Intracystic Papillary Carcinoma of Breast: Interrelationship With in Situ and Invasive Carcinoma and a Proposal of Pathogenesis

Array Comparative Genomic Hybridization Study of 14 Cases

Thaer Khoury; Qiang Hu; Song Liu; Jianmin Wang

Disclosures

Mod Pathol. 2014;27(2):194-203. 

In This Article

Discussion

Classifying intracystic papillary carcinoma under ductal carcinoma in situ or invasive ductal carcinoma is still a matter of debate and the biology and pathogenesis are still unclear. It was proposed that intracystic papillary carcinoma is a form of invasive ductal carcinoma for the following reasons: (1) Lack of the surrounding myoepithelial cell layer at the tumor–stromal interface in intracystic papillary carcinoma, a known feature for invasive ductal carcinoma.[3,5] However, lack of myoepithelial cell layer is not always indicative of invasion; microglandular adenosis is a good example.[4,16] (2) Axillary lymph node or distant metastases in intracystic papillary carcinoma that lacks unequivocal invasion is well known.[17] However, up to 7% of ductal carcinoma in situ could develop nodal metastases with no evidence of invasion.[18] In addition, tumor local invasion may be related to tumor displacement[19] and lymph node metastasis may be the result of synchronous separate occult invasive ductal carcinoma.[20] Therefore, local recurrence and/or lymph node metastases do not make intracystic papillary carcinoma an invasive carcinoma. In conclusion and based on the known indolent behavior of this tumor, the WHO task force recommended that intracystic papillary carcinoma be staged and treated like ductal carcinoma in situ.[1]

There have been few studies that have investigated the biology of intracystic papillary carcinoma. Using markers of invasion, Rakha et al[21] found that intracystic papillary carcinoma exhibited an expression pattern of invasion-associated markers between ductal carcinoma in situ and invasive ductal carcinoma, concluding that this tumor has unique biological features. Duprez et al[9] investigated whether papillary carcinomas would constitute an entity distinct from invasive ductal carcinoma at the genomic level, or whether they would merely constitute a morphological variant of estrogen receptor-positive invasive ductal carcinoma. They included 39 intracystic papillary carcinomas and compared with grade- and estrogen receptor-matched non-concurrent invasive ductal carcinoma. They suggested that papillary carcinoma belongs to a spectrum of estrogen receptor-positive invasive ductal carcinoma. They also hypothesized that papillary carcinoma may merely constitute a final stage of development of a ductal carcinoma in situ within a pre-existing papilloma.

Our study is the first to compare between intracystic papillary carcinoma with concurrent ductal carcinoma in situ and invasive ductal carcinoma on the genomic level. Although intracystic papillary carcinoma clustered with concurrent ductal carcinoma in situ and invasive ductal carcinoma, it had more complex gene aberration than the other two. Duprez et al[9] found that papillary carcinomas harbor significantly fewer gene copy number aberration than matched invasive ductal carcinoma. The reason for this discordance may be related to one or many reasons: (1) they included all types of papillary carcinomas, and also intracystic papillary carcinoma type; (2) the studied invasive ductal carcinoma was not concurrent; (3) we studied a relatively small number of cases; and finally, (4) most of our intracystic papillary carcinoma cases were of grade II and invasive ductal carcinoma cases of grade I. They also found that papillary carcinoma (invasive ductal carcinoma and intracystic papillary carcinoma) did not form a separate cluster from matching invasive ductal carcinoma. The pattern of genomic aberration found in papillary carcinoma was similar to that of grade- and estrogen receptor-matched invasive ductal carcinoma.[9] Our study had similar results, as intracystic papillary carcinoma clustered with invasive ductal carcinoma as well as with concurrent ductal carcinoma in situ with minimal aberration. However, it is worth noting that the number of studied cases is very small, particularly invasive ductal carcinoma, for the results to be definitive.

We found that intracystic papillary carcinoma clustered with two of two invasive ductal carcinoma and four of five ductal carcinoma in situ. This finding would suggest that intracystic papillary carcinoma is closer to invasive ductal carcinoma than to ductal carcinoma in situ. However, the number of studied cases was too small, particularly for invasive ductal carcinoma (n=2) to draw this conclusion. Therefore, we relied on the total number and percentage of genomic changes present in all three entities. The order of genomic change was as follows: intracystic papillary carcinoma, ductal carcinoma in situ and then invasive ductal carcinoma with borderline statistical significance.

There are three major theories about the tumorigenesis and progression of different tumors. The first theory is that the proposed tumors each start with different clone/clones. Thereafter, they acquire mutations and evolve into different tumors independently and in parallel. The second theory is that the normal cell acquires single or set of clones, which produces a type of tumor. Thereafter, this tumor acquires another set of mutations producing a completely different tumor type. The third theory is that the tumors arise from the same ancestral clone. At certain stage, they split into different tumor types and acquire extra changes.[22,23]

In the light of these theories and our findings of variable gene aberrations in different types of intracystic papillary carcinoma, we believe that intracystic papillary carcinoma is genetically not one disease. Intracystic papillary carcinoma with concurrent invasive ductal carcinoma might have different progression pathway and pathogenesis than intracystic papillary carcinoma with concurrent ductal carcinoma in situ (Figure 7). When we compared intracystic papillary carcinoma with ductal carcinoma in situ vs without ductal carcinoma in situ, we found that the latter had gain in small loci 5q35.3, 8q24.3 and 21q13.2. Similarly, when we compared intracystic papillary carcinoma with invasive ductal carcinoma vs without invasive ductal carcinoma, we found that the latter had two changes, 11q22.1–23.3 loss and chr5 gain.

Figure 7.

Proposed pathogenesis of intracystic papillary carcinoma (ICPC) with relation to ductal carcinoma in situ (DCIS) and invasive ductal carcinoma (IDC).

Although intracystic papillary carcinoma with or without ductal carcinoma in situ had similar major chromosomal changes (16p gain, 16q loss, 1q gain and 7q loss), intracystic papillary carcinoma with invasive ductal carcinoma, in addition, had chr5 gain. Therefore, we believe that when intracystic papillary carcinoma develops through acquired gains/losses of major chromosomes, it may start with a set of major changes (16p gain, 16q loss, 1q gain and 7q loss) and then later acquires another major change (chr5 gain) when it progresses into intracystic papillary carcinoma with invasive ductal carcinoma. The other possibility is that intracystic papillary carcinoma without invasive ductal carcinoma started separately and in parallel with intracystic papillary carcinoma with invasive ductal carcinoma.

Interestingly, we found that intracystic papillary carcinoma with invasive ductal carcinoma had 11q22.1–23.3 gain when compared with intracystic papillary carcinoma without invasive ductal carcinoma. In this region, there are multiple MMPs genes. These genes are known to have major role in tumor invasiveness.[24] These genes might be responsible for giving intracystic papillary carcinoma the capability for invasion and forming invasive ductal carcinoma. However, we could not compare intracystic papillary carcinoma vs invasive ductal carcinoma because of the small number of invasive ductal carcinoma cases that had successful gene array.

For the pathogenesis of pure intracystic papillary carcinoma vs intracystic papillary carcinoma with ductal carcinoma in situ, we believe that they both start with major chromosomal changes (16p gain, 16q loss, 1q gain and 7q loss). They differ in additional minor chromosomal changes (Figure 7).

When intracystic papillary carcinoma was compared with concurrent ductal carcinoma in situ, we found that the former had 1q21.3–23 gain. In this region, there are two genes of interest, PRCC and NTRK1, which are altered in papillary carcinomas of the kidney and thyroid, respectively.[25,26] Chromosomal rearrangements involving the NTRK1 gene are found in approximately 10% of thyroid papillary carcinoma. They result from the fusion of the tyrosine kinase domain of TRK to 5′-terminal sequence of the tropomyosin (TMP3), or TPR genes also located on chromosome 1q, or of the TFG gene on chromosome 3.[26]PRCC is one of several genes that are involved in the renal cell carcinoma type that is associated with Xp11.2 translocations/TFE3 gene fusions.[25] One of the interesting points of papillary carcinoma of the kidney and the thyroid is that these tumors have more indolent behavior compared with their organ-related counterparts.[25,26] It is known that intracystic papillary carcinoma has an indolent behavior as well.[1] Therefore, this frequent change in intracystic papillary carcinoma but not ductal carcinoma in situ could explain the papillary morphology and its indolent clinical behavior. Duprez et al[9] did not find a determinant of the papillary phenotype of papillary carcinoma of the breast (invasive ductal carcinoma and intracystic papillary carcinoma). They suggested that this determination could be related to other changes including genetic aberrations other than gene copy number aberrations (ie copy number silent loss of heterozygosity events, somatic mutations or fusion genes), epigenetic changes or distinctive tumor–microenvironment interactions. Although we found amplification of these two genes (through 1q21.3–23 gain), rather than translocation (like in PRCC) or rearrangement (like NTRK1), this finding could be useful for future investigation. Although there have been no studies to correlate between PRCC and breast cancer, Dolle et al[27] have demonstrated an autocrine NGF/NTRK1 loop involvement in breast cancer progression.

When ductal carcinoma in situ is compared with concurrent intracystic papillary carcinoma, we found that the former had a set of copy number changes including 8q22.2–22.3 gain, 8q24.3 gain, 11q25 loss and 12q24.31–24.33 loss. We found a gene set of interest that are all involved in cell adhesion and cell motility. Table 2 shows the genes names and their description. These genes differ in their function, pro- or antiadhesion.[28–33] Ductal carcinoma in situ has variable expression of these genes (gain or loss). Therefore, we believe that the function of cell-to-cell adhesion is significantly altered. We suggest that the motility and adhesion of cell molecular changes permit the neoplastic cells to migrate from intracystic papillary carcinoma and colonize in the surrounding ducts, but do not give it the required mechanism for stromal invasion.

We conclude that intracystic papillary carcinoma is genetically closer to ductal carcinoma in situ than to invasive ductal carcinoma, which may explain the indolent behavior of this tumor. Intracystic papillary carcinoma shares major copy number variation with minimal but significant genetic variation that may have a critical role in the pathogenesis and biology of this tumor. Further studies with larger number of cases are needed to verify our results.

processing....