in rescuing a surgical complication in elderly patients, possibly as a result of comorbidities and diminished physiologic reserves.³³

For all appendiceal patients who had a CRS-HIPEC for PSD, approximately 10% ultimately have a repeat cytoreduction to treat recurrent disease. Patients selected in the authors’ institution for repeat HIPEC are those who have maintained an ECOG 0 to 1 functional status, have adequate nutritional reserves, and have had an interval between the 2 procedures lasting a least a year. The survival of these patients depends predominantly on the completion status of the second CRS and the interval between the 2 operations.⁶ In the authors’ experience, each additional month between cytoreductions was related to a reduction of the risk of cancer-related death by 2.6%, while the median survival was 1.3 years for less than a 1-year interval, 3.7 years for a 1- to 2-year interval, and 7 years for an interval greater than 2 years interval, P < .001. Incomplete (R2a,b) first cytoreduction is not an absolute contraindication to an attempted second cytoreduction for low-grade appendiceal primaries, because 20% of these patients achieve a complete R0/R1 repeat cytoreduction. Successful complete repeat CRS/HIPEC for recurrent carcinomatosis from HGA is a rare event with a handful of cases recorded in the authors’ database over the last 26 years.³⁴

Despite similar morbidity and mortality between repeat CRS/HIPEC procedures, it is the authors’ experience that each surgical exploration alters the mechanical properties of the LGA peritoneal disease, with much more sclerotic features and plane obliteration upon subsequent surgeries.

FUTURE DIRECTIONS

The variability of outcomes of appendiceal primaries cannot be predicted by the current classification systems that in a nutshell are based on the presence or not of mucin and grade. By using hierarchical clustering analysis of tumor expression profiles, the authors had previously identified a 139-gene cassette that distinguished 2 LGA molecular subtypes (based on low vs high expression of the gene cassette) that correlated with survival. In multivariate analysis inclusive of molecular subtype, grade, R status of resection, ECOG, and age, only the molecular subtype (P = .0007), grade (P<.001), and ECOG (P = .007) remained significant, while R status of resection was not.³⁵ This raises the question of which patient will not benefit from a CRS, because of either indolent disease or too aggressive disease.

The authors are working with their biomedical engineering department and reconstructing individual patients’ own appendiceal tumors in the form of 3-dimensional organoids that are inclusive of stroma and tumor cells. They replicate the tumor microenvironment by microengineering 3-dimensional tumor organoids directly from fresh appendiceal tumors with biofabrication time of less than 1 week and a take rate of approximately 90%. These organoids are housed in microfluidic devices – resulting in tumor-on-a-chip system – enabling parallel real-time screening of multiple chemotherapy drugs while linking multiple patient-derived tissues and tumor together in a systems approach^36,37 (Fig. 3).

Fig. 3.

(A) LGA organoids from patient derived tumor specimen. (left) 2-dimensional image of an LGA organoid. (Right) 3 dimensional (B) Subsequent LGA and HGA organoid chemosensitivity results.

These systems are currently lacking clinical correlation data. Upon establishment of clinical correlation, they can potentially match patients, based on in vitro efficacy, with the best available drug or combination of drugs, in the neoadjuvant, adjuvant, or intraperitoneal setting. Another possibility is identifying patients who will not benefit from surgery or chemotherapy, therefore sparing them from unnecessary morbidity while reducing health care cost at the single patient level.

KEY POINTS.

• The early symptoms of appendiceal cancer may mimic the clinical picture of appendicitis.

• Most patients are diagnosed incidentally during surgical exploration or late when peritoneal or systemic dissemination has already occurred, as colonoscopy rarely will diagnose an appendiceal cancer.

• Systemic/extraperitoneal metastases are distinctly unusual for appendiceal mucinouslesions.