September 2021

Synopsis: Most primitive multicellular organisms have a central body cavity housing critical organs as well as resident leukocytes. In lower organisms these leukocytes not only patrol for pathogens and clear debris, but also form multicellular aggregates at sites of injury. While peritoneal macrophages perform a similar role clearing dead cells and debris in mammals, their role in responding to injuries within the peritoneal cavity remains unclear.

The authors of this study developed an intravital microscopy model to visualize and study the dynamics and molecular mechanisms of resident peritoneal macrophage recruitment after peritoneal injury. Using a focused high-power laser beam to generate a localized injury to the peritoneal lining, the authors were able to image the immune response to peritoneal injury in real time in mice. Peritoneal macrophages were recruited to sites of injury first through the formation of an initial tether of macrophages to the site of injury, which was followed by formation of an aggregate of cells that resembles a thrombus. Macrophage aggregation at these sites of injury closely resembled that of platelets in adjacent vascular injuries.

Aggregation of peritoneal macrophages at sites of injury was dependent on primordial scavenger receptor cysteine-rich (SRCR) domains rather than classical adhesion molecules such as integrins and selectins. These macrophage aggregates physically sealed injuries in the peritoneal lining and promoted repair of these lesions. However, using a model of surgical injury, where a laparotomy was performed and polypropylene suture material introduced into the abdominal cavity, macrophages formed dense aggregates around the suture material that ultimately led to the development of intra-abdominal adhesions. These super aggregates led to subsequent recruitment of mesothelial cells, and finally to collagen deposition with formation of dense adhesions by 7 days after injury. The authors also show that macrophage depletion or scavenger receptor inhibition with Poly-(I) dramatically reduces peritoneal adhesion formation.

Bottom line: Peritoneal macrophages play a critical role in responding to sites of injury by forming aggregates that promote wound healing. In the setting of surgical injury, these macrophages form super aggregates that drive peritoneal adhesion formation. Macrophage scavenger receptors may be a promising therapeutic target for the prevention of adhesion formation after surgery.

Synopsis: Neoadjuvant chemoradiotherapy followed by surgical resection and subsequent adjuvant chemotherapy is currently the standard of care for patients with locally advanced rectal cancer (LARC). Recently, total neoadjuvant therapy (TNT), in which patients receive all therapy before surgery, has emerged as a potential therapeutic alternative for these patients. TNT has numerous hypothesized benefits over the traditional treatment paradigm, including the potential for earlier administration of effective systemic therapy, downstaging before surgery to increase the rates of margin-negative and sphincter-sparing resections, and increasing the rates of complete pathologic response (pCR) upon resection, which is associated with lower rates of locoregional relapse and improved survival outcomes. Both standard and TNT protocols for LARC routinely use 5-fluorouracil and platinum based chemotherapeutic agents. However, thus far, studies have not been able to demonstrate superior disease-free or overall survival rates for LARC with TNT. In this context, there is growing interest in incorporating novel targeted molecular and immune checkpoint blockade agents into TNT treatment protocols. The discovery of programmed death-ligand 1 expression on colorectal tumor cells following neoadjuvant chemoradiotherapy has led some to consider whether the addition of anti-programmed death 1 therapies, such as pembrolizumab, may further improve the efficacy of TNT.

In this paper, Rahma et al. evaluated the efficacy and safety of pembrolizumab therapy concomitant with traditional TNT. The authors performed a phase 2 randomized clinical trial using an intent-to-treat analysis comparing neoadjuvant rectal scores (NAR), pCR rates, and R0 and sphincter-sparing surgery rates between patients who received traditional TNT alone versus those who received traditional TNT plus pembrolizumab. The primary outcome, the NAR score, incorporates a number of post-neoadjuvant therapy pathologic metrics to evaluate rectal cancer response to neoadjuvant treatment, and has been shown to correlate better with disease-free and overall survival than pCR alone. Lower NAR scores indicate a stronger response to neoadjuvant therapy.

Patients in the control arm (CA) received induction FOLFOX, followed by chemoradiotherapy with 5-FU and 4500 cGY of radiation plus a 540-cGY boost, followed by surgical resection. Patients in the pembrolizumab arm (PA) received the same treatment as the CA arm, with the addition of up to 200mg of pembrolizumab during the chemoradiotherapy phase of treatment. A total of 95 and 90 patients were randomized to the CA and PA arms, respectively. Eighty-one of the 90 patients in the PA arm were able to start radiotherapy and thus be eligible to receive pembrolizumab, although only 37 (45.7%) of these patients completed the full 6 dose regimen. Ultimately, 137 patients (68 in the CA and 69 in the PA) completed TNT and underwent surgical resection with the required pathologic data to compute NAR scores. The mean NAR scores for PA (11.55, standard deviation 12.43) and CA (14.08, standard deviation 13.82) did not significantly differ (p=0.26). The two groups did not significantly differ in their rates of pCR either: 31.9% in the PA versus 29.4% in the CA, p=0.75. The rate of sphincter-sparing surgical resections was lower for PA compared to CA, though the difference was not statistically significant (59.4% vs. 71.0%, p=0.15). Finally, the rates of R0 resections were similar between the two groups (94.0% PA vs. 89.4% CA, p=0.36). Thirty-five patients (43.2%) in the PA developed immune-related adverse events, although only 3 (3.7%) developed grade 3 events, and no grade 4 or 5 events occurred.

Overall, this trial demonstrated that concomitant administration of pembrolizumab with traditional TNT was both safe (acceptable rates of immune-related adverse events) and feasible (equivalent proportions of patients within the PA and CA groups made it to surgery), but not efficacious in improving NAR scores. Further studies are needed to determine the ultimate role of novel immune checkpoint blockade or targeted molecular agents to current treatment regimens for locally advanced rectal cancer.

Synopsis: Emergency General Surgery (EGS) accounts for more than 2 million hospitalizations per year and is estimated to cost $28.4 billion per year in the US. Despite this, and despite robust systems for quality improvement in elective and trauma surgery (NSQIP, TQIP), no such system exists to monitor and improve EGS outcomes. Some, including our group here at Penn, have suggested that tracking EGS outcomes in this way would be worthwhile. These authors also make this point by using observed-to-expected mortality rates to calculate excess EGS deaths resulting from hospital-level variability.

Nine years’ worth of HCUP Nationwide Emergency Department Sample (NEDS) data (2006-2014) was used to perform a retrospective analysis; inclusion criteria were adult patients (≥18 years) admitted with an EGS diagnosis (as defined by the AAST). Hospitals missing >20% of mortality data and hospitals reporting <100 EGS patients over the course of 9 years were excluded. “Operative EGS” and “High-Burden Operative EGS” (based on previous work describing common EGS operations accounting for the majority of the EGS burden in the US) subsets were created. The authors then used logistic regression with reliability adjustment to adjust for hospital- and patient-level characteristics and benchmark hospitals into mortality quintiles. Finally, excess deaths for each quintile were calculated relative to the highest-performing quintile.

The NEDS is a 20% stratified sample of all US ED visits; when weighted appropriately the sample was representative of 26.3 million discharges over the study period, about 25% of which were operative. In each of the three cohorts described above (all, operative, high-burden operative), the worst-performing quintile had an adjusted mortality rate greater than two-fold that of the best performing quintile. Overall, the authors calculate that there are about 17,000 excess EGS deaths per year in the US, nearly half of which occur at hospitals in the worst-performing quintile.

This paper is an important addition to the literature, as it brings attention to a long-neglected topic, the importance of which is only recently starting to be recognized. The primary flaw, generally speaking, is the incomplete clinical picture that is obtained from an administrative data set, such as the NEDS. While the authors have done their best to control for measured covariates, they lack information on acuity of illness and physiologic derangements at the time of presentation. Interestingly, they find that only 27% of excess deaths occurred in operative patients, which would seem to simultaneously underscore the importance of studying non-operative EGS outcomes and suggest that there may be some confounding by indication (i.e., sick patients may not have been operative candidates). Nonetheless, it is hard to deny that there is at least some indication that there is significant inter-hospital variability in EGS outcomes, supporting the position that dedicated EGS QI efforts are needed.

Synopsis: One of the defining features of Penn Surgery is the introduction and utilization of advanced adjuncts in the care of surgical patients. This is most clearly seen then in the realm of intraoperative molecular imaging (IMI). In a very recent manuscript published by JAMA Surgery, Dr. Kennedy et al. describe their experience with this technology in the surgical treatment of 40 patients with T1 lung lesions suspicious for malignancy and with features consistent with a nonpalpable lesion. The selection of this specific patient population and understanding their “typical” surgical treatment is key. If a surgeon cannot properly localize the tumor, more parenchyma (typically an entire segment or lobe in the case of lung) must be resected to maximize the rate of a complete (R0) resection. In a patient population with a high rate of pre-existing lung disease, this extra parenchyma may be the difference between a suboptimal and optimal patient outcome. While the technology underlying IMI is complex, the basic idea is rather simple. About 90% of all adenocarcinomas and 70% of squamous cell carcinomas in the lung overexpress folate receptor alpha. Patients receive a preoperative infusion of a folate analog conjugated to the fluorescent dye which can then be visualized by a special thoracoscope intraoperatively.

Out of the 40 lesions which were suspected to be non-palpable based on preoperative work up, 22 were able to be localized by conventional surgical methods. Out of the other 18 lesions, 15 were visualized by IMI. These 15 patients, without IMI, would have required either a thoracotomy or an extended resection of lung parenchyma. Another benefit of IMI is the ability to “back table” the specimen with the same imaging device to assess margin distance. The back table margin measurements were nearly identical to final pathology margins with a median difference of 1.3 mm. Decreased anesthesia time afforded by knowing real-time margin status, as opposed to waiting for frozen sections, is a very real consideration in a subset of this patient population. 

Overall, about 38% (15/40) of the cohort were ultimately helped with IMI by changing their surgical management. It remains to be seen whether long-term outcomes of these patients hold up with other national trends of T1 lung lesions, although it is well established that R0 resection is the most important prognostic factor. While not addressed in this study specifically, it will also be important to clarify if non-palpable lung lesions differ biologically from palpable lung lesions and possibly should always receive more extended resection based on possibly more aggressive biology. 

Although the study only used 40 patients, the implications and future directions of this technology is endless. It remains to be seen what other cancers can utilize this technology. Dr. Kennedy and the rest of the Singhal lab continue to be leaders in this field and have set the stage for the widespread adoption of this technique with the entirety of the surgical management of cancer impacted by these advances.