Pharma Patent Search – A Challenging Task

This post was first published on 15th March 2011.
It is beyond contention that prior art search is an absolutely critical step in acquiring a patent, irrespective of the technology domain the invention relates to. Although considerable amount of time and effort needs to be spent for conducting a comprehensive patent search or any invention, in case of inventions pertaining to the pharmaceutical technology, the time and efforts triplicate. Here is why.
The technology space in the pharmaceutical sector is extremely crowded which in turn crowds the patent space. Although this is very encouraging for researchers, it is a nightmare for patent analysts who carry out patent search and analysis to earn their bread and butter.  When a comprehensive patent search in a non pharmaceutical sector would take 4-7 days, it would take 15-17 days in a pharmaceutical sector. Conducting a patent search for a pharmaceutical moiety is time consuming and is an extremely tedious because designing around patents is a very common feature in Pharmaceutical research and  hence it is very normal to find thousands of patents, which have a single point of difference between each other. To conduct a comprehensive patent search one needs to preferably conduct search in 4 steps.
Step 1: Keyword search
To conduct a keyword search it is necessary to know the biology of the disease the invention targets.
A list of keywords has to be generated considering the target the molecule modulates and the spectrum of indications it is effective for. For example, if my invention pertains to a new chemical entity is a cancer treatment drug that modulates p53,  my main focus for generating keywords would be on 1. p53 and 2. Cancer.
One must incorporate synonyms and alternatives for the target and indications. In this case fro example the alternatives for p53 are tumor protein 53, TP53, LFS1 and TRP53.
Once the list of keywords is generated, a query is created to include patent specifications comprising anyone of the keywords. In this case the query would be (p53 OR TP53 OR LFS1……) AND (Cancer OR Neoplasm OR malignant OR tumor OR tumour….).
Step 2: Classification Search
The International Patent Classification system has classified pharmaceutical patents under the A61K31 class as several subclasses based on the chemical moiety present in the pharmaceutical molecule. Each subclass may comprise of 300 to 20,000 patents.. For example if my invention has gluconic acid as one of the moieties, then the relevant class would be A61K 31/191 (This subclass comprises of patent documents pertaining to inventions having Acyclic acids having two or more hydroxy groups, e.g. gluconic acid.)
On identifying the right class for the molecule that needs to be searched for, all the patents of the relevant classes are looked through in greater detail to shortlist the relevant documents for further analysis. Some irrelevant patents can be discarded based on abstract and claims, whereas in some cases one may have to go through the entire specification to determine its relevance.
Step 3: Structure Search
Various databases such as STN, Scifinder, Reaxys, Surechem and so on enable structure search. These databases are linked to Chemdraw or Chemsketch which are softwares that enable drawing the chemical structure which can then be exported to the search database for carrying out the patent search.
Step 4: Top up search
Since to get to this step it would have taken a patent analyst at least 12 days, it is necessary to carry out a top up search using the same technique as provided in Step1 to 3 from the date of the search till date to ensure that no relevant patent is missed which has been published during the period of search.
If one goes through all these three stages meticulously, the search can be considered to be 90% efficient.  However, considering the size of prior art, the scope for human error is high and due care must be taken to minimize the same.
Authored by Ms. Vinita Radhakrishnan

Leave a comment