EOC patients with NLNM are classified as FIGO stage IV and are expected to have poor survival outcomes. However, our study revealed a subgroup of EOC patients with NLNM who demonstrated a distinct clinical manifestation with relatively good survival outcomes. These patients received only surgical removal of primary tumors and adjuvant chemotherapy. Although recurrence rates after treatment in these patients were as high those of patients in advanced stages, one patient remained in complete remission for a long period after treatment.
Timing of NLNM identification is a key predictor for patient survival. We found that patients with simultaneous NLNM at initial surgical intervention of primary ovarian lesions had better survival outcomes than patients with NLNM discovered at cancer recurrence. However, among the patients who had NLNM at initial surgical intervention, only those patients who had tumors confined to the pelvic cavity had better survival rates. Patients with ovarian lesions that had spread to the abdominal cavity at NLNM identification had very poor outcomes. Similarly, poor survival rates were observed in EOC patients who had retroperitoneal lymph node metastasis. Patients with retroperitoneal lymph node metastasis without peritoneal carcinomatosis were reported to have an overall survival rate of 58% to 84% compared with 18% to 36% for those with macroscopic peritoneal carcinomatosis, depending on variations in the cancer treatment [2–6]. It could be inferred that NLNM is an extension of retroperitoneal lymph node metastasis and therefore demonstrates similar clinical features and survival outcomes as patients with retroperitoneal lymph node metastasis.
Our findings suggest that there are two distinct biological types of invasive EOC: one is very aggressive and prefers peritoneal invasion, and one that has limited capabilities of peritoneal invasion. In both Groups A and B, some patients appeared to have very aggressive clinical courses at detection of NLNM. They were more likely to be non-serous histology, chemoresistant, presented with early cancer recurrence accompanied with distant metastasis (including brain metastasis), and fatal outcome. However, the exact biological factors that contribute to this unique entity of EOC are unknown and require further research.
On the other hand, a few patients in Group B had ILNR at the time of NLNM. Particularly, all patients in Group A that had tumors confined to the pelvis showed ILNR at the time of cancer recurrence; one of them even showed complete cancer remission. All these patients had serous histology and were good responders to chemotherapy; many of them had complete remission of neck lymph nodes after chemotherapy. Based on the relevant literature, ILNR occurs in about 4% to 6% of patients with EOC [7, 8], and represents approximately 10% of overall recurrences . Although it is relatively rare in EOC, it shows amazingly good survival outcomes [7–12]. No obvious predisposing factors have been reported in association with the occurrence of ILNR. We found patients with ILNR were more likely from Group A where EOC tumors were found confined to the pelvic cavity at the time of NLNM. However, after long follow-up periods, peritoneal carcinomatosis eventually occurred in 4 patients, resulting in the death of two patients due to colon obstruction at 41 and 44 months after the discovery of NLNM. These results again support our observation that patients who have confined pelvic tumors during initial NLNM identification belong to a unique EOC group that is biologically less aggressive. Tumors with a limited capability of peritoneal spreading showed ILNR during cancer recurrence. After a longer period, peritoneal carcinogenesis somehow developed and led to patient death. Likewise, Legge et al. observed the occurrence of peritoneal spreading in 25% of their 32 ILNR patients at a median follow-up of 22.5 months (range: 7–96) and subsequent rapid fatal outcomes .
Although we observed some commonalities between our NLNM patients and the reported ILNR patients, seven of our patients did not have systemic retroperitoneal lymphatic node metastasis during NLNM. Five of these patients were diagnosed solely by imaging studies, and one patient showed a positive imaging study but pathologically negative retroperitoneal lymph nodes. It has been reported that the diagnostic sensitivity of CT for pelvic and abdominal lymph nodes is low (43%), but the specificity is high (92%) . Even so, the concordance of radiologic staging with surgical staging in ovarian cancer is as high as 78% . However, 15.8% of extra-abdominal lymph node metastasis is unpredictable by PET/CT imaging studies . Therefore, the true incidence of retroperitoneal lymphatic metastasis in our cases could be higher. It is possible that some of our cases might have microscopic retroperitoneal lymphatic metastases that were not detected by radiologic studies or were misdiagnosed without complete or radical lymph node dissection. Even so, lymphatic cancer spreading might not occur systematically along the lymphatic tract. Kleppe et al. reported that in EOC that appeared confined to the pelvis at the time of initial exploration, the mean incidence of lymph node metastases was 14.2% (range: 6.1–29.6%) . In these cases, lymph node metastasis was found to be 4.3% (range 0.0–14.8%) in both the pelvic and para-aortic region, 7.1% (range 3.0–13.0%) only in the para-aortic region, and 2.9% (range 0.0–11.1%) only in the pelvic region . High grade and serous histology were the two risk factors for lymph node metastasis in these apparently early-stages of EOC . This suggested the skipping nature of lymphatic spreading in EOC.
One other limitation of our study was that only enlarged neck lymph nodes that were biopsied and pathologically proven were included. This might reduce the true incidence of NLNM in EOC patients, particularly in cases with neck lymph node enlargement identified during multiple sites of ovarian cancer recurrence (i.e. Group B patients). Our long study period might overcome this limitation.
By survival analysis, non-serous histology of EOC was the strongest predictor for patient survival. All patients with non-serous histology in the two groups had cancer progression or recurrence shortly after treatment. Cancers that regressed completely after treatment were only in patients with serous histology. Brain metastasis was the next strongest predictor for patient survival. Brain metastasis has been reported as uncommon with an incidence of less than 2% in EOC patients [17, 18]. The incidence of brain metastasis in our cases was very high. Five patients developed brain metastasis: one in Group A and four in Group B. The high incidence of brain metastasis in NLNM could be explained by the close approximation of the neck lymph nodes to the brain; it also highlights the very aggressive nature of these forms of cancer. Our patients had a mean survival period of 4.4 months after brain metastasis and 2 had clear cell carcinoma histology. Similarly, high mortality rates from brain metastasis have been reported, and brain metastasis are believed to be a late manifestation of ovarian cancer in advanced stages .
Several other factors, such as small sample size, varying histology of ovarian cancer, and different causes of death, could have confounded the analysis results. Although NLNM is a rare clinical manifestation in ovarian cancer, the relative small sample size for a retrospective study limits the conclusions. Future studies with larger sample sizes and more rigorous research designs will help to clarify the prognostic value of NLNM.
Our study also suggested that EOC tumors that displayed the lymphatic route of spreading could be more chemosensitive than those with intraperitoneal spreading. In observing that some of our patients with ILNR eventually developed peritoneal carcinomatosis after a long follow-up period, a more aggressive treatment should be considered for patients with ILNR. From literature reports on patients with ILNR, survival outcome was found to be longer in cases that underwent surgical removal of metastatic lymph nodes in addition to chemotherapy [10–12]. When possible, surgical removal of metastatic neck lymph nodes in patients with confined pelvic mass during initial surgical intervention of primary tumor would likely result in a higher chance of complete eradication of the cancer.