Background

A standard therapeutic approach for locally advanced distal rectal cancer that has reached the dentate line or infiltrates the levator muscles is preoperative radiochemotherapy followed by extralevatoric abdominoperineal resection [1]. Over the last decade evidence has grown that patients with a clinical complete response (cCR) after radiochemotherapy can safely omit surgery and instead be monitored closely and undergo salvage surgery in case of local regrowth [2]. While initially only a minority of patients qualified for this nonoperative management, strategies to maximize this number are the focus of ongoing clinical trials [3,4,5]. As such, the German CAO/ARO/AIO-16 trial conducted by the German Rectal Cancer Study Group will investigate the potential of a novel and innovative radiochemotherapy protocol to increase the clinical complete response rate (NCT03561142). However concerns about the costs associated with required imaging and endoscopic studies have repeatedly been expressed in the past [6]. The current study therefore aimed to simulate the costs arising for this surveillance in the upcoming CAO/ARO/AIO-16 study and compared them with the costs of immediate radical surgery.

Methods

The analysis is from a third party payer’s perspective. The following assumptions were made for the present cost analysis.

Clinical scenario

Patients with locally advanced distal rectal cancer who would otherwise undergo abdominalperineal resection but achieved cCR after radiochemotherapy will be managed nonoperatively and undergo an intensive follow-up regimen incorporating magnetic resonance imaging (MRI) and rectoscopy at the intervals as in our upcoming CAO/ARO/AIO-16 trial to ensure timely detection of local regrowths and successful salvage surgery. Patients completing the entire 5‑year follow-up without local regrowth will undergo 15 pelvic MRIs and rectoscopies.

We assumed that patients who achieve a cCR will develop local regrowths with a likelihood of 25% within 3 years, with the majority of local regrowths developing within the first year [2, 3, 7, 8]. After salvage surgery the follow-up regimen will be equal to the one in patients who underwent immediate surgery. Patients who undergo immediate surgery despite a cCR will not develop local failures prompting surgical or systemic treatment [9]. A total of 12% of patients with an initial cCR will develop distant failure and receive no further surveillance for local regrowth by MRI or endoscopy [10].

Only patients developing isolated local regrowths will undergo salvage surgery, patients with combined local regrowth and distant failure will receive systemic treatment and no salvage surgery. Transitions between health states (cCR, local regrowth, distant failure) will occur in yearly intervals. Two additional scenarios with a 25% increased cost for salvage surgery compared with immediate surgery and a doubled isolated local regrowth rate were simulated.

Fig. 1 provides an overview of the follow-up strategy in both treatment arms and the transition.

Fig. 1
figure 1

Flowchart depicting transition probabilities between health states. For simplification a virtual cohort of 100 patients is shown. cCR Clinical complete response, T transition point, MR magnetic resonance imaging, RS Rectoscopy, FU Follow-up

Full size image

Economic assumptions

Costs for surgical resection and follow-up procedures (MRI and rectoscopy) were discounted at a yearly rate of 3.5% based on the previously published recommendations on cost-effectiveness analysis [11].

For the purpose of this analysis costs for examinations in the outpatient setting were calculated according to the “Gebührenordnung für Ärzte” (GOÄ) [12]. The GOÄ system is the billing system for private patients and charges higher fees than the equivalent catalogue for nonprivate patients, turning the following cost calculations into a “worst-case” scenario. Rectoscopy with consultation, intravenous short anesthesia and monitoring is reimbursed by 63.8 € per exam, pelvic MRI with 383.53 € including consultation, intravenous contrast and computer-based analysis. For abdominoperineal resection using a combined open and laparoscopic approach the diagnosis-related group G‑16B (DRG) system generates costs of a total of 14,510.52 € based on the national average base case value of 3376 € in 2017 [13]. An overview of the applied GOÄ and DRG codes is provided in Table 1. We further assume that costs for systemic treatments or additional surgical therapies for oligometastatic disease do not differ between patients who were included in a wait-and-see protocol or received immediate surgery after achieving a cCR. For the same reason costs of radiotherapy before response assessment were not part of this analysis. Simulations were performed using a Microsoft Excel template (Microsoft, Redmond, WA, USA).

Table 1 Overview of GOÄ (“Gebührenordnung für Ärzte”) and OPS codes (“Operationen- und Prozedurenschlüssel”) and corresponding fees for MRI, rectoscopy and surgery
Full size table

Results

Costs associated with MRI and rectoscopy studies in a patient who does not develop any regrowth or recurrence within 5 years of follow-up sum up to a total of 6344 €. Compared to immediate radical surgery this result in a cost saving of 8167 €. Assuming a 25% likelihood for local regrowth with subsequent salvage surgery and a 2% risk for concomitant distant failure that will not prompt resection as described above, the average costs per patient planned for nonoperative management will be 8299 € over a 5-year time frame including the costs for salvage surgery and therefore cost-saving by 6212 € per patient. According to our model nonoperative management will remain cost-saving compared with immediate surgery as long as summed costs for MRI and rectoscopy visit remain below 955 € per follow-up (Fig. 2).

Fig. 2
figure 2

Average costs per patient for the nonoperative approach with selective surgery in case of local regrowth compared with primary surgery in dependence of the costs for magnetic resonance imaging (MRI) and rectoscopy per visit. Cost-savingness by the nonoperative approach is given as long as costs for MRI and rectoscopy remain below 955 € per visit

Full size image

To account for increased costs that might arise after salvage surgery due to surgical difficulty and postoperative complications, the simulation was repeated with costs for salvage surgery increased by 25% (18,139 €) while leaving the costs for immediate surgery unchanged (14,511 €). In this scenario the nonoperative management still remains cost-saving as long as the costs for diagnostics (MRI and rectoscopy) remain below 885 € per time point. In another sensitivity analysis we doubled the incidence of isolated local regrowths from 23% to 46% within the first 3 years of follow-up. In this scenario the average cost saving will be 3822 € per patient.

Discussion

In our comparative cost simulation between the costs associated with additional imaging and endoscopic follow-up procedures and immediate surgery we found that the nonoperative approach will be cost saving in all simulated scenarios.

The optimal follow-up strategy for patients managed nonoperatively after radiochemotherapy still has to be established. Rectoscopy is the diagnostic tool of choice for the early detection of local regrowths and MRI is generally preferred for the surveillance of mesorectal lymph nodes that are also a potential source of regrowth [14]. The regimen chosen for CAO/ARO/AIO-16 trial and thus for this simulation can be considered rather intensive in particular towards the later years since local regrowths after the third year of follow-up are a rare event [2, 7]. However, as long-term follow-up data on patients managed with a wait-and-see strategy is still limited, surveillance procedures for local regrowths should extend beyond the first 3 years.

A key assumption of our simulations was the equivalence of the wait-and-see strategy and immediate surgery in terms extrapelvic failures and associated costs for systemic or local treatments, which is supported by a variety of studies [3, 7, 15]. One of the scenarios in our model included a 25% higher reimbursement for salvage surgery compared with immediate surgery to cover for postoperative complications and secondary treatments of the latter. Here, conflicting results exist regarding the surgical difficulty and postoperative complications when surgery is delayed beyond the commonly practiced 6 weeks. The recently published GRECCAR-6 trial showed a significantly higher percentage of suboptimal mesorectal specimens and an increased rate of postoperative complications in patients who underwent surgery after 11 vs 7 weeks after the end of radiochemotherapy [16]. However, this data is challenged by the results of the TIMING trial where surgeons did not report an increase in surgical difficulty despite a higher rate of fibrosis when surgery is delayed as far as 20 weeks after the end of radiochemotherapy [17].

The third scenario doubled the incidence of isolated local regrowths to 46% within the first 3 years of follow-up, which is higher than any rate reported in a recent review [18]. In this scenario almost half of the patients will receive both imaging studies at least for one year and salvage surgery. Despite this very unfavorable scenario the nonoperative approach remained cost saving. The local regrowth rate clearly is a function of the definition of the cCR after radiochemotherapy and will vary accordingly. For instance, Maas et al. report a local regrowth rate as low as approximately 5% with a very strict definition incorporating not only endoscopic but also MRI-based criteria for the response assessment of the primary tumor [3]. Lowering the threshold to define a clinical complete response will qualify a lot more patients but inevitably result in a higher local regrowth rate as seen in a recent report of the OnCoRe registry where the actuarial local regrowth rate was 38% after 3 years. However, salvage rates and oncological outcome were excellent favoring a rather generous cCR definition.

Advanced imaging techniques will open doors to tailored treatments with the selective use of treatment modalities such as surgery, chemotherapy or radiotherapy based on imaging parameters [19]. The costs associated with these strategies will have to be considered in the context of the health benefit achieved for the patient. An example for this is the positron-emission tomography (PET) guided selective use of cervical neck dissection after definitive radiochemotherapy, which was recently shown not only to be cost saving compared with immediate neck dissection but also result in an additional lifetime gain of 0.13 quality adjusted life years (QALYs) [20]. Utility data that could be used for cost-effectiveness analysis in rectal cancer patients after nonoperative management is very limited. In a study comparing quality of life (QoL) by using the EQ-5D index patients had a significant 10 point gain in QoL when a primary anastomosis was performed after perforated diverticulitis compared with a Hartmann procedure and stoma placement [21]. Together with other reports of acceptable rectal function and physical functioning it appears most likely that the nonoperative approach of rectal cancer after radiochemotherapy is not only cost saving but also cost efficient.

One of the challenges to maximize the number of patients that can be treated with a nonoperative approach is to establish treatment protocols that can increase the complete response rates. These strategies will generate costs that are not reflected in this analysis, yet the most promising approach described in the literature, namely “total neoadjuvant therapy” with consolidation chemotherapy and delayed surgery will not produce high costs since the drugs used in the commonly applied FOLFOX regimen are generically available and have a favorable toxicity profile [22, 23]. Other parameters not considered in our model include costs associated with the life-long use of colostomy bags in patients undergoing immediate surgery or potential endoscopic interventions for late toxicity of radiotherapy in nonresected rectal tissue.

In summary our work suggests that the costs of the intensive follow-up regimen in the upcoming CAO/ARO/AIO-16 will not exceed the costs of immediate radical surgery. The trial will also assess utility data in a prospective manner to allow cost-utility analysis.