| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Department of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada.
Correspondence: Address correspondence to: David R. McCready, MD, MSc, FRCSC, FACS, Princess Margaret Hospital, Department of Surgical Oncology, 610 University Ave., Toronto, Ontario, Canada M5G 2M9; Fax: 416-946-6590; E-mail: david.mccready{at}uhn.on.ca
As the sentinel lymph node biopsy (SLNB) becomes included in the standard treatment armamentarium for breast cancer, results from this diagnostic procedure must be incorporated into the overall therapeutic plan. Most patients with an SLNB positive for metastatic disease undergo completion axillary lymphadenectomy because the number of lymph nodes with metastases is of prognostic value, helps guide selection of adjuvant treatment, and has potential therapeutic benefit.1 Data regarding the practice of completion lymphadenectomy following a positive SLNB are accumulating. It has been incorporated into the investigational arm of the National Surgical Adjuvant Breast and Bowel Project B-32 randomized trial comparing sentinel node biopsy with level I/II axillary dissection and is the standard arm of the American College of Surgeons Oncology Group Z-011 trial comparing observation versus completion lymphadenectomy for patients with a positive SLNB.
There are two common approaches in the timing of SLNB and, if positive, a completion lymphadenectomy: (1) plan for SLNB only and perform a completion axillary dissection at a later date if the final SLNB pathology records metastatic disease; and (2) perform completion lymphadenectomy concurrently if an intraoperative sentinel node evaluation is positive and at a later date if the intraoperative evaluation was falsely negative compared with the reference standard. Both approaches have their advocates.
The first approach ignores the possibility of intraoperative evaluation of the sentinel node because there may not be the available expertise to perform an accurate evaluation, the surgical or anesthetic environment may not be equipped to perform the completion lymphadenectomy, the chance of finding metastases is so low that the extra surgical planning is deemed unwarranted, or the patient may want to proceed with SLNB alone and explore options only after the final pathology is available.
The second approach is currently common because, from most patients’ perspective, surgery is ideally performed for an appropriate indication as one efficient complication-free event. Reoperative surgery in the axilla can be difficult depending on the amount of the previous sentinel lymph node (SNL) dissection and inherent scarring tendencies. Concurrent completion lymphadenectomy seems to be easier and less costly for both surgeon and patient. Yet, one must still be cognizant of other factors, including a false-positive intraoperative result. If the concurrent "unnecessary" axillary dissection is accompanied by significant postoperative arm symptoms, then that patient has not been well served. This interactive approach requires accurate intraoperative evaluation of the SLNB with a minimum of false results in a timely, cost-efficient manner.
Recently, an intraoperative technique using a complete and final frozen section evaluation of the whole node with serial sections and immunohistochemistry has been described.2 However, this methodology is not recommended because, even if one had the time and expertise to perform this Herculean effort, the value of such intensive staging is still unknown.3–5
In this issue of the Annals of Surgical Oncology, Henry-Tillman et al.6 have elegantly documented the University of Arkansas experience with intraoperative cytologic evaluation of SLNs. Since March 1996, 479 sentinel nodes were evaluated perioperatively using touch preparation cytology, and, in their hands, the technique was accurate (99%) and specific (99%), albeit with one false-positive result. The rate of finding nodal metastases when such nodal disease was present (sensitivity) was 94%, and its corollary the false-negative rate (1-sensitivity) was 6%. Metastatic disease was present in 14% of the analyzed sentinel nodes, they found that a negative cytologic touch preparation was correct 99% of the time (negative predictive value), and the finding of cytologically malignant cells correctly predicted nodal positivity in 98.5% (positive predictive value).
Although the present study confirms and extends the groups’ previous work7 on touch preparation cytology, it also contrasts the results to their own data on frozen section analysis of sentinel nodes. The parameters of concurrent frozen section were slightly worse, and thus, they recommended that the faster, cheaper cytologic touch preparation be used for intraoperative SLN evaluation. Should this methodology be immediately incorporated into clinical practice and thus, consign the cryostat to the junk heap? Well, there are a few considerations.
Surgical pathologists are generally more comfortable with intraoperative frozen section evaluations as they have had more experience with this modality. It must also be remembered that this article does not prove frozen section is an inferior technique. Although the reported false-negative rate is higher for the frozen sections, this was not a study designed for comparison. Frozen sections were performed at the "discretion of the pathologist" (possibly on more difficult cases or earlier on in his/her experience) following touch preparation and thus, are unlikely to be considered independent evaluations. Furthermore comparisons between their reported sensitivities of the touch preparations and frozen sections (64/68 = .94; 16/19 = .84, respectively) suffer from a paucity of frozen section events and overlap of their confidence limits.
The reference or gold standard in this series seems to be one hematoxylin and eosin section per 2–3 mm thick block of lymph node in accordance with a recent consensus statement.3 As recommended by others,8 both methods in the present series were compared with this same hematoxylin and eosin standard. However, comparisons among different centers are difficult to interpret because of disparate methodologies.5,6,9–12 On review of this literature, it seems that touch preparation cytology and frozen section analysis produce similar results, but because the frozen section method suffers from an inherent loss of tissue, there is a theoretical decrease in the ultimate detection of metastatic disease. Furthermore, neither technique has proven itself better in the detection of micrometastatic disease. In this series of touch preparations, three quarters of the false negatives were in nodes with micrometastatic deposits, and frozen section has been reported to have a 17% sensitivity for micrometastases.9 If serial sectioning and routine immunohistochemistry are used in the final pathologic assessment, then more micrometastatic disease will surely be found,12 and the number of false intraoperative evaluations, by definition, will increase.
Evaluation of the cytologic touch preparation methodology quite rightly proceeds node by node. However, incorporating this process into clinical practice involves transforming the data into a patient-oriented format and then checking whether the resultant information is generalizable. Using the data supplied in Table 3, one can quickly determine two things: (1) the sensitivity, specificity, and accuracy (43/47 = .91; 199/200 = .99; 242/247 = .98, respectively) are still excellent; and (2) the prevalence of node positivity is relatively low (47/247 = .19) in their overall patient population and in the T1 subgroup (14/134 = .10).
This prevalence of nodal disease is probably the result of both patient selection (well-differentiated, nonpalpable cancers in a well-screened population who are candidates for sentinel node biopsies) and the avoidance of upstaging by intensive final pathologic assessment. It is important to acknowledge the influence of prevalence on the predictive values, but even if there were twice as many node-positive patients as in the current study, touch preparation cytology, performed in the same accurate manner, would still have a negative predictive value of 95% (if number of node positive patients is doubled from 47 to 94 and sensitivity remains at 91%, this yields 8 false negative (FN) and 152 true-negative (TN) patients. Negative predictive value = FN/(TN+FN) = 8/(152 + 8) = .95).
Cytologic evaluation of the sentinel node seems to be a viable, accurate, and inexpensive method to determine nodal status intraoperatively. It requires cytologic expertise, but given its availability, it is equal and possibly better than frozen section analysis because it conserves tissue. Whether this extra conserved is clinically important will depend on how the reference standard for SLNB is influenced by the data from the prospective clinical trials. The sensitivity and value of intraoperative evaluation of low-risk tumors will decrease if multiple sections, routine immunohistochemistry, and genetic analysis become the next reference standard. However, if a concurrent completion lymphadenectomy continues to be in the best interest of the patient, then touch preparation cytology seems to be an ideal tool for intraoperative evaluation.
Received for publication February 27, 2002. Accepted for publication March 8, 2002.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  D. Lardinois, T. Brack, A. Gaspert, T. Spahr, D. Schneiter, H.C. Steinert, and W. Weder Bronchoscopic radioisotope injection for sentinel lymph-node mapping in potentially resectable non-small-cell lung cancer Eur. J. Cardiothorac. Surg., May 1, 2003; 23(5): 824 - 827. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
