Clinicopathological features and prognosis of triple negative breast cancer in Kuwait: A comparative/perspective analysis

Abstract

Aim

The aim of this study was to determine the incidence of TNBC in Kuwait, to analyze the clinicopathologic features and prognosis of this type of breast cancer, and compare it with reports from other regions of the world.

Background

Triple negative breast cancer (TNBC) is defined as a subtype that is negative for estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 (HER2). There is a growing evidence of the heterogeneity of such entity on the molecular level that may cause discrete outcomes.

Methods

We analyzed the clinicopathologic features of 363 TNBC cases which were diagnosed in Kuwait from July 1999 to June 2009. The disease-free survival (DFS) and overall survival (OS) were analyzed by Kaplan–Meier method. Comparison was done with reports from USA, Europe, Middle and Far East.

Results

Among 2986 patients diagnosed with breast cancer in Kuwait, 363 patients (12.2%) were TNBC. The median age was 48 years, 57.2% had lymph nodes (LN) metastasis, 56.9% were of grade III tumor and 41.9% had stage II disease. 81% developed recurrences and 75% of deaths occurred by 2.5 years after treatment. There is marked variation of clinicopathologic features according to country of patients’ cohort.

Conclusion

The incidence of TNBC in our study is similar to other studies. TNBC patients showed an early major recurrence surge peaking at approximately year 2.5. Regional variation of clinicopathologic features indicates a need for molecular studies to define underlying molecular features and its impact on survival.

1. Background

Breast cancer (BC) is increasingly recognized as a heterogeneous disease exhibiting substantial differences with regard to biological behavior and requiring distinct therapeutic interventions. Steroid hormone receptors (HR) such as estrogen receptor (ER) and progesterone receptor (PR) in concert with the oncogene ErbB-2/human epidermal growth factor receptor 2(HER-2) are critical determinants of these BC subtypes.^1–6 Triple-negative breast cancer (TNBC) is characterized by a lack of expression of both ER and PR as well as HER-2. A recent analysis indicates that TNBC carries a distinct molecular profile when compared with HR-positive BC.

A breast cancer classification emerged in the scientific scene based on gene expression profiles. The subgroups (luminal, HER2, normal breast and basal-like; BLBC) have distinct gene expression patterns and phenotypical characteristics. TNBC shares phenotypical features with basal-like breast cancer, which is in turn the most aggressive and with worse outcome.^7–11 However, the molecular classification of breast cancer has not led to changes in treatment recommendations and should yet be considered investigational, as the clinicopathological entities defined by the use of common immunohistochemistry (IHC) methods still represent the base for such recommendations.¹²

In Kuwait, a trend of presentation of breast cancer at earlier age was documented.¹³ This carries a risk of having more patients with TNBC (that is known to be more common at young ages). Many reports from different countries across the world documented different clinicopathological features that may be different by ethnicity.

2. Aim

The aim of this study was to determine the incidence of TNBC in Kuwait. In addition, analysis of the clinical and pathologic features of TNBC patients in Kuwait as well as the prognosis of this type of breast cancer was documented. Third, these findings were compared with reports representing different ethnic and demographic populations over the world.

3. Patients and methods

This is a retrospective analysis of the patients who attended the Kuwait Cancer Control Center (KCCC) for treatment or follow-up of breast cancer. Based on the Hospital cancer registry, medical files of all patients diagnosed with breast cancer in the period from July 1999 to June 2009 were reviewed. For all patients, the pathology was reviewed to confirm the diagnosis and the hormonal receptor and Her-2-neu status. Pathological diagnosis was based on biopsy from the primary breast lesion even in the context of cases presented with metastasis. We analyzed the clinicopathologic features of 363 triple negative cases which were diagnosed in this period. The TNM staging was based on pathologic findings in patients who had undergone upfront surgical treatment, while it was clinical and radiologic staging in patients who had received neoadjuvant chemotherapy. The disease-free survival (DFS) and overall survival (OS) were analyzed by the Kaplan–Meier method.

ER and PR were assessed using immunohistochemical staining for quantitative and qualitative assessment. Negativity was defined as absent IHC stain in all the examined tissue, i.e. 0%. Her-2-neu scores of 0 and 1 were considered negative, and a score of 3 was considered positive.¹⁴ Score of 2 was considered equivocal and FISH was considered. Ki67 was considered positive if it was more than 10%.

Follow up of patients was scheduled to be every 4 months for the first 2 years post-treatment, every 6 months for the next 3 years then annually. Follow up was by clinical examination and annual mammography. Follow up duration was 41.9 months in average (range 1–131 months). Studies from the USA, Europe, Lebanon, Korea and Japan were reviewed.^23–30

4. Results

4.1. Clinical features

During the period from July 1999 to June 2009, 2980 patients were documented to have breast cancer in KCCC. Out of them, 363 patients (12.2%) were confirmed to have a triple negative disease. Patient characteristics are summarized in Table 1. The mean age was 48 ± 11.7 years for the study population. Median age at the first birth was 23 ± 5.3 years.

Table 1.

Patient and tumor characteristics.

	n	%
Ethnicity (n = 363)
Caucasian	298	82
Asian	65	18
Age (n = 363)
Bellow 40	88	24
40–60	220	61
Above 60	55	15
Menopausal status (n = 363)
Premenopausal	221	61
Postmenopausal	142	39
Family history of cancer (n = 349)a
Breast	75	21
Other gynecologic	14	4
Others	7	2
History of breast cancer (n = 344)a
Positive	12	3
Negative	332	97
Smoking (n = 298)a
Yes	23	8
No	275	92
Parity (n = 299)a
More than 3	146	49
History of hormone therapy (n = 210)a
Yes	92	44
No	118	56
Breast feeding (n = 208)a
Yes	132	63
No	76	37
Laterality of breast cancer (n = 363)
Right	191	53
Left	169	46
Bilateral	3	1
Localization (n = 363)
UOQ	167	46
Multicentric	38	10
Histological type (n = 363)
IDC	291	80
ILC	13	4
Medullary	35	10
Mucinous and tubular	9	2
Others	12	4
Grade (n = 320)a
I	33	10
II	105	33
III	182	57
T stage (n = 353)a
T1	63	18
T2	156	44
T3	43	12
T4	91	26
N stage (n = 353)a
N0	151	43
N1	85	24
N2	94	27
N3	23	6
Metastatic at diagnosis (n = 25)b
Bone	15	60
Lung	10	40
Liver	8	32
Stage (n = 353)a
I	49	14
II	148	42
III	131	37
IV	25	7
Ki67 (n = 198)a
Positive	128	65
Negative	70	35
LVI (n = 315)a
Present	90	29
Absent	225	71
DCIS (n = 301)
Present	77	26
Absent	224	74

^aOnly available data in cohort cases could be analyzed.

^bSome patients were presented with two or more sites of metastasis.

4.2. Pathological features

Tumor characteristics are summarized in Table 1. Sixty two percent had T1-2 tumor and the mean tumor volume was 3 cm. Majority of cases had node negative (42.8%) and 24.1% had N1 disease. N2-3 disease was the rule in 33.1%. Her-2-neu was negative by IHC staining in 87.6% while FISH was needed to confirm diagnosis in 45 patients.

Treatment modalities are summarized in Table 2. It should be mentioned that, in our patient cohort, 51 patients had not received chemotherapy at all during their treatment course.

Table 2.

Treatment modality (n = 363).

	N	%
Surgery (n = 315)
MRM	201	64
Lumpectomy	114	36
Axillary surgery (n = 309)
Clearance	295	95
Sampling	14	5
Dissected LN (339)
0–9	74	22
10–20	165	49
More than 20	100	29
Chemotherapy (n = 304)
Neoadjuvant	42	14
Adjuvant	194	64
Both	51	17
Palliative	17	5
Type of chemotherapy (n = 304)a
Anthracycline-based	266	87
Taxane-based	144	47

^aSome patients received both agents.

4.3. Pattern of failure Table 3

Table 3.

Pattern of failure (n = 100).

Site of recurrence		% of all cases
Local recurrence	39	10.7
Bone metastasis	28	7.7
Lung metastasis	28	7.7
Liver metastasis	19	5.2
Brain metastasis	14	3.8
Contra-lateral breast cancer	8	2.2

There were 100 documented recurrences in our study population. The most common site of recurrence was local recurrence in 10.7% of cases (39% of all recurrences) followed by bone and lung metastasis that occurred in 7.7% of cases each (28% of all recurrences each). Liver metastasis was documented in 5.2%, brain metastasis in 3.8%, and contra-lateral breast recurrence in 2.2%, (19%, 14%, 8% of all recurrences, respectively) (Table 3).

4.4. DFS and OS (Fig. 1)

Fig. 1.

DFS and OS of all study cohor (cumulative and by stage).

Excluding patients presented with metastatic disease, most of recurrences happened in the first 2–3 years i.e. 81% of recurrences occurred by 2.5 years after treatment. DFS was 76.6% and 74.4% at 3 and 5 years of follow-up, respectively. OS was 83.2% and 81.5% at 3 and 5 years of diagnosis, respectively.

5. Comparison to other reports from different regions of the world

5.1. Epidemiology and risk factors (Table 4)

Table 4.

Patient characteristics in different studies.

	Prevalence (%)	Mean age	Premenopausal (%)	Parity (more than 3) (%)	OCP (%)	Smoking history (%)	Family history of cancer (%)
Kuwaita	12.2	48	61	50.2	44	7.7	20
Lebanon26	9.3	52	48				10
Turkey29	10.6	44	70		35
Singapore23	11	53
Korea27	16	45
Japan24		56
Mayo Clinic34		59.7
Bauer et al.16		50		64.6
Dent et al.28		50
Phipps et al.37				57	55		28
Kwan et al.36				34	72	49	20

^aCurrent study.

There have been no prospective studies specifically designed to examine the risk factors associated with specific molecular subtypes of breast cancer. However, several retrospective studies of large population-based cohorts have attempted to answer this question. The epidemiological risk factors for TNBC compared with non-TNBC appear to differ significantly.^15–22 Also, it was observed that the epidemiological and clinicopathologic features of breast cancer vary across the globe. Based on our results analysis, we can confirm that variation also applies to TNBC. We compared our results with results from published peer-reviewed results from the USA, Europe, Turkey, Lebanon, Singapore, Korea and Japan.^23–30 Table 4.

5.2. Prevalence

Overall, the prevalence of TNBC in large unselected breast cancer patient cohorts is 11–20%.^{10,17,18,20,29,31–35} In the compared studies, Korea seems to have the highest and Lebanon the lowest^27,28 prevalence. In our study, the prevalence of TNBC was 12.2%.

5.3. Age and reproductive history

Compared to other breast cancer subtypes, TNBC develops earlier in life, and consequently more often in pre-menopausal women.^18,29,36

As documented, 63% of patients with triple-negative tumors were diagnosed before the age of 60 compared with fewer than half among those with other tumor types.¹⁷ High average age was noticed in a study from MayoClinic (59.7 years).³⁵ In a meta-analysis of 1997 patients by Yang et al., the mean age was 56.²⁶

However, many studies documented 50 years as the mean age at diagnosis of TNBC.^17,23,29,37 In our study, the mean age was 48 years, younger than the mean age of breast cancer in general, independent of phenotype (52 years in Kuwait).¹³ Korea and Turkey have the youngest cohort of TNBC patients (44–45),^28,30 while Japan has the oldest (mean age was 56 years).²⁵ Fifteen percent were younger than 40 years in the study from Singapore versus 24% in our study.²⁴

Premenopausal status varied from 70% of patients in Turkey,³⁰ 48% in Lebanon²⁷ to 61% of TNBC patients in our study. Parity was considered as a characteristic in TNBC that deserve reporting in many studies (Table 5).^17,37,38 In our study, only 4% were nulliparous (vs. 13% in Yang meta-analysis²⁶). This may be related to cultural reasons rather than a fertility issue. Breast feeding was not documented in 41% of our cohort. However, 37.4% of the documented cases lack breast feeding similar to that documented by Millikan et al.²⁰

Table 5.

Tumor characteristics in TNBC in different studies.

	IDC (%)	High grade (%)	Ki67 (%)	LVI (%)	Mean tumor size (cm)	T2 lesion (%)	Stages I and II	LN + ve (%)
Kuwaita	81	57	65	28	3	43	56	58
Lebanon26	85	63	50				64	50
Turkey29		53		19				29.4
Singapore23	93	77				70
Korea27		53.7					90
Japan24	95	92				43	86.5	34
Mayo Clinic34	88			18
Bauer et al.16
Dent et al.28		63	50		3			54.6
Phipps et al.37
Kwan et al.36					2			48
Kansas, US22		75		33

^aCurrent study.

5.4. History of hormone Intake and smoking

Forty-four percent gave history of oral contraceptives(OCP) in our cohort as compared by 72% in Kwan et al.³⁷ study, 55% in Phipps et al. study,³⁸ and 35% in the Turkish study.³⁰ Only 7.7% of our study cohort gave history of smoking versus 49% in Kwan et al.³⁷

5.5. Family history

A positive family history was found in 10% of patients with TNBC in Lebanon compared with 1% of patients with breast cancer when all phenotypes are included.^27,39 In our study, 20% gave that history, similar to that documented by Kwan et al.³⁷ It was 28% in Phipps et al.³⁸

In Yang meta-analysis,²⁶ having a positive family history of breast cancer was more frequent in case of patients with basal-type tumors compared with patients with ER+/HER2- or PR+/HER2-tumors (25% vs. 22%) in spite of the absence of this difference when classification was based on IHC only (19% vs. 18%). On the other hand, the family history of breast cancer was the same between TNBC and non-TNBC in the Turkish study.³⁰ It seems that genetic inheritance and family history is an important risk factor in TNBC in Kuwait.

5.6. Tumor characteristics Table 5

5.6.1. Histological presentation of TNBC

More than 90% of TNBCs exhibit an invasive ductal histology (IDC).^32,40–44 In current study, 81% were IDC. IDC was most predominant in Singapore and Japan (93% and 95%, respectively).^24,25 ILC was interestingly documented in 2% of patients in Singapore²⁴ vs. about 4% in our current study. This was lately highlighted in an Italian study as 2.3% of their patients expressed the ILC phenotype.⁴⁵ This may represent the pleomorphic subtype of ILC.⁴⁶ We were very concerned about these 13 patients and pathology review was done.

Additional characteristics of TNBC are frequent metaplastic elements and medullary/atypical medullary features.^40,41,47,48 In our study, 12.9% had metaplastic and/or medullary carcinomatous features of their tumors.

Triple negative cancers are predominately of high histological grade.^{10,28,29,40,49–53} In our study, similar to many others, 57% were of high histological grade.^{29,47,48,54–58} In Singapore, they documented a higher rate of high grade (77%),²⁴ similar to a study from Kansas.²³

But Japan showed the highest predominance of high grade tumors (92%).²⁵

5.7. Ki67

The high frequency of more aggressive expression profile with low Bcl-2 but high p53 and Ki67 expression was documented in many studies.^{15,21,40,49–51,53,52} In our study, 65% of examined specimens showed high mitotic index. Both the Lebanese study and Dent et al.²⁹ study documented Ki-67 positivity in 50% of patients.²⁷

5.8. Lymphovascular invasion (LVI)

In the study from Mayo Clinic, LVI was less common in TNBC when compared with HER2+ (18% and 24%, respectively), but both had higher rates than that seen with ER+ (15%, P = 0.006).³⁵ However, in a Turkish trial, there was no difference among both groups.³⁰

In our cohort, LVI was documented in 28% of examined specimens; a much higher rate than that in both studies but close to that recorded in the Kansas study (33%).²³

5.9. Size of tumor

At diagnosis, TNBCs are commonly of larger tumor size.^{35,40,49–51,53,52} In our study, the mean tumor size was 3 cm in TNBC group; similar to that in Dent et al.²⁹ study and larger than other non-TNBC group in the same study (3.0 versus 2.1 cm, respectively; P < 0.0001). It was smaller (2 cm) in Tawfik et al.²³ study. Most of tumors were T2 in Singapore study (70%).²⁴ In both our study and in Japan, T2 represents about 43% of cases of TNBC.²⁵

5.10. Stage

Ghosn et al.²⁷ compared the staging of TNBC as found in the Lebanese study with the staging of breast cancer of all phenotypes in Lebanon (stages I and II, 64% versus 75%; stage III, 24% versus 20%. They concluded that TNBC is more often locally advanced at diagnosis.^59,60 In our study, stages I and II represents 56% while stage III represents 37% of TNBC in Kuwait; which means that TNBC patients are diagnosed at a later stage in our community. In Japan, 86.5% have stages I and II while only 10.3% has stage III,²⁵ while it is 90% and 10%, respectively, in Korea,²⁸ which may result from more strict and nation-wide screening systems.

In a series of 1263 women diagnosed with invasive breast cancer in the Henry Ford Health System, women with advanced stage (stages III and IV) were 16 times more likely to have triple-negative tumors than those with early-stage [odds ratio (OR) 16.4; 95% confidence interval (CI): 7.8–34.2].⁵⁴

5.11. Lymph node metastasis

The presence of positive lymph nodes in the Lebanese study was 50%, while it was detected in 58% in our study. With almost a similar percentage, Dent et al.²⁹ found that the rate of node positivity was slightly higher in the triple-negative group compared with the other group (54.6% versus 45.6%, respectively; P = 0.02). In contrast, non-significant difference was documented by Tawfik et al.²³ in the study from Kansas (48 vs. 41%). Interestingly, the ratio was reversed in the Turkish study (29.4 in TNBC vs. 38.5% in non-TNBC)^11,30 found it to be higher in their study (68%) while it was much lower in Japanese patients (34%).²⁵

5.12. Survival

5.12.1. DFS and OS (Fig. 1)

TNBC accounts for a disproportionate number of BC deaths; the majority of studies indicate a negative impact of a TN.^{10,18,20,21,29,61–65} Patients with TNBC typically have a high risk of early recurrence that sharply increases in years 1–4 after diagnosis, and the majority of deaths occur in the first 5 years after treatment.

However, differences between TNBC and non-TNBC regarding overall survival (OS) wear off at 10 years of follow-up.⁶⁶

The risk for late recurrences (i.e. beyond 5 years of diagnosis) is decreased by 50% compared with HR-positive disease.⁶⁷ In other words, for women with triple negative disease, a substantial number appear to be cured if they remain recurrence free for the first five years after diagnosis.⁶⁸

Follow-up of about 200 patients diagnosed with TNBC in Toronto between 1987 and 1997 showed a peak of recurrence rate much greater than that of non-TNBC tumors during the first and third years, as well as a higher 5-year mortality rate.²⁹ This was subsequently confirmed in patients treated with neoadjuvant therapy at M.D. Anderson,⁶⁹ who showed a higher 3-year relapse and mortality rates.

In both the Lebanese ²⁷ and US studies (Dent's ²⁹ and Haffty et al.⁷⁰ studies), the peak of recurrence occurred after 6–18 months. In our study, this peak was noticed later; at 30 months (81% of recurrence).

In the Mayo Clinic Study, local or regional recurrence developed significantly more often with TN tumors (5.7%) vs. 10.7% in our study. In the same study, at 5 years after surgery, TN patients exhibited an overall survival of 85% vs. 81.5 in our study.³⁵

In Korea, during the median 73.3 months of follow-up, the 5-year relapse rates among TNBC was 30.1% vs. 26% in our study, and the 5-year OS rate was 83.1%²⁸ vs. 77% in the California study.¹⁷

5.12.2. Aggressiveness

The poor prognosis of high-grade TNBC relates to poor disease-free interval in the adjuvant setting,^29,71 shortened progression-free survival in the metastatic setting,^29,72 and the lack of targeted therapy. However, not all TNBCs are associated with a poor prognosis.⁷³

5.12.3. Locoregional recurrence

In a study by Haffty et al.⁷⁰, they found that the locoregional relapse rate for TNBC appears to be identical to that of other molecular subgroups after conservative surgical management; however, the TN phenotype appears to be associated with a higher rate of distant metastases.

There was no difference between the Lebanese²⁷ and Haffty et al.⁷⁰ studies regarding the sites of metastasis, which were, in order of frequency, the lungs, brain, then liver. Dent and colleagues ²⁹ found few cases in which local recurrence preceded distant metastases; these, in turn, are more common in the viscera and soft tissues than in bone.^74,75

Several studies have supported a significantly increased rate of visceral versus bone metastasis^57,69 among patients with TNBC compared with non-TNBC. In the largest report to date, data on 12,858 patients indicate an increased risk for lung [odds ratio (OR) 2.27 and brain (OR 5.32) metastasis as first sites of recurrence and lower risk for bone recurrence (OR 0.23) in patients with TNBC.³²

However, in our cohort, the most common site of recurrence was local recurrence in 10.7% of cases (39% of all recurrences). This was followed by bone and lung metastasis that occurred in 7.7% of cases each (28% of all recurrences each). Then comes liver metastasis in 5.2%, brain metastasis 3.8%, and contra-lateral breast recurrence in 2.2%, (19%, 14%, 8% of all recurrences, respectively).

5.12.4. Central nervous system metastases

Patients with TNBC compared with other subtypes reportedly experience an increased risk of central nervous system metastases (CM) of 6–46% of those experiencing metastatic spread of disease.^75–77 Similarly, in a single-institution study among 3193 patients, a significantly elevated risk of CM among patients with TNBC and HER-2-positive BC compared with other phenotypes was reported (HR 4.5 and 4.9 for TNBC and HER-2+, respectively)⁷⁶; the risk of CM was particularly pronounced among young patients with node positive disease: the incidence of CM among patients <50 years of age and node positive was 20.0% for TNBC compared with 4.8% for HER-2 positive. In our study, CM represents the site of metastasis in 14% of the patients experienced metastasis.

Recently, a nomogram has been indicated⁷⁸ to calculate the probability for developing cerebral metastasis, particularly for patients with TNBC, the clinical implications and validation of which, however, remain unclear.

6. Conclusion

From the above overview, it seems that TNBC has another aspect of heterogeneity; that is the demographic characteristics of patients.

• •A step forward toward genetic studies for TNBC is needed as phenotypic and molecular classifications lake explanation for variable responses and prognosis.
• •Regional collaborative studies should be performed in TNBC to define underlying molecular unique characteristics that may affect treatment options in every ethnicity or population.
• •Another point is the proper evaluation of the TNBC problem in relation to the whole breast cancer patients in view of cost-effectiveness; especially in middle and low-income countries.

Conflict of interest

None declared.

Disclaimer

None.

Financial disclosure

None declared.