In this study we have investigated the differences in the number of lymph nodes harvested between ELAPE and conventional APR. The median number of LN yielded from conventional APR is 15. Authors such as Shen et al. 2009 [19] retrospectively inspected 69 cases and had reported the same median number of 15. The absolute LN count in our ELAPE group is higher but not significantly compared to the APR group, and is comparable to the 13.7 nodes from the ELAPE results in the large scale (n = 519) Swedish population study [20]. Neither our study nor the Swedish cohorts found significant differences between APR and ELAPE groups. However, reported number of LN harvest from APEs in literature varies greatly. In general LN counts for abdominoperineal excisions are lower than other type of colorectal resections, and can be as low as 9 LN [21].Though comparable, the number of lymph nodes harvested at our unit appears higher compared to the reported literature. This may be explained by the contemporaneousness of our data and higher proportion of laparoscopic surgery in our study compared with established evidence. Dolan et al. [22], in a prospective study of 896 patients spanning 20 years (1997–2016), had found significant independent correlation between later operative date, increasing prevalence of laparoscopic surgery and higher lymph node harvests.
12-LN minimal requirement
Secondly, we have found a reduction in the rate of resections failing to meet the 12-LN minimal requirement under the ELAPE technique compared with traditional APR. The differences are considerable and almost statistically significant. Hitherto; there has been a lack of evidence comparing quality of LN harvests between the two techniques despite emphases of its importance by many authors [13, 23–25] and guidelines including the National Cancer Institute [16]. A minimum of 12 lymph nodes was recommended [23] as below this cut-off value there is a high risk of false-negatives in reporting lymph node metastases due to inadequate sampling [25]. The 12-node standard has been endorsed by other researches for reasons of “diminishing returns” beyond the examination of 12–17 nodes [18]. In our study, 60% and 77% of specimens in the APR and ELAPE groups respectively met this standard. When compared with known literature from authors who specifically investigated the 12-LN criteria among rectal cancers, our cohort of patients have achieved higher rates in general. Field et al. [21] reported 50%, while Gurawalia et al. [26] achieved 52%, and Baxter et al. [27] had a 46.4% attainment rate. Our higher success rate could be explained again by the contemporaneousness of our cases, and a higher prevalence of laparoscopic surgery in our cohort, both considered to be independent determinant of meeting the 12-LN standard as found by Dolan et al. [22]
Node Negative Disease
The British reviewers Ong and Schofield [28] have summarised that node-negative colorectal cancer patients have a 5-year survival rate of 70%-80% in contrast to 30%-60% for their nodal-positive counterparts. Survival can be improved in the latter group by adjuvant chemotherapy [14]. The 20%-30% disease recurrence in apparently completely excised tumours without lymph node metastases is thought to be due to occult lymph node disease [29]. If this subset of patients could be identified by better lymph node staging, they might also benefit from adjuvant chemotherapy. Nodal-positive resections, irrespective of the number of LN harvested, would indicate adjuvant chemotherapy. Therefore substantial researches have investigated how the number of LN examined affects the accuracy of colorectal cancer staging and prognosis among node-negative individuals [12, 30, 31]. As aforementioned, Fielding et al. [23] advocated a minimum of 12 lymph nodes since below this cut-off value there is a high risk of false-negative reporting of lymph node involvement due to inadequate sampling. Based on this evidence, we have conducted a separate analysis of nodal-negative cases investigating whether the minimum 12-LN standard has been met in this subset. We have found significantly lower rates of failures in meeting the 12-LN requirement with ELAPE when compared to APEs. This suggests that ELAPE may be superior at staging rectal cancers, minimising false-negatives in apparently negative nodal disease, leading to more appropriate decision-making on adjuvant treatments. The mechanism underlying the observed improvement with ELAPE is not entirely understood, since anatomically the mesorectum tapers as it adjoins the pelvic floor. However as Holms et al. eluded to in their research [6], while other techniques (e.g. intersphincteric, extrasphincteric dissections) are available, ELAPE offers a standardised approach for surgeons to perform abdominoperineal excisions, leading to more consistent quality of resections which in turn reduces the rates of sub-standard TMEs.
CRM Involvement
The overall CRM rate of our cohort is 16.7%, which is relatively high but comparable with published data of 16.6% from Great Britain [32] and 16.7% from Canada [33]. Our study has showed a decreased rate of CRM involvement among the ELAPE group compared with conventional APR, though not quite to a statistically significant level. Most of the studies [20, 34, 35] have also failed to show statistically significant superiority of ELAPE in CRM clearance. However the Danish study [35] suggests a magnitude of reduction (OR 0.386) that is considerable and comparable to our data (OR of 0.365). The more recent, yet small (n = 34) randomised controlled trial, nevertheless, demonstrated a significantly improved CRM and zero intraoperative perforations in their ELAPE arm of patients [36]. Our investigation suggests that the apparent benefit of ELAPE may be reproducible in a district general hospital setting among 8 different local colorectal surgeons who have performed these procedures at our unit.
Potential Confounders
It is established that the quality of surgery and pathologic examination are both major determinants of LN harvests [37, 38]. However, multiple clinicopathologic factors may also affect the number of lymph node retrievals. In particular, a significant reduction in the mean LN yield was found in patients who received neoadjuvant chemoradiotherapy, from 17 to 13[ 39], and from an average of 19 to 16 [40].Other authors such as Field et al. [21] have also found that young, female patients and higher T stages of cancers to be correlated with higher LN yield. There was also significant link between laparoscopic surgery and higher rate of attaining the 12-LN standard [22]. On this basis, we have conducted regression analyses taking into account the above predictors of LN yield as potential confounders. This has not changed the correlations found in our results. The rate of failure to achieving 12LN standard is significantly lower in the ELAPE group among the nodal negative cohorts. The reduction in CRM involvement is still present but not significant. Our analyses have demonstrated superiority of ELAPE in these pathological aspects independently of other influencing factors.
Limitations
As a retrospective observational study, our investigation is subject to the usual limitations of selection and recall biases. The aim of our research was to establish whether ELAPE leads to a better pathological outcome. Although our results have suggested it, there remains uncertainty whether this would necessarily translate to an improvement in clinical outcomes i.e. local recurrence rate (LR) and disease survival. Several systematic reviews found no differences in either CRM or LR with ELAPE [34, 41]. Some meta-analyses [42] showed that despite ELAPE significantly lowering the rates of CRM involvement, there has been no benefit in the LR, while others [43], did demonstrate a significant reduction in LR (OR 0.30, P < 0.01). Even if the LR were found to be improved with ELAPE, some authors have found no difference in overall survival, disease free survival or disease progression in a prospective controlled study [44], and from a randomised controlled trial with median follow-up of 20 months [45]. However these studies have been marred by their small recruitment numbers (n = 69 and n = 67 respectively) and short follow-up periods.
It is less disputable, nevertheless, that the retrieval and analysis of a larger number of lymph nodes result in a survival advantage [30]. This was initially thought to be due to upstaging of “missed” positive lymph nodes. However, more recent studies suggest that this phenomenon cannot be attributed to staging migration alone. A systematic review [18] found improved survivals correlating with higher lymph node harvests in stage III (4 of 6 studies) as well as stage II (16 of 17 studies) diseases. Furthermore, lymph node sampling past a certain point does not appear to improve disease staging [46]. Interestingly Joseph et al.[47] Found that improvement in colorectal cancer survival was associated with greater LN harvest irrespective of patients’ nodal statuses (N1 or N0). Tumour-host interactions may be a plausible explanation for this, as higher LN yield may reflect a stronger host immune response[48].
Among other factors that could have influenced our results was the lack of pathological reporting of the lengths of our specimens. This may be because there has not been a standardised length of resection for APEs in general. The amount of mesentery associated with specimen length undoubtedly correlates with the number of LNs found [19].
Secondly, authors such as Dolan et al. [22] have noticed greater number of LN yield to be associated with more contemporaneous operations when comparing cases before and after year 2007. Since our study design is based on the adoption before and after the adoption of ELAPE in year 2014, the latter date of the surgery in the ELAPE group may be a potential bias. Are we simply becoming better surgeons and pathologists as a whole rather than observing improvement due to the merits of the new technique?
Despite our effort to improve LN harvests, no internationally recognised standards of practice have been developed for the histopathological processing of lymph nodes in specimens [49]. A notable Canadian study showed that only 58% of pathologists were aware of current guidelines and that only 25% recognized that a minimum of 12 LNs was necessary for accurate designation of node negativity [50]. On the other hand, the concern with rigidly fixing a recommended number (i.e. 12) is that once this figure is met, the lab search for LNs may end at that point regardless of how many potentially positive nodes are left in the sample [27].
The apparent advantages of ELAPE also need to be balanced with the morbidity this new technique brings. Some authors have reported significantly more post-operative wound infections after ELAPE (20.5%) than for APR (12%) [20]. A prospective multicentred trial [51] also suggested higher rates of sexual dysfunction, urinary retention, and perineal complications including chronic perineal pain associated with ELAPE. However, a meta-analysis [52] found no differences in complication rates between the two techniques, and some of these potential complications can be mitigated with reconstructions using meshes or plastic surgery. Our enthusiasm for ELAPE should nevertheless be tempered with caution and consideration of its potentially higher complexity and morbidity.