Barriers to Offering Organ Preservation for Rectal Cancer in a Predominantly Hispanic Safety Net Hospital

Abstract

Background:

Total neoadjuvant therapy (TNT) for locally advanced rectal cancer (LARC) has shown promise in achieving pathologic complete response (pCR) and enabling organ preservation through watch-and-wait (WW) strategies. However, implementation of WW protocols in diverse patient populations and safety-net hospitals faces unique challenges.

This objective of this study is to evaluate TNT outcomes and identify barriers to WW implementation in a predominantly Hispanic safety-net hospital in South Texas.

Methods:

A retrospective review of 40 LARC patients at an academic tertiary referral center in South Texas who were treated with TNT between 2018 and 2023 was conducted. Patient demographics, disease characteristics, and pathologic complete response (pCR) rates were analyzed. A survey of multidisciplinary providers assessed perceived institutional and patient-related barriers to WW implementation.

Results:

The cohort was 70% Hispanic, with a median age of 54 years. Most patients had advanced disease at diagnosis (57.5% T4, 65% N2). The pCR rate was 18.5% (5/27) among patients undergoing surgery. Re-review of MRIs for pCR patients revealed that 2/5 had minimal residual disease. The provider survey identified MRI quality variability, lack of dedicated treatment coordinators, and concerns about patient compliance and financial barriers as key obstacles to WW implementation.

Conclusions:

Despite advanced disease presentation in a predominantly Hispanic population, TNT achieved pCR rates comparable to international trials. Institutional and patient-level barriers to WW were identified, informing the development of a tailored WW protocol for this unique patient population.

Introduction

Locally-advanced rectal cancer (LARC) has traditionally been treated with chemoradiotherapy followed by total mesorectal excision (TME) with or without adjuvant chemotherapy (Bahadoer RR 2021). Recently, evidence has emerged that foregoing surgical removal of the rectum is a possibility for select patients following completion of total neoadjuvant therapy (TNT) ((Garcia-Aguilar J 2022), (Dattani M 2018), (van der Valk MJM 2018), (Appelt AL 2015)). This approach consists of intense surveillance to monitor for local recurrence but can result in organ preservation rates as high as 77.9% at three years (Dattani M 2018). Results from the Organ Preservation in Patients with Rectal Adenocarcinoma Treated with Total Neoadjuvant Therapy (OPRA) Trial showed that patients who demonstrated a clinical complete response (cCR) following completion of TNT were able to enter a watch-and-wait (WW) pathway without compromising disease-free or overall survival ((Dattani M 2018)). Watch-and-wait pathway consists of intensive surveillance via routine physical examination (digital rectal examination), proctoscopy and MRI of the pelvis. This is performed every four months for the first three years and every six months for the following two years. Furthermore, studies have demonstrated successful salvage resections in cases of local regrowth among patients in the WW pathway (Dattani M 2018), (van der Valk MJM 2018), (Martens MH 2016).

Hispanic and Latinx patients develop colorectal cancer at earlier ages and are diagnosed at later stages when compared to non-Hispanic white individuals (Rahman R 2015). Our facility, a safety net hospital serving the South Texas region, sees a large number of patients who identify as Hispanic or Latinx. We have been selectively offering TNT to rectal cancer patients since 2019 but have not offered WW, despite interest in doing so both amongst providers and patients. The benefits of avoiding surgery are numerous but are especially evident for working age adults in which both time off for surgery and ostomies present significant challenges (Couwenberg AM 2020), (Custers PA 2023), (R. Colbran 2024). Given the morbidity and quality of life concerns in these young patients, the ability to offer WW to patients who appropriately respond to TNT is of great importance to our patient population.

We sought to investigate our institution’s outcomes with LARC to determine if WW was feasible in our patient population. We also investigated barriers to offering WW to patients following completion of TNT by performing a survey of institutional providers who contribute to medical decision making by these patients. Our goal with this study was to identify variables that could be targeted with specific strategies to improve our ability to offer organ preservation to patients with LARC in a safety net hospital setting.

Materials and methods

Data source:

After institutional review board approval was obtained, retrospective and prospective chart review was performed on enrolled patients. Demographic data was collected including age, race, ethnicity, occupation, insurance status, comorbidities, history of colon cancer screening, and personal or family history of any cancer. Disease data was collected including clinical tumor, node, metastasis (TNM) stage at diagnosis, pathology results from biopsy, date of diagnosis, and whether the malignancy was identified as a result of symptoms (e.g. rectal bleeding) or on screening colonoscopy. Treatment data was collected including method and duration of neoadjuvant chemotherapy, method and duration of neoadjuvant radiotherapy, whether the patient was determined to have clinical response to treatment, whether the patient underwent surgical resection, date of surgical resection, extent of resection performed (i.e. low anterior resection [LAR] or APR and whether an ostomy was created), additional surgical procedures (e.g. partial hepatectomy or metastatectomy), surgical complications, hospital length of stay (LOS), and surgical pathology results including pathologic TNM staging. Use of adjuvant chemotherapy or radiotherapy was also noted.

Patient population:

Our institution is a tertiary referral center that serves a large area of South Texas. TNT therapy was initiated in 2018 at our institution. The patients included in our study were treated at our institution between 2018–2023. Patients were identified via our institution’s multidisciplinary tumor board which consists of surgeons, radiation oncologists, medical oncologists, pathologists, radiologists, and care coordinators. Patients were eligible for inclusion if they had LARC for which the tumor board recommended TNT. Exclusion criteria included age less than 18 years, patients incarcerated at time of treatment, pregnant individuals, and patients who did not receive TNT.

Outcomes:

The primary outcome of this study was pathologic complete response (pCR). We defined pCR as no residual tumor cells at the primary tumor site or in regional lymph nodes, in accordance with prior studies (Garcia-Aguilar J 2011). We also analyzed treatment outcomes including length and type of treatment provided, surgical and pathologic findings, and disease-free and overall survival.

Survey:

A survey was conducted among providers who care for patients with rectal cancer, including physicians, physician assistants, and nurse practitioners, via RedCap (Harris PA 2009). The survey asked participants to identify their field of practice (surgery, radiation oncology, radiology, medical oncology, or other). Participants were then asked to rank institutional and patient-related barriers in terms of perceived significance in their clinical decision making, using a Likert scale from 1 (not very significant) to 5 (very significant) (Table 1). The barriers listed in the survey were based on what various providers have observed while caring for patients at our institution. Free text boxes were available for participants to list any other factors they deemed to be significant in their clinical decision making. Factors were weighted based on perceived significance (rate of perceived significance multiplied by number of responses at that rating) then divided by the total number of respondents to give an average rating.

Table 1.

Prompts and responses included in the provider survey. Participants were asked to rank each response on a Likert scale.

Think about institutional barriers that would steer you away from offering or recommending organ preservation to a patient. Please rate the following in terms of how important they are for your clinical decision making, with 1 being not very significant and 5 being very significant.

Lack of dedicated radiologist to read all rectal cancer MRIs

Lack of dedicated treatment coordinator

Lack of endoscopy capabilities in clinic

Different facilities for clinic follow up and endoscopy (i.e. Mays Cancer Center and MARC)

MRIs being performed at different locations leading to varying quality

Lack of a standardized protocol for reading MRIs (i.e. no clinical correlation)

Think about patient-related barriers that would steer you away from offering or recommending organ preservation to a patient. Please rate the following in terms of how important they are for your clinical decision making, with 1 being not very significant and 5 being very significant.

Concerns regarding medical literacy

Concerns regarding access to transportation to follow-up appointments

Concerns regarding ability to maintain insurance or pay for ongoing surveillance

Questionable motivation/ willingness to comply with recommended surveillance

Concerns regarding lack of social or family support

Statistical analysis:

Statistical analysis was performed in R (version 4.3.1). Categorical variables were summarized as frequencies and percentages while continuous variables were summarized as medians and interquartile ranges (IQR). The institutional survey was treated as a qualitative metric as valid statistical analysis would not be possible due to the limited number of participants.

Results

Patient characteristics:

Forty patients with LARC were treated with TNT and met inclusion criteria for our study. Demographics and disease features at time of diagnosis are shown in Table 2. Thirty patients (75%) were male, and the median age at diagnosis was 54 (interquartile range [IQR] 49.8–63.3). Twenty-eight (70%) patients identified as Hispanic or Latinx, ten (25%) identified as non-Hispanic white, and two (5%) identified as Black or African American. Median body mass index was 25.9 (IQR 22–30.4) and 14 patients (35%) used tobacco products at time of treatment. Thirty patients (75%) had a family history of any cancer, and nine (22.5%) had a family history of colorectal cancer. At time of chart review, three patients (7.5%) were uninsured; 26 (65%) had private insurance or Medicare/ Medicaid; and 11 (27.5%) were participating in an income-based discount and repayment plan coordinated by our institution.

Table 2.

Patient demographics and disease features at time of diagnosis.

n	40
Male sex, n (%)	30 (75.0)
Age in years, median (IQR)	54 (49.8–63.3)
Race/ ethnicity, n (%)
White, non-Hispanic	10 (25.0)
Hispanic or Latinx	28 (70.0)
African-American	2 (5.0)
Tumor size in cm, median (IQR)	6.9 (5.0–8.7)
T-stage at diagnosis, n (%)
T2	2 (5.0)
T3	14 (35.0)
T4	24 (60.0)
N-stage at diagnosis, n (%)
N0	2 (5.0)
N1	12 (30)
N2	26 (65.0)

Most patients (39, 97.5%) were diagnosed with LARC after a diagnostic endoscopy for symptoms such as rectal bleeding. 33 patients were eligible for screening colonoscopy yet, 28 (84.5%) had never had one before symptom onset. Only one patient (2.5%) had their cancer found on screening colonoscopy. Most patients had advanced disease at time of diagnosis, with 23 patients (57.5%) having T4 and 26 patients (65%) having N2 disease. Three patients (7.5%) had metastatic disease at time of diagnosis but underwent TNT and resection. Median tumor size was 6.9cm (IQR 5.0–8.7). Median tumor distance from the anal verge was 7.1cm (IQR 3.8–9.5). Mid-rectal (5–10 cm from anal verge) cancer was most common (17 patients, 42.5%) followed by low (0–5 cm from anal verge, (15 patients, 37.5%)) and high (10–15 cm from anal verge (eight patients, 20%)).

Treatment outcomes:

All 40 patients underwent neoadjuvant chemoradiotherapy. The most common chemotherapy regimens were FOLFOX (n=17, 42.5%) and XELOX (n=17, 42.5%). The median number of chemotherapy cycles was six (IQR 4.25–9.0). Most patients (25, 62.5%) were treated with short-course radiotherapy at 25 Gy and the remaining 15 patients (37.5%) were treated with long-course radiotherapy with capecitabine. Twenty-three patients (57.5%) underwent consolidation chemotherapy.

Criteria for determining clinical response has been described using a combination of physical exam, endoscopy, and MRI (Garcia-Aguilar J 2022). We found that endoscopy was not routinely used in post-TNT reevaluation at our institution, so clinical response was determined based on post-TNT MRI interpretation and physical exam as noted in the surgical clinic note. Twenty-seven patients (67.5%) ultimately underwent surgery with curative intent between December 2019 and February 2024. Median time from end of TNT to undergoing surgery was 12.1 weeks. Of the 13 patients who did not undergo surgery, six (46.2%) were lost to follow-up; three (23.1%) developed metastatic disease; two (15.4%) declined surgical intervention or further treatment; one (7.7%) sought treatment at a different facility; and one (7.7%) was determined to have unresectable disease.

Of the 27 patients who underwent surgery, most (20, 74.1%) underwent low anterior resection with diverting loop ileostomy. Six patients (22.2%) underwent abdominoperineal resection with end colostomy and one (3.7%) underwent total proctocolectomy for a diagnosis of Familial Adenomatous Polyposis. Three patients (11.1%) underwent concurrent liver resection in combination with one of the above procedures. The most common approach was robotic (15, 55.6%) followed by open (8 patients, 29.6%) and laparoscopic (4, 14.8%). Pathologic complete response was seen in five patients (18.5%). Nineteen patients (70.4%) had partial pathologic response and three (11.1%) had no pathologic response. Pathologic tumor and nodal stages are shown in Table 3.

Table 3.

Treatment outcomes for patients undergoing surgical resection with curative intent.

Patients undergoing surgical resection with curative intent, n	27
Pathologic T-stage, n (%)
ypT0 (no cancer detected)	6 (22.2)
ypT1	0
ypT2	8 (29.6)
ypT3	10 (37.1)
ypT4	3 (11.1)
Pathologic N-stage, n (%)
ypN0	21
ypN1	4
ypN2	2
Pathologic response, n (%)
Absent	3 (11.1)
Partial	19 (70.4)
Complete	5 (18.5)

Survey:

Five surgeons, four medical oncologists, and two radiation oncologists participated in our survey. Average scores of perceived significance for each barrier are shown in Table 4. The institutional barrier of highest concern was MRIs being performed at different locations leading to varying quality of images. In the free text box, participants highlighted “quality of the imaging and the reads are a real problem”, “lack of clinic support to ensure protocols are being correctly followed”, and “complexity of moving patients through our systems with two MyCharts [online patient portals] that patients need to toggle between to know where they need to be when...we need more oncology nurse navigators on the GI service lines so patients can have more personalized assistance moving through complex treatment plans involving multiple disciplines.”

Table 4.

Results from provider survey regarding barriers to offering watch-and-wait therapy.

Institutional barriers	Average score
Lack of dedicated radiologist to read all rectal cancer MRIs	3.0
Lack of dedicated treatment coordinator	3.3
Lack of endoscopy capabilities in clinic	3.3
Different facilities for clinic follow up and endoscopy (i.e. Mays Cancer Center and MARC*)	2.9
MRIs being performed at different locations leading to varying quality	3.8
Lack of a standardized protocol for reading MRIs (i.e. no clinical correlation)	3.4
Patient-related barriers
Concerns regarding medical literacy	3.7
Concerns regarding access to transportation to follow-up appointments	4.3
Concerns regarding ability to maintain insurance or pay for ongoing surveillance	5
Questionable motivation/ willingness to comply with recommended surveillance	3.8
Concerns regarding lack of social or family support	3.5

^*Medical Arts and Research Center: Clinic site and outpatient endoscopy center

The physician-reported, patient-related barrier of highest concern was ability to maintain insurance or pay for ongoing surveillance, closely followed by concerns regarding access to transportation for follow-up appointments. Participants additionally stated “follow up and compliance are an issue with many patients”, “medical literacy to understand the need for surveillance”, “gauging level of comprehension can be difficult”, and “access and maintaining insurance or CareLink [income-based repayment plan] is a huge factor”.

Discussion

In this retrospective single-institution study we demonstrated a pCR rate of 18.5% which compares favorably with a systemic review and meta-analysis done by Kasi et al., which found pCR ranged from 17.2%–38.5% in international studies [13] (Kasi A 2020). The cohort that our study observed has some distinct differences that merit being highlighted. 70% of our patients were Hispanic, 20% of whom did not speak English, compared to the OPRA trial which only had 5.5% Hispanic participants (Garcia-Aguilar J 2022). 35% of our patients were either uninsured or on a subsidized medical payment plan available only to residents of Bexar County. This is significant because the population examined in our trial is one that is usually not the majority in other major trials. This finding suggests that even in safety net hospitals, comparable pCR rates are achievable and thus, WW therapy can be offered. Unfortunately, almost 85% of our patients >45 years of age had never undergone a screening colonoscopy. Regarding disease severity, 57.5% of our patients had T4 disease and 65% of patients had N2 disease, compared to the OPRA trial which had the bulk of their patients with T3 disease but with similar node positivity (Garcia-Aguilar J 2022). Despite the severity of disease at time of detection and noncompliance with screening, the pCR rate was 18.5%. This demonstrates that a significant amount of these patients would likely qualify for organ preservation at restaging.

The findings of the survey conducted by our group revealed that variation in imaging quality performed at multiple institutions leads to the inability to definitively conclude whether there is imaging evidence of cCR. Flexible sigmoidoscopy to assess for cCR was also not offered due to difficulty in maintaining adequate endoscopy suite availability. Along with this, inadequate ancillary support to stay in consistent communication with our patients, many of whom do not speak English, is a perceived impeding factor. These are all modifiable factors that our institution can correct to aid in the feasibility of WW therapy.

Limitations of our study include the retrospective observational method of study, leading to some forms of bias in data collection, interpretation and the survey findings. We were not able to report survival outcomes due to almost half of patients having no long-term follow-up. This highlights the point that to safely utilize WW, there must be an established institutional protocol and support from the institution to support the more comprehensive follow-up required. To employ WW, providers must know their patients have a consistent and reliable method of following up often for clinic appointments, imaging and colonoscopies. Over a third of our patients were uninsured or enrolled in a payment plan at the time of chart review, making WW surveillance difficult to ensure.

In conclusion, we demonstrate that TNT for LARC results in a 18.5% pCR rate at our institution, despite our patient population having advanced disease and socioeconomic disparities. The hallmark concern from a provider standpoint is the ability to offer WW therapy safely and reliably. We identified barriers to offering WW at our institution that have led to the development of a WW protocol that will allow us to safely conduct surveillance in our patient population. Our protocol includes numerous adjustments to prior WW protocols, including radiation, care coordination, radiographic, and educational components that we hypothesize will allow for successful WW implementation at our cancer center. We plan to study the efficacy of our protocol in a phase II trial starting this year.