Quadruple Score: A Novel Scoring System for Diagnosing Breast Lump – A Retrospective Pilot Study

Abstract

Background:

Triple test was a breakthrough in the initial screening of breast cancer due to its high sensitivity, but it has few limitations. To overcome those limitations, we developed a novel quadruple score which included physical examination, ultrasound, mammography, and fine-needle aspiration cytology (FNAC).

Aim:

The aim of this study was to test the sensitivity, specificity, negative predictive value (NPV), and positive predictive value of a novel quadruple score for the initial screening of breast cancer.

Methodology:

This retrospective study was conducted on 320 patients having breast lump to the Breast Cancer Outpatient Department of AIIMS, Bhopal. The physical examination findings, breast sonography findings, mammography, and FNAC findings were recorded in the set pro forma, these were later analyzed, and a cumulative scoring was done categorizing patients in the benign, dubious, and malignant category according to the QTS score. These findings were then validated and compared by histopathology.

Results:

Statistical results show a positive covariance σ suggesting the positive relation between them; also, the correlation coefficient was 0.9996 suggesting a strong linear correlation. On comparing the accuracy of all four variables as physical examination, USG, mammography and FNAC individually and QTS, the accuracy of QTS was highest i.e 99.44%.

Conclusion:

QTS is accurate and reliable diagnostic approach for evaluation of breast lumps. By using the quadruple scoring system, management of palpable breast lump will become more streamlined, providing a platform for managing discordant results, which can be followed universally.

INTRODUCTION

Breast lump is the most common presentation of breast diseases which can be of various etiologies, benign or malignant, management and prognosis of which depends on early diagnosis and treatment. Although benign breast lump is more common, malignant breast lump is more dreaded as cancer of the breast, and is one of the most frequently encountered cancers recognized in 2.58 per million women taking mortality to 1.27 per million women.[1] Thirteen percent will have encroaching breast cancer in total life span, and the death rate due to carcinoma breast will be 1 woman out of 39 (3%) women.[1]

The diagnostic approach of triple assessment of palpable breast lumps to evaluate and distinguish between benign and malignant lumps in outpatient clinics, is a quick, cost-effective, least invasive method with accuracy comparable to open biopsy.[2]

The accuracy of diagnosis of breast cancer even in the most experienced hands-on physical examination is only 70%;[3] for mammography alone, it is 82%,[4] and for fine-needle aspiration cytology (FNAC); it is 78%.[4] Later on, a new score called modified triple score incorporating breast sonography, instead of mammography, was made, taking into consideration that mammography has low specificity in the young female population owing to dense breast and 30% of breast carcinoma is present in younger population. When all the components of triple test assessment (TTS) or modified TTS point to one possibility (are concordant), then the diagnosis is almost certain and management can be confidently planned in such a situation. However, about 30% of cases were still discordant that is differing in their interpretation of the breast lump.[4] In QTS, sonography and mammography are taken as separate investigating modalities to improve upon the diagnostic accuracy of the score.

In QTS, sonography when used in conjunction with physical, mammographic examination and FNAC is a more accurate diagnostic tool even in younger females. By bringing in the quadruple scoring system, management of palpable breast lump will become more streamlined, providing a platform for managing discordant results, which can be followed universally, thus empowering surgeon to go ahead in managing breast lump effectively and confidently. The present study was performed to determine the role of quadruple test in palpable breast lesions for providing one-stop diagnostic tool and to evaluate performance characteristics of its components individually and combined keeping histopathological examination as a reference standard for streamlining the diagnostic protocol for palpable breast lump.

METHODOLOGY

The present study was undertaken as a retrospective pilot study with the aim to estimate quadruple score in preoperative diagnosis of perceptible breast lesions for developing a standard protocol for management of breast lump and to test the validity of the Quadruple Score in making a pre-procedural diagnosis of palpable breast lumps so that an prospective trial can be done.

Medical records of patients aged 20 years or above with clinically palpable breast lump from September 1, 2017, to August 31, 2018, including history/physical examination, ultrasonography (USG), mammography, and FNAC in association with tru-cut or excisional pathology results, were retrospectively analyzed.

Inclusion criteria

1. All patients of age >20 years with palpable breast lump who underwent triple assessment were selected to participate in the present trial.

Criteria for exclusion

1. Patients with the known presence of breast carcinoma before attending breast clinic

2. Patients whose results of all three elements of triple test were not clearly recorded.

Scoring background

The scoring was done by keeping the following scoring criteria's for physical examination, USG, mammography, and biopsy of the specimen. As per Best practice guidelines for patients presenting with breast symptoms and American College of Radiology Breast Imaging Reporting and Data System (BI-RADS).[5,6] The malignancy level will be recorded using 1–5 scale.[5] The scoring of each of the component of QTS score was done separately which was later added to achieve the designed cumulative QTS score which could finally classify the lump as benign, dubious, or malignant. The histopathology report of specimens obtained by core-cut or excisional biopsy of the breast mass was correlated with quadruple assessment of breast lump.[6] All the scores are mentioned in Table 1.

Table 1.

Proposed quadruple test screening score

Clinical exam	USG	Mammography	FNAC/biopsy	Result
P-1=Normal	U-1=Normal (BIRADS-I)	M-1=Normal (BIRADS-I)	C-1/B1 unsatisfactory/normal tissue	4-7
P-2=Benign	U-2=Benign (BIRADS-II)	M-2=Benign (BIRADS-II)	C-2/B2-benign	8-11
P-3=Uncertain	U-3=Uncertain (BIRADS-III)	M-3=Probably benign (BIRADS-III)	C-3/B3-probably benign	12-15
P-4=Dubious	U-4=Dubious (BIRADS-IV)	M-4=Probably malignant-(BIRADS-IV)	C-4/B4 dubious of malignancy	16-19
P-5=Malignant	U-5=Malignant (BIRADS-V)	M-5-malignant (BIRADS-V)	C-5/B-5 malignant	20

Interpretation of score: Benign=Score 4-11, Dubious=Score 12-19, Malignant=Score >20. FNAC: Fine-needle aspiration cytology, BIRADS: Breast imaging-reporting and data system, USG: Ultrasonography

Statistical analysis

The results of the study were then analyzed individually and as a combination using standard statistical software. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of quadruple score are calculated according to histopathologic results. A positive covariance means that the two variables are positively related and they have the same direction, and a correlation coefficient uses values from − 1 to 1; correlation with greater absolute value shows a stronger linear correlation. Comparison between the age groups was done by Tukey's honestly significant difference test.

RESULTS

Totally 320 patients were included in the study, out of which 7 were males and 313 were females. All the patients were divided into 6 (A–F): 5 (A–E) groups according to the age [Table 2].

Table 2.

The demographic data for the study participants

Age groups (years)	Total number	Sex		Urban/rural		Average age of menarche±SD (females)
		Male	Female	Urban	Rural
15-19 (A)	35	0	35	22	13	16.23±0.85
20-29 (B)	103	5	98	70	33	14.016±1.43
30-39 (C)	88	0	88	63	25	13.78±1.68
40-49 (D)	63	0	63	31	32	14.87±1.68
50-59 (E)	18	2	16	14	4	13.86±1.36
≥60 (F)	13	0	13	10	3	4.33±0.47
Total	320	7	313	210	110

SD: Standard deviation

In our study, individuals were divided into six groups according to the age and demographic data were recorded. Table 3 shows the clinical examination scoring in the studied age groups.

Table 3.

The clinical examination scoring in the studied age groups

Age groups (years)	15-19 (A)	20-29 (B)	30-39 (C)	40-49 (D)	50-59 (E)	≥60 (F)	Total
Clinical exam score
P1	5	24	24	17	4	2	76
P2	12	28	25	22	5	2	94
P3	18	48	22	12	7	3	110
P4	0	3	12	6	1	3	25
P5	0	0	5	6	1	3	15

Tukey’s HSD P value ranged from 0.1393051 to −0.8999947, which is statistically insignificant age-wise clinical examination. HSD: Honestly significant difference

Table 4 shows the scores of USG. It was found that 81, 106, 88, 23, and 22 patients were categorized in U1, U2, U3, U4, and U5, respectively.

Table 4.

The ultrasonography scores

Age groups (years)	15-19 (A)	20-29 (B)	30-39 (C)	40-49 (D)	50-59 (E)	≥60 (F)	Total
USG score
U1	4	25	26	19	4	3	81
U2	10	38	30	19	7	2	106
U3	18	29	22	14	4	1	88
U4	2	9	7	2	0	3	23
U5	1	2	3	9	3	4	22

Tukey’s HSD P value ranged from 0.1119184 to −0.8999947, which is statistically insignificant age-wise USG scores. USG: Ultrasonography, HSD: Honestly significant difference

As per Table 5, a statistically nonsignificant difference was found when the different age groups were compared as per the clinical examination scoring, suggesting that no age group was more inclined toward any category.

Table 5.

The scoring of proposed quadruple test screening

Quadruple score (age groups)	15-19 (A)	20-29 (B)	30-39 (C)	40-49 (D)	50-59 (E)	≥60 (F)	Total
Benign (4-11)	35	102	85	57	15	7	301
Dubious (12-19)		1	2	4	2	5	14
Malignant (≥20)			1	2	1	1	5

QTS: Quadruple test screening

Table 6 shows the histopathology of all the studied patients. According to this 301 were benign and 5 were malignant 14 were of uncertain malignant potential

Table 6.

The scores of histopathology

Age groups (years)	15-19 (A)	20-29 (B)	30-39 (C)	40-49 (D)	50-59 (E)	≥60 (F)	Total
Histopathology
B unsatisfactory/normal tissue							0
B2 for benign	35	103	87	61	17	12	315
B3 for uncertain			1				1
B4 for dubious					1		1
B5 for malignant				1	1	1	3

Table 7 shows the correlation of QTS score with histopathology. Statistical results show a positive covariance σ suggesting the positive relation between them, and also, the correlation coefficient was 0.9996 suggesting a strong linear correlation.

Table 7.

Correlation between quadruple test screening and histopathology

Quad score	Histopathology
	Benign	Dubious and uncertain malignant potential	Malignant	Total
Benign	301			301
Dubious and uncertain malignant potential	12	1	1	14
Malignant	2	1	2	5
Total	315	2	3	320

Covariance σ: 20241.5556 A positive covariance means that the two variables are positively related and they have the same direction. “Correlation coefficient” 0.9996 (correlation coefficient uses values from −1 to 1, correlation with greater absolute value has a stronger linear correlation)

Table 8 shows the comparisons of the variables and accuracy of physical examination score, USG, mammography, FNAC, and QTS. We found that the accuracy of QTS was highest among all four, i.e. 99.44%.

Table 8.

The comparison of specificity, sensitivity, positive predictive value and negative predictive value of physical examination score, ultrasonography, mammography, fine-needle aspiration cytology, and quadruple test score

Measure	PE (%)	USG (%)	Mammography (%)	FNAC (%)	QTS value (%)
Sensitivity	100	100	100	100	100.00
Specificity	96.2	93.99	94.3	96.23	99.37
PPV	78.22	69.40	70.54	78.3	157.50
NPV	100	100	100	100	0.00
Positive likelihood ratio	26.33	16.63	17.56	26.35	12.00
Negative likelihood ratio	0.00	0.00	0.00	0.00	95.55
Accuracy	96.66	94.71	94.99	96.6	99.44

PPV: Positive predictive value, NPV: Negative predictive value, FNAC: Fine-needle aspiration cytology, USG: Ultrasonography, QTS: Quadruple test screening, PE: Physical examination

DISCUSSION

The breakthrough in breast cancer screening came with the introduction of triple assessment, but still, this system had some drawbacks and so a modified triple assessment was introduced, even with this modified system about 30% of cases remained discordant, hence the need for improvising leads us to introduce this quadruple test score (QTS).[7]

In our study also, the patients presenting with lump in breast were more from urban areas as compared to the rural population.

It is evident by many studies that mammography screening[8] can early diagnose and hence it is helpful to reduce mortality in postmenopausal women. FNAC is a reliable tool with a good sensitivity and positive predictive value (PPV) but again dependent upon the experience of a cytopathologist. The accuracy of triple assessment can be increased if it is modified to QTS as shown in our study.

As per our clinical examination scoring, 110, 25, and 19 patients were in P3, P4, and P5 categories, respectively, although when comparing the age groups, no significant difference was observed, suggesting that the clinical examination score was independent of age although the malignancy was detected in the 30 years and above age group in histopathology. As per a study conducted by Anders et al.,[9] approximately 7% of the patients diagnosed with breast cancer are below 40 years of age, hence validating our study. This suggests that diagnosis only on the basis of physical examination has poor value.

In the USG examination, 88, 23, and 22 patients were in U3, U4, and U5 categories, respectively. As per the American Cancer Society, mammography in addition with the USG in comparison to only mammography can be more fruitful than alone. As per Rebolj et al.,[10] supplementary ultrasound screening increased detection of breast cancer.

As per our QTS score, 301 cases were benign, 14 were dubious, and 5 were malignant. When this result was compared with the histopathology, it showed that our QTS did not miss a single case. When the histopathology results were compared to the QTS results, we found that both the results were in a linear and positive relation validating our scoring system.

All 301 cases which were tested benign by QTS were benign in the histopathology as well. According to this 301 were benign and 2 were malignant 12 were of uncertain malignant potential. Five cases were reported malignant by QTS; out of them three were malignant in histopathology report.

The sensitivity of our score was 100%, suggesting that not even single potentially malignant case was missed, but the specificity was low 96.2%, suggesting that those tested positive for malignancy should be confirmed by histopathology because likelihood of false positivity is high with this score. The purpose of developing this score was for initial screening. The diagnostic tools for initial screening are required to be more sensitive rather than specific and our QTS was successful in that. The PPV of this proposed QTS tool was 26.23, but the negative predictive value (NPV) was 100%. These results are in consistent with the triple test score in breast carcinoma, a study conducted by Kharkwal et al.[11] to test the validity of triple test score, suggesting the diagnostic value of our tool.

A previous study conducted by Khalid[12] discovered QTS accuracy up to 99.44%, whereas the accuracy of triple score was only 83.3%. Our study also implicates similar results, suggesting that QTS provides a user-friendly and objective evaluation tool for screening breast cancer.

CONCLUSION

QTS, a novel scoring tool for breast cancer, was developed which is based on physical examination, ultrasound, mammography, and FNAC. This retrospective pilot study was conducted to test the validity of this score as an initial screening tool. We found that the sensitivity and NPV was 100%, emphasizing that the error or limitation of one modality was compensated by the other elements. When clinical signs and symptoms, USG and mammography and FNAC were combined which when correlated with true cut findings, diagnostic accuracy was 99.4%. This study shows that QTS is accurate and reliable diagnostic approach for evaluation of breast lumps.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.