Clinical usefulness of KRAS mutational analysis in the diagnosis of pancreatic adenocarcinoma by means of endosonography-guided fine-needle aspiration biopsy
Dr A. Castells, Department of Gastroenterology, Institut de Malalties Digestives, Hospital Clínic, Villarroel 170, 08036 Barcelona, Catalonia, Spain.
Aim : To establish the usefulness of KRAS mutational analysis in the diagnosis of pancreatic adenocarcinoma by comparing this technique with conventional cytology in aspirates obtained by endosonography-guided fine-needle aspiration.
Methods : All consecutive patients with pancreatic focal lesions undergoing endosonography-guided fine-needle aspiration were included. Samples were obtained with the concourse of an attendant cytopathologist. Detection of codon-12 KRAS mutations was performed by the restriction fragment length polymorphism-polymerase chain reaction method. The effectiveness of conventional cytology, KRAS mutational analysis and their combination was established with respect to the definitive diagnosis. A cost-effectiveness analysis was also performed.
Results : Thirty-three patients had pancreatic adenocarcinoma and 24 patients had other lesions. A total of 136 samples was obtained. In patients in whom specimens were adequate (93% for cytology; 100% for mutational analysis), the specificity of both techniques was 100%, whereas the sensitivity favoured cytology (97% vs. 73%). When inadequate samples were considered as misdiagnosed, a combination of both techniques reached the highest overall accuracy (cytology, 91%; mutational analysis, 84%; combination of both, 98%).
Conclusions : Cytology from aspirates obtained by endosonography-guided fine-needle aspiration is the most precise single technique for the diagnosis of pancreatic adenocarcinoma. However, when adequate specimens are not available to reach a cytological diagnosis, the addition of KRAS mutational analysis represents the best strategy.
Cancer of the exocrine pancreas is the fourth most common cause of cancer deaths in Western countries. Although there has been a tremendous advance in imaging techniques, the diagnosis of this disease remains difficult. Pre-operative pathological confirmation is hampered by the retroperitoneal location of the gland and specific tumour characteristics, such as the presence of indurated tissue, cystic compounds and surrounding inflammatory infiltrate.1, 2 In this scenario, the detection of mutations in the KRAS gene, a molecular event present in up to 80% of pancreatic adenocarcinomas,3 has been proposed as a useful approach to increase the diagnostic efficacy of conventional cytology from percutaneous fine-needle aspirates1, 2 or pancreatic juice.4, 5 Previous studies have reported an increase in overall accuracy of up to 20% when KRAS mutational analysis is combined with conventional cytology in specimens obtained by percutaneous biopsy, especially when the cytology report is not conclusive due to the presence of suspicious cells, insufficient cellular material or normal-appearing duct cells.2, 6
Recently, endosonography-guided fine-needle aspiration biopsy has been identified as a feasible, safe and highly effective diagnostic technique for pancreatic cancer.7–14 Although endosonography-guided fine-needle aspiration biopsy has been reported to be superior to percutaneous biopsy performed under ultrasonography or computed tomography guidance,15–17 and its specificity generally reaches 100%, the amount of aspirated specimen is sometimes not sufficient to obtain a pathological diagnosis (sensitivity, 80–90%).11 Different approaches have been proposed to overcome this limitation, including the use of needles with a larger diameter or tru-cut capability,18, 19 increasing the number of passes, the presence of a cytopathologist to improve the yield of satisfactory specimens20, 21 and the introduction of KRAS mutational analysis.22 It has been suggested that, when a pathologist is present, the diagnostic yield increases by 10%.21 On the other hand, the usefulness of KRAS mutational analysis has only been evaluated partially by Tada et al.,22 suggesting that it might increase the effectiveness of conventional cytology. However, it should be noted that the cytology results were poorer than those reported previously,7–11 probably reflecting the fact that this study was performed without on-site examination and with a limited number of passes.22
The present prospective investigation was aimed at establishing the actual usefulness of KRAS mutational analysis in the diagnosis of pancreatic adenocarcinoma by comparing this technique with conventional cytology in aspirates obtained by endosonography-guided fine-needle aspiration biopsy. To reach this overall goal, the analysis addressed the following issues: (i) did KRAS mutational analysis render a more effective diagnosis than conventional cytology with the presence of a cytopathologist?; (ii) did KRAS mutational analysis increase the effectiveness of conventional cytology when both techniques were combined?; (iii) would these results have been reproduced without the involvement of an attendant cytopathologist?; (iv) which was the most cost-effective strategy?.
Between September 2001 and March 2002, all consecutive patients with pancreatic focal lesions undergoing endosonography-guided fine-needle aspiration biopsy were included in the study. The exclusion criterion was refusal to participate in the study. The protocol was approved by the Institutional Ethics of Research Committee, and informed consent was obtained from each patient.
Endosonography-guided fine-needle aspiration biopsy was carried out as part of the diagnostic protocol of our unit. This procedure was performed under conscious sedation by a single, fully trained endoscopist (AG). Prophylactic antibiotics were given in patients with cystic lesions, as well as for endocarditis prophylaxis when appropriate. Evaluation of the target lesion and staging of the tumour was initially performed with a radial scanning echoendoscope (GF UM20, Olympus America Inc., Melville, NY, USA). Endosonography-guided fine-needle aspiration biopsy was then carried out using a curved linear array echoendoscope (GF UM30P, Olympus America Inc., Melville, NY, USA) with Doppler capability and a scanning plane in the long axis of the instrument. A 22G needle-catheter system (Wilson-Cook Medical Inc., Winston-Salem, NC, USA) was inserted through the working channel and advanced into the lesion under real-time ultrasound control, taking care not to pass through intervening vessels. Following removal of the stylet, a 10-mL syringe was connected to the hub of the needle and a 3–4-mL suction was applied as the needle was moved back and forth within the lesion. Endosonography-guided fine-needle aspiration biopsy was either transgastric or transduodenal, depending on the location of the lesion. In each patient, the number of passes required to reach a final diagnosis and the order in which they were obtained were registered.
For the purpose of the present study, the results of conventional cytology and KRAS mutational analysis were compared with respect to the definitive diagnosis, which was established by either pathological examination of the resected specimen or clinical follow-up in those cases not submitted to surgery. In the latter group, lesions were considered to be malignant in the case of clinical progression of the disease or response to chemotherapy or radiation therapy. On the contrary, lesions were considered to be benign when spontaneous resolution or a lack of progression was observed on imaging studies after a minimum follow-up of 9 months. Cytological and mutational evaluations were performed in a blind fashion with respect to each other, as well as to the definitive diagnosis.
The sensitivity, specificity and positive and negative predictive values of conventional cytology, KRAS mutational analysis and the combination of both techniques were calculated with respect to the definitive diagnosis in those patients in whom adequate material was obtained. In addition, the overall accuracy of these approaches was also calculated by considering inadequate samples as misdiagnosed in order to establish their effectiveness in a clinical setting.
The aspirated material was sprayed on to glass slides and partially preserved with Diff-Quick stain (Stat Laboratory Medical Products, Lewisville, TX, USA) for immediate review by an on-site cytopathologist, who verified the adequacy of the specimen or advised on the need for additional passes. The rest of the aspirated material was preserved on glass slides in ethanol for further Papanicolaou staining and in saline for processing into a cell block. Cytological material was considered to be adequate when the attendant cytopathologist reported that there were a sufficient number of representative cells from the target lesion for a diagnosis to be made or the presence of malignant cells was confirmed. In cytologically benign pancreatic masses, the decision on when to cease making needle passes was established by taking into account meaningful clinical factors, such as the degree of clinical suspicion of underlying malignancy, the clinical impact of a non-diagnostic aspirate, the cytological appearance of the aspirated material and the total number of passes.21
KRAS mutational analysis
Material from the needle was washed out with saline and collected for mutational analysis. DNA was extracted following standard procedures. Mutations at codon 12 of the KRAS gene were detected by a primer-mediated restriction fragment length polymorphism (RFLP) method after in vitro amplification by polymerase chain reaction (PCR).6, 23 In brief, first-round amplification of exon 1 of the KRAS gene was performed using the primer DD5P (5′-TCATGAAAATGGTCAGAGAA-3′) and the mutant primer KRAS 5′ (5′-ACTGAATATAAACTTGTGGTAGTTGGACCT-3′) for 10 cycles (44 °C for 15 s, 72 °C for 15 s and 92 °C for 15 s) in a Hibaid Omnigene Thermalcycler (Teddington, UK). The mutant primer KRAS 5′ was used to create the restriction site for the enzyme BstNI [CCTGG], which is lost whenever a mutation occurs at codon 12. PCR was performed in a volume of 50 µL containing PCR buffer, 1.5 mmol MgCl2, 0.2 µmol deoxynucleotides (Promega Corporation, Madison, WI, USA), 1 U of Taq polymerase (Gibco BRL Life Technologies Inc., Gaithersburg, MD, USA) and 150 ng of the PCR primers. As an internal control of the enzymatic digestion, samples were re-amplified using the mutant primer KRAS 3′ (5′-TCAAAGAATGGTCCTGGACC-3′) and KRAS 5′ mentioned above for 35 cycles (54 °C for 15 s, 72 °C for 15 s and 92 °C for 15 s). RFLP was performed using the BstNI (New England Biolabs Inc., Beverly, MA, USA) restriction enzyme. After enzymatic digestion, samples were electrophoresed on a 6% polyacrylamide gel. The 143-base pair fragment depicts the presence of the mutant allele, whereas a 114-base pair band depicts the normal allele. The sensitivity of detection was a mutant to normal allele ratio of 102. Controls prepared with different proportions of mutant allele, as well as positive and negative controls, were included in each experiment.
In order to increase the sensitivity of KRAS mutational analysis, continuous enrichment by an enzymatic digestion RFLP/PCR-based method, able to detect mutants in 104 wild-type alleles, was also applied. This enriched approach combined the two-step RFLP/PCR method described and its continuous enrichment for the mutant allele by enzymatic digestion during the first PCR on the basis of the thermoestability of the BstNI enzyme. This technique has been described extensively elsewhere.24
Analysis of the usefulness of on-site cytological examination
In order to establish the effectiveness of the presence of a cytopathologist during endosonography-guided fine-needle aspiration biopsy, the overall accuracy of conventional cytology, KRAS mutational analysis and the combination of both techniques was calculated after particular numbers of passes. This was carried out by evaluating all labelled slides and DNA samples from each patient in the same order as they were obtained. The results achieved in the presence of a pathologist were compared with those that would have been obtained if a particular number of passes had been performed without on-site evaluation.
The outcome measure for the cost-effectiveness analysis was a correct diagnosis. For this analysis, the overall accuracy of conventional cytology, KRAS mutational analysis and a combination of both techniques was calculated for each patient.
The costs included the salaries of the endoscopist, pathologist and technicians, as well as the materials for both conventional cytology and mutational analysis. Physician and technician fees were calculated assuming a 1-year full salary of US$47 000 and US$18 000, respectively, according to Spanish national health system rates, and an average time to obtain the specimen and to perform on-site examination of 15 min per sample. The costs of endosonography imaging, fine-needle aspiration (materials, nursing, anaesthesiology requirements, overnight admission, etc.) and those derived from procedure-related complications were not considered in the analysis because they were assumed to be identical in both techniques. Taking all of these considerations into account, the costs of cytological evaluation without on-site examination, cytological evaluation with an attendant cytopathologist and mutational analysis were estimated to be US$11, US$16 and US$17 per sample, respectively. These costs were factored by the number of samples required to achieve a correct diagnosis in each approach.
Continuous variables were expressed as the mean ± standard deviation. Correlations between qualitative variables were evaluated by the chi-squared test, applying Yates' correction when needed. A P value of less than 0.05 was considered to be statistically significant.
Fifty-seven consecutive patients with either solid or cystic pancreatic lesions were included in the study. According to the definitive diagnosis, 33 (58%) patients had a pancreatic adenocarcinoma and 24 (42%) patients had other lesions (neuroendocrine tumour, n = 6; low-grade intraductal papillary mucinous tumour, n = 5; cystadenoma, n = 5; pseudotumoral chronic pancreatitis, n = 5; pancreatic metastasis, n = 2; and lymphoma, n = 1) and constituted the control group. The baseline characteristics of these patients are detailed in Table 1.
Table 1. Baseline characteristics of the patients included in this study
|Age (years)||63 ± 12||55 ± 15|
|Size (mm)||35 ± 8||35 ± 3|
|Criteria for definitive diagnosis (surgery/follow-up)||10/23||9/15|
A total of 136 samples was obtained from the 57 patients included in the study, thus representing 2.4 passes per patient on average (range, 1–5 passes). Sixty-seven of the 136 (49%) samples were considered to be adequate to reach a cytological diagnosis, and KRAS mutational analysis was performed in 133 of the 136 (98%) samples (χ2 = 79.8; P < 0.001). Accordingly, aspirates obtained by endosonography-guided fine-needle aspiration biopsy were sufficient to reach a mutational diagnosis in all patients (100%), but this figure was limited to 53 (93%) patients for conventional cytology (χ2 = 2.3; P = 0.12). In those patients in whom the material was adequate for diagnosis, conventional cytology revealed carcinoma cells in 31 of the 32 (97%) patients with pancreatic adenocarcinoma and in none of the patients from the control group. Codon 12 KRAS mutations were detected in 24 of the 33 (73%) patients with pancreatic adenocarcinoma, and in none of the patients from the control group. Taking into account these results, conventional cytology was considered to be more accurate than mutational analysis in terms of sensitivity and negative predictive value, whereas the specificity and positive predictive value were 100% for both techniques (Table 2). The combination of both techniques slightly increased the effectiveness of conventional cytology for the diagnosis of pancreatic carcinoma in terms of the negative predictive value (Table 2).
Table 2. Performance characteristics of conventional cytology, KRAS mutational analysis and their combination for the diagnosis of pancreatic adenocarcinoma
|Conventional cytology||31/32 (97%)||21/21 (100%)||31/31 (100%)||21/22 (95%)|
|Standard KRAS mutational analysis*||24/33 (73%)||24/24 (100%)||24/24 (100%)||24/33 (73%)|
|Combination of conventional cytology and standard KRAS mutational analysis*||32/33 (97%)||24/24 (100%)||32/32 (100%)||24/25 (96%)|
|Enriched KRAS mutational analysis†||26/33 (79%)||22/24 (92%)||26/28 (93%)||22/29 (76%)|
|Combination of conventional cytology and enriched KRAS mutational analysis†||32/33 (97%)||22/24 (92%)||32/34 (94%)||22/23 (96%)|
The use of an enriched method to detect KRAS mutations increased the sensitivity of mutational analysis, but decreased the specificity. Indeed, KRAS mutations were found in 26 of the 33 (79%) patients with pancreatic carcinoma, but were also observed in two of the 24 patients from the control group (one patient with a serous cystadenoma and one patient with a neuroendocrine tumour). Moreover, the combination of this enriched method with conventional cytology did not represent any advantage (Table 2).
In order to establish the actual effectiveness of conventional cytology and KRAS mutational analysis in a clinical setting, the analysis was repeated considering inadequate samples as misdiagnosed. Conventional cytology obtained a correct diagnosis in 52 of the 57 patients (overall accuracy, 91%), whereas mutational analysis obtained a correct diagnosis in 48 of the 57 patients (overall accuracy, 84%). Interestingly, the combination of both techniques allowed a correct classification of 56 of the 57 patients (overall accuracy, 98%).
Analysis of the usefulness of on-site cytological examination
The overall accuracy of conventional cytology, KRAS mutational analysis and the combination of the two techniques was calculated after particular numbers of passes (Table 3). The effectiveness of the combination of the two techniques was higher than that of the individual techniques in any particular number of passes. In addition, although the overall accuracy of conventional cytology progressively increased with the number of samples obtained (χ2 = 48.9, P < 0.0001 for trend), this figure reached the highest value after the second pass in the mutational analysis (χ2 = 2.8, P < 0.59 for trend) (Table 3).
Table 3. Overall accuracy of conventional cytology, KRAS mutational analysis and their combination for the diagnosis of pancreatic adenocarcinoma according to the number of passes
|1 pass||24 (42%)||43 (75%)||48 (84%)|
|2 passes||36 (63%)||44 (77%)||51 (89%)|
|3 passes||45 (79%)||48 (84%)||55 (96%)|
|4 passes||51 (89%)||48 (84%)||56 (98%)|
|5 passes||52 (91%)||48 (84%)||56 (98%)|
In order to establish the usefulness of the presence of a cytopathologist, results obtained with this strategy were compared with those that would have been obtained with a particular number of passes without on-site evaluation (Table 3). When the results were referred to conventional cytology alone, the overall accuracy of four passes (89%) was similar to the figure obtained under pathologist guidance (91%) with a mean of 2.4 passes per patient. Similarly, when the results were referred to the combination of conventional cytology and mutational analysis, the overall accuracy of three passes (96%) was similar to that obtained with on-site examination (98%) (Table 3) with 2.4 passes in average as explained before.
The costs of the above-mentioned strategies with equivalent efficacy are depicted in Table 4. Strategies performed with the assistance of an attendant cytopathologist were cost-effective with respect to the corresponding strategies without on-site examination. Furthermore, as a whole, the most cost-effective strategy corresponded to that based on conventional cytology alone performed under the guidance of an attendant cytopathologist (Table 4).
Table 4. Cost-effectiveness analysis
|Combination of mutational analysis and cytology (without attendant pathologist)||55/57 (96%)||171||1881||2907||4788||87|
|Combination of mutational analysis and cytology (with attendant pathologist)||56/57 (98%)||136||2176||2312||4488||80|
|Cytology alone (without attendant pathologist)||51/57 (89%)||228||2508|| ||2508||49|
|Cytology alone (with attendant pathologist)||52/57 (91%)||136||2176|| ||2176||42|
The results of the current investigation indicate that both conventional cytology and KRAS mutational analysis are highly specific, highly sensitive techniques for the diagnosis of pancreatic carcinoma by endosonography-guided fine-needle aspiration biopsy. When the analysis was restricted to patients in whom adequate samples were obtained, cytological examination of aspirates was superior to mutational analysis, meaning that the combination of the two techniques offered minimal advantage. Nevertheless, when all patients were considered, the addition of KRAS mutational analysis to conventional cytology constituted the most accurate approach. This apparent contradiction was due to the fact that adequate specimens to reach a cytological diagnosis were not available in all cases, whereas amplification of the KRAS gene was always feasible. It is important to emphasize, however, that the excellent results of conventional cytology obtained in our study may actually reflect the involvement of an attendant cytopathologist; this allowed the number of samples needed to achieve a correct diagnosis to be reduced, thus making the approach more cost-effective and diminishing potential morbidity. Thus, the importance of our study lies in the analysis of the contribution of a cytology–mutational approach, as well as on-site cytopathological examination, in the maximization of the effectiveness of endosonography-guided fine-needle aspiration biopsy for the diagnosis of pancreatic focal lesions.
The high prevalence of KRAS mutations in patients with pancreatic carcinoma has provided the rationale for evaluating this molecular event as a tumour marker. Indeed, several studies have shown that the detection of KRAS mutations improves the effectiveness of conventional cytology in aspirates obtained by percutaneous biopsy1, 2 and in pancreatic juice.4, 5 This technique is favoured by its high proficiency as well as by the simple and objective interpretation of the results. However, the diagnostic utility of KRAS mutational analysis is hampered by both false-positive and false-negative results. Indeed, KRAS mutations have been found in hyperplastic duct cells of normal-appearing pancreas and in pre-malignant lesions, such as intraductal papillary mucinous tumours and chronic pancreatitis.25–28 To overcome this limitation, Tada et al. proposed a semi-quantitative mutational analysis which, however, has been shown to be more useful in pancreatic juice than in specimens obtained by endosonography-guided fine-needle aspiration biopsy.22 In our study, using the standard method to detect KRAS mutations, no false-positive results were observed in spite of the fact that five patients with intraductal papillary mucinous tumours were included in the control group. On the other hand, up to 20–30% of pancreatic carcinomas do not carry mutations on the KRAS gene, and a heterogeneous intratumour distribution of this genetic event has been reported.2, 28, 29 Both circumstances explain the relatively high proportion of false-negative results, which prompted us to evaluate more sensitive strategies. In our study, however, the use of an enriched method to detect KRAS mutations provided a minimal advantage in terms of sensitivity and impaired the specificity.
The results of our study confirm that the cytological evaluation of aspirates obtained by endosonography-guided fine-needle aspiration biopsy constitutes the most precise single strategy for the diagnosis of pancreatic carcinoma. In fact, the effectiveness of this technique, in accordance with that previously reported (70–95%),7–11 is significantly higher than that obtained by other procedures, such as percutaneous biopsy (55–70%)15–17 or endoscopic retrograde cholangiopancreatography (30–45%).30, 31 The superiority of endosonography-guided fine-needle aspiration biopsy may be due to better accessibility to the gland, the possibility of directing the needle under real-time control and the presence of an on-site cytopathologist for immediate review of the specimens. Nevertheless, patients in whom aspirates are not adequate for cytological diagnosis may benefit from the detection of KRAS mutations. In the current study, the contribution of mutational analysis seemed less evident because this circumstance occurred in only 7% of cases. However, it should be emphasized that the proportion of undiagnosed patients by conventional cytology in this study was low when compared with other studies,7–11, 20–22 thus reinforcing the putative beneficial effect of a combined cytology–mutation approach.
The excellent results of conventional cytology obtained in our study may be due, at least in part, to the presence of a fully trained cytopathologist. The actual clinical usefulness of this strategy, which may lead to an increase in the yield of satisfactory specimens and is recommended by several groups,7, 8, 11, 20, 21 has never been evaluated properly. In this investigation, we compared the results achieved with on-site examination with those that would have been obtained if a particular number of passes had been performed without an attendant pathologist. Although mutational analysis reached a maximum steady-state accuracy after the second pass, the accuracy of conventional cytology progressively increased, reaching its best value up to the fourth pass, in accordance with previous reports.7, 21 In addition, a combination of cytology and mutational analysis showed the highest accuracy regardless of the number of passes performed, thus confirming the clinical usefulness of this approach. Taking into account all of these considerations, the results of our investigation demonstrate that, for those centres in which KRAS mutational analysis is available, the involvement of an attendant cytopathologist provides a limited benefit, as it does not allow a substantial reduction in the number of passes (from 2.4 passes, on average, with on-site evaluation to three passes without). On the contrary, when mutational analysis is not available, as in most centres in which endosonography-guided fine-needle aspiration biopsy is carried out, the presence of a pathologist almost halves the number of passes needed to achieve a correct cytological diagnosis (from 2.4 passes, on average, with on-site evaluation to four passes without). The relevance of this reduction is not limited to the fact that it minimizes the risk of procedure-related complications; it also represents a cost-effective strategy, at least in our media. Indeed, for both cytology alone and the cytology–mutation combination, the involvement of an attendant cytopathologist does not increase the overall cost of the cytological examination. We are aware, however, that this ratio can suffer noteworthy deviations depending on the country and health system considered, mainly due to differences in salary rates. Accordingly, the extension of cost-effectiveness results to other centres or medical organizations will require the re-calculation of these figures for their particular conditions, an estimation that our data can facilitate.
In conclusion, the results of this study demonstrate that conventional cytology from aspirates obtained by endosonography-guided fine-needle aspiration biopsy is the most precise single strategy for the diagnosis of pancreatic carcinoma, especially with on-site examination. However, when considering the proportion of cases in which adequate specimens are not available to reach a cytological diagnosis, the addition of KRAS mutational analysis renders almost absolute accuracy and, accordingly, should be recommended.
This work was supported in part by grants from the Agència d'Avaluació de Tecnologia i Recerca Mèdiques of the Generalitat de Catalunya (026/16/2000), the Fondo de Investigaciones Sanitarias (FIS 01/0104-02), the Ministerio de Ciencia y Tecnología (SAF00-0038) and the Instituto de Salud Carlos III (C03/02, C03/10, G03/156). Maria Pellisé and Francisco Rodriguez-Moranta are research fellows from the Hospital Clínic, and Sergi Castellví-Bel, Virgínia Piñol and Glòria Fernàndez-Esparrach are research fellows from the Institut d'Investigaciones Biomèdiques August Pi i Sunyer (IDIBAPS).