Cholecystectomy continues to be one of the most frequent general surgery procedures and is performed for a variety of indications that are predominantly benign in nature. Historical dogma suggests that resected gallbladder specimens should be evaluated postoperatively by a pathologist; however, this tradition is not reflected in strong evidence or formalized guidelines [1]. This screening modality has been shown to detect incidental gallbladder carcinoma (iGBC) [2]. Prior studies have shown iGBC in 0.3–0.9 % of gallbladders following cholecystectomy [3]. Detection of iGBC (also referred to as unexpected gallbladder cancer in the literature) can potentially alter clinical management by necessitating a more extensive hepatic resection and lymphadenectomy [4–7]. It has shown to have improved outcomes over gallbladder carcinoma diagnosed via symptoms; however, the outcomes remain poor [3]. Little evidence exists in the literature on the patient value conferred with pathological evaluation during routine benign cases.

While gallbladder carcinoma is the most common biliary tract malignancy, it remains quite rare in Western populations with an incidence in USA of 1.2/100,000 [8, 9]. It is estimated that half of all gallbladder cancers are now found incidentally during presumed benign resections, with the other half diagnosed via local symptoms or upon metastasis [7, 10]. Unfortunately, GBC is very difficult to diagnose early by any other means and over 75 % of GBC is not resectable at the time of diagnosis due to its aggressive pathophysiology [11]. iGBC as a subset of this population boasts better outcomes due to a higher likelihood of early-stage detection [12]. However, no evidence to date justifies the global screening of benign gallbladder specimens as a cost-effective screening modality. The principal aim of this study is to review the pathology results from gallbladder specimens sent for routine pathology review and to determine whether excised and resected specimens should be sent for postoperative histopathological analysis.

Methods

The University of Vermont Institutional Review Board approved this study. Patient data were collected in a secure fashion to preserve patient confidentiality.

This is a retrospective analysis of all cholecystectomies performed at a single academic medical center (University of Vermont Medical Center) from 2009 to 2014, for a total of 2153 patients. We then excluded patients who underwent cholecystectomy for a known or suspected malignancy (n = 119, patients with a known gallbladder polyp preoperatively (n = 24) or patients who underwent cholecystectomy for a non-biliary indication such as traumatic injury (n = 26)). This resulted in a total of 1984 included patients (Fig. 1).

Fig. 1
figure 1

Schematic for study design, including relevant inclusion and exclusion criteria. Study period was from June 2009 until June 2014

Full size image

The baseline characteristics regarding the patient population preoperative data were collected, which included age, sex, surgeon, preoperative diagnosis, race/ethnicity, smoking status, ASA class, comorbidities, and body mass index (Table 1). We also collected information relevant to their perioperative course, including elective or emergent nature of operation, surgeon, and surgery type (laparoscopic, open, laparoscopic converted to open, intraoperative cholangiography).

Table 1 Baseline characteristics of study population with comparison to the characteristics of patients within population found to have iGBC
Full size table

Pathology results were tabulated and quantified. Overall, a total of 55 unique pathological findings were collected and tabulated for each patient—when a finding was included in the report summary, it was included in the patient’s analysis and data set. Results were quantified as proportions with 95 % confidence intervals determined with binomial exact distribution.

Histopathological analysis of gallbladder specimens was performed by the department of surgical pathology per an existing protocol. This involves gross examination of the intraluminal and external surfaces and microscopic/histological examination of three representative slices: gallbladder fundus, middle of body, and cystic duct margin (infundibulum). If carcinoma, dysplasia, or other unanticipated findings are noted on any of the three sample slices, the pathologist proceeds with complete histological evaluation of the gallbladder. Gallbladder cancer, when discovered, was staged using the TNM staging implemented by the AJCC [13].

Patients found to have iGBC had their postoperative course followed for key outcome variables, such as follow-up extended resection or lymphadenectomy, neoadjuvant chemo-radiation, and length of hospital stay. Pathological variables gathered included tumor size, grade, stage, margin status, and concurrent pathological features. Key outcome variables collected included second operation for wide resection and lymphadenectomy, adjuvant chemotherapy and/or chemo-radiation, survival status (alive or dead at present), and cumulative survival time.

Primary outcomes used for calculation of life years (LYs) included survival time post-initial cholecystectomy compared to expected survival with simple cholecystectomy alone based on prior data. Cost data regarding the expense of pathological testing of the specimen were obtained via the Jeffords Institute for Quality and are specific to our medical center.

Baseline data for patients found to have iGBC (n = 5) were compared to patients with benign findings (n = 1979) utilizing Fisher’s exact test for categorical variables and Wilcoxon’s rank sum test for interval variables.

Results

The 1984 patients were categorized into two groups: those with benign findings (n = 1979) and those with iGBC (n = 5) (Table 1). The iGBC group had a statistically significant higher rate of open procedure (p = 0.03) and longer operative time (p = 0.01), implying greater intraoperative difficulty during cholecystectomy. The iGBC group trended toward being older (average age 61.4 vs. 48.4) and less likely elective (71.5 vs. 60 %); however, these findings were not statistically significant. We did not find differences in race or ethnicity, though traditionally gallbladder carcinoma is found to be more common in Native American and Asian populations [14]. These groups are under-represented in our hospital’s area of referral.

The observed rate of iGBC was 0.25 % of the total population (confidence interval 0.08, 0.59). The rate of dysplasia was 0.71 % (95 % CI 0.39, 1.20). More common findings included chronic cholecystitis in 89 % of gallbladder (87.4, 90.3), cholelithiasis in 81 % (79.1, 82.6), acute cholecystitis in 20.2 % (18.5, 22.0), and cholesterolosis in 17.9 % (16.3, 19.7) (Fig. 2). A comprehensive list of the incidence of all 55 noted unique pathological findings is provided in the supplemental material (Supplemental Table 1).

Fig. 2
figure 2

Common pathological findings of gallbladder after cholecystectomy. This figure illustrates the most common or relevant pathological findings to the discussion of the paper. Error bars represent 95 % CI as determined by binomial exact distribution. A comprehensive list of pathological findings is included in the supplemental figures

Full size image

Of the 5 patients found to have iGBC, they varied in stage, grade, and histopathological features. The least advanced tumor stage was pT2N0, and the most advanced was pT3N1pM1. The two patients diagnosed at stage T2 have achieved ongoing disease-free survival (35+ months), and the three patients with stage T3 disease had a mean survival of 4.3 months (with one declining further surgery) (Table 2). The two long-term survivors had characteristics favorable for complete surgical cure, such as relatively early stage (T2) and minimal to no residual disease on radical re-resection.

Table 2 Characteristics of incidental gallbladder patient pathology
Full size table

Hospital cost for pathological evaluation per gallbladder (including hospital and provider costs, both direct and indirect) totaled $76.92 at this institution. The total cost of screening 1984 patients with presumed benign gallbladders was $152,609.28. Cumulative survival after iGBC was 84 months (7.0 years) (see Table 2). Based on previous data in the literature, expected survival without further intervention would be 16 months for T2 tumors and 8 months for T3 tumors [15], with a cumulative expected survival with simple cholecystectomy alone in this cohort being 56 months. Total charges per LY gained to date (assuming cumulative 28-month gained) were $65,403.98. The number needed to screen (NNS) to save one life from GBC was 992.

Discussion

Laparoscopic cholecystectomy has become an increasingly common treatment modality for various maladies of the biliary tract, most notably symptomatic cholelithiasis, cholecystitis, and biliary dyskinesia.

Previous literature reports the incidence of iGBC to range from 0.25 to 3 % [16]. Depending on the stage at the time of diagnosis, this unexpected discovery may mandate further surgical resection, including hepatic resection of segments IVb and V, as well as pericholedochic and hepatoduodenal lymphadenectomy [16].

Gallbladder carcinoma continues to have an extremely poor prognosis, with 5-year survival below 5 % [11, 17]. Further, most long-term survivors are those with early-stage carcinoma confined to the gallbladder itself. Evidence does suggest a better prognosis in patients with iGBC, rather than patients with symptomatic or preoperatively diagnosed GBC [18].

Surgical management of GBC varies depending on stage of diagnosis and remains somewhat controversial. The consensus among hepatopancreatobiliary surgeons seems to suggest that T2 and above tumors require further resection of the adjacent liver parenchyma in the gallbladder fossa (segments V and IVb), as well as lymphadenectomy to obtain regional control. T1a tumors are usually well controlled with simple cholecystectomy (assuming no bile spillage), and T1b management remains controversial. R0 resection remains key to long-term survival and disease management [10, 19]. While T3 and T4 tumors also benefit from radical resection, T2 tumors have the best outcomes after radical repeat surgery [7, 19]. We demonstrate similar findings in our data, with the T2-staged patients achieving the longest survival after definitive surgery. Previous data suggest improved outcomes when minimal or no carcinoma is present on repeat operation [7], which concurs with the pathological findings of our long-term survivors.

Little investigation has occurred thus far to determine the quality of pathological evaluation of the gallbladder as a screening modality for gallbladder carcinoma. As healthcare measures become increasingly cost-conscious, it is critical to evaluate whether interventions we provide as surgical teams (particular ones applied in a ubiquitous manner) are conferring value to the patient. Common simple measurements used to determine quality in a screening modality, such as life years (LYs) gained, have not been applied prior to this study.

Our calculation yielded a cost of $65,404 per LY gained with pathological screening. This calculation, however, makes a key assumption that almost certainly overestimates the cost. Two patients are currently still alive with no evidence of recurrence. If they are able to achieve further long-term disease-free survival, the total LY’s gained will certainly increase and cost/LY will decrease dramatically. Therefore, we can estimate that cost of screening/LY gained is at most $65,404 and likely a lower value. While the value of one LY is controversial and varies within the literature, one acceptable standard is considered to be $129,000 per QALY (quality-adjusted life year), the estimated cost of maintaining a life on hemodialysis [11, 19]. However, this figure is of debate, as others cite lower values of US $50,000/QALY as a metric of value [9].

We did not utilize more complex measures to determine values including quality-adjusted life years (QALYs) and disability-adjusted life years (DALYs). Due to the extensive morbidity of hepatic resection and adjuvant chemo-radiation, leading to a low quality of life during treatment for gallbladder carcinoma, it is likely that these figures will be significantly lower than the more simple and commonly utilized LY measurement tool [9]. We also did not incorporate the economic benefit of the detection of dysplasia or adenoma (found in 14 patients within the cohort). As gallbladder cancer has known familial clustering [20], the presence of precancer within a specimen would certainly be valuable information to confer to the patient.

The data presented here appear to narrowly justify, based on LY gained per healthcare dollar spent, the universal screening of gallbladder specimens with pathological analysis. Despite this, an argument could be made for the utilization of selective screening of specimens based on patient risk factors for potential GBC (including age, sex, race/ethnicity, or intraoperative characteristics).

Routine pathological analysis will likely remain a component of benign biliary care, as it continues to be the only effective means of detecting GBC at a point in which meaningful intervention remains possible. Improved screening modalities for GBC in high-risk populations may ultimately prove more economically efficient.

Limitations of the study include the patient population associated with the single-center study, which may not reflect the demographics of other regions, nations, and medical centers. Thus, the results may not be applicable to patient populations dissimilar to our study. A second limitation is the size of the patient population—which is not optimal for calculation of LYs as the number of patients in the affected group was low (n = 5). Further, as two patients remain alive at this point, the potential life years may be underestimated. Future investigations should be directed toward applying these principles of quality to gallbladder pathological evaluation on a national (or multi-institutional) level to provide greater statistical power to the study. Finally, as no early-stage (T1a/b) disease was found in our cohort, the three histological slices protocol utilized by our pathology department may be insufficient to detect early disease.

Conclusions

The incidence of gallbladder carcinoma found on pathology following cholecystectomies performed for benign indications varies within the literature. Our results indicate that in our patient population, gallbladder carcinoma may be more rare than previously described patient pools. Despite the rarity of findings and the cost of screening, the huge benefit conferred with radical resection of tumor (particularly in T2 stage) makes pathological analysis a worthwhile endeavor.

Ultimately, the decision to send the resected specimen for pathological analysis should be left to the surgeon and the patient. Patients at higher risk of potential gallbladder carcinoma, including older patients or cases converted to open, should certainly be evaluated by a pathologist. Surgeons should strongly consider on-table evaluation of specimen to assure that there are no gross abnormalities visually and by palpation [21]. Targeted pathological analysis based on epidemiological risk factors, intraoperative findings, and on-table evaluation could be a feasible and more cost-effective alternative to universal screening. Future cost–benefit analysis studies may elucidate further whether routine pathological evaluation should be universally sought.