I will give some warnings for your benefit and knowledge along with a direct answer to the question.
You will need to stop thinking in terms of keywords. You will suffer search performance as a result.
Google, for example, is a semantics based search engine. It does not make linear keyword matches, but evaluates HTML DOM elements and content using semantics. It also uses semantics to try and assess the search query intent and matches results based upon intent and not keywords. The SEO online world has done you a disservice in making you believe that this is how search works when in fact, it does not.
How semantic works:
Keyword A - keyword 1 does nothing for you from a semantics point of view. For example, What about keyword A and keyword 1? Are you selling them?, Reviewing them?, Explaining them?, etc.
What works and gives clues to what your page is about are sentences using semantic clues such as the subject, predicate, and object. A simple example is Bob threw the ball. where Bob is the subject, threw is the predicate, and ball is the object. It does not have to be this simple. However, even a simple sentence such as this can be analyzed in different ways such as The ball was thrown by Bob., Bob possessed the ball., Bob no longer possesses the ball., etc. While this still seems too simplistic, taking sentences together can create an amazing wealth of knowledge not directly stated in the context of the content or HTML DOM elements. For example, if the next sentence is Sally caught the ball., what can be inferred is that Bob threw the ball to Sally., Bob is close to Sally., Bob can see Sally., Bob has some sort of relationship to Sally., as well as the reverse also being true Sally is close to Bob., Sally can see Bob., and finally Sally has some sort of relationship to Bob., etc. In fact, the more that is available for semantic analysis the greater the inferred knowledge can be.
The results are staggering. One simple example I give in a paper I wrote results in a 20:1 (+/-) concrete knowledge increase using semantics over the original content. It is this knowledge increase that also influences search results. This form of semantics is known as a knowledge base and the link graph is known as a knowledge graph. Google not only makes use of this knowledge graph for sidebar information, but the knowledge graph has also taken over as a major part of determining search intent. Lest you think my example rather silly, I warn you that real knowledge and intent can be extracted in the same way and really does provide solid factual results. This technology has existed since the early to mid '70s and is being perfected and expanded by Google daily.
Why did I tell you this?
Because if you look hard, you will see sites that rank for topics and terms that are not even mentioned. They are inferred using semantic analysis. Again, it goes back to matching intent.
The more semantic clues you give, the more search results you will receive. Simple use of keywords defeats this effect.
You are always better off using a sentence or something near a sentence in title tags, link text, description meta-tags, header tags, etc. (taken in order of importance according to Google). These elements are analyzed along with the content to determine not only what the page is about, but what potential the page has to satisfy the users query. The longer and more semantic based the query, the more likely that Google will be able to make an ideal match.
As well, there is the click-trough rate (CTR) as a factor. You are not motivating or compelling the user to click. Again What about keyword A and keyword 1? You need to tell the user what you have and how it benefits the user. If your SERP link receives little CTR or higher bounce rates, Google uses the search query log and metrics such as CTR, bounce rate, time spent on page, time spent on site, etc. to determine how the SERP link performed. This also informs the semantic engine how to rank semantic results. You may see that Google will cluster similar results and that SERP links will often follow themes. This is based upon past performance and what the past user base has found to be useful for the query made. Your goal is to not only influence the semantic link graph, but also the result performance metrics for queries where your site may receive a SERP link.
It is in precisely this way you will improve your sites performance. Not using keywords, tricks, influencers, etc.
Now to the duplicate content part of your question:
Duplicate content applies to content and very little to HTML DOM elements. HTML DOM elements do not provide enough information to make this determination in of themselves.
Duplicate content is determined using a term index with proximity and plain 'ole semantic analysis. Each page of content will score topics and terms based upon not only the term, but proximity to other terms, phrases, the semantic definition of the term, it's relative position from the beginning of the content, it's relative position within each content block, it's relative position within clusters of content blocks as defined as being between header tags, the proficiency of the content, the grade level of the content in evaluating audience level, etc. There are a lot of scores made.
As well, the citations found by the citation engine, any misspelling, grammar, linguistic usage, etc., can determine authorship, similarity, origination, etc. to help evaluate if content was copied and modified.
It is the comparison of these scores made from page to page that determines if two pages are duplicates or near duplicates.
In this analysis, the title tag, link text, and description meta-tag are used, but score fairly low and do not really influence the duplicate scoring enough to matter much unless there is a wider indication of duplication through other means.
In other words, it is the content that matters most in determining duplication, however, it is best to avoid duplicate title tags as Google recommends and pages without duplicate title tags should perform better. I rather suspect that Google does give a lower score not only for semantics, but for title duplication. It goes to quality at this point and from a search engines perspective, duplicate title tags could be considered as unuseful or of little value and likely score lower or potentially ignored entirely- if not now, perhaps later.