Predictors of Nonoperative Management Failure and Recurrence in Adults with Acute Appendicitis: A Single-Center Retrospective Study

[Purpose] Failure and recurrence are concerns in nonoperative management (NOM) for acute appendicitis (AA) and interval appendectomy (IA). The goal of the study was to identify predictors of failure and recurrence in patients with AA who underwent NOM. [Methods] A total of 348 adults with appendicitis were treated in our department from April 2016 to October 2021. Among these patients, 337 who did not undergo emergency surgery were divided into those with failed NOM and unplanned surgery (n=28) and successful NOM (n=309). These groups were compared to identify predictive factors for NOM failure. In the NOM-success group, 195 patients were divided into non-recurrence (n=166) and recurrence (n=29) groups to identify predictive factors for recurrence of appendicitis.


Introduction
Acute appendicitis (AA) is a common gastroenterological disease that often requires urgent surgery.
However, nonoperative management (NOM) for children with uncomplicated AA appears to be effective based on less morbidity, fewer disability days, and lower costs compared to surgery 1,2 . There are also reports of NOM for uncomplicated AA in adults without use of appendectomy [3][4][5] . However, failure of NOM is a concern in pediatric and adult patients, with failure rates of 8.5-10% in patients with uncomplicated appendicitis [5][6][7] , an increased failure risk in patients with an appendicolith 6 , and failure in complicated appendicitis involving abscess formation or perforation. Thus, there is a need to identify patients at high risk for NOM failure to avoid unnecessary observation for such patients.
After NOM for appendicitis, there is a need to consider whether patients should undergo interval appendectomy (IA). One concern with IA is recurrence after NOM, since a recent meta-analysis found a recurrence rate of symptoms within one year of 27.4% 8 . However, this also indicates that about 70% of adult patients do not have recurrent appendicitis after NOM. Therefore, careful selection of patients with a low risk of recurrence of AA and requirement for IA may avoid unnecessary surgery 9 . We usually perform NOM for AA cases, except for those with diffuse peritonitis or severe conditions such as septic shock, disseminated intravascular coagulopathy (DIC), or refractory pain control. We then perform IA for cases with appendicitis with fecal stones or abscess formation and for recurrence cases. Herein, we analyzed short-term outcomes of IA and risk factors for NOM failure and recurrence after NOM to evaluate the feasibility of NOM in adult patients with AA.

Patients
Following approval by our institutional review board, a retrospective analysis was performed for 348 adult patients with AA who were treated at Aomori City Hospital from April 2016 to October 2021. Of these patients, 11 underwent emergency appendectomy and 337 were initially treated with NOM. Of the NOM cases, 309 patients were successfully treated and discharged (NOM-success group), whereas general or abdominal conditions deteriorated in 28 patients and unplanned appendectomy (UA) was performed (NOM-failure group).
Analysis of recurrence was performed in the 300 patients in the NOM-success group, after excluding 6 patients diagnosed with appendix or cecum cancer and 3 patients who wanted to undergo surgery at other institutions. We recommended elective IA for patients in whom a fecal stone or abscess formation was detected in the clinical course and for those with recurrence. A total of 105 patients initially underwent IA. No further treatment was initially performed in 195 patients, of whom 166 had no recurrence (non-recurrence group) and 29 had recurrent appendicitis (recurrence group) and subsequently underwent further NOM. Recurrence was de ned as a diagnosis of AA after initial treatment and rehospitalization that required treatment, without considering the period until recurrence.
NOM failed in one patient in the recurrence group and UA was performed. NOM was successful in the other 28 patients with recurrence, of whom 16 underwent subsequent IA and 12 patients refused IA. Thus, in total, 150 patients were treated with appendectomy (excluding emergency appendectomy), including 121 who underwent IA (IA group), and 29 who underwent UA (UA group) ( Figure 1).

Diagnosis of appendicitis
Diagnosis of appendicitis was based on clinical symptoms, such as elevated body temperature, nausea, anorexia, abdominal pain, tenderness in the right lower quadrant, and in ammation in laboratory ndings. A contrast CT scan was also routinely performed, and appendicitis was diagnosed if appendiceal dilatation >5 mm, wall thickening or enhancement, and periappendiceal fat stranding were found. Cases were also divided into uncomplicated and complicated types, with complicated appendicitis de ned based on the presence of an appendicolith or abscess formation.
Nonoperative management (NOM) protocol All patients were started with withholding of food with intravenous uid support and administration of cefmetazole or piperacillin/tazobactam, and addition of metronidazole in cases with abscess formation.
Clinical condition, including mainly abdominal ndings, was evaluated twice daily, and blood tests were usually performed on hospital days 1 and 3. If the patient did not improve without diffuse peritonitis, treatment was changed by addition or a switch to more broad-spectrum antibiotics, such as meropenem.
If conditions worsened despite the change of antibiotics, urgent appendectomy was performed. Patients with an good clinical response to antibiotic therapy, as proved by adequate enteral intake, were discharged. IA was recommended for patients with fecal stones or abscess formation and those with recurrence, who were instructed to visit our hospital after discharge.

Results
Safety of interval appendectomy (IA) The surgical outcomes in the IA and UA groups are shown in  Figure 2).

Malignant lesions
Appendiceal tumors, including low-grade appendiceal neoplasm (LAMN), and colon cancers (cecum and ascending colon) were identi ed in 2.6% (9/348) of cases. Of the 309 patients in the NOM-success group, 6 (1.9%) were diagnosed with malignancy after NOM and then underwent ileocecal resection or right hemicolectomy for cancer. In three patients, a malignant lesion was detected pathologically after IA. Two LAMN cases were kept under observation without additional surgery. One patient was diagnosed with appendix cancer after appendectomy and underwent subsequent ileocecal resection.

Risk factors for recurrence
The recurrence rate was 14.9% and the median age in recurrence cases was 45.0 (range 16-93) years in the patients in the recurrence and non-recurrence groups (

Discussion
This study investigated the feasibility of NOM for uncomplicated and complicated appendicitis, and the outcomes of IA for high-risk patients. The results showed differences between the IA and UA groups, NOM-success and NOM-failure groups, and recurrence and non-recurrence groups, with three main results. First, the short-term surgical outcomes (operation time, blood loss, complication rate, and postoperative hospital stay) were signi cantly better in the IA group than in the UA group. Second, posttreatment CRP level, and fecal stones and abscess formation were independent risk factors for NOM failure. Finally, the initial CRP level was signi cantly lower and the % decrease in CRP was signi cantly higher in the recurrence group than in the non-recurrence group.
The feasibility of NOM and subsequent IA remains controversial in terms of morbidity, e cacy, recurrence, and cost. In 2021, a Japanese nationwide study showed that emergency surgery for complicated AA places the patient at relatively higher risk, and that the risk associated with elective appendectomy is signi cantly lower than that for other methods 11 . A systematic review and metaanalysis showed that immediate surgery is associated with higher morbidity than nonsurgical treatment 12 , and a meta-analysis suggested that, for both uncomplicated and complicated adult AA, NOM with antibiotics was associated with signi cantly fewer complications and a shorter hospital stay 13 .
Thus, the bene ts of NOM and IA seem to be safe surgery with a lower complication rate, as also seen in our results.
There are also reports that indicate concerns with NOM, including one of the above studies showing lower e cacy and higher relapse with NOM compared to appendectomy 13 . An open-label randomized controlled trial in patients with AA in 2007 demonstrated that amoxicillin plus clavulanic acid was noninferior to emergency appendectomy based on rates of 30-day postintervention peritonitis, unplanned appendectomy, and recurrence 14 . A retrospective study of 1081 patients who underwent appendectomy for AA in 2006 showed that prompt appendectomy is necessary based on the risk of developing advanced pathology, and that postoperative complications increase with time 15 . However, these two studies were performed more than ten years ago and modern conservative treatment strategies including new antibiotics may be more effective.
One advantage of NOM is the potential to explore other possible malignancies after NOM. We routinely perform total colonoscopy after NOM for patients >40 years old and we have detected appendiceal tumors or colon cancers in 1.9% (6/309) of successful NOM cases, with 2.5% (3/121) diagnosed incidentally after IA. A retrospective review of IA for complicated appendicitis in 402 patients found a 9% rate for appendiceal neoplasms, with this rate rising to 11% in patients ≥30 years old and 16% in patients ≥50 years old 16 . Other reports have shown rates of 10-29% [17][18][19][20][21] , but the neoplasm rate in patients with complicated appendicitis is thought to be higher 19,22,23 . Therefore, complicated appendicitis, such as that with abscess or mass formation, may be better treated as conservatively as possible from an oncological perspective. NOM and subsequent total colonoscopy are helpful for cancer surgery with complete lymph node dissection for possible appendix or colon cancer.
Regarding NOM for appendicitis, including complicated appendicitis, the two main concerns are failure and recurrence. A randomized controlled study suggested that fever at initial presentation, high presenting serum CRP, and an intraluminal fecalith were risk factors for NOM failure 14 . Another retrospective study showed that the combination of elevated CRP and a fecalith predicted failure of antibiotic therapy in 224 patients who received initial antibiotic therapy 24 . From a radiological perspective, incarceration of an appendicolith and periappendiceal uid on CT images were suggested as predictive factors for failure of NOM in a retrospective study 25 . In our study, the post-treatment CRP level and abscess formation with a fecalith were identi ed as independent risk factors for failure of NOM; in particular, the combination of the presence of a fecalith and abscess formation resulted in an extremely high odds ratio for NOM failure.
The recurrence rate of AA has been reported as 14-35% 8,14,26,27 and recurrence risk factors have been identi ed in several studies. In 2000, a retrospective study in 60 patients with appendicitis suggested an appendix diameter of >8 mm measured by ultrasonography was a risk for recurrence 28 . Recently, a retrospective study in 12,235 patients who underwent NOM for appendicitis found that recurrence was independently associated with young age, male sex, percutaneous abscess drainage, and medical center admission in multivariate analysis 29 . An initial CRP level >10.3 mg/dl was found to be an independent risk factor for recurrence in children with appendicitis with appendiceal masses in a retrospective study in 2017 30 . However, our results suggested that the initial CRP level in the recurrence group was signi cantly lower than that in the non-recurrence group, and that the % CRP decrease was higher in the recurrence group.
Good surgical outcomes were obtained in patients who underwent IA. This may be due to stable general conditions and vital signs and a low level of in ammation in the IA group. In contrast, in the UA group, hyperbacteremia due to severe appendicitis leading to possible abscess formation or perforation or even peritonitis may have made surgery challenging both intra-and postoperatively. Emergency appendectomy is technically demanding due to distorted anatomy, bowel adhesive loops, and di culty closing the appendiceal stump because of in amed tissues 31 . Under such conditions, early laparoscopic appendectomy for AA may be converted to open appendectomy, ileocecal resection, or right hemicolectomy 32 , as found in this study. There may also be complications postoperatively, including infection and organ failure. Emergency surgery should be avoided when possible because of these intraand postoperative di culties, but conservative treatment also has a risk for mortality.
We found that that the post-treatment CRP level and a combination of fecal stones and abscess formation were associated with NOM failure. CRP is broadly known as a marker that indicates an in ammatory response, and the post-treatment CRP level re ects overall in ammation severity and response to antibiotics. The % CRP decrease was not a signi cant factor in multivariate analysis, which may be due to an insu cient effect of antibiotics due to severe in ammation or a result of raising the CRP level belatedly. The presence of abscess formation may also re ect the intensity of in ammation, and there is a need to consider the relationship between fecal stones and in ammation severity, and the location and size of the stone. As mentioned above, incarceration of fecal stones has been suggested to be an independent risk factor for NOM failure 25 and such incarceration may exacerbate in ammation.
This study showed that the initial CRP level was signi cantly lower and the % CRP decrease was signi cantly higher in the recurrence group than in the non-recurrence group. These results are challenging to interpret because the recurrence rate has been found to increase with more severe in ammation in initial appendicitis in other reports 29,30 . Our results suggest that the severity of initial appendicitis may not be related to recurrence. However, the greater % CRP decrease might suggest that patients with recurrence are more vulnerable to infection, including appendicitis. Other factors that may be associated with recurrence include microbiota 33 , daily bowel movement 34 , and morphological characteristics of the appendix 35 .
This study has several limitations, including the small number of patients and the retrospective, nonrandomized design without a control group. Also, the de nition of complicated appendicitis has not been established and target cases differ among studies, which makes it di cult to compare the present results with those from previous studies.
In conclusion, the post-treatment CRP level and fecal stones with abscess formation may be predictors of NOM failure. However, recurrence of appendicitis does not seem to be affected by the severity of previous appendicitis, and prediction of recurrence is di cult based on in ammation severity.