Outcomes of surgical resection and intraoperative electron radiotherapy for patients with para-aortic recurrences of gastrointestinal and gynecologic malignancies

Treatment of recurrent lymph node metastases in the para-aortic region is challenging because local therapies are limited. Surgical resection may leave microscopic- or gross-positive margins and high dose external beam radiation may be too toxic given proximity to nearby sensitive structures (e.g., bowel, kidney). IORT affords the advantage of high-dose irradiation while uninvolved critical structures are physically displaced away from the treatment field. However, there are limited data on efficacy and toxicity with this technique. We report outcomes for patients with para-aortic recurrent malignancy treated with surgery and IORT. The majority of patients in our series had a primary diagnosis of colorectal cancer, with a high risk of local and distant progression.

Recurrence and survival

Oncologic outcomes for patients treated with surgery and IORT for para-aortic lymph node recurrences have been reported but these were mostly patients with primary gynecologic malignancies with the exception of the reports by Haddock et al. [7,8,9] Not surprisingly, we report a statistically significant worse LR in patients with positive or close margins compared to negative margins, which has been shown in many cancers but also in recurrent colorectal cancer treated with IORT [10]. IORT may help improve local control in patients known to have a high risk of LR, such as those with gross or microscopic residual disease [9, 11,12,13]. However, this benefit is difficult to quantify in the absence of randomized data.

Haddock et al. reported a similar series of patients, treated from 1981 to 2000, with colorectal cancer and recurrent para-aortic or mesenteric lymph node disease treated with surgery and IORT. They reported 81% LC at 3 years and 5-year OS of 34%9. Calvo et al. reported a LR rate of 14% at median follow up of 39 months and 2-year OS of 57% for patients with a variety of recurrent malignancies treated with surgery and IORT (54% had para-aortic recurrences treated with IORT) [7]. We report patients treated from 2008 to 2020 with similar OS and LC as evidenced by our reported median OS, median time to recurrence, and Kaplan-Meier curves (Figs. 1 and 2). However, significant patient heterogeneity and relatively small number of patients make it difficult to compare results. We acknowledge the limitations of a Kaplan-Meier analysis with few patients available during longer interval follow-up.

Multiple systematic reviews have been published reporting outcomes of surgery without IORT for para-aortic lymph node recurrences in colorectal cancer. Zizzo et al. performed a systemic review for patients with colorectal primary tumors which included 9 studies over the last 30 years. They reported a mean second recurrence rate of 62%, 14–24 months median disease-free survival (DFS), and 3-year OS of 53–88% from the pooled population [14]. 16–29% of re-recurrences were in the para-aortic lymph nodes [14]. Ho et al. also reported results from a systematic review containing 6 larger series for colorectal primary tumors which included a median DFS between 17 and 21 months with a recurrence rate of 67–100% and overall operative morbidity rate of 17–33% [15]. However, many of these recurrences were distant with the exception of those reported by Shibata et al. who reported local recurrences for 11/15 patients who underwent resection with negative margins [12].

The absolute benefit of lymph node dissection and resection for para-aortic recurrences is unclear in this population without prospective randomized data. However, these reviews reported relatively high rates of recurrence and low rates of long-term OS suggesting outcomes may be improved with improved local and systemic therapy. Our in-field recurrence rate of 32% for patients with colorectal primary tumors is similar to previously cited recurrence rates. However, there are many differences including neoadjuvant and adjuvant therapies, comorbidities, cancer burden and biology, and selection bias that make comparisons between studies difficult.

Toxicity

IORT toxicity has been studied in large retrospective series and a systematic review for patients with primary colorectal cancer [16]. The majority of sites treated with IORT in these cases were in the pelvis. Mirnezami et al. performed a systematic review of locally advanced and locally recurrent rectal cancers treated with surgery +/- IORT. A number of these patients also received EBRT and/or chemotherapy. They found an increased risk of wound complications (delayed healing, wound infection, dehiscence, and “non-specific wound related problems”) in patients who received IORT compared to those who did not (OR 1.86; 95% CI = 1.03–3.38; p = 0.049)13. Calvo et al. reported minimal toxicity in patients treated with IORT to para-aortic sites with 1 patient developing a pancreatic fistula who was treated with IORT to the celiac axis [7].

Wound infections and pelvic abscesses have otherwise been reported in approximately 25% of patients but determining the increase in risk due to the addition of IORT to the baseline risk of these complications from surgery is impossible [16]. 15% (4/26 patients) of patients were diagnosed with wound or infectious complications in our study which is similar to previously reported results above. Determining the added side effects of IORT can be difficult in the presence of confounding factors such as surgery and systemic therapy. We felt it was best to report rates of diarrhea and AKI in this group but do not feel that these should all be attributed to IORT.

Studies of IORT in animals have been used to estimate the tolerance of organs and anatomic structures to irradiation. Sindelar et al. studied the effects of IORT to retroperitoneal structures in 14 foxhounds with doses up to 45 Gy [17]. In relation to nerve injury, they reported lower extremity weakness in 1 foxhound approximately 3 years following IORT to 30 Gy [17]. Another study in canines by LeCouteur et al. randomized 85 beagles to receive increasing doses of IORT, EBRT, or both and measured nerve function by electrophysiology or physical exam. They reported a numerical correlation between rates of neuropathy and IORT dose starting at 15 Gy [18]. Notably, nerves within the field were within the 90% isodose line for nearly all subjects in the study [18]. This is not typically true when treating human patients.

Many of our patients experienced peripheral neuropathy due to chemotherapy (oxaliplatin). This and other factors such as nerve-related pain masquerading as musculoskeletal pain may obscure accurate counting of IORT-induced nerve injury. However, no patients in our series suffered from isolated peripheral nerve injury that was thought to be due to IORT. Low rates of neuropathy may also suggest that para-aortic locations for IORT may be favorable as it pertains to IORT-induced peripheral nerve injury which could be due to more accommodating anatomy. The actual radiation dose received by the nerve was also likely much lower than the prescription dose. Thus, we would expect a very low rate of neuropathy due to IORT given these factors and the relatively high-dose tolerance demonstrated by canine studies.

Our study reports a toxicity rate lower than what is typically reported in the literature suggesting IORT to para-aortic sites may cause little or no additional unacceptable toxicity compared to better-studied pelvic sites. One other explanation for the lower toxicity rate compared to other studies in the literature is the number of patients that were followed by their community oncologist after having surgery and IORT at our center. Patients lost to follow-up or followed by physicians outside our health system may have reported complications not recorded in our electronic health record. Surgical techniques and quality have also improved over time which could be responsible for lower toxicity rates compared to studies with patients treated decades ago. Our reported rate of wound complications is in line with previously reported studies of IORT treating non-para-aortic regions [16, 19].

Limitations

Limitations include the retrospective nature of the study and relatively small cohort. The precise anatomic contents of the IORT field can be difficult to determine based on simple descriptors such as “para-aortic” which can make determining IORT field overlap in patients with prior radiation difficult. The lack of IORT field images makes the discussion of toxicity more difficult. Patients also received EBRT at different times in their disease course (initial definitive management, for a separate recurrence, and preoperatively with surgery and IORT) which can affect recurrence patterns. Many patients in this study were lost to follow-up, often because they were followed at another institution, and this hinders complete collection of oncologic outcomes and toxicities.

留言 (0)

沒有登入
gif