when a mastectomy is the operativechoice. Studies need to focus onfinding clinical/prognostic factorsthat will allow better prediction ofthe presence of DCISM or invasivecancer on final pathology and guidethe need for sentinel lymph nodebiopsy at the initial operation.

C. Laronga, MDJ. M. Pimiento, MD

References1. de Mascarel I, MacGrogan G,

Mathoulin-Pelissier S, Soubeyran I,Picot V, Coindre JM. Breast ductalcarcinoma in situ with microinvasion:a definition supported by a long-termstudy of 1248 serially sectionedductal carcinomas. Cancer. 2002;94:2134-2142.

2. Cox CE, Nguyen K, Gray RJ, et al.Importance of lymphatic mapping in

Breast Dis

ductal carcinoma in situ (DCIS): whymap DCIS? Am Surg. 2001;67:513-519.

3. Yi M, Krishnamurthy S, Kuerer HM,et al. Role of primary tumorcharacteristics in predicting sentinellymph nodes in patients with ductalcarcinoma in situ or microinvasivebreast cancer. Am J Surg. 2008;196:81-87.

Predictors to Assess Non-Sentinel Lymph Node Status inBreast Cancer Patients withSentinel Lymph NodeMetastasis

Jinno H, Sakata M, Asaga S, et al (KeioUniv School of Medicine, Tokyo, Japan;et al)

Breast J 14:551-555, 2008

The next step of sentinel lymphnode biopsy (SLNB) in breast canceris to determine which patients needaxillary lymph node dissection(ALND) following a positive SLNB.A prospective database of 239 patientswho underwent SLNB followed bycomplete ALND at Keio UniversityHospital from January 2001 to June2005 was reviewed. A total of 131patients with one or more positivesentinel lymph nodes (SLNs) werefurther analyzed. A univariate analysisshowed a significant correlationbetween non-SLN involvement andlymphatic invasion, vascular invasion,number of tumor-involved SLNs,radioactivity of SLNs, and size of

SLN metastasis (p¼ 0.0002,p¼ 0.004, p¼ 0.006, p¼ 0.04,p¼ 0.03, respectively). By multivariateanalysis, lymphatic invasion and thenumber of tumor-involved SLNsremained significant predictors ofnon-SLN involvement. In breast cancerpatients with a positive SLN, lymphaticinvasion and the number of tumor-involved SLNs were both independentpredictors of non-SLN involvement.

In this article by Jinno andcolleagues, an analysis of 131 breastcancer patients with at least 1 positiveSLN was performed to identify factorsthat predicted additional non-SLNinvolvement with metastasis. Onmultivariate analysis, only lymphaticinvasion and the number of positiveSLNs were found to be significantpredictive factors. Similar analyseshave been reported by several groups,including our own, and indeed,lymphatic invasion and SLN burdenhave been frequently reported to bepredictive of additional nodalinvolvement.1-4

Interestingly, the current studydid not find that the size of the primary

tumor was significantly associatedwith non-SLN positivity, as has beenreported by others.1-4 The reason forthis discrepancy is unclear; however,this study did exclude larger tumors(>3 cm) and thus was biased towardsmaller tumors. Furthermore, the sizeof the SLN metastases was asignificant predictor in the univariateanalysis but did not appear to besignificant on multivariate analysis.As Jinno and colleagues noted, thisfactor tends to be one of the bestpredictors of non-SLN metastasis3,5,6

and is not included as a factor inthe nomogram by Van Zee andcolleagues.4 Accurate measurementof SLN metastatic deposits can betime consuming and difficult andthus is not routinely performedby all pathologists. However,when available, this informationcan be extremely helpful for theclinician in counseling patientsregarding the utility of completionALND.

R. F. Hwang, MD

eases: A Year Book� Quarterly 417Vol 20 No 4 2010

References1. Coutant C, Olivier C, Lambaudie E,

et al. Comparison of models topredict nonsentinel lymph node statusin breast cancer patients withmetastatic sentinel lymph nodes:a prospective multicenter study.J Clin Oncol. 2009;27:2800-2808.

2. Degnim AC, Griffith KA, Sabel MS,et al. Clinicopathologic features ofmetastasis in nonsentinel lymphnodes of breast carcinoma patients.Cancer. 2003;98:2307-2315.

418 Breast Diseases: A Year Book� Quart

Vol 20 No 4 2010

3. Hwang RF, Krishnamurthy S,Hunt KK, et al. Clinicopathologicfactors predicting involvement ofnonsentinel axillary nodes in womenwith breast cancer. Ann Surg Oncol.2003;10:248-254.

4. Van Zee KJ, Manasseh DM,Bevilacqua JL, et al. A nomogram forpredicting the likelihood of additionalnodal metastases in breast cancerpatients with a positive sentinel nodebiopsy. Ann Surg Oncol. 2003;10:1140-1151.

erly

5. Saidi RF, Dudrick PS, Remine SG,Mittal VK. Nonsentinel lymph nodestatus after positive sentinel lymphnode biopsy in early breast cancer.Am Surg. 2004;70:101-105.

6. Stitzenberg KB, Meyer AA, Stern SL,et al. Extracapsular extension of thesentinel lymph node metastasis:a predictor of nonsentinel node tumorburden. Ann Surg. 2003;237:607-612.

Prognostic Implications ofPositive Nonsentinel LymphNodes Removed DuringSelective SentinelLymphadenectomy for BreastCancer

Lang JE, Liu L-C, Lu Y, et al (UCSFComprehensive Cancer Ctr, San Francisco,CA; UCSF Helen Diller Family ComprehensiveCancer Ctr, San Francisco, CA)

Breast J 15:242-246, 2009

Nonsentinel lymph nodes (SLNs)are commonly removed at the time ofselective sentinel lymphadenectomy(SSL). Their predictive value for therest of the nodal basin is unknown. Aretrospective review of 436 breastcancer patients who underwent SSLbetween 12/97 and 04/03 at a singleinstitution. One-hundred nineteenpatients had non-SLNs removed atSSL; eight were positive (6.7%). Posi-tive non-SLNs predicted that SLNswould also be positive (p¼ 0.008).There was no difference in rates ofadditional positive nodes found oncompletion axillary node dissectionbetween the non-SLN and SLNpositive patients (p¼ 0.62). After

adjustment for covariates, the presenceof positive non-SLNs was not associ-ated with poorer disease free survival(p¼ 0.24), time to systemic recurrence(p¼ 0.57), or overall survival(p¼ 0.70). Positive non-SLNsremoved during SSL are not a signifi-cant risk factor for additional positivenodes on completion axillary nodaldissection (CALND) or for worsesurvival than positive SLNs.

Here, Lang and colleaguesaddressed the subject of non-SLNs in436 breast cancer patients who hadSLN biopsy; non-SLNs were removed in119 patients (27%), were positive in 8(2%), and were positive without SLNmetastasis in 2 (0.5%). I find theirconclusion (‘‘removal of non-SLN inpatients with a clinically negative axillais not beneficial’’) reasonable, althoughnot clearly supported by their data.

First, Lang and colleagues definedSLNs as nodes that were blue and/orhot, but the authors did not clearlydistinguish between non-SLNs thatwere removed incidentally and thosethat were removed on purpose (ie, werepalpably suspicious). Second, theauthors tell us nothing aboutintraoperative assessment (frozen

section or touch prep) of SLNs and non-SLNs: was it done, and how did it affectthe decision for ALND? Third, they tellus that 2 of 8 patients with positive non-SLNs had additional positive nodes onALND, but they do not tell us if thesewere the same patients who also hadpositive SLNs. Fourth, they do not give(a) the proportion of patients withpositive SLNs who had ALND, (b) theproportion of patients with positivenon-SLNs who had ALND, (c) the exactresults of the ALND, and (d) whetherthis changed treatment. Fifth, all oftheir data on pathology findings inSLN, non-SLN, and ALND specimenswere given as uninformative mediansrather than the exact numbers ofpositive nodes; the number of positivenodes is of course the mostprognostically relevant variable: forpatients with a single positive node, itshould make no difference whether thiswas an SLN, a non-SLN, or anotheraxillary node. Finally, because only 8patients had positive non-SLNs (6 ofwhom also had positive axillary nodes),the univariate Kaplan-Meier curves forrecurrence-free survival, overallsurvival, time to systemic relapse, andtime to local recurrence arequestionable, and one must ask