The Prognostic Impact Of Nodal Micrometastases

Nodal status is generally considered an important prognostic factor. Several studies have highlighted that the volume of metastatic nodal load also represents a prognostic parameter beyond the node-positive versus node-negative status. At one end of the spectrum, metastasis in a large number of lymph nodes is associated with a worse prognosis than the involvement of a few lymph nodes,52,53 or the involvement of a larger proportion of the examined lymph nodes is worse than a lower ratio of metastatic lymph nodes.54,55 When only a few lymph nodes are involved, small metastases lumped into the category of micrometastasis may also represent a smaller prognostic disadvantage than larger metastases.56

The prognostic impact of micrometastases in breast cancer is largely disputed. Some authors found no survival disadvantage for micrometastatic nodal involvement; others have reported a worse associated outcome, but only for disease-free survival and not for overall survival; and still others described a definite disadvantage (Table 6.1). Occult metastases are often admixed with micro-metastases in the quoted studies, especially in the earlier ones. The methods of pathological evaluation are also different from study to study and many would be considered less than optimal with our current knowledge; most studies were aimed at detecting some occult metastases and were not devised to exclude all occult metastases of a given size. It is likely that no attempt was ever made to exclude the possibility of the identified micrometastases being the tip of a larger metastasis "iceberg" (Figure 6.3). As shown in Table 6.1, the micrometastatic group often comprised only a few cases and the follow-up was also limited. Therefore, many of the series have insufficient power to demonstrate a minor benefit.

After reviewing many of the series published, it was concluded that micrometastases are of prognostic impact, but only large series with long follow-up can demonstrate this.78,79 The need for a long follow-up, especially for better differentiated tumors, was also stressed by the analysis of the Survival Epidemiology and End Results database.80 With the advent of SLNB, nodal micrometastases are not only found more frequently28 but smaller micrometastases and ITCs are also increasingly discovered. Theoretically, this may suggest an even lesser survival disadvantage, which would be backed up by most studies on the predictive role of these small metastases, but some recent studies report a worse disease-free survival even with ITCs (Table 6.1).14,74,75 These later studies are in keeping with the hypothesis that metastases may develop early in the neoplastic process as determined by the molecular signature of the tumor.81,82

Table 6.1 Studies on the survival effect (prognostic significance) of low-volume metastases

First author

Method of detection

Number of patients/details

Type of metastasis

Prognostic significance

Table 6.1 Studies on the survival effect (prognostic significance) of low-volume metastases

First author

Method of detection

Number of patients/details

Type of metastasis

Prognostic significance

Pickren3

Single HE vs SS

40 pNO vs 11 pNO upstaged to pN+ by SS

Occult

No (minimum 5-year follow-up)

Huvos1

NI

164 pNO vs 18 pNlmi (<2mm)

Micrometastasis

No (minimum 8-year follow-up)

Fisher6'

Single HE vs SS

59 pNO vs 19 pNO upstaged to pN+ by SS

Occult

No (mean follow-up: 61 months)

Fisher68

Single HE

287 pNO vs 21 pNlmi (<2mm)

Micrometastasis

No (mean follow-up: 49 months)

Rosen69

NI

77 macrometastatic vs 70 pNl mi

Micrometastasis

Yes for DFS of T1 tumors (minimum 10 year follow-up)

Wilkinson60

HE vs SS

441 pNO vs 84 pNO upstaged to pN+ by SS

Occult

No (minimum follow-up: 5 years)

Trojani61

HE vs IHC of same sections

129 pNO vs 21 pNO upstaged to pN+ by IHC

Occult

Yes for DFS and OS (mean follow-up: 10 years)

Friedman62

Serial HE

637 pNO vs 41 with sinusoidal low-volume metastases in a single lymph node

"Occult", sinusoidal

Yes (relative risk of distant metastasis at least 1.7 for occult metastases)

Sedmak63

HE vs new HE+IHC

36 pNO vs 9 pNO upstaged to pN+ by IHC

Occult

Yes for OS (minimum follow-up 10 years)

IBCSG64

Single level vs SS

838 pNO vs 83 pNO upstaged to pN+ by SS

Occult

Yes for DFS and OS (median follow-up:

5 years)

Galea66

HE vs IHC

89 pNO vs 9 pNO upstaged to pN+ by IHC (8 micrometastasis and 1 macrometastasis)

Occult

No (follow-up out to 14 years)

De Mascarel66

HE vs IHC of same sections

168 pNO vs 50 pNO upstaged to pN+ by IHC; same series as Trojani et al61 but longer follow-up

Occult

Yes - for DFS and ductal cancers (median follow up: 15.6 for ductal cancers)

Hainsworth6'

1 HE vs 1 IHC

302 pNO vs 41 pNO upstaged to pN+ by IHC (31 micrometastasis and 10 macrometastasis)

Occult

Yes - only for DFS, but not for OS (follow-up: median 79 months)

Nasser68

HE vs SS + IHC

151pN0 vs 50 pNO upstaged to pN+ by SS + IHC

Occult

No for all. Yes for both DFS and OS for occult metastases >0.2 mm (mean follow-up: 11 years)

McGuckin69

HE vs SS (4 levels) + IHC

155 pNO vs 53 pNO upstaged to pN+ by SS + IHC

Occult

Yes - only for DFS, but not for OS (median follow-up: 92 months)

Clare70

HE + IHC (5 levels each)

75 pNO vs 11 pNO upstaged to pN+ by SS + IHC

Occult

Yes - for disease-specific 5-year survival (median follow-up: 80 months)

Table 6.1 (Continued)

First author

Method of detection

Number of patients/details

Type of metastasis

Prognostic significance

Table 6.1 (Continued)

First author

Method of detection

Number of patients/details

Type of metastasis

Prognostic significance

Cote'1

HE vs 6 levels HE and 1 level IHC

588 pNO vs 148 pNO upstaged to pN+ by IHC; same series as IBCSG,64 but IHC added and longer follow-up

Occult

Yes - for both DFS and OS in postmenopausal women (median follow-up: 12 years)

Colpaert'2

HE vs 2 further levels IHC

80 pNO vs 24 pNO upstaged to pN+ by IHC (17 ITC and 7 pNlmi)

Occult (ITC and micrometastasis)

No (median follow-up for patients with and without relapse: 25 months and 91.5 months, respectively)

Millis'3

HE vs IHC

4l7pN0 vs 60 pNO upstaged to pN+ by IHC

Occult

No (median follow-up: 18.9 years)

Kuijt66

No data (registry data)

4377 pNO vs 179 pNlmi

Micrometastasis

Yes for 10 year OS (minimum follow-up: 52 months)

Mullenix'4

HE + IHC

175 pN0(i—)(sn) vs 8 pN0(i+)(sn) or pNlmi(sn)

ITC and micrometastasis

Yes for DFS, but insufficient data (median follow-up: 25 months)

Colleoni'6

No data, but step FS and IHC of suspicious cases on the basis of other reports from the same institution31'48

1400 pNO (i—) (sn) vs 232 pN0(i+)(sn) or pNlmi(sn)

ITC and micrometastasis

Yes for DFS, but not for OS (median follow-up: 50 months)

Imoto'6

HE+IHC

147 pNO(i-) (sn) vs 17 pNO(sn) upstaged to pN0(i+) (sn)

ITC

No (minimum follow-up: 5 years)

Nagashima"

HE/2 mm slice (+ IHC in suspicious cases)

241 pNO(sn) vs 19 pNlmi(sn)

Micrometastasis

No (median follow-up: 30 months)

Querzoli14

(HE + IHC) + (HE + IHC) pairs being separated by 100]im/both half

328 pNO(i-) vs 49 pNO upstaged to pN + (24 pN0(i+) and 25 pNlmi)

Occult (ITC and micrometastasis)

Yes for DFS (median follow-up: 8 years)

DFS, Disease-free survival; HE, hematoxylin and eosin; IBCSG, International Breast Cancer Study Group; IHC, immunohistochemistry; ITC, isolated tumor cells; NI, no information available; OS, overall survival; pN0(i+), isolated tumor cells (as defined in the articles in question); pN0(i—), no nodal involvement identified with multilevel assessment or immunostains; pNlmi, micrometastasis (<2mm, unless otherwise stated); pN+, any nodal involvement identified; (sn), symbol for nodal status established on the basis of sentinel lymph node biopsy without axillary clearance; SS, multiple level assessment by serial sectioning or step sectioning.

Figure 6.3 The "tip of the iceberg" phenomenon. The sentinel lymph node containing the 90 |m large cluster of cells (classified as "isolated tumor cells") shown in the inset was further sectioned at 250 |m, and turned out to be involved by a micrometastasis measuring 390 |m shown on the main field. Due to the fact of sampling, the lesion might have been somewhat larger in one of the unsampled levels. It should be accepted that size measurements are generally not perfect and despite objective measures they represent only the best approximation we can make. Similarly, micrometastases may be upstaged to macrometastases. Bars, 0.2 mm; inset 400x; main picture 100x.

Figure 6.3 The "tip of the iceberg" phenomenon. The sentinel lymph node containing the 90 |m large cluster of cells (classified as "isolated tumor cells") shown in the inset was further sectioned at 250 |m, and turned out to be involved by a micrometastasis measuring 390 |m shown on the main field. Due to the fact of sampling, the lesion might have been somewhat larger in one of the unsampled levels. It should be accepted that size measurements are generally not perfect and despite objective measures they represent only the best approximation we can make. Similarly, micrometastases may be upstaged to macrometastases. Bars, 0.2 mm; inset 400x; main picture 100x.

On the whole, the data relating to the prognostic significance of micrometastases and ITCs are nonconclusive, and partly due to the lack of similar pathology methods (including sampling of all lymph nodes, sectioning them, using IHC, interpreting the findings, etc) they are also incomparable. It can be expected that the search for low-volume metastases in the SLNs will result in a stage migration83 by diluting the traditional node-positive group with the "better prognosis" micrometastatic cases, and by concentrating the traditional node-negative group by taking out of it some cases with occult nodal involvement, resulting in a virtual improvement of prognosis in both new groups of node-positive and node-negative cases. This phenomenon makes the new results even less comparable with the historical reports listed in Table 6.1. The clinical trials on SLNB have not yet had sufficient long-term follow-up for definite conclusions about survival with micrometastases, although they support SLNB as a low morbidity staging procedure. It may however be hypothesized that patients with nodal micrometastases receiving adjuvant systemic treatment would derive some benefit from this treatment.56,77

One should not forget that when systemic treatment decisions are based on nodal status, this is considered as a marker of metastatic dissemination - albeit a rather imperfect marker. Indeed, node-negative patients may die of disease and a subset of node-positive cases may survive with locoregional treatment alone. Therefore, it is not surprising that the role of nodal status in influencing adjuvant treatment decisions has lost some weight, or that other prognostic factors have gained significance in outcome prediction and therapeutic planning. Although lymph node involvement is still considered an important prognosticator, according to the newer St Gallen consensus statements, it does not automatically define high risk and should be considered with other risk factors.84,85 Interestingly enough, these guidelines suggest that both ITCs and micrometastases should be ignored in risk allocation and treatment decisions.84 The latter recommendation is made despite the contradictory evidences outlined in Table 6.1, with greater favor being given to the significance of micrometastases than to their lack. This is also in opposition with the arbitrary TNM segregation of ITCs and micrometastases into the pN0 and pN1 categories, respectively.

SENTINEL LYMPH NODE MICROMETASTASES AND THE PREDICTION OF NONSENTINEL NODE INVOLVEMENT

When SLNB is performed, the finding of a metastasis is generally perceived as an indication for axillary dissection or radiotherapy. The detailed histological or molecular analysis of the SLN and non-SLN status has led to the recognition that many breast cancer patients have metastases limited to the SLNs, in keeping with the theory that the SLNs are the most likely sites of regional nodal metastasis.86-91 It has also been a common finding that the frequency of non-SLN metastases is dependent on SLN metastasis size92 and some other factors. In this context, micrometastases are often reported to be associated with such a low rate, and therefore risk of non-SLN involvement, that several reports have concluded that no further axillary treatment is necessary in case of micrometastatic SLN pos-itivity. In contrast, other studies have concluded that even ITCs in the SLNs are associated with a rate of non-SLN positivity which justifies axillary clearance as a general treatment option for any SLN involvement, and such findings also support the use of cytokeratin IHC in the evaluation of SLNs (Table 6.2).

The relevant studies are rather heterogeneous from several aspects, and therefore their meta-analysis may give more idea about the risk of non-SLN metastases than any of the individual reports.117 Until the relevant clinical trials reach a reasonable follow-up time for relevant conclusions, a 10-15% overall risk may be the best estimation of non-SLN involvement for cases with SLN micrometastases.117 This risk is very similar to the 5-10% false-negative rate of SLNB itself.

It must also be considered that micrometa-stases, as currently defined, are heterogeneous and this also relates to their size: they form a continuum. Larger ones at the macrometastasis edge probably reflect more harm than the smaller ones at the ITC edge. Likewise, some authors found that the rate of non-SLN involvement is significantly higher in patients with SLN micrometastases >1 mm than in those with smaller SLN micrometas-tases.111,114 Another study identified 1.3 mm as a possible cut-off for higher (>10%) and smaller non-SLN metastasis risk,118 whereas an Austrian paper suggested that micrometastases <0.5 mm were those in which the rate of non-SLN positivity could be considered negligible.115 It must also be remembered that studies which found it sufficient to identify ITCs in a single IHC-stained slide of a 3 mm thick slice of an SLN, and failed to exclude a larger metastasis underlying this, must be considered with caution as regards the size of the

SLN involvement and the associated rate of non-SLN positivity. On the other hand, studies using an enhanced method of metastasis detection for SLNs, but a standard one to a few H&E levels approach to non-SLNs (a very rational, acceptable and practical approach to nodal staging), are likely to underestimate the risks of overall non-SLN involvement, but not of macrometastatic non-SLN involvement. (Pathological methods used are briefly summarized in Table 6.2.)

Another consideration which should be taken into account is the contribution of other factors in models assessing the risk of non-SLN involvement. Tumor size and lym-phovascular invasion (LVI) are often reported as parameters influencing this risk, along with the number of SLNs involved, the number of SLNs found, or alternately the ratio of SLNs involved.

Therefore, despite the fact that the overall risk of non-SLN positivity associated with SLN micrometastases and ITCs found in different publications is generally low, and the estimate of 10-15% quoted previously117 may generally work well, there may be combinations of other factors increasing or lowering this risk. For example, patients with in situ carcino-mas,119,120 small pT1a and pT1b (up to 1 cm large) tumors without demonstrable LVI116,119 or with some special type tumors of good prognosis (like tubular or cribriform carcino-mas)116,121 would most certainly not benefit from axillary dissection after the finding of a micrometastatic SLN.

The interaction of different parameters has generated the search for predictive tools such as: the nomograms created at the Memorial Sloan-Kettering Cancer Center (MSKCC)122 and at the Mayo Clinic;123 the scoring systems generated by the MD Anderson Cancer Center,97 the Tenon Hospital124 and the Louisville sentinel lymph node study;125 or the decision table stemming from an Australian study.126 Some of these tools have been validated on independent datasets, but it seems that the group with the lowest risk of non-SLN involvement is very small, and this is where the predictive tools may perform more

Table 6.2 Studies on SLNB-based prediction of non-SLN metastasis including a minimum of 25 relevant cases oo

First author

Number of patients with small metastases (non-SLN+; %) Findings

Pathology ofSLNs; comments

Reynolds93 27 Nlmi or ITC (6; 22%)

Wong94 28 only IHC-positive (3; 11 %)

Abdessalam99 35 Nlmi or ITC (7; 20%); 5 only IHC-positive (1; 20%)

Freneaux100 35 only IHC-positive (1; 3%)

T and SLN metastasis size (micro vs macro) are independent preditors of non-SLN+ T1 and pNlmi(sn) have low risk of non-SLN+

pT category and number of positive SLNs are independent factors predicting non-SLN+ A small subset of patients with only IHC-positive SLNs have obvious non-SLN metastases

Patients with micrometastatic SLN have a risk of non-SLN involvement, that might be minimal for ductal-type T1 tumors without LVI

Non-SLN positivity differs significantly between cases with SLN micrometastasis and macrometastasis Risk-benefit assessment is needed for decisions on ALND

Patients with Tla tumors or SLN micrometastases 5HE + IHC; Nlmi and ITC considered as one category; <1 mm2 still have a risk of non-SLN+ micrometastasis definition: <1 mm2

4 HE + IHC; Nlmi and ITC considered as one category

HE/2mm + IHC performed on only 49% of the cases

The influence of SLN metastasis size could not be assessed reliably; probably mainly ITC identified

1-7 HE+ IHC (whole thickness at 0.5 mm steps); Nlmi and ITC considered as one category

3 level HE (+IHC not specified); Nlmi and ITC considered as separate categories

SLN metastasis size (micro/ITC vs macro), pT (pTl vs >2 cm) and LVI are the independent factors influencing non-SLN+

T (T1 vs > 2 cm) and histological grade are associated with non-SLN+

ALND is supported for low-volume SLN

involvement

SLN involvement, macrometastatic SLNs, LVI, extracapsular nodal involvement all increase the likelihood of non-SLN+

IHC detected occult metastases are rarely associated with non-SLN+

1 HE/2-3 mm slice - subset: 3 serial HE and 1 IHC; Nlmi and ITC probably considered as one category

4 HE + 4 IHC at 250 |lm/slice - non-SLNs also 3-10 IHC; Nlmi and ITC considered as separate categories; ITC limited to 3 cells

6FS + SS (HE) + IHC; Nlmi and ITC considered as one category

1-4HE+ (l-4)x6IHC; Nlmi and ITC considered as one category; overlap with Houvenaeghel et al116

Table 6.2 (Continued)

First author

Number of patients with small metastases (non-SLN+; %)

Findings

Pathology o/SLNs; comments

Zgajnar101 31 Nlmi and 5 ITC (4; 11% for all, 13% for Nlmi, 0% for ITC)

Cserni102

Kamath

Carcoforo

Calhoun1' Menes106

Jakub1'

No non-SLN macrometastases are likely if the ultrasound examination of the axilla is negative and the SLN contains only micrometastases or ITC

Tumor size, SLN metastasis size, SLN+ ratio, extracapsular spread are predictors of non-SLN+

SLN micrometastases detected by IHC only are associated with a low risk of non-SLN+

pT (T1 vs >2 cm), LVI, and MIB-1 proliferation index (>10% vs <10%) are factors influencing non-SLN+

• Patients with micrometastatic SLNs but without LVI, MIB1 index <10%, and T1 tumors can probably avoid ALND

61 only IHC/ITC-positive (3; 5%) • ALND is not supported for ITC involved SLNs 4-6 HE + 2IHC; overlap with Turner et al112

43 Nlmi or ITC (5; 12% for all, 4/39 for Nlmi, 1/4 for ITC)

46 Nlmi or ITC (7; 15%); 26 only IHC-positive (2; 8%)

HE SS + IHC; Nlmi and ITC considered as separate categories

HE SS at 50-100 or 250 |lm + multiple IHC; Nlmi and ITC considered as separate categories

SS(HE) +IHC; Nlmi and ITC considered as one category, divided by method of detection; overlap with Jakub et al10'

HE SS + IHC; ITCs not included in the SLN-positive group

Di Tommaso109 62 Nlmi or ITC (10; 16% for all;

7/25 (28%) for Nlmi >1 mm, 3/37 (8%) for Nlmi <1 mm; 9/31 (29% ) for intraparenchymal Nlmi and 1/31 (3%) for intrasinusoidal Nlmi

SLN metastasis size (micro/ITC vs macro), positive SLN ratio are the only factors associated with non-SLN+

High risk of residual disease if ALND is not done for micrometastatic or ITC involved SLNs

Both IHC of SLNs and ALND for SLNs positive with IHC only are justified

Only LVI associated with non-SLN+ ALND is indicated for minimal SLN involvement too

5HE and 2IHC/SLN; Nlmi and ITC considered as separate categories

HE + IHC/2-3 mm slice; Nlmi and ITC not distinguished, probably mainly ITCs considered; overlap with Kamath et al103

HE + IHC/l-2mm slice (later cases also step sectioned at 0.5 or 0.2 mm); Nlmi and ITC considered as separate categories

Micrometastases are less likely to be associated HE SS on FS, no IHC; Nlmi and ITC considered as one with non-SLN involvement than macrometastases. category

Table 6.2 (Continued)

00 o

First author

Number of patients with small metastases (non-SLN+; %) Findings

Pathology ofSLNs; comments

Mignotte110 68 Nlmi or ITC (15; 22%); 44

Leidenius111 84 Nlmi (22; 26% for all, 37% for

22%);* 39 only IH&positive (5; 13%); (10; 26%)*

Weiser113 93 Nlmi or ITC (17; 18%)

Viale114 109 Nlmi (24; 22% for all; 12/33

(36% ) for pNlmi >1 mm, 12/77 (16%)or pNlmi <1 mm)

Schrenk116 122 Nlmi or ITC (22; 18% for all;

Houvenaeghel116 700 Nlmi or ITC (94; 13%)

No subset with very low risk of non-SLN+ could be identified

Ratio of positive SLNs <0.2 makes the risk of non-SLN+ negligible

High overall risk of residual disease in the axilla if no completion ALND is done for micrometastatic SLNs

SLN metastasis size (micro/ITC vs macro), pT (pTla-b vs >1 cm) and LVI are the independent factors influencing non-SLN+ Micrometastatic Tl-2 tumors without LVI, hilar extracapsular nodal spread have low risk of non-SLN+

SLN metastasis size (micro/ITC vs macro), pT (pTla-b vs >1 cm) and LVI are the independent factors influencing non-SLN+ Micrometastatic Tla-b tumors without LVI, have low risk of non-SLN+

Omission of ALND may be considered for micrometastases <1 mm

Size of pNlmi and LVI significantly associated with non-SLN+

pT, detection method (IHC vs HE), LVI are factors influencing non-SLN+ pTla-b and some special type pTla-b-c cancers have low risk of non-SLN+

SS at 1-2mm (HE) +6 IHC; Nlmi and ITC considered as one category, divided by method of detection; overlap with Houvenaeghel et al116

FS + 2HE + IHC/1-1.5 mm slice; ITC not mentioned, Nlmi and ITC probably considered as one category

et al106

overalp with Calhoun

FS + SS + IHC; Nlmi and ITC considered as one category

HE SS on FS + IHC; Nlmi and ITC considered as one category

SS at 250|lm + IHC; Nlmi and ITC considered as separate categories

HE + IHC; Nlmi and ITC considered as separate categories

ALND, Axillary lymph node dissection; FS, frozen sections; HE, hematoxylin and eosin; IHC, immunohistochemistry; ITC, isolated tumor cells; LVI, lymphovascular invasion; non-SLN+, positive nonsentinel lymph node(s); pNlmi: micrometastasis (<2mm, unless otherwise stated); (p)T, tumor size as reflected by the T or pT categories of the TNM classification; SLN, sentinel lymph node; SLNB, sentinel lymph node biopsy; (sn), symbol for nodal status established on the basis of SLNB without axillary clearance; SS, multiple level assessment by serial sectioning or step sectioning; *, (results gained when) non-SLN also assessed with enhanced histopathology; TNM, tumor node metastasis classification of malignant tumors.

♦Results gained when nonsentinal lymph nodes were also assessed with enhanced histopathology.

Figure 6.4 Extracapsular extension of a nodal micrometa-stasis from a tubulolobular carcinoma.

poorly.127-129 It has also been found that the MSKCC nomogram is unreliable when SLNs are only micrometastatic.130 Interestingly, several of the mentioned predictive tools do not include the size of the SLN metastasis as a variable.122,125 The Louisville clinical prediction rule uses only factors which can be available during the SLNB procedure itself, to allow an intraoperative decision,125 whereas the MSKCC nomogram reflects metastasis size only by its detection method (frozen section vs serial HE vs IHC).122

Many publications have documented that extracapsular spread of the SLN metastasis is a factor associated with increased risk of non-SLN involvement.99,102,112,131,132 Although this phenomenon is perceived as a sign of nodal obliteration,133 and therefore massive involvement, some micrometastases also show this feature,134 especially in tumors with tubular histology (Figure 6.4). In such instances, the presence of extracapsular extension may be less associated with non-SLN metastases than in general.121,134

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment